diff --git a/test/CodeGen/2007-06-18-SextAttrAggregate.c b/test/CodeGen/2007-06-18-SextAttrAggregate.c
index f548951..92171e2 100644
--- a/test/CodeGen/2007-06-18-SextAttrAggregate.c
+++ b/test/CodeGen/2007-06-18-SextAttrAggregate.c
@@ -1,5 +1,5 @@
 // RUN: %clang_cc1 %s -o - -emit-llvm | FileCheck %s
-// XFAIL: aarch64
+// XFAIL: aarch64, arm64
 
 // PR1513
 
diff --git a/test/CodeGen/2009-02-13-zerosize-union-field-ppc.c b/test/CodeGen/2009-02-13-zerosize-union-field-ppc.c
index 2bcc7c3..eabf463 100644
--- a/test/CodeGen/2009-02-13-zerosize-union-field-ppc.c
+++ b/test/CodeGen/2009-02-13-zerosize-union-field-ppc.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 %s -triple powerpc-pc-linux -emit-llvm -o - | grep {i32 32} | count 3
 // XFAIL: *
 //  Every printf has 'i32 0' for the GEP of the string; no point counting those.
diff --git a/test/CodeGen/2010-01-13-MemBarrier.c b/test/CodeGen/2010-01-13-MemBarrier.c
index c2b0acd..74d0cb0 100644
--- a/test/CodeGen/2010-01-13-MemBarrier.c
+++ b/test/CodeGen/2010-01-13-MemBarrier.c
@@ -1,5 +1,4 @@
 // RUN: %clang_cc1 %s -emit-llvm -o - | FileCheck %s
-// XFAIL: sparc
 // rdar://7536390
 
 typedef unsigned __INT32_TYPE__ uint32_t;
diff --git a/test/CodeGen/2010-06-17-asmcrash.c b/test/CodeGen/2010-06-17-asmcrash.c
index 1b5efd3..048e429 100644
--- a/test/CodeGen/2010-06-17-asmcrash.c
+++ b/test/CodeGen/2010-06-17-asmcrash.c
@@ -1,5 +1,6 @@
-// REQUIRES: x86-64-registered-target
-// RUN: %clang_cc1 -triple x86_64-unknown-unknown -O1 -S -o - %s | FileCheck %s
+// REQUIRES: x86-registered-target
+// RUN: %clang_cc1 -triple x86_64-unknown-unknown -O1 -no-integrated-as \
+// RUN:     -S -o - %s | FileCheck %s
 
 typedef long long int64_t;
 typedef unsigned char uint8_t;
diff --git a/test/CodeGen/PR4611-bitfield-layout.c b/test/CodeGen/PR4611-bitfield-layout.c
index a383f34..4f9abea 100644
--- a/test/CodeGen/PR4611-bitfield-layout.c
+++ b/test/CodeGen/PR4611-bitfield-layout.c
@@ -1,6 +1,6 @@
 // RUN: %clang_cc1 -triple i386-unknown-unknown %s -emit-llvm -o - | FileCheck %s
 //
-// CHECK: struct.object_entry = type { [4 x i8] }
+// CHECK: struct.object_entry = type { i32 }
 
 struct object_entry {
        unsigned int type:3, pack_id:16, depth:13;
diff --git a/test/CodeGen/PR8880.c b/test/CodeGen/PR8880.c
new file mode 100644
index 0000000..e03d2a4
--- /dev/null
+++ b/test/CodeGen/PR8880.c
@@ -0,0 +1,173 @@
+// RUN: %clang_cc1 -Wno-gcc-compat -emit-llvm -o - %s | FileCheck %s
+
+void pr8880_cg_1(int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_1(
+  int i, j;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+  for (i = 2; i != 10 ; i++ )
+// CHECK: [[OUTER_COND]]
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    for (j = 3 ; j < 22; (void)({ ++j; break; j;})) {
+// CHECK: [[INNER_COND]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+      *iptr = 7;
+// CHECK: store i32 7,
+// CHECK: br label %[[INNER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_INC]]
+
+// break in 3rd expression of inner loop causes branch to end of inner loop
+
+// CHECK: br label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_END]]
+    }
+// CHECK: br label %[[OUTER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_INC]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret
+}
+
+void pr8880_cg_2(int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_2(
+  int i, j;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+  for (i = 2; i != 10 ; i++ )
+// CHECK: [[OUTER_COND]]
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    for (j = 3 ; j < 22; (void)({ ++j; continue; j;})) {
+// CHECK: [[INNER_COND]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+      *iptr = 7;
+// CHECK: store i32 7,
+// CHECK: br label %[[INNER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_INC]]
+
+// continue in 3rd expression of inner loop causes branch to inc of inner loop
+
+// CHECK: br label %[[INNER_INC]]
+// CHECK: [[INNER_END]]
+    }
+// CHECK: br label %[[OUTER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_INC]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret
+}
+
+void pr8880_cg_3(int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_3(
+  int i, j;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+  for (i = 2 ; i != 10 ; i++ )
+// CHECK: [[OUTER_COND]]
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    for (j = 3 ; ({break; j;}); j++) {
+
+// break in 2nd expression of inner loop causes branch to end of inner loop
+
+// CHECK: [[INNER_COND]]
+// CHECK: br label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+      *iptr = 7;
+// CHECK: store i32 7,
+// CHECK: br label %[[INNER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_INC]]
+// CHECK: br label %[[INNER_COND]]
+    }
+// CHECK: [[INNER_END]]
+// CHECK: br label %[[OUTER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_INC]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret
+}
+
+void pr8880_cg_4(int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_4(
+  int i, j;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+  for (i = 2 ; i != 10 ; i++ )
+// CHECK: [[OUTER_COND]]
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    for (j = 3 ; ({continue; j;}); j++) {
+
+// continue in 2nd expression of inner loop causes branch to inc of inner loop
+
+// CHECK: [[INNER_COND]]
+// CHECK: br label %[[INNER_INC:[0-9A-Za-z$._]+]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+      *iptr = 7;
+// CHECK: store i32 7,
+// CHECK: br label %[[INNER_INC]]
+// CHECK: [[INNER_INC]]
+// CHECK: br label %[[INNER_COND]]
+    }
+// CHECK: [[INNER_END]]
+// CHECK: br label %[[OUTER_INC:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_INC]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret
+}
+
+void pr8880_cg_5(int x, int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_5(
+  int y = 5;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_COND]]
+  while(--x) {
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    while(({ break; --y; })) {
+// CHECK: [[INNER_COND]]
+// CHECK: br label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+      *iptr = 7;
+// CHECK: store i32 7,
+    }
+// CHECK: br label %[[INNER_COND]]
+  }
+// CHECK: [[INNER_END]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret void
+}
+
+void pr8880_cg_6(int x, int *iptr) {
+// CHECK-LABEL: define void @pr8880_cg_6(
+  int y = 5;
+// CHECK: br label %[[OUTER_COND:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_COND]]
+  while(--x) {
+// CHECK: label %[[OUTER_BODY:[0-9A-Za-z$._]+]], label %[[OUTER_END:[0-9A-Za-z$._]+]]
+// CHECK: [[OUTER_BODY]]
+// CHECK: br label %[[INNER_BODY:[0-9A-Za-z$._]+]]
+// CHECK: [[INNER_BODY]]
+    do {
+// CHECK: store i32 7,
+      *iptr = 7;
+// CHECK: br label %[[INNER_COND:[0-9A-Za-z$._]+]]
+    } while(({ break; --y; }));
+// CHECK: [[INNER_COND]]
+// CHECK: br label %[[INNER_END:[0-9A-Za-z$._]+]]
+// CHECK: label %[[INNER_BODY:[0-9A-Za-z$._]+]], label %[[INNER_END]]
+  }
+// CHECK: [[INNER_END]]
+// CHECK: br label %[[OUTER_COND]]
+// CHECK: [[OUTER_END]]
+// CHECK: ret void
+}
diff --git a/test/CodeGen/a15.c b/test/CodeGen/a15.c
deleted file mode 100644
index e4986d8..0000000
--- a/test/CodeGen/a15.c
+++ /dev/null
@@ -1,5 +0,0 @@
-// RUN: %clang -target armv7-none-linux-gnueabi -mcpu=cortex-a15 -emit-llvm -S %s  -o /dev/null
-
-int main() {
-  return 0;
-}
diff --git a/test/CodeGen/a5.c b/test/CodeGen/a5.c
deleted file mode 100644
index b342d35..0000000
--- a/test/CodeGen/a5.c
+++ /dev/null
@@ -1,5 +0,0 @@
-// RUN: %clang -target armv7-none-linux-gnueabi -mcpu=cortex-a5 -emit-llvm -S %s  -o /dev/null
-
-int main() {
-  return 0;
-}
diff --git a/test/CodeGen/aarch64-arguments.c b/test/CodeGen/aarch64-arguments.c
index a70dfb1..f875886 100644
--- a/test/CodeGen/aarch64-arguments.c
+++ b/test/CodeGen/aarch64-arguments.c
@@ -190,5 +190,10 @@
 void variadic(int a, ...);
 void f43(__fp16 *in) {
   variadic(42, *in);
-// CHECK: call void @variadic(i32 42, double
+// PCS: call void (i32, ...)* @variadic(i32 42, double
 }
+
+// Checking: `double' and `long double' have different machine types, so cannot both be in an HFA
+struct s44 { long double a; double b; };
+// PCS: define void @f44(%struct.s44*
+struct s44 f44() {}
diff --git a/test/CodeGen/aarch64-neon-2velem.c b/test/CodeGen/aarch64-neon-2velem.c
index 03f7df7..19c9b16 100644
--- a/test/CodeGen/aarch64-neon-2velem.c
+++ b/test/CodeGen/aarch64-neon-2velem.c
@@ -1,1698 +1,2457 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int16x4_t test_vmla_lane_s16(int16x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmla_lane_s16
+  // CHECK-LABEL: test_vmla_lane_s16
   return vmla_lane_s16(a, b, v, 3);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int16x8_t test_vmlaq_lane_s16(int16x8_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlaq_lane_s16
+  // CHECK-LABEL: test_vmlaq_lane_s16
   return vmlaq_lane_s16(a, b, v, 3);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vmla_lane_s32(int32x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmla_lane_s32
+  // CHECK-LABEL: test_vmla_lane_s32
   return vmla_lane_s32(a, b, v, 1);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlaq_lane_s32(int32x4_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlaq_lane_s32
+  // CHECK-LABEL: test_vmlaq_lane_s32
   return vmlaq_lane_s32(a, b, v, 1);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int16x4_t test_vmla_laneq_s16(int16x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmla_laneq_s16
+  // CHECK-LABEL: test_vmla_laneq_s16
   return vmla_laneq_s16(a, b, v, 7);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int16x8_t test_vmlaq_laneq_s16(int16x8_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlaq_laneq_s16
+  // CHECK-LABEL: test_vmlaq_laneq_s16
   return vmlaq_laneq_s16(a, b, v, 7);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int32x2_t test_vmla_laneq_s32(int32x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmla_laneq_s32
+  // CHECK-LABEL: test_vmla_laneq_s32
   return vmla_laneq_s32(a, b, v, 3);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlaq_laneq_s32(int32x4_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlaq_laneq_s32
+  // CHECK-LABEL: test_vmlaq_laneq_s32
   return vmlaq_laneq_s32(a, b, v, 3);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int16x4_t test_vmls_lane_s16(int16x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmls_lane_s16
+  // CHECK-LABEL: test_vmls_lane_s16
   return vmls_lane_s16(a, b, v, 3);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int16x8_t test_vmlsq_lane_s16(int16x8_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsq_lane_s16
+  // CHECK-LABEL: test_vmlsq_lane_s16
   return vmlsq_lane_s16(a, b, v, 3);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vmls_lane_s32(int32x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmls_lane_s32
+  // CHECK-LABEL: test_vmls_lane_s32
   return vmls_lane_s32(a, b, v, 1);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlsq_lane_s32(int32x4_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsq_lane_s32
+  // CHECK-LABEL: test_vmlsq_lane_s32
   return vmlsq_lane_s32(a, b, v, 1);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int16x4_t test_vmls_laneq_s16(int16x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmls_laneq_s16
+  // CHECK-LABEL: test_vmls_laneq_s16
   return vmls_laneq_s16(a, b, v, 7);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int16x8_t test_vmlsq_laneq_s16(int16x8_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsq_laneq_s16
+  // CHECK-LABEL: test_vmlsq_laneq_s16
   return vmlsq_laneq_s16(a, b, v, 7);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int32x2_t test_vmls_laneq_s32(int32x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmls_laneq_s32
+  // CHECK-LABEL: test_vmls_laneq_s32
   return vmls_laneq_s32(a, b, v, 3);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlsq_laneq_s32(int32x4_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsq_laneq_s32
+  // CHECK-LABEL: test_vmlsq_laneq_s32
   return vmlsq_laneq_s32(a, b, v, 3);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int16x4_t test_vmul_lane_s16(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vmul_lane_s16
+  // CHECK-LABEL: test_vmul_lane_s16
   return vmul_lane_s16(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int16x8_t test_vmulq_lane_s16(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vmulq_lane_s16
+  // CHECK-LABEL: test_vmulq_lane_s16
   return vmulq_lane_s16(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vmul_lane_s32(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vmul_lane_s32
+  // CHECK-LABEL: test_vmul_lane_s32
   return vmul_lane_s32(a, v, 1);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmulq_lane_s32(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vmulq_lane_s32
+  // CHECK-LABEL: test_vmulq_lane_s32
   return vmulq_lane_s32(a, v, 1);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 uint16x4_t test_vmul_lane_u16(uint16x4_t a, uint16x4_t v) {
-  // CHECK: test_vmul_lane_u16
+  // CHECK-LABEL: test_vmul_lane_u16
   return vmul_lane_u16(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 uint16x8_t test_vmulq_lane_u16(uint16x8_t a, uint16x4_t v) {
-  // CHECK: test_vmulq_lane_u16
+  // CHECK-LABEL: test_vmulq_lane_u16
   return vmulq_lane_u16(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 uint32x2_t test_vmul_lane_u32(uint32x2_t a, uint32x2_t v) {
-  // CHECK: test_vmul_lane_u32
+  // CHECK-LABEL: test_vmul_lane_u32
   return vmul_lane_u32(a, v, 1);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 uint32x4_t test_vmulq_lane_u32(uint32x4_t a, uint32x2_t v) {
-  // CHECK: test_vmulq_lane_u32
+  // CHECK-LABEL: test_vmulq_lane_u32
   return vmulq_lane_u32(a, v, 1);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int16x4_t test_vmul_laneq_s16(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vmul_laneq_s16
+  // CHECK-LABEL: test_vmul_laneq_s16
   return vmul_laneq_s16(a, v, 7);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int16x8_t test_vmulq_laneq_s16(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vmulq_laneq_s16
+  // CHECK-LABEL: test_vmulq_laneq_s16
   return vmulq_laneq_s16(a, v, 7);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int32x2_t test_vmul_laneq_s32(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vmul_laneq_s32
+  // CHECK-LABEL: test_vmul_laneq_s32
   return vmul_laneq_s32(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmulq_laneq_s32(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vmulq_laneq_s32
+  // CHECK-LABEL: test_vmulq_laneq_s32
   return vmulq_laneq_s32(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 uint16x4_t test_vmul_laneq_u16(uint16x4_t a, uint16x8_t v) {
-  // CHECK: test_vmul_laneq_u16
+  // CHECK-LABEL: test_vmul_laneq_u16
   return vmul_laneq_u16(a, v, 7);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 uint16x8_t test_vmulq_laneq_u16(uint16x8_t a, uint16x8_t v) {
-  // CHECK: test_vmulq_laneq_u16
+  // CHECK-LABEL: test_vmulq_laneq_u16
   return vmulq_laneq_u16(a, v, 7);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 uint32x2_t test_vmul_laneq_u32(uint32x2_t a, uint32x4_t v) {
-  // CHECK: test_vmul_laneq_u32
+  // CHECK-LABEL: test_vmul_laneq_u32
   return vmul_laneq_u32(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 uint32x4_t test_vmulq_laneq_u32(uint32x4_t a, uint32x4_t v) {
-  // CHECK: test_vmulq_laneq_u32
+  // CHECK-LABEL: test_vmulq_laneq_u32
   return vmulq_laneq_u32(a, v, 3);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 float32x2_t test_vfma_lane_f32(float32x2_t a, float32x2_t b, float32x2_t v) {
-  // CHECK: test_vfma_lane_f32
+  // CHECK-LABEL: test_vfma_lane_f32
   return vfma_lane_f32(a, b, v, 1);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 float32x4_t test_vfmaq_lane_f32(float32x4_t a, float32x4_t b, float32x2_t v) {
-  // CHECK: test_vfmaq_lane_f32
+  // CHECK-LABEL: test_vfmaq_lane_f32
   return vfmaq_lane_f32(a, b, v, 1);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float32x2_t test_vfma_laneq_f32(float32x2_t a, float32x2_t b, float32x4_t v) {
-  // CHECK: test_vfma_laneq_f32
+  // CHECK-LABEL: test_vfma_laneq_f32
   return vfma_laneq_f32(a, b, v, 3);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 float32x4_t test_vfmaq_laneq_f32(float32x4_t a, float32x4_t b, float32x4_t v) {
-  // CHECK: test_vfmaq_laneq_f32
+  // CHECK-LABEL: test_vfmaq_laneq_f32
   return vfmaq_laneq_f32(a, b, v, 3);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 float32x2_t test_vfms_lane_f32(float32x2_t a, float32x2_t b, float32x2_t v) {
-  // CHECK: test_vfms_lane_f32
+  // CHECK-LABEL: test_vfms_lane_f32
   return vfms_lane_f32(a, b, v, 1);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 float32x4_t test_vfmsq_lane_f32(float32x4_t a, float32x4_t b, float32x2_t v) {
-  // CHECK: test_vfmsq_lane_f32
+  // CHECK-LABEL: test_vfmsq_lane_f32
   return vfmsq_lane_f32(a, b, v, 1);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float32x2_t test_vfms_laneq_f32(float32x2_t a, float32x2_t b, float32x4_t v) {
-  // CHECK: test_vfms_laneq_f32
+  // CHECK-LABEL: test_vfms_laneq_f32
   return vfms_laneq_f32(a, b, v, 3);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 float32x4_t test_vfmsq_laneq_f32(float32x4_t a, float32x4_t b, float32x4_t v) {
-  // CHECK: test_vfmsq_laneq_f32
+  // CHECK-LABEL: test_vfmsq_laneq_f32
   return vfmsq_laneq_f32(a, b, v, 3);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 float64x2_t test_vfmaq_lane_f64(float64x2_t a, float64x2_t b, float64x1_t v) {
-  // CHECK: test_vfmaq_lane_f64
+  // CHECK-LABEL: test_vfmaq_lane_f64
   return vfmaq_lane_f64(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float64x2_t test_vfmaq_laneq_f64(float64x2_t a, float64x2_t b, float64x2_t v) {
-  // CHECK: test_vfmaq_laneq_f64
+  // CHECK-LABEL: test_vfmaq_laneq_f64
   return vfmaq_laneq_f64(a, b, v, 1);
   // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[1]
 }
 
 float64x2_t test_vfmsq_lane_f64(float64x2_t a, float64x2_t b, float64x1_t v) {
-  // CHECK: test_vfmsq_lane_f64
+  // CHECK-LABEL: test_vfmsq_lane_f64
   return vfmsq_lane_f64(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float64x2_t test_vfmsq_laneq_f64(float64x2_t a, float64x2_t b, float64x2_t v) {
-  // CHECK: test_vfmsq_laneq_f64
+  // CHECK-LABEL: test_vfmsq_laneq_f64
   return vfmsq_laneq_f64(a, b, v, 1);
   // CHECK: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[1]
 }
 
+float32_t test_vfmas_laneq_f32(float32_t a, float32_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vfmas_laneq_f32
+  return vfmas_laneq_f32(a, b, v, 3);
+  // CHECK: fmla {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
+}
+
+float64_t test_vfmsd_lane_f64(float64_t a, float64_t b, float64x1_t v) {
+  // CHECK-LABEL: test_vfmsd_lane_f64
+  return vfmsd_lane_f64(a, b, v, 0);
+  // CHECK: {{fmls d[0-9]+, d[0-9]+, v[0-9]+\.d\[0\]|fmsub d[0-9]+, d[0-9]+, d[0-9]+}}
+}
+
+float32_t test_vfmss_laneq_f32(float32_t a, float32_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vfmss_laneq_f32
+  return vfmss_laneq_f32(a, b, v, 3);
+  // CHECK: fmls {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
+}
+
+float64_t test_vfmsd_laneq_f64(float64_t a, float64_t b, float64x2_t v) {
+  // CHECK-LABEL: test_vfmsd_laneq_f64
+  return vfmsd_laneq_f64(a, b, v, 1);
+  // CHECK: fmls {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
+}
+
 int32x4_t test_vmlal_lane_s16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlal_lane_s16
+  // CHECK-LABEL: test_vmlal_lane_s16
   return vmlal_lane_s16(a, b, v, 3);
   // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlal_lane_s32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlal_lane_s32
+  // CHECK-LABEL: test_vmlal_lane_s32
   return vmlal_lane_s32(a, b, v, 1);
   // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlal_laneq_s16(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlal_laneq_s16
+  // CHECK-LABEL: test_vmlal_laneq_s16
   return vmlal_laneq_s16(a, b, v, 7);
   // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlal_laneq_s32(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlal_laneq_s32
+  // CHECK-LABEL: test_vmlal_laneq_s32
   return vmlal_laneq_s32(a, b, v, 3);
   // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlal_high_lane_s16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlal_high_lane_s16
+  // CHECK-LABEL: test_vmlal_high_lane_s16
   return vmlal_high_lane_s16(a, b, v, 3);
   // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlal_high_lane_s32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlal_high_lane_s32
+  // CHECK-LABEL: test_vmlal_high_lane_s32
   return vmlal_high_lane_s32(a, b, v, 1);
   // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlal_high_laneq_s16(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlal_high_laneq_s16
+  // CHECK-LABEL: test_vmlal_high_laneq_s16
   return vmlal_high_laneq_s16(a, b, v, 7);
   // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlal_high_laneq_s32(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlal_high_laneq_s32
+  // CHECK-LABEL: test_vmlal_high_laneq_s32
   return vmlal_high_laneq_s32(a, b, v, 3);
   // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlsl_lane_s16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_lane_s16
+  // CHECK-LABEL: test_vmlsl_lane_s16
   return vmlsl_lane_s16(a, b, v, 3);
   // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlsl_lane_s32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_lane_s32
+  // CHECK-LABEL: test_vmlsl_lane_s32
   return vmlsl_lane_s32(a, b, v, 1);
   // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlsl_laneq_s16(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_laneq_s16
+  // CHECK-LABEL: test_vmlsl_laneq_s16
   return vmlsl_laneq_s16(a, b, v, 7);
   // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlsl_laneq_s32(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_laneq_s32
+  // CHECK-LABEL: test_vmlsl_laneq_s32
   return vmlsl_laneq_s32(a, b, v, 3);
   // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlsl_high_lane_s16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_high_lane_s16
+  // CHECK-LABEL: test_vmlsl_high_lane_s16
   return vmlsl_high_lane_s16(a, b, v, 3);
   // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlsl_high_lane_s32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_high_lane_s32
+  // CHECK-LABEL: test_vmlsl_high_lane_s32
   return vmlsl_high_lane_s32(a, b, v, 1);
   // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlsl_high_laneq_s16(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_high_laneq_s16
+  // CHECK-LABEL: test_vmlsl_high_laneq_s16
   return vmlsl_high_laneq_s16(a, b, v, 7);
   // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlsl_high_laneq_s32(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_high_laneq_s32
+  // CHECK-LABEL: test_vmlsl_high_laneq_s32
   return vmlsl_high_laneq_s32(a, b, v, 3);
   // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlal_lane_u16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlal_lane_u16
+  // CHECK-LABEL: test_vmlal_lane_u16
   return vmlal_lane_u16(a, b, v, 3);
   // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlal_lane_u32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlal_lane_u32
+  // CHECK-LABEL: test_vmlal_lane_u32
   return vmlal_lane_u32(a, b, v, 1);
   // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlal_laneq_u16(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlal_laneq_u16
+  // CHECK-LABEL: test_vmlal_laneq_u16
   return vmlal_laneq_u16(a, b, v, 7);
   // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlal_laneq_u32(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlal_laneq_u32
+  // CHECK-LABEL: test_vmlal_laneq_u32
   return vmlal_laneq_u32(a, b, v, 3);
   // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlal_high_lane_u16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlal_high_lane_u16
+  // CHECK-LABEL: test_vmlal_high_lane_u16
   return vmlal_high_lane_u16(a, b, v, 3);
   // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlal_high_lane_u32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlal_high_lane_u32
+  // CHECK-LABEL: test_vmlal_high_lane_u32
   return vmlal_high_lane_u32(a, b, v, 1);
   // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlal_high_laneq_u16(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlal_high_laneq_u16
+  // CHECK-LABEL: test_vmlal_high_laneq_u16
   return vmlal_high_laneq_u16(a, b, v, 7);
   // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlal_high_laneq_u32(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlal_high_laneq_u32
+  // CHECK-LABEL: test_vmlal_high_laneq_u32
   return vmlal_high_laneq_u32(a, b, v, 3);
   // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlsl_lane_u16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_lane_u16
+  // CHECK-LABEL: test_vmlsl_lane_u16
   return vmlsl_lane_u16(a, b, v, 3);
   // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlsl_lane_u32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_lane_u32
+  // CHECK-LABEL: test_vmlsl_lane_u32
   return vmlsl_lane_u32(a, b, v, 1);
   // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlsl_laneq_u16(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_laneq_u16
+  // CHECK-LABEL: test_vmlsl_laneq_u16
   return vmlsl_laneq_u16(a, b, v, 7);
   // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlsl_laneq_u32(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_laneq_u32
+  // CHECK-LABEL: test_vmlsl_laneq_u32
   return vmlsl_laneq_u32(a, b, v, 3);
   // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmlsl_high_lane_u16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_high_lane_u16
+  // CHECK-LABEL: test_vmlsl_high_lane_u16
   return vmlsl_high_lane_u16(a, b, v, 3);
   // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmlsl_high_lane_u32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_high_lane_u32
+  // CHECK-LABEL: test_vmlsl_high_lane_u32
   return vmlsl_high_lane_u32(a, b, v, 1);
   // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmlsl_high_laneq_u16(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_high_laneq_u16
+  // CHECK-LABEL: test_vmlsl_high_laneq_u16
   return vmlsl_high_laneq_u16(a, b, v, 7);
   // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmlsl_high_laneq_u32(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_high_laneq_u32
+  // CHECK-LABEL: test_vmlsl_high_laneq_u32
   return vmlsl_high_laneq_u32(a, b, v, 3);
   // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmull_lane_s16(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vmull_lane_s16
+  // CHECK-LABEL: test_vmull_lane_s16
   return vmull_lane_s16(a, v, 3);
   // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmull_lane_s32(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vmull_lane_s32
+  // CHECK-LABEL: test_vmull_lane_s32
   return vmull_lane_s32(a, v, 1);
   // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 uint32x4_t test_vmull_lane_u16(uint16x4_t a, uint16x4_t v) {
-  // CHECK: test_vmull_lane_u16
+  // CHECK-LABEL: test_vmull_lane_u16
   return vmull_lane_u16(a, v, 3);
   // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 uint64x2_t test_vmull_lane_u32(uint32x2_t a, uint32x2_t v) {
-  // CHECK: test_vmull_lane_u32
+  // CHECK-LABEL: test_vmull_lane_u32
   return vmull_lane_u32(a, v, 1);
   // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmull_high_lane_s16(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vmull_high_lane_s16
+  // CHECK-LABEL: test_vmull_high_lane_s16
   return vmull_high_lane_s16(a, v, 3);
   // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vmull_high_lane_s32(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vmull_high_lane_s32
+  // CHECK-LABEL: test_vmull_high_lane_s32
   return vmull_high_lane_s32(a, v, 1);
   // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 uint32x4_t test_vmull_high_lane_u16(uint16x8_t a, uint16x4_t v) {
-  // CHECK: test_vmull_high_lane_u16
+  // CHECK-LABEL: test_vmull_high_lane_u16
   return vmull_high_lane_u16(a, v, 3);
   // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 uint64x2_t test_vmull_high_lane_u32(uint32x4_t a, uint32x2_t v) {
-  // CHECK: test_vmull_high_lane_u32
+  // CHECK-LABEL: test_vmull_high_lane_u32
   return vmull_high_lane_u32(a, v, 1);
   // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vmull_laneq_s16(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vmull_laneq_s16
+  // CHECK-LABEL: test_vmull_laneq_s16
   return vmull_laneq_s16(a, v, 7);
   // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmull_laneq_s32(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vmull_laneq_s32
+  // CHECK-LABEL: test_vmull_laneq_s32
   return vmull_laneq_s32(a, v, 3);
   // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 uint32x4_t test_vmull_laneq_u16(uint16x4_t a, uint16x8_t v) {
-  // CHECK: test_vmull_laneq_u16
+  // CHECK-LABEL: test_vmull_laneq_u16
   return vmull_laneq_u16(a, v, 7);
   // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
 }
 
 uint64x2_t test_vmull_laneq_u32(uint32x2_t a, uint32x4_t v) {
-  // CHECK: test_vmull_laneq_u32
+  // CHECK-LABEL: test_vmull_laneq_u32
   return vmull_laneq_u32(a, v, 3);
   // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vmull_high_laneq_s16(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vmull_high_laneq_s16
+  // CHECK-LABEL: test_vmull_high_laneq_s16
   return vmull_high_laneq_s16(a, v, 7);
   // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vmull_high_laneq_s32(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vmull_high_laneq_s32
+  // CHECK-LABEL: test_vmull_high_laneq_s32
   return vmull_high_laneq_s32(a, v, 3);
   // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 uint32x4_t test_vmull_high_laneq_u16(uint16x8_t a, uint16x8_t v) {
-  // CHECK: test_vmull_high_laneq_u16
+  // CHECK-LABEL: test_vmull_high_laneq_u16
   return vmull_high_laneq_u16(a, v, 7);
   // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 uint64x2_t test_vmull_high_laneq_u32(uint32x4_t a, uint32x4_t v) {
-  // CHECK: test_vmull_high_laneq_u32
+  // CHECK-LABEL: test_vmull_high_laneq_u32
   return vmull_high_laneq_u32(a, v, 3);
   // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vqdmlal_lane_s16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vqdmlal_lane_s16
+  // CHECK-LABEL: test_vqdmlal_lane_s16
   return vqdmlal_lane_s16(a, b, v, 3);
   // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmlal_lane_s32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vqdmlal_lane_s32
+  // CHECK-LABEL: test_vqdmlal_lane_s32
   return vqdmlal_lane_s32(a, b, v, 1);
   // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmlal_high_lane_s16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vqdmlal_high_lane_s16
+  // CHECK-LABEL: test_vqdmlal_high_lane_s16
   return vqdmlal_high_lane_s16(a, b, v, 3);
   // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmlal_high_lane_s32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vqdmlal_high_lane_s32
+  // CHECK-LABEL: test_vqdmlal_high_lane_s32
   return vqdmlal_high_lane_s32(a, b, v, 1);
   // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmlsl_lane_s16(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vqdmlsl_lane_s16
+  // CHECK-LABEL: test_vqdmlsl_lane_s16
   return vqdmlsl_lane_s16(a, b, v, 3);
   // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmlsl_lane_s32(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vqdmlsl_lane_s32
+  // CHECK-LABEL: test_vqdmlsl_lane_s32
   return vqdmlsl_lane_s32(a, b, v, 1);
   // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmlsl_high_lane_s16(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vqdmlsl_high_lane_s16
+  // CHECK-LABEL: test_vqdmlsl_high_lane_s16
   return vqdmlsl_high_lane_s16(a, b, v, 3);
   // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmlsl_high_lane_s32(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vqdmlsl_high_lane_s32
+  // CHECK-LABEL: test_vqdmlsl_high_lane_s32
   return vqdmlsl_high_lane_s32(a, b, v, 1);
   // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmull_lane_s16(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqdmull_lane_s16
+  // CHECK-LABEL: test_vqdmull_lane_s16
   return vqdmull_lane_s16(a, v, 3);
   // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmull_lane_s32(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqdmull_lane_s32
+  // CHECK-LABEL: test_vqdmull_lane_s32
   return vqdmull_lane_s32(a, v, 1);
   // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmull_laneq_s16(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vqdmull_laneq_s16
+  // CHECK-LABEL: test_vqdmull_laneq_s16
   return vqdmull_laneq_s16(a, v, 3);
   // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmull_laneq_s32(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vqdmull_laneq_s32
+  // CHECK-LABEL: test_vqdmull_laneq_s32
   return vqdmull_laneq_s32(a, v, 3);
   // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 int32x4_t test_vqdmull_high_lane_s16(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqdmull_high_lane_s16
+  // CHECK-LABEL: test_vqdmull_high_lane_s16
   return vqdmull_high_lane_s16(a, v, 3);
   // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int64x2_t test_vqdmull_high_lane_s32(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqdmull_high_lane_s32
+  // CHECK-LABEL: test_vqdmull_high_lane_s32
   return vqdmull_high_lane_s32(a, v, 1);
   // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmull_high_laneq_s16(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vqdmull_high_laneq_s16
+  // CHECK-LABEL: test_vqdmull_high_laneq_s16
   return vqdmull_high_laneq_s16(a, v, 7);
   // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
 }
 
 int64x2_t test_vqdmull_high_laneq_s32(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vqdmull_high_laneq_s32
+  // CHECK-LABEL: test_vqdmull_high_laneq_s32
   return vqdmull_high_laneq_s32(a, v, 3);
   // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 int16x4_t test_vqdmulh_lane_s16(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqdmulh_lane_s16
+  // CHECK-LABEL: test_vqdmulh_lane_s16
   return vqdmulh_lane_s16(a, v, 3);
   // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int16x8_t test_vqdmulhq_lane_s16(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqdmulhq_lane_s16
+  // CHECK-LABEL: test_vqdmulhq_lane_s16
   return vqdmulhq_lane_s16(a, v, 3);
   // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vqdmulh_lane_s32(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqdmulh_lane_s32
+  // CHECK-LABEL: test_vqdmulh_lane_s32
   return vqdmulh_lane_s32(a, v, 1);
   // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqdmulhq_lane_s32(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqdmulhq_lane_s32
+  // CHECK-LABEL: test_vqdmulhq_lane_s32
   return vqdmulhq_lane_s32(a, v, 1);
   // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int16x4_t test_vqrdmulh_lane_s16(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqrdmulh_lane_s16
+  // CHECK-LABEL: test_vqrdmulh_lane_s16
   return vqrdmulh_lane_s16(a, v, 3);
   // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
 }
 
 int16x8_t test_vqrdmulhq_lane_s16(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqrdmulhq_lane_s16
+  // CHECK-LABEL: test_vqrdmulhq_lane_s16
   return vqrdmulhq_lane_s16(a, v, 3);
   // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vqrdmulh_lane_s32(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqrdmulh_lane_s32
+  // CHECK-LABEL: test_vqrdmulh_lane_s32
   return vqrdmulh_lane_s32(a, v, 1);
   // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int32x4_t test_vqrdmulhq_lane_s32(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqrdmulhq_lane_s32
+  // CHECK-LABEL: test_vqrdmulhq_lane_s32
   return vqrdmulhq_lane_s32(a, v, 1);
   // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float32x2_t test_vmul_lane_f32(float32x2_t a, float32x2_t v) {
-  // CHECK: test_vmul_lane_f32
+  // CHECK-LABEL: test_vmul_lane_f32
   return vmul_lane_f32(a, v, 1);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 
 float64x1_t test_vmul_lane_f64(float64x1_t a, float64x1_t v) {
-  // CHECK: test_vmul_lane_f64
+  // CHECK-LABEL: test_vmul_lane_f64
   return vmul_lane_f64(a, v, 0);
-  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+\.d\[0\]|d[0-9]+}}
 }
 
 
 float32x4_t test_vmulq_lane_f32(float32x4_t a, float32x2_t v) {
-  // CHECK: test_vmulq_lane_f32
+  // CHECK-LABEL: test_vmulq_lane_f32
   return vmulq_lane_f32(a, v, 1);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float64x2_t test_vmulq_lane_f64(float64x2_t a, float64x1_t v) {
-  // CHECK: test_vmulq_lane_f64
+  // CHECK-LABEL: test_vmulq_lane_f64
   return vmulq_lane_f64(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float32x2_t test_vmul_laneq_f32(float32x2_t a, float32x4_t v) {
-  // CHECK: test_vmul_laneq_f32
+  // CHECK-LABEL: test_vmul_laneq_f32
   return vmul_laneq_f32(a, v, 3);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 float64x1_t test_vmul_laneq_f64(float64x1_t a, float64x2_t v) {
-  // CHECK: test_vmul_laneq_f64
+  // CHECK-LABEL: test_vmul_laneq_f64
   return vmul_laneq_f64(a, v, 1);
   // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 
 float32x4_t test_vmulq_laneq_f32(float32x4_t a, float32x4_t v) {
-  // CHECK: test_vmulq_laneq_f32
+  // CHECK-LABEL: test_vmulq_laneq_f32
   return vmulq_laneq_f32(a, v, 3);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 float64x2_t test_vmulq_laneq_f64(float64x2_t a, float64x2_t v) {
-  // CHECK: test_vmulq_laneq_f64
+  // CHECK-LABEL: test_vmulq_laneq_f64
   return vmulq_laneq_f64(a, v, 1);
   // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[1]
 }
 
 float32x2_t test_vmulx_lane_f32(float32x2_t a, float32x2_t v) {
-  // CHECK: test_vmulx_lane_f32
+  // CHECK-LABEL: test_vmulx_lane_f32
   return vmulx_lane_f32(a, v, 1);
   // CHECK: fmulx {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 float32x4_t test_vmulxq_lane_f32(float32x4_t a, float32x2_t v) {
-  // CHECK: test_vmulxq_lane_f32
+  // CHECK-LABEL: test_vmulxq_lane_f32
   return vmulxq_lane_f32(a, v, 1);
   // CHECK: fmulx {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float64x2_t test_vmulxq_lane_f64(float64x2_t a, float64x1_t v) {
-  // CHECK: test_vmulxq_lane_f64
+  // CHECK-LABEL: test_vmulxq_lane_f64
   return vmulxq_lane_f64(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float32x2_t test_vmulx_laneq_f32(float32x2_t a, float32x4_t v) {
-  // CHECK: test_vmulx_laneq_f32
+  // CHECK-LABEL: test_vmulx_laneq_f32
   return vmulx_laneq_f32(a, v, 3);
   // CHECK: fmulx {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
 }
 
 float32x4_t test_vmulxq_laneq_f32(float32x4_t a, float32x4_t v) {
-  // CHECK: test_vmulxq_laneq_f32
+  // CHECK-LABEL: test_vmulxq_laneq_f32
   return vmulxq_laneq_f32(a, v, 3);
   // CHECK: fmulx {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
 }
 
 float64x2_t test_vmulxq_laneq_f64(float64x2_t a, float64x2_t v) {
-  // CHECK: test_vmulxq_laneq_f64
+  // CHECK-LABEL: test_vmulxq_laneq_f64
   return vmulxq_laneq_f64(a, v, 1);
   // CHECK: fmulx {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[1]
 }
 
 int16x4_t test_vmla_lane_s16_0(int16x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmla_lane_s16_0
+  // CHECK-LABEL: test_vmla_lane_s16_0
   return vmla_lane_s16(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmlaq_lane_s16_0(int16x8_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlaq_lane_s16_0
+  // CHECK-LABEL: test_vmlaq_lane_s16_0
   return vmlaq_lane_s16(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmla_lane_s32_0(int32x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmla_lane_s32_0
+  // CHECK-LABEL: test_vmla_lane_s32_0
   return vmla_lane_s32(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlaq_lane_s32_0(int32x4_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlaq_lane_s32_0
+  // CHECK-LABEL: test_vmlaq_lane_s32_0
   return vmlaq_lane_s32(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vmla_laneq_s16_0(int16x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmla_laneq_s16_0
+  // CHECK-LABEL: test_vmla_laneq_s16_0
   return vmla_laneq_s16(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmlaq_laneq_s16_0(int16x8_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlaq_laneq_s16_0
+  // CHECK-LABEL: test_vmlaq_laneq_s16_0
   return vmlaq_laneq_s16(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmla_laneq_s32_0(int32x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmla_laneq_s32_0
+  // CHECK-LABEL: test_vmla_laneq_s32_0
   return vmla_laneq_s32(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlaq_laneq_s32_0(int32x4_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlaq_laneq_s32_0
+  // CHECK-LABEL: test_vmlaq_laneq_s32_0
   return vmlaq_laneq_s32(a, b, v, 0);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vmls_lane_s16_0(int16x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmls_lane_s16_0
+  // CHECK-LABEL: test_vmls_lane_s16_0
   return vmls_lane_s16(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmlsq_lane_s16_0(int16x8_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsq_lane_s16_0
+  // CHECK-LABEL: test_vmlsq_lane_s16_0
   return vmlsq_lane_s16(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmls_lane_s32_0(int32x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmls_lane_s32_0
+  // CHECK-LABEL: test_vmls_lane_s32_0
   return vmls_lane_s32(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsq_lane_s32_0(int32x4_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsq_lane_s32_0
+  // CHECK-LABEL: test_vmlsq_lane_s32_0
   return vmlsq_lane_s32(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vmls_laneq_s16_0(int16x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmls_laneq_s16_0
+  // CHECK-LABEL: test_vmls_laneq_s16_0
   return vmls_laneq_s16(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmlsq_laneq_s16_0(int16x8_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsq_laneq_s16_0
+  // CHECK-LABEL: test_vmlsq_laneq_s16_0
   return vmlsq_laneq_s16(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmls_laneq_s32_0(int32x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmls_laneq_s32_0
+  // CHECK-LABEL: test_vmls_laneq_s32_0
   return vmls_laneq_s32(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsq_laneq_s32_0(int32x4_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsq_laneq_s32_0
+  // CHECK-LABEL: test_vmlsq_laneq_s32_0
   return vmlsq_laneq_s32(a, b, v, 0);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vmul_lane_s16_0(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vmul_lane_s16_0
+  // CHECK-LABEL: test_vmul_lane_s16_0
   return vmul_lane_s16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmulq_lane_s16_0(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vmulq_lane_s16_0
+  // CHECK-LABEL: test_vmulq_lane_s16_0
   return vmulq_lane_s16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmul_lane_s32_0(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vmul_lane_s32_0
+  // CHECK-LABEL: test_vmul_lane_s32_0
   return vmul_lane_s32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmulq_lane_s32_0(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vmulq_lane_s32_0
+  // CHECK-LABEL: test_vmulq_lane_s32_0
   return vmulq_lane_s32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 uint16x4_t test_vmul_lane_u16_0(uint16x4_t a, uint16x4_t v) {
-  // CHECK: test_vmul_lane_u16_0
+  // CHECK-LABEL: test_vmul_lane_u16_0
   return vmul_lane_u16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 uint16x8_t test_vmulq_lane_u16_0(uint16x8_t a, uint16x4_t v) {
-  // CHECK: test_vmulq_lane_u16_0
+  // CHECK-LABEL: test_vmulq_lane_u16_0
   return vmulq_lane_u16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 uint32x2_t test_vmul_lane_u32_0(uint32x2_t a, uint32x2_t v) {
-  // CHECK: test_vmul_lane_u32_0
+  // CHECK-LABEL: test_vmul_lane_u32_0
   return vmul_lane_u32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmulq_lane_u32_0(uint32x4_t a, uint32x2_t v) {
-  // CHECK: test_vmulq_lane_u32_0
+  // CHECK-LABEL: test_vmulq_lane_u32_0
   return vmulq_lane_u32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vmul_laneq_s16_0(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vmul_laneq_s16_0
+  // CHECK-LABEL: test_vmul_laneq_s16_0
   return vmul_laneq_s16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vmulq_laneq_s16_0(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vmulq_laneq_s16_0
+  // CHECK-LABEL: test_vmulq_laneq_s16_0
   return vmulq_laneq_s16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vmul_laneq_s32_0(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vmul_laneq_s32_0
+  // CHECK-LABEL: test_vmul_laneq_s32_0
   return vmul_laneq_s32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmulq_laneq_s32_0(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vmulq_laneq_s32_0
+  // CHECK-LABEL: test_vmulq_laneq_s32_0
   return vmulq_laneq_s32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 uint16x4_t test_vmul_laneq_u16_0(uint16x4_t a, uint16x8_t v) {
-  // CHECK: test_vmul_laneq_u16_0
+  // CHECK-LABEL: test_vmul_laneq_u16_0
   return vmul_laneq_u16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 uint16x8_t test_vmulq_laneq_u16_0(uint16x8_t a, uint16x8_t v) {
-  // CHECK: test_vmulq_laneq_u16_0
+  // CHECK-LABEL: test_vmulq_laneq_u16_0
   return vmulq_laneq_u16(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 uint32x2_t test_vmul_laneq_u32_0(uint32x2_t a, uint32x4_t v) {
-  // CHECK: test_vmul_laneq_u32_0
+  // CHECK-LABEL: test_vmul_laneq_u32_0
   return vmul_laneq_u32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmulq_laneq_u32_0(uint32x4_t a, uint32x4_t v) {
-  // CHECK: test_vmulq_laneq_u32_0
+  // CHECK-LABEL: test_vmulq_laneq_u32_0
   return vmulq_laneq_u32(a, v, 0);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vfma_lane_f32_0(float32x2_t a, float32x2_t b, float32x2_t v) {
-  // CHECK: test_vfma_lane_f32_0
+  // CHECK-LABEL: test_vfma_lane_f32_0
   return vfma_lane_f32(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmaq_lane_f32_0(float32x4_t a, float32x4_t b, float32x2_t v) {
-  // CHECK: test_vfmaq_lane_f32_0
+  // CHECK-LABEL: test_vfmaq_lane_f32_0
   return vfmaq_lane_f32(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vfma_laneq_f32_0(float32x2_t a, float32x2_t b, float32x4_t v) {
-  // CHECK: test_vfma_laneq_f32_0
+  // CHECK-LABEL: test_vfma_laneq_f32_0
   return vfma_laneq_f32(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmaq_laneq_f32_0(float32x4_t a, float32x4_t b, float32x4_t v) {
-  // CHECK: test_vfmaq_laneq_f32_0
+  // CHECK-LABEL: test_vfmaq_laneq_f32_0
   return vfmaq_laneq_f32(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vfms_lane_f32_0(float32x2_t a, float32x2_t b, float32x2_t v) {
-  // CHECK: test_vfms_lane_f32_0
+  // CHECK-LABEL: test_vfms_lane_f32_0
   return vfms_lane_f32(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmsq_lane_f32_0(float32x4_t a, float32x4_t b, float32x2_t v) {
-  // CHECK: test_vfmsq_lane_f32_0
+  // CHECK-LABEL: test_vfmsq_lane_f32_0
   return vfmsq_lane_f32(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vfms_laneq_f32_0(float32x2_t a, float32x2_t b, float32x4_t v) {
-  // CHECK: test_vfms_laneq_f32_0
+  // CHECK-LABEL: test_vfms_laneq_f32_0
   return vfms_laneq_f32(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmsq_laneq_f32_0(float32x4_t a, float32x4_t b, float32x4_t v) {
-  // CHECK: test_vfmsq_laneq_f32_0
+  // CHECK-LABEL: test_vfmsq_laneq_f32_0
   return vfmsq_laneq_f32(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vfmaq_laneq_f64_0(float64x2_t a, float64x2_t b, float64x2_t v) {
-  // CHECK: test_vfmaq_laneq_f64_0
+  // CHECK-LABEL: test_vfmaq_laneq_f64_0
   return vfmaq_laneq_f64(a, b, v, 0);
   // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float64x2_t test_vfmsq_laneq_f64_0(float64x2_t a, float64x2_t b, float64x2_t v) {
-  // CHECK: test_vfmsq_laneq_f64_0
+  // CHECK-LABEL: test_vfmsq_laneq_f64_0
   return vfmsq_laneq_f64(a, b, v, 0);
   // CHECK: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 int32x4_t test_vmlal_lane_s16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlal_lane_s16_0
+  // CHECK-LABEL: test_vmlal_lane_s16_0
   return vmlal_lane_s16(a, b, v, 0);
   // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_lane_s32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlal_lane_s32_0
+  // CHECK-LABEL: test_vmlal_lane_s32_0
   return vmlal_lane_s32(a, b, v, 0);
   // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_laneq_s16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlal_laneq_s16_0
+  // CHECK-LABEL: test_vmlal_laneq_s16_0
   return vmlal_laneq_s16(a, b, v, 0);
   // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_laneq_s32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlal_laneq_s32_0
+  // CHECK-LABEL: test_vmlal_laneq_s32_0
   return vmlal_laneq_s32(a, b, v, 0);
   // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_high_lane_s16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlal_high_lane_s16_0
+  // CHECK-LABEL: test_vmlal_high_lane_s16_0
   return vmlal_high_lane_s16(a, b, v, 0);
   // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_high_lane_s32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlal_high_lane_s32_0
+  // CHECK-LABEL: test_vmlal_high_lane_s32_0
   return vmlal_high_lane_s32(a, b, v, 0);
   // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_high_laneq_s16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlal_high_laneq_s16_0
+  // CHECK-LABEL: test_vmlal_high_laneq_s16_0
   return vmlal_high_laneq_s16(a, b, v, 0);
   // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_high_laneq_s32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlal_high_laneq_s32_0
+  // CHECK-LABEL: test_vmlal_high_laneq_s32_0
   return vmlal_high_laneq_s32(a, b, v, 0);
   // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_lane_s16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_lane_s16_0
+  // CHECK-LABEL: test_vmlsl_lane_s16_0
   return vmlsl_lane_s16(a, b, v, 0);
   // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_lane_s32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_lane_s32_0
+  // CHECK-LABEL: test_vmlsl_lane_s32_0
   return vmlsl_lane_s32(a, b, v, 0);
   // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_laneq_s16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_laneq_s16_0
+  // CHECK-LABEL: test_vmlsl_laneq_s16_0
   return vmlsl_laneq_s16(a, b, v, 0);
   // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_laneq_s32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_laneq_s32_0
+  // CHECK-LABEL: test_vmlsl_laneq_s32_0
   return vmlsl_laneq_s32(a, b, v, 0);
   // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_high_lane_s16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_high_lane_s16_0
+  // CHECK-LABEL: test_vmlsl_high_lane_s16_0
   return vmlsl_high_lane_s16(a, b, v, 0);
   // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_high_lane_s32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_high_lane_s32_0
+  // CHECK-LABEL: test_vmlsl_high_lane_s32_0
   return vmlsl_high_lane_s32(a, b, v, 0);
   // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_high_laneq_s16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_high_laneq_s16_0
+  // CHECK-LABEL: test_vmlsl_high_laneq_s16_0
   return vmlsl_high_laneq_s16(a, b, v, 0);
   // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_high_laneq_s32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_high_laneq_s32_0
+  // CHECK-LABEL: test_vmlsl_high_laneq_s32_0
   return vmlsl_high_laneq_s32(a, b, v, 0);
   // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_lane_u16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlal_lane_u16_0
+  // CHECK-LABEL: test_vmlal_lane_u16_0
   return vmlal_lane_u16(a, b, v, 0);
   // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_lane_u32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlal_lane_u32_0
+  // CHECK-LABEL: test_vmlal_lane_u32_0
   return vmlal_lane_u32(a, b, v, 0);
   // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_laneq_u16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlal_laneq_u16_0
+  // CHECK-LABEL: test_vmlal_laneq_u16_0
   return vmlal_laneq_u16(a, b, v, 0);
   // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_laneq_u32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlal_laneq_u32_0
+  // CHECK-LABEL: test_vmlal_laneq_u32_0
   return vmlal_laneq_u32(a, b, v, 0);
   // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_high_lane_u16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlal_high_lane_u16_0
+  // CHECK-LABEL: test_vmlal_high_lane_u16_0
   return vmlal_high_lane_u16(a, b, v, 0);
   // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_high_lane_u32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlal_high_lane_u32_0
+  // CHECK-LABEL: test_vmlal_high_lane_u32_0
   return vmlal_high_lane_u32(a, b, v, 0);
   // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlal_high_laneq_u16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlal_high_laneq_u16_0
+  // CHECK-LABEL: test_vmlal_high_laneq_u16_0
   return vmlal_high_laneq_u16(a, b, v, 0);
   // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlal_high_laneq_u32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlal_high_laneq_u32_0
+  // CHECK-LABEL: test_vmlal_high_laneq_u32_0
   return vmlal_high_laneq_u32(a, b, v, 0);
   // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_lane_u16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_lane_u16_0
+  // CHECK-LABEL: test_vmlsl_lane_u16_0
   return vmlsl_lane_u16(a, b, v, 0);
   // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_lane_u32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_lane_u32_0
+  // CHECK-LABEL: test_vmlsl_lane_u32_0
   return vmlsl_lane_u32(a, b, v, 0);
   // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_laneq_u16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_laneq_u16_0
+  // CHECK-LABEL: test_vmlsl_laneq_u16_0
   return vmlsl_laneq_u16(a, b, v, 0);
   // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_laneq_u32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_laneq_u32_0
+  // CHECK-LABEL: test_vmlsl_laneq_u32_0
   return vmlsl_laneq_u32(a, b, v, 0);
   // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_high_lane_u16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vmlsl_high_lane_u16_0
+  // CHECK-LABEL: test_vmlsl_high_lane_u16_0
   return vmlsl_high_lane_u16(a, b, v, 0);
   // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_high_lane_u32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vmlsl_high_lane_u32_0
+  // CHECK-LABEL: test_vmlsl_high_lane_u32_0
   return vmlsl_high_lane_u32(a, b, v, 0);
   // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmlsl_high_laneq_u16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
-  // CHECK: test_vmlsl_high_laneq_u16_0
+  // CHECK-LABEL: test_vmlsl_high_laneq_u16_0
   return vmlsl_high_laneq_u16(a, b, v, 0);
   // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmlsl_high_laneq_u32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
-  // CHECK: test_vmlsl_high_laneq_u32_0
+  // CHECK-LABEL: test_vmlsl_high_laneq_u32_0
   return vmlsl_high_laneq_u32(a, b, v, 0);
   // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmull_lane_s16_0(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vmull_lane_s16_0
+  // CHECK-LABEL: test_vmull_lane_s16_0
   return vmull_lane_s16(a, v, 0);
   // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmull_lane_s32_0(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vmull_lane_s32_0
+  // CHECK-LABEL: test_vmull_lane_s32_0
   return vmull_lane_s32(a, v, 0);
   // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmull_lane_u16_0(uint16x4_t a, uint16x4_t v) {
-  // CHECK: test_vmull_lane_u16_0
+  // CHECK-LABEL: test_vmull_lane_u16_0
   return vmull_lane_u16(a, v, 0);
   // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 uint64x2_t test_vmull_lane_u32_0(uint32x2_t a, uint32x2_t v) {
-  // CHECK: test_vmull_lane_u32_0
+  // CHECK-LABEL: test_vmull_lane_u32_0
   return vmull_lane_u32(a, v, 0);
   // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmull_high_lane_s16_0(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vmull_high_lane_s16_0
+  // CHECK-LABEL: test_vmull_high_lane_s16_0
   return vmull_high_lane_s16(a, v, 0);
   // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmull_high_lane_s32_0(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vmull_high_lane_s32_0
+  // CHECK-LABEL: test_vmull_high_lane_s32_0
   return vmull_high_lane_s32(a, v, 0);
   // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmull_high_lane_u16_0(uint16x8_t a, uint16x4_t v) {
-  // CHECK: test_vmull_high_lane_u16_0
+  // CHECK-LABEL: test_vmull_high_lane_u16_0
   return vmull_high_lane_u16(a, v, 0);
   // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 uint64x2_t test_vmull_high_lane_u32_0(uint32x4_t a, uint32x2_t v) {
-  // CHECK: test_vmull_high_lane_u32_0
+  // CHECK-LABEL: test_vmull_high_lane_u32_0
   return vmull_high_lane_u32(a, v, 0);
   // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmull_laneq_s16_0(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vmull_laneq_s16_0
+  // CHECK-LABEL: test_vmull_laneq_s16_0
   return vmull_laneq_s16(a, v, 0);
   // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmull_laneq_s32_0(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vmull_laneq_s32_0
+  // CHECK-LABEL: test_vmull_laneq_s32_0
   return vmull_laneq_s32(a, v, 0);
   // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmull_laneq_u16_0(uint16x4_t a, uint16x8_t v) {
-  // CHECK: test_vmull_laneq_u16_0
+  // CHECK-LABEL: test_vmull_laneq_u16_0
   return vmull_laneq_u16(a, v, 0);
   // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 uint64x2_t test_vmull_laneq_u32_0(uint32x2_t a, uint32x4_t v) {
-  // CHECK: test_vmull_laneq_u32_0
+  // CHECK-LABEL: test_vmull_laneq_u32_0
   return vmull_laneq_u32(a, v, 0);
   // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vmull_high_laneq_s16_0(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vmull_high_laneq_s16_0
+  // CHECK-LABEL: test_vmull_high_laneq_s16_0
   return vmull_high_laneq_s16(a, v, 0);
   // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vmull_high_laneq_s32_0(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vmull_high_laneq_s32_0
+  // CHECK-LABEL: test_vmull_high_laneq_s32_0
   return vmull_high_laneq_s32(a, v, 0);
   // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 uint32x4_t test_vmull_high_laneq_u16_0(uint16x8_t a, uint16x8_t v) {
-  // CHECK: test_vmull_high_laneq_u16_0
+  // CHECK-LABEL: test_vmull_high_laneq_u16_0
   return vmull_high_laneq_u16(a, v, 0);
   // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 uint64x2_t test_vmull_high_laneq_u32_0(uint32x4_t a, uint32x4_t v) {
-  // CHECK: test_vmull_high_laneq_u32_0
+  // CHECK-LABEL: test_vmull_high_laneq_u32_0
   return vmull_high_laneq_u32(a, v, 0);
   // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmlal_lane_s16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vqdmlal_lane_s16_0
+  // CHECK-LABEL: test_vqdmlal_lane_s16_0
   return vqdmlal_lane_s16(a, b, v, 0);
   // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmlal_lane_s32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vqdmlal_lane_s32_0
+  // CHECK-LABEL: test_vqdmlal_lane_s32_0
   return vqdmlal_lane_s32(a, b, v, 0);
   // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmlal_high_lane_s16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vqdmlal_high_lane_s16_0
+  // CHECK-LABEL: test_vqdmlal_high_lane_s16_0
   return vqdmlal_high_lane_s16(a, b, v, 0);
   // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmlal_high_lane_s32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vqdmlal_high_lane_s32_0
+  // CHECK-LABEL: test_vqdmlal_high_lane_s32_0
   return vqdmlal_high_lane_s32(a, b, v, 0);
   // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmlsl_lane_s16_0(int32x4_t a, int16x4_t b, int16x4_t v) {
-  // CHECK: test_vqdmlsl_lane_s16_0
+  // CHECK-LABEL: test_vqdmlsl_lane_s16_0
   return vqdmlsl_lane_s16(a, b, v, 0);
   // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmlsl_lane_s32_0(int64x2_t a, int32x2_t b, int32x2_t v) {
-  // CHECK: test_vqdmlsl_lane_s32_0
+  // CHECK-LABEL: test_vqdmlsl_lane_s32_0
   return vqdmlsl_lane_s32(a, b, v, 0);
   // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmlsl_high_lane_s16_0(int32x4_t a, int16x8_t b, int16x4_t v) {
-  // CHECK: test_vqdmlsl_high_lane_s16_0
+  // CHECK-LABEL: test_vqdmlsl_high_lane_s16_0
   return vqdmlsl_high_lane_s16(a, b, v, 0);
   // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmlsl_high_lane_s32_0(int64x2_t a, int32x4_t b, int32x2_t v) {
-  // CHECK: test_vqdmlsl_high_lane_s32_0
+  // CHECK-LABEL: test_vqdmlsl_high_lane_s32_0
   return vqdmlsl_high_lane_s32(a, b, v, 0);
   // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmull_lane_s16_0(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqdmull_lane_s16_0
+  // CHECK-LABEL: test_vqdmull_lane_s16_0
   return vqdmull_lane_s16(a, v, 0);
   // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmull_lane_s32_0(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqdmull_lane_s32_0
+  // CHECK-LABEL: test_vqdmull_lane_s32_0
   return vqdmull_lane_s32(a, v, 0);
   // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmull_laneq_s16_0(int16x4_t a, int16x8_t v) {
-  // CHECK: test_vqdmull_laneq_s16_0
+  // CHECK-LABEL: test_vqdmull_laneq_s16_0
   return vqdmull_laneq_s16(a, v, 0);
   // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmull_laneq_s32_0(int32x2_t a, int32x4_t v) {
-  // CHECK: test_vqdmull_laneq_s32_0
+  // CHECK-LABEL: test_vqdmull_laneq_s32_0
   return vqdmull_laneq_s32(a, v, 0);
   // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmull_high_lane_s16_0(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqdmull_high_lane_s16_0
+  // CHECK-LABEL: test_vqdmull_high_lane_s16_0
   return vqdmull_high_lane_s16(a, v, 0);
   // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmull_high_lane_s32_0(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqdmull_high_lane_s32_0
+  // CHECK-LABEL: test_vqdmull_high_lane_s32_0
   return vqdmull_high_lane_s32(a, v, 0);
   // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmull_high_laneq_s16_0(int16x8_t a, int16x8_t v) {
-  // CHECK: test_vqdmull_high_laneq_s16_0
+  // CHECK-LABEL: test_vqdmull_high_laneq_s16_0
   return vqdmull_high_laneq_s16(a, v, 0);
   // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int64x2_t test_vqdmull_high_laneq_s32_0(int32x4_t a, int32x4_t v) {
-  // CHECK: test_vqdmull_high_laneq_s32_0
+  // CHECK-LABEL: test_vqdmull_high_laneq_s32_0
   return vqdmull_high_laneq_s32(a, v, 0);
   // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vqdmulh_lane_s16_0(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqdmulh_lane_s16_0
+  // CHECK-LABEL: test_vqdmulh_lane_s16_0
   return vqdmulh_lane_s16(a, v, 0);
   // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vqdmulhq_lane_s16_0(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqdmulhq_lane_s16_0
+  // CHECK-LABEL: test_vqdmulhq_lane_s16_0
   return vqdmulhq_lane_s16(a, v, 0);
   // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vqdmulh_lane_s32_0(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqdmulh_lane_s32_0
+  // CHECK-LABEL: test_vqdmulh_lane_s32_0
   return vqdmulh_lane_s32(a, v, 0);
   // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqdmulhq_lane_s32_0(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqdmulhq_lane_s32_0
+  // CHECK-LABEL: test_vqdmulhq_lane_s32_0
   return vqdmulhq_lane_s32(a, v, 0);
   // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 int16x4_t test_vqrdmulh_lane_s16_0(int16x4_t a, int16x4_t v) {
-  // CHECK: test_vqrdmulh_lane_s16_0
+  // CHECK-LABEL: test_vqrdmulh_lane_s16_0
   return vqrdmulh_lane_s16(a, v, 0);
   // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
 }
 
 int16x8_t test_vqrdmulhq_lane_s16_0(int16x8_t a, int16x4_t v) {
-  // CHECK: test_vqrdmulhq_lane_s16_0
+  // CHECK-LABEL: test_vqrdmulhq_lane_s16_0
   return vqrdmulhq_lane_s16(a, v, 0);
   // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
 }
 
 int32x2_t test_vqrdmulh_lane_s32_0(int32x2_t a, int32x2_t v) {
-  // CHECK: test_vqrdmulh_lane_s32_0
+  // CHECK-LABEL: test_vqrdmulh_lane_s32_0
   return vqrdmulh_lane_s32(a, v, 0);
   // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 int32x4_t test_vqrdmulhq_lane_s32_0(int32x4_t a, int32x2_t v) {
-  // CHECK: test_vqrdmulhq_lane_s32_0
+  // CHECK-LABEL: test_vqrdmulhq_lane_s32_0
   return vqrdmulhq_lane_s32(a, v, 0);
   // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vmul_lane_f32_0(float32x2_t a, float32x2_t v) {
-  // CHECK: test_vmul_lane_f32_0
+  // CHECK-LABEL: test_vmul_lane_f32_0
   return vmul_lane_f32(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vmulq_lane_f32_0(float32x4_t a, float32x2_t v) {
-  // CHECK: test_vmulq_lane_f32_0
+  // CHECK-LABEL: test_vmulq_lane_f32_0
   return vmulq_lane_f32(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vmul_laneq_f32_0(float32x2_t a, float32x4_t v) {
-  // CHECK: test_vmul_laneq_f32_0
+  // CHECK-LABEL: test_vmul_laneq_f32_0
   return vmul_laneq_f32(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float64x1_t test_vmul_laneq_f64_0(float64x1_t a, float64x2_t v) {
-  // CHECK: test_vmul_laneq_f64_0
+  // CHECK-LABEL: test_vmul_laneq_f64_0
   return vmul_laneq_f64(a, v, 0);
   // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
 float32x4_t test_vmulq_laneq_f32_0(float32x4_t a, float32x4_t v) {
-  // CHECK: test_vmulq_laneq_f32_0
+  // CHECK-LABEL: test_vmulq_laneq_f32_0
   return vmulq_laneq_f32(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vmulq_laneq_f64_0(float64x2_t a, float64x2_t v) {
-  // CHECK: test_vmulq_laneq_f64_0
+  // CHECK-LABEL: test_vmulq_laneq_f64_0
   return vmulq_laneq_f64(a, v, 0);
   // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float32x2_t test_vmulx_lane_f32_0(float32x2_t a, float32x2_t v) {
-  // CHECK: test_vmulx_lane_f32_0
+  // CHECK-LABEL: test_vmulx_lane_f32_0
   return vmulx_lane_f32(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vmulxq_lane_f32_0(float32x4_t a, float32x2_t v) {
-  // CHECK: test_vmulxq_lane_f32_0
+  // CHECK-LABEL: test_vmulxq_lane_f32_0
   return vmulxq_lane_f32(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vmulxq_lane_f64_0(float64x2_t a, float64x1_t v) {
-  // CHECK: test_vmulxq_lane_f64_0
+  // CHECK-LABEL: test_vmulxq_lane_f64_0
   return vmulxq_lane_f64(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float32x2_t test_vmulx_laneq_f32_0(float32x2_t a, float32x4_t v) {
-  // CHECK: test_vmulx_laneq_f32_0
+  // CHECK-LABEL: test_vmulx_laneq_f32_0
   return vmulx_laneq_f32(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vmulxq_laneq_f32_0(float32x4_t a, float32x4_t v) {
-  // CHECK: test_vmulxq_laneq_f32_0
+  // CHECK-LABEL: test_vmulxq_laneq_f32_0
   return vmulxq_laneq_f32(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vmulxq_laneq_f64_0(float64x2_t a, float64x2_t v) {
-  // CHECK: test_vmulxq_laneq_f64_0
+  // CHECK-LABEL: test_vmulxq_laneq_f64_0
   return vmulxq_laneq_f64(a, v, 0);
   // CHECK: fmulx {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 int32x4_t test_vmull_high_n_s16(int16x8_t a, int16_t b) {
-  // CHECK: test_vmull_high_n_s16
+  // CHECK-LABEL: test_vmull_high_n_s16
   return vmull_high_n_s16(a, b);
-  // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vmull_high_n_s32(int32x4_t a, int32_t b) {
-  // CHECK: test_vmull_high_n_s32
+  // CHECK-LABEL: test_vmull_high_n_s32
   return vmull_high_n_s32(a, b);
-  // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 uint32x4_t test_vmull_high_n_u16(uint16x8_t a, uint16_t b) {
-  // CHECK: test_vmull_high_n_u16
+  // CHECK-LABEL: test_vmull_high_n_u16
   return vmull_high_n_u16(a, b);
-  // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 uint64x2_t test_vmull_high_n_u32(uint32x4_t a, uint32_t b) {
-  // CHECK: test_vmull_high_n_u32
+  // CHECK-LABEL: test_vmull_high_n_u32
   return vmull_high_n_u32(a, b);
-  // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 int32x4_t test_vqdmull_high_n_s16(int16x8_t a, int16_t b) {
-  // CHECK: test_vqdmull_high_n_s16
+  // CHECK-LABEL: test_vqdmull_high_n_s16
   return vqdmull_high_n_s16(a, b);
-  // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vqdmull_high_n_s32(int32x4_t a, int32_t b) {
-  // CHECK: test_vqdmull_high_n_s32
+  // CHECK-LABEL: test_vqdmull_high_n_s32
   return vqdmull_high_n_s32(a, b);
-  // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 int32x4_t test_vmlal_high_n_s16(int32x4_t a, int16x8_t b, int16_t c) {
-  // CHECK: test_vmlal_high_n_s16
+  // CHECK-LABEL: test_vmlal_high_n_s16
   return vmlal_high_n_s16(a, b, c);
-  // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vmlal_high_n_s32(int64x2_t a, int32x4_t b, int32_t c) {
-  // CHECK: test_vmlal_high_n_s32
+  // CHECK-LABEL: test_vmlal_high_n_s32
   return vmlal_high_n_s32(a, b, c);
-  // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 uint32x4_t test_vmlal_high_n_u16(uint32x4_t a, uint16x8_t b, uint16_t c) {
-  // CHECK: test_vmlal_high_n_u16
+  // CHECK-LABEL: test_vmlal_high_n_u16
   return vmlal_high_n_u16(a, b, c);
-  // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 uint64x2_t test_vmlal_high_n_u32(uint64x2_t a, uint32x4_t b, uint32_t c) {
-  // CHECK: test_vmlal_high_n_u32
+  // CHECK-LABEL: test_vmlal_high_n_u32
   return vmlal_high_n_u32(a, b, c);
-  // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 int32x4_t test_vqdmlal_high_n_s16(int32x4_t a, int16x8_t b, int16_t c) {
-  // CHECK: test_vqdmlal_high_n_s16
+  // CHECK-LABEL: test_vqdmlal_high_n_s16
   return vqdmlal_high_n_s16(a, b, c);
-  // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vqdmlal_high_n_s32(int64x2_t a, int32x4_t b, int32_t c) {
-  // CHECK: test_vqdmlal_high_n_s32
+  // CHECK-LABEL: test_vqdmlal_high_n_s32
   return vqdmlal_high_n_s32(a, b, c);
-  // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 int32x4_t test_vmlsl_high_n_s16(int32x4_t a, int16x8_t b, int16_t c) {
-  // CHECK: test_vmlsl_high_n_s16
+  // CHECK-LABEL: test_vmlsl_high_n_s16
   return vmlsl_high_n_s16(a, b, c);
-  // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vmlsl_high_n_s32(int64x2_t a, int32x4_t b, int32_t c) {
-  // CHECK: test_vmlsl_high_n_s32
+  // CHECK-LABEL: test_vmlsl_high_n_s32
   return vmlsl_high_n_s32(a, b, c);
-  // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 uint32x4_t test_vmlsl_high_n_u16(uint32x4_t a, uint16x8_t b, uint16_t c) {
-  // CHECK: test_vmlsl_high_n_u16
+  // CHECK-LABEL: test_vmlsl_high_n_u16
   return vmlsl_high_n_u16(a, b, c);
-  // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 uint64x2_t test_vmlsl_high_n_u32(uint64x2_t a, uint32x4_t b, uint32_t c) {
-  // CHECK: test_vmlsl_high_n_u32
+  // CHECK-LABEL: test_vmlsl_high_n_u32
   return vmlsl_high_n_u32(a, b, c);
-  // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 int32x4_t test_vqdmlsl_high_n_s16(int32x4_t a, int16x8_t b, int16_t c) {
-  // CHECK: test_vqdmlsl_high_n_s16
+  // CHECK-LABEL: test_vqdmlsl_high_n_s16
   return vqdmlsl_high_n_s16(a, b, c);
-  // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+\.h\[0\]|v[0-9]+\.8h}}
 }
 
 int64x2_t test_vqdmlsl_high_n_s32(int64x2_t a, int32x4_t b, int32_t c) {
-  // CHECK: test_vqdmlsl_high_n_s32
+  // CHECK-LABEL: test_vqdmlsl_high_n_s32
   return vqdmlsl_high_n_s32(a, b, c);
-  // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+\.s\[0\]|v[0-9]+\.4s}}
 }
 
 float32x2_t test_vmul_n_f32(float32x2_t a, float32_t b) {
-  // CHECK: test_vmul_n_f32
+  // CHECK-LABEL: test_vmul_n_f32
   return vmul_n_f32(a, b);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vmulq_n_f32(float32x4_t a, float32_t b) {
-  // CHECK: test_vmulq_n_f32
+  // CHECK-LABEL: test_vmulq_n_f32
   return vmulq_n_f32(a, b);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vmulq_n_f64(float64x2_t a, float64_t b) {
-  // CHECK: test_vmulq_n_f64
+  // CHECK-LABEL: test_vmulq_n_f64
   return vmulq_n_f64(a, b);
   // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float32x2_t test_vfma_n_f32(float32x2_t a, float32x2_t b, float32_t n) {
-  // CHECK: test_vfma_n_f32
+  // CHECK-LABEL: test_vfma_n_f32
   return vfma_n_f32(a, b, n);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmaq_n_f32(float32x4_t a, float32x4_t b, float32_t n) {
-  // CHECK: test_vfmaq_n_f32
+  // CHECK-LABEL: test_vfmaq_n_f32
   return vfmaq_n_f32(a, b, n);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float32x2_t test_vfms_n_f32(float32x2_t a, float32x2_t b, float32_t n) {
-  // CHECK: test_vfms_n_f32
+  // CHECK-LABEL: test_vfms_n_f32
   return vfms_n_f32(a, b, n);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vfmsq_n_f32(float32x4_t a, float32x4_t b, float32_t n) {
-  // CHECK: test_vfmsq_n_f32
+  // CHECK-LABEL: test_vfmsq_n_f32
   return vfmsq_n_f32(a, b, n);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
+
+int16x4_t test_vmul_n_s16(int16x4_t a, int16_t b) {
+  // CHECK-LABEL: test_vmul_n_s16
+  return vmul_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int16x8_t test_vmulq_n_s16(int16x8_t a, int16_t b) {
+  // CHECK-LABEL: test_vmulq_n_s16
+  return vmulq_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+int32x2_t test_vmul_n_s32(int32x2_t a, int32_t b) {
+  // CHECK-LABEL: test_vmul_n_s32
+  return vmul_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vmulq_n_s32(int32x4_t a, int32_t b) {
+  // CHECK-LABEL: test_vmulq_n_s32
+  return vmulq_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+uint16x4_t test_vmul_n_u16(uint16x4_t a, uint16_t b) {
+  // CHECK-LABEL: test_vmul_n_u16
+  return vmul_n_u16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint16x8_t test_vmulq_n_u16(uint16x8_t a, uint16_t b) {
+  // CHECK-LABEL: test_vmulq_n_u16
+  return vmulq_n_u16(a, b);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+uint32x2_t test_vmul_n_u32(uint32x2_t a, uint32_t b) {
+  // CHECK-LABEL: test_vmul_n_u32
+  return vmul_n_u32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmulq_n_u32(uint32x4_t a, uint32_t b) {
+  // CHECK-LABEL: test_vmulq_n_u32
+  return vmulq_n_u32(a, b);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+int32x4_t test_vmull_n_s16(int16x4_t a, int16_t b) {
+  // CHECK-LABEL: test_vmull_n_s16
+  return vmull_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vmull_n_s32(int32x2_t a, int32_t b) {
+  // CHECK-LABEL: test_vmull_n_s32
+  return vmull_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmull_n_u16(uint16x4_t a, uint16_t b) {
+  // CHECK-LABEL: test_vmull_n_u16
+  return vmull_n_u16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint64x2_t test_vmull_n_u32(uint32x2_t a, uint32_t b) {
+  // CHECK-LABEL: test_vmull_n_u32
+  return vmull_n_u32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vqdmull_n_s16(int16x4_t a, int16_t b) {
+  // CHECK-LABEL: test_vqdmull_n_s16
+  return vqdmull_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vqdmull_n_s32(int32x2_t a, int32_t b) {
+  // CHECK-LABEL: test_vqdmull_n_s32
+  return vqdmull_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int16x4_t test_vqdmulh_n_s16(int16x4_t a, int16_t b) {
+  // CHECK-LABEL: test_vqdmulh_n_s16
+  return vqdmulh_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int16x8_t test_vqdmulhq_n_s16(int16x8_t a, int16_t b) {
+  // CHECK-LABEL: test_vqdmulhq_n_s16
+  return vqdmulhq_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+int32x2_t test_vqdmulh_n_s32(int32x2_t a, int32_t b) {
+  // CHECK-LABEL: test_vqdmulh_n_s32
+  return vqdmulh_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vqdmulhq_n_s32(int32x4_t a, int32_t b) {
+  // CHECK-LABEL: test_vqdmulhq_n_s32
+  return vqdmulhq_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+int16x4_t test_vqrdmulh_n_s16(int16x4_t a, int16_t b) {
+  // CHECK-LABEL: test_vqrdmulh_n_s16
+  return vqrdmulh_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int16x8_t test_vqrdmulhq_n_s16(int16x8_t a, int16_t b) {
+  // CHECK-LABEL: test_vqrdmulhq_n_s16
+  return vqrdmulhq_n_s16(a, b);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+int32x2_t test_vqrdmulh_n_s32(int32x2_t a, int32_t b) {
+  // CHECK-LABEL: test_vqrdmulh_n_s32
+  return vqrdmulh_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vqrdmulhq_n_s32(int32x4_t a, int32_t b) {
+  // CHECK-LABEL: test_vqrdmulhq_n_s32
+  return vqrdmulhq_n_s32(a, b);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+int16x4_t test_vmla_n_s16(int16x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vmla_n_s16
+  return vmla_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int16x8_t test_vmlaq_n_s16(int16x8_t a, int16x8_t b, int16_t c) {
+  // CHECK-LABEL: test_vmlaq_n_s16
+  return vmlaq_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+int32x2_t test_vmla_n_s32(int32x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vmla_n_s32
+  return vmla_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vmlaq_n_s32(int32x4_t a, int32x4_t b, int32_t c) {
+  // CHECK-LABEL: test_vmlaq_n_s32
+  return vmlaq_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+uint16x4_t test_vmla_n_u16(uint16x4_t a, uint16x4_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmla_n_u16
+  return vmla_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint16x8_t test_vmlaq_n_u16(uint16x8_t a, uint16x8_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmlaq_n_u16
+  return vmlaq_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+uint32x2_t test_vmla_n_u32(uint32x2_t a, uint32x2_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmla_n_u32
+  return vmla_n_u32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmlaq_n_u32(uint32x4_t a, uint32x4_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmlaq_n_u32
+  return vmlaq_n_u32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4s, w0
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+int32x4_t test_vmlal_n_s16(int32x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vmlal_n_s16
+  return vmlal_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vmlal_n_s32(int64x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vmlal_n_s32
+  return vmlal_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmlal_n_u16(uint32x4_t a, uint16x4_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmlal_n_u16
+  return vmlal_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint64x2_t test_vmlal_n_u32(uint64x2_t a, uint32x2_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmlal_n_u32
+  return vmlal_n_u32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vqdmlal_n_s16(int32x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vqdmlal_n_s16
+  return vqdmlal_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vqdmlal_n_s32(int64x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vqdmlal_n_s32
+  return vqdmlal_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int16x4_t test_vmls_n_s16(int16x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vmls_n_s16
+  return vmls_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int16x8_t test_vmlsq_n_s16(int16x8_t a, int16x8_t b, int16_t c) {
+  // CHECK-LABEL: test_vmlsq_n_s16
+  return vmlsq_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+int32x2_t test_vmls_n_s32(int32x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vmls_n_s32
+  return vmls_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vmlsq_n_s32(int32x4_t a, int32x4_t b, int32_t c) {
+  // CHECK-LABEL: test_vmlsq_n_s32
+  return vmlsq_n_s32(a, b, c);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+uint16x4_t test_vmls_n_u16(uint16x4_t a, uint16x4_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmls_n_u16
+  return vmls_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint16x8_t test_vmlsq_n_u16(uint16x8_t a, uint16x8_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmlsq_n_u16
+  return vmlsq_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.8h, w0
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+uint32x2_t test_vmls_n_u32(uint32x2_t a, uint32x2_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmls_n_u32
+  return vmls_n_u32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmlsq_n_u32(uint32x4_t a, uint32x4_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmlsq_n_u32
+  return vmlsq_n_u32(a, b, c);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+int32x4_t test_vmlsl_n_s16(int32x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vmlsl_n_s16
+  return vmlsl_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vmlsl_n_s32(int64x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vmlsl_n_s32
+  return vmlsl_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vmlsl_n_u16(uint32x4_t a, uint16x4_t b, uint16_t c) {
+  // CHECK-LABEL: test_vmlsl_n_u16
+  return vmlsl_n_u16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint64x2_t test_vmlsl_n_u32(uint64x2_t a, uint32x2_t b, uint32_t c) {
+  // CHECK-LABEL: test_vmlsl_n_u32
+  return vmlsl_n_u32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32x4_t test_vqdmlsl_n_s16(int32x4_t a, int16x4_t b, int16_t c) {
+  // CHECK-LABEL: test_vqdmlsl_n_s16
+  return vqdmlsl_n_s16(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.4h, w0
+  // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+int64x2_t test_vqdmlsl_n_s32(int64x2_t a, int32x2_t b, int32_t c) {
+  // CHECK-LABEL: test_vqdmlsl_n_s32
+  return vqdmlsl_n_s32(a, b, c);
+  // CHECK: dup {{v[0-9]+}}.2s, w0
+  // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint16x4_t test_vmla_lane_u16_0(uint16x4_t a, uint16x4_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmla_lane_u16_0
+  return vmla_lane_u16(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+uint16x8_t test_vmlaq_lane_u16_0(uint16x8_t a, uint16x8_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_u16_0
+  return vmlaq_lane_u16(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+uint32x2_t test_vmla_lane_u32_0(uint32x2_t a, uint32x2_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmla_lane_u32_0
+  return vmla_lane_u32(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+uint32x4_t test_vmlaq_lane_u32_0(uint32x4_t a, uint32x4_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_u32_0
+  return vmlaq_lane_u32(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+uint16x4_t test_vmla_laneq_u16_0(uint16x4_t a, uint16x4_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmla_laneq_u16_0
+  return vmla_laneq_u16(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+uint16x8_t test_vmlaq_laneq_u16_0(uint16x8_t a, uint16x8_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_u16_0
+  return vmlaq_laneq_u16(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+uint32x2_t test_vmla_laneq_u32_0(uint32x2_t a, uint32x2_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmla_laneq_u32_0
+  return vmla_laneq_u32(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+uint32x4_t test_vmlaq_laneq_u32_0(uint32x4_t a, uint32x4_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_u32_0
+  return vmlaq_laneq_u32(a, b, v, 0);
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqdmlal_laneq_s16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlal_laneq_s16_0
+  return vqdmlal_laneq_s16(a, b, v, 0);
+  // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+int64x2_t test_vqdmlal_laneq_s32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlal_laneq_s32_0
+  return vqdmlal_laneq_s32(a, b, v, 0);
+  // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqdmlal_high_laneq_s16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlal_high_laneq_s16_0
+  return vqdmlal_high_laneq_s16(a, b, v, 0);
+  // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+int64x2_t test_vqdmlal_high_laneq_s32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlal_high_laneq_s32_0
+  return vqdmlal_high_laneq_s32(a, b, v, 0);
+  // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+uint16x4_t test_vmls_lane_u16_0(uint16x4_t a, uint16x4_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmls_lane_u16_0
+  return vmls_lane_u16(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+uint16x8_t test_vmlsq_lane_u16_0(uint16x8_t a, uint16x8_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_u16_0
+  return vmlsq_lane_u16(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+uint32x2_t test_vmls_lane_u32_0(uint32x2_t a, uint32x2_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmls_lane_u32_0
+  return vmls_lane_u32(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+uint32x4_t test_vmlsq_lane_u32_0(uint32x4_t a, uint32x4_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_u32_0
+  return vmlsq_lane_u32(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+uint16x4_t test_vmls_laneq_u16_0(uint16x4_t a, uint16x4_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmls_laneq_u16_0
+  return vmls_laneq_u16(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+uint16x8_t test_vmlsq_laneq_u16_0(uint16x8_t a, uint16x8_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_u16_0
+  return vmlsq_laneq_u16(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+uint32x2_t test_vmls_laneq_u32_0(uint32x2_t a, uint32x2_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmls_laneq_u32_0
+  return vmls_laneq_u32(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+uint32x4_t test_vmlsq_laneq_u32_0(uint32x4_t a, uint32x4_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_u32_0
+  return vmlsq_laneq_u32(a, b, v, 0);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqdmlsl_laneq_s16_0(int32x4_t a, int16x4_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlsl_laneq_s16_0
+  return vqdmlsl_laneq_s16(a, b, v, 0);
+  // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+int64x2_t test_vqdmlsl_laneq_s32_0(int64x2_t a, int32x2_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlsl_laneq_s32_0
+  return vqdmlsl_laneq_s32(a, b, v, 0);
+  // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqdmlsl_high_laneq_s16_0(int32x4_t a, int16x8_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlsl_high_laneq_s16_0
+  return vqdmlsl_high_laneq_s16(a, b, v, 0);
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+int64x2_t test_vqdmlsl_high_laneq_s32_0(int64x2_t a, int32x4_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlsl_high_laneq_s32_0
+  return vqdmlsl_high_laneq_s32(a, b, v, 0);
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+int16x4_t test_vqdmulh_laneq_s16_0(int16x4_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmulh_laneq_s16_0
+  return vqdmulh_laneq_s16(a, v, 0);
+  // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+int16x8_t test_vqdmulhq_laneq_s16_0(int16x8_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmulhq_laneq_s16_0
+  return vqdmulhq_laneq_s16(a, v, 0);
+  // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+int32x2_t test_vqdmulh_laneq_s32_0(int32x2_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmulh_laneq_s32_0
+  return vqdmulh_laneq_s32(a, v, 0);
+  // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqdmulhq_laneq_s32_0(int32x4_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmulhq_laneq_s32_0
+  return vqdmulhq_laneq_s32(a, v, 0);
+  // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+int16x4_t test_vqrdmulh_laneq_s16_0(int16x4_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqrdmulh_laneq_s16_0
+  return vqrdmulh_laneq_s16(a, v, 0);
+  // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[0]
+}
+
+int16x8_t test_vqrdmulhq_laneq_s16_0(int16x8_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqrdmulhq_laneq_s16_0
+  return vqrdmulhq_laneq_s16(a, v, 0);
+  // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[0]
+}
+
+int32x2_t test_vqrdmulh_laneq_s32_0(int32x2_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqrdmulh_laneq_s32_0
+  return vqrdmulh_laneq_s32(a, v, 0);
+  // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+int32x4_t test_vqrdmulhq_laneq_s32_0(int32x4_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqrdmulhq_laneq_s32_0
+  return vqrdmulhq_laneq_s32(a, v, 0);
+  // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+uint16x4_t test_vmla_lane_u16(uint16x4_t a, uint16x4_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmla_lane_u16
+  return vmla_lane_u16(a, b, v, 3);
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
+}
+
+uint16x8_t test_vmlaq_lane_u16(uint16x8_t a, uint16x8_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_u16
+  return vmlaq_lane_u16(a, b, v, 3);
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
+}
+
+uint32x2_t test_vmla_lane_u32(uint32x2_t a, uint32x2_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmla_lane_u32
+  return vmla_lane_u32(a, b, v, 1);
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+}
+
+uint32x4_t test_vmlaq_lane_u32(uint32x4_t a, uint32x4_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_u32
+  return vmlaq_lane_u32(a, b, v, 1);
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+}
+
+uint16x4_t test_vmla_laneq_u16(uint16x4_t a, uint16x4_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmla_laneq_u16
+  return vmla_laneq_u16(a, b, v, 7);
+  // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+uint16x8_t test_vmlaq_laneq_u16(uint16x8_t a, uint16x8_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_u16
+  return vmlaq_laneq_u16(a, b, v, 7);
+  // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+uint32x2_t test_vmla_laneq_u32(uint32x2_t a, uint32x2_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmla_laneq_u32
+  return vmla_laneq_u32(a, b, v, 3);
+  // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+uint32x4_t test_vmlaq_laneq_u32(uint32x4_t a, uint32x4_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_u32
+  return vmlaq_laneq_u32(a, b, v, 3);
+  // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqdmlal_laneq_s16(int32x4_t a, int16x4_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlal_laneq_s16
+  return vqdmlal_laneq_s16(a, b, v, 7);
+  // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+int64x2_t test_vqdmlal_laneq_s32(int64x2_t a, int32x2_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlal_laneq_s32
+  return vqdmlal_laneq_s32(a, b, v, 3);
+  // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqdmlal_high_laneq_s16(int32x4_t a, int16x8_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlal_high_laneq_s16
+  return vqdmlal_high_laneq_s16(a, b, v, 7);
+  // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+int64x2_t test_vqdmlal_high_laneq_s32(int64x2_t a, int32x4_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlal_high_laneq_s32
+  return vqdmlal_high_laneq_s32(a, b, v, 3);
+  // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+uint16x4_t test_vmls_lane_u16(uint16x4_t a, uint16x4_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmls_lane_u16
+  return vmls_lane_u16(a, b, v, 3);
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[3]
+}
+
+uint16x8_t test_vmlsq_lane_u16(uint16x8_t a, uint16x8_t b, uint16x4_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_u16
+  return vmlsq_lane_u16(a, b, v, 3);
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[3]
+}
+
+uint32x2_t test_vmls_lane_u32(uint32x2_t a, uint32x2_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmls_lane_u32
+  return vmls_lane_u32(a, b, v, 1);
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+}
+
+uint32x4_t test_vmlsq_lane_u32(uint32x4_t a, uint32x4_t b, uint32x2_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_u32
+  return vmlsq_lane_u32(a, b, v, 1);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+}
+
+uint16x4_t test_vmls_laneq_u16(uint16x4_t a, uint16x4_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmls_laneq_u16
+  return vmls_laneq_u16(a, b, v, 7);
+  // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+uint16x8_t test_vmlsq_laneq_u16(uint16x8_t a, uint16x8_t b, uint16x8_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_u16
+  return vmlsq_laneq_u16(a, b, v, 7);
+  // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+uint32x2_t test_vmls_laneq_u32(uint32x2_t a, uint32x2_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmls_laneq_u32
+  return vmls_laneq_u32(a, b, v, 3);
+  // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+uint32x4_t test_vmlsq_laneq_u32(uint32x4_t a, uint32x4_t b, uint32x4_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_u32
+  return vmlsq_laneq_u32(a, b, v, 3);
+  // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqdmlsl_laneq_s16(int32x4_t a, int16x4_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlsl_laneq_s16
+  return vqdmlsl_laneq_s16(a, b, v, 7);
+  // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+int64x2_t test_vqdmlsl_laneq_s32(int64x2_t a, int32x2_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlsl_laneq_s32
+  return vqdmlsl_laneq_s32(a, b, v, 3);
+  // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqdmlsl_high_laneq_s16(int32x4_t a, int16x8_t b, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmlsl_high_laneq_s16
+  return vqdmlsl_high_laneq_s16(a, b, v, 7);
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+int64x2_t test_vqdmlsl_high_laneq_s32(int64x2_t a, int32x4_t b, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmlsl_high_laneq_s32
+  return vqdmlsl_high_laneq_s32(a, b, v, 3);
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+int16x4_t test_vqdmulh_laneq_s16(int16x4_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmulh_laneq_s16
+  return vqdmulh_laneq_s16(a, v, 7);
+  // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+int16x8_t test_vqdmulhq_laneq_s16(int16x8_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqdmulhq_laneq_s16
+  return vqdmulhq_laneq_s16(a, v, 7);
+  // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+int32x2_t test_vqdmulh_laneq_s32(int32x2_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmulh_laneq_s32
+  return vqdmulh_laneq_s32(a, v, 3);
+  // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqdmulhq_laneq_s32(int32x4_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqdmulhq_laneq_s32
+  return vqdmulhq_laneq_s32(a, v, 3);
+  // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+int16x4_t test_vqrdmulh_laneq_s16(int16x4_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqrdmulh_laneq_s16
+  return vqrdmulh_laneq_s16(a, v, 7);
+  // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.h[7]
+}
+
+int16x8_t test_vqrdmulhq_laneq_s16(int16x8_t a, int16x8_t v) {
+  // CHECK-LABEL: test_vqrdmulhq_laneq_s16
+  return vqrdmulhq_laneq_s16(a, v, 7);
+  // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.h[7]
+}
+
+int32x2_t test_vqrdmulh_laneq_s32(int32x2_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqrdmulh_laneq_s32
+  return vqrdmulh_laneq_s32(a, v, 3);
+  // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+int32x4_t test_vqrdmulhq_laneq_s32(int32x4_t a, int32x4_t v) {
+  // CHECK-LABEL: test_vqrdmulhq_laneq_s32
+  return vqrdmulhq_laneq_s32(a, v, 3);
+  // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
diff --git a/test/CodeGen/aarch64-neon-3v.c b/test/CodeGen/aarch64-neon-3v.c
new file mode 100644
index 0000000..5251690
--- /dev/null
+++ b/test/CodeGen/aarch64-neon-3v.c
@@ -0,0 +1,489 @@
+// REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu -S -O3 -o - %s | FileCheck %s
+
+// Test new aarch64 intrinsics and types
+
+#include <arm_neon.h>
+
+int8x8_t test_vand_s8(int8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_vand_s8
+  return vand_s8(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int8x16_t test_vandq_s8(int8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_vandq_s8
+  return vandq_s8(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int16x4_t test_vand_s16(int16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_vand_s16
+  return vand_s16(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int16x8_t test_vandq_s16(int16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_vandq_s16
+  return vandq_s16(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int32x2_t test_vand_s32(int32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_vand_s32
+  return vand_s32(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int32x4_t test_vandq_s32(int32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_vandq_s32
+  return vandq_s32(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int64x1_t test_vand_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_vand_s64
+  return vand_s64(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int64x2_t test_vandq_s64(int64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_vandq_s64
+  return vandq_s64(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint8x8_t test_vand_u8(uint8x8_t a, uint8x8_t b) {
+  // CHECK-LABEL: test_vand_u8
+  return vand_u8(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_vandq_u8(uint8x16_t a, uint8x16_t b) {
+  // CHECK-LABEL: test_vandq_u8
+  return vandq_u8(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_vand_u16(uint16x4_t a, uint16x4_t b) {
+  // CHECK-LABEL: test_vand_u16
+  return vand_u16(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint16x8_t test_vandq_u16(uint16x8_t a, uint16x8_t b) {
+  // CHECK-LABEL: test_vandq_u16
+  return vandq_u16(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint32x2_t test_vand_u32(uint32x2_t a, uint32x2_t b) {
+  // CHECK-LABEL: test_vand_u32
+  return vand_u32(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint32x4_t test_vandq_u32(uint32x4_t a, uint32x4_t b) {
+  // CHECK-LABEL: test_vandq_u32
+  return vandq_u32(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint64x1_t test_vand_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_vand_u64
+  return vand_u64(a, b);
+  // CHECK: and {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint64x2_t test_vandq_u64(uint64x2_t a, uint64x2_t b) {
+  // CHECK-LABEL: test_vandq_u64
+  return vandq_u64(a, b);
+  // CHECK: and {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int8x8_t test_vorr_s8(int8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_vorr_s8
+  return vorr_s8(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int8x16_t test_vorrq_s8(int8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_vorrq_s8
+  return vorrq_s8(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int16x4_t test_vorr_s16(int16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_vorr_s16
+  return vorr_s16(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int16x8_t test_vorrq_s16(int16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_vorrq_s16
+  return vorrq_s16(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int32x2_t test_vorr_s32(int32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_vorr_s32
+  return vorr_s32(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int32x4_t test_vorrq_s32(int32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_vorrq_s32
+  return vorrq_s32(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int64x1_t test_vorr_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_vorr_s64
+  return vorr_s64(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int64x2_t test_vorrq_s64(int64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_vorrq_s64
+  return vorrq_s64(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint8x8_t test_vorr_u8(uint8x8_t a, uint8x8_t b) {
+  // CHECK-LABEL: test_vorr_u8
+  return vorr_u8(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_vorrq_u8(uint8x16_t a, uint8x16_t b) {
+  // CHECK-LABEL: test_vorrq_u8
+  return vorrq_u8(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_vorr_u16(uint16x4_t a, uint16x4_t b) {
+  // CHECK-LABEL: test_vorr_u16
+  return vorr_u16(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint16x8_t test_vorrq_u16(uint16x8_t a, uint16x8_t b) {
+  // CHECK-LABEL: test_vorrq_u16
+  return vorrq_u16(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint32x2_t test_vorr_u32(uint32x2_t a, uint32x2_t b) {
+  // CHECK-LABEL: test_vorr_u32
+  return vorr_u32(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint32x4_t test_vorrq_u32(uint32x4_t a, uint32x4_t b) {
+  // CHECK-LABEL: test_vorrq_u32
+  return vorrq_u32(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint64x1_t test_vorr_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_vorr_u64
+  return vorr_u64(a, b);
+  // CHECK: orr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint64x2_t test_vorrq_u64(uint64x2_t a, uint64x2_t b) {
+  // CHECK-LABEL: test_vorrq_u64
+  return vorrq_u64(a, b);
+  // CHECK: orr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int8x8_t test_veor_s8(int8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_veor_s8
+  return veor_s8(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int8x16_t test_veorq_s8(int8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_veorq_s8
+  return veorq_s8(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int16x4_t test_veor_s16(int16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_veor_s16
+  return veor_s16(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int16x8_t test_veorq_s16(int16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_veorq_s16
+  return veorq_s16(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int32x2_t test_veor_s32(int32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_veor_s32
+  return veor_s32(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int32x4_t test_veorq_s32(int32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_veorq_s32
+  return veorq_s32(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int64x1_t test_veor_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_veor_s64
+  return veor_s64(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int64x2_t test_veorq_s64(int64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_veorq_s64
+  return veorq_s64(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint8x8_t test_veor_u8(uint8x8_t a, uint8x8_t b) {
+  // CHECK-LABEL: test_veor_u8
+  return veor_u8(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_veorq_u8(uint8x16_t a, uint8x16_t b) {
+  // CHECK-LABEL: test_veorq_u8
+  return veorq_u8(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_veor_u16(uint16x4_t a, uint16x4_t b) {
+  // CHECK-LABEL: test_veor_u16
+  return veor_u16(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint16x8_t test_veorq_u16(uint16x8_t a, uint16x8_t b) {
+  // CHECK-LABEL: test_veorq_u16
+  return veorq_u16(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint32x2_t test_veor_u32(uint32x2_t a, uint32x2_t b) {
+  // CHECK-LABEL: test_veor_u32
+  return veor_u32(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint32x4_t test_veorq_u32(uint32x4_t a, uint32x4_t b) {
+  // CHECK-LABEL: test_veorq_u32
+  return veorq_u32(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint64x1_t test_veor_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_veor_u64
+  return veor_u64(a, b);
+  // CHECK: eor {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint64x2_t test_veorq_u64(uint64x2_t a, uint64x2_t b) {
+  // CHECK-LABEL: test_veorq_u64
+  return veorq_u64(a, b);
+  // CHECK: eor {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int8x8_t test_vbic_s8(int8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_vbic_s8
+  return vbic_s8(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int8x16_t test_vbicq_s8(int8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_vbicq_s8
+  return vbicq_s8(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int16x4_t test_vbic_s16(int16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_vbic_s16
+  return vbic_s16(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int16x8_t test_vbicq_s16(int16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_vbicq_s16
+  return vbicq_s16(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int32x2_t test_vbic_s32(int32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_vbic_s32
+  return vbic_s32(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int32x4_t test_vbicq_s32(int32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_vbicq_s32
+  return vbicq_s32(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int64x1_t test_vbic_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_vbic_s64
+  return vbic_s64(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int64x2_t test_vbicq_s64(int64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_vbicq_s64
+  return vbicq_s64(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint8x8_t test_vbic_u8(uint8x8_t a, uint8x8_t b) {
+  // CHECK-LABEL: test_vbic_u8
+  return vbic_u8(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_vbicq_u8(uint8x16_t a, uint8x16_t b) {
+  // CHECK-LABEL: test_vbicq_u8
+  return vbicq_u8(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_vbic_u16(uint16x4_t a, uint16x4_t b) {
+  // CHECK-LABEL: test_vbic_u16
+  return vbic_u16(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint16x8_t test_vbicq_u16(uint16x8_t a, uint16x8_t b) {
+  // CHECK-LABEL: test_vbicq_u16
+  return vbicq_u16(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint32x2_t test_vbic_u32(uint32x2_t a, uint32x2_t b) {
+  // CHECK-LABEL: test_vbic_u32
+  return vbic_u32(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint32x4_t test_vbicq_u32(uint32x4_t a, uint32x4_t b) {
+  // CHECK-LABEL: test_vbicq_u32
+  return vbicq_u32(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint64x1_t test_vbic_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_vbic_u64
+  return vbic_u64(a, b);
+  // CHECK: bic {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint64x2_t test_vbicq_u64(uint64x2_t a, uint64x2_t b) {
+  // CHECK-LABEL: test_vbicq_u64
+  return vbicq_u64(a, b);
+  // CHECK: bic {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int8x8_t test_vorn_s8(int8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_vorn_s8
+  return vorn_s8(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int8x16_t test_vornq_s8(int8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_vornq_s8
+  return vornq_s8(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int16x4_t test_vorn_s16(int16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_vorn_s16
+  return vorn_s16(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int16x8_t test_vornq_s16(int16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_vornq_s16
+  return vornq_s16(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int32x2_t test_vorn_s32(int32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_vorn_s32
+  return vorn_s32(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int32x4_t test_vornq_s32(int32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_vornq_s32
+  return vornq_s32(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+int64x1_t test_vorn_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_vorn_s64
+  return vorn_s64(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+int64x2_t test_vornq_s64(int64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_vornq_s64
+  return vornq_s64(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint8x8_t test_vorn_u8(uint8x8_t a, uint8x8_t b) {
+  // CHECK-LABEL: test_vorn_u8
+  return vorn_u8(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_vornq_u8(uint8x16_t a, uint8x16_t b) {
+  // CHECK-LABEL: test_vornq_u8
+  return vornq_u8(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_vorn_u16(uint16x4_t a, uint16x4_t b) {
+  // CHECK-LABEL: test_vorn_u16
+  return vorn_u16(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint16x8_t test_vornq_u16(uint16x8_t a, uint16x8_t b) {
+  // CHECK-LABEL: test_vornq_u16
+  return vornq_u16(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint32x2_t test_vorn_u32(uint32x2_t a, uint32x2_t b) {
+  // CHECK-LABEL: test_vorn_u32
+  return vorn_u32(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint32x4_t test_vornq_u32(uint32x4_t a, uint32x4_t b) {
+  // CHECK-LABEL: test_vornq_u32
+  return vornq_u32(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint64x1_t test_vorn_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_vorn_u64
+  return vorn_u64(a, b);
+  // CHECK: orn {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint64x2_t test_vornq_u64(uint64x2_t a, uint64x2_t b) {
+  // CHECK-LABEL: test_vornq_u64
+  return vornq_u64(a, b);
+  // CHECK: orn {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
diff --git a/test/CodeGen/aarch64-neon-across.c b/test/CodeGen/aarch64-neon-across.c
index 257b839..17ac3ef 100644
--- a/test/CodeGen/aarch64-neon-across.c
+++ b/test/CodeGen/aarch64-neon-across.c
@@ -1,271 +1,274 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int16_t test_vaddlv_s8(int8x8_t a) {
-  // CHECK: test_vaddlv_s8
+  // CHECK-LABEL: test_vaddlv_s8
   return vaddlv_s8(a);
   // CHECK: saddlv {{h[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 int32_t test_vaddlv_s16(int16x4_t a) {
-  // CHECK: test_vaddlv_s16
+  // CHECK-LABEL: test_vaddlv_s16
   return vaddlv_s16(a);
   // CHECK: saddlv {{s[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 uint16_t test_vaddlv_u8(uint8x8_t a) {
-  // CHECK: test_vaddlv_u8
+  // CHECK-LABEL: test_vaddlv_u8
   return vaddlv_u8(a);
   // CHECK: uaddlv {{h[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 uint32_t test_vaddlv_u16(uint16x4_t a) {
-  // CHECK: test_vaddlv_u16
+  // CHECK-LABEL: test_vaddlv_u16
   return vaddlv_u16(a);
   // CHECK: uaddlv {{s[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 int16_t test_vaddlvq_s8(int8x16_t a) {
-  // CHECK: test_vaddlvq_s8
+  // CHECK-LABEL: test_vaddlvq_s8
   return vaddlvq_s8(a);
   // CHECK: saddlv {{h[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 int32_t test_vaddlvq_s16(int16x8_t a) {
-  // CHECK: test_vaddlvq_s16
+  // CHECK-LABEL: test_vaddlvq_s16
   return vaddlvq_s16(a);
   // CHECK: saddlv {{s[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 int64_t test_vaddlvq_s32(int32x4_t a) {
-  // CHECK: test_vaddlvq_s32
+  // CHECK-LABEL: test_vaddlvq_s32
   return vaddlvq_s32(a);
   // CHECK: saddlv {{d[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 uint16_t test_vaddlvq_u8(uint8x16_t a) {
-  // CHECK: test_vaddlvq_u8
+  // CHECK-LABEL: test_vaddlvq_u8
   return vaddlvq_u8(a);
   // CHECK: uaddlv {{h[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 uint32_t test_vaddlvq_u16(uint16x8_t a) {
-  // CHECK: test_vaddlvq_u16
+  // CHECK-LABEL: test_vaddlvq_u16
   return vaddlvq_u16(a);
   // CHECK: uaddlv {{s[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 uint64_t test_vaddlvq_u32(uint32x4_t a) {
-  // CHECK: test_vaddlvq_u32
+  // CHECK-LABEL: test_vaddlvq_u32
   return vaddlvq_u32(a);
   // CHECK: uaddlv {{d[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 int8_t test_vmaxv_s8(int8x8_t a) {
-  // CHECK: test_vmaxv_s8
+  // CHECK-LABEL: test_vmaxv_s8
   return vmaxv_s8(a);
   // CHECK: smaxv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 int16_t test_vmaxv_s16(int16x4_t a) {
-  // CHECK: test_vmaxv_s16
+  // CHECK-LABEL: test_vmaxv_s16
   return vmaxv_s16(a);
   // CHECK: smaxv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 uint8_t test_vmaxv_u8(uint8x8_t a) {
-  // CHECK: test_vmaxv_u8
+  // CHECK-LABEL: test_vmaxv_u8
   return vmaxv_u8(a);
   // CHECK: umaxv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 uint16_t test_vmaxv_u16(uint16x4_t a) {
-  // CHECK: test_vmaxv_u16
+  // CHECK-LABEL: test_vmaxv_u16
   return vmaxv_u16(a);
   // CHECK: umaxv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 int8_t test_vmaxvq_s8(int8x16_t a) {
-  // CHECK: test_vmaxvq_s8
+  // CHECK-LABEL: test_vmaxvq_s8
   return vmaxvq_s8(a);
   // CHECK: smaxv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 int16_t test_vmaxvq_s16(int16x8_t a) {
-  // CHECK: test_vmaxvq_s16
+  // CHECK-LABEL: test_vmaxvq_s16
   return vmaxvq_s16(a);
   // CHECK: smaxv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 int32_t test_vmaxvq_s32(int32x4_t a) {
-  // CHECK: test_vmaxvq_s32
+  // CHECK-LABEL: test_vmaxvq_s32
   return vmaxvq_s32(a);
   // CHECK: smaxv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 uint8_t test_vmaxvq_u8(uint8x16_t a) {
-  // CHECK: test_vmaxvq_u8
+  // CHECK-LABEL: test_vmaxvq_u8
   return vmaxvq_u8(a);
   // CHECK: umaxv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 uint16_t test_vmaxvq_u16(uint16x8_t a) {
-  // CHECK: test_vmaxvq_u16
+  // CHECK-LABEL: test_vmaxvq_u16
   return vmaxvq_u16(a);
   // CHECK: umaxv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 uint32_t test_vmaxvq_u32(uint32x4_t a) {
-  // CHECK: test_vmaxvq_u32
+  // CHECK-LABEL: test_vmaxvq_u32
   return vmaxvq_u32(a);
   // CHECK: umaxv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 int8_t test_vminv_s8(int8x8_t a) {
-  // CHECK: test_vminv_s8
+  // CHECK-LABEL: test_vminv_s8
   return vminv_s8(a);
   // CHECK: sminv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 int16_t test_vminv_s16(int16x4_t a) {
-  // CHECK: test_vminv_s16
+  // CHECK-LABEL: test_vminv_s16
   return vminv_s16(a);
   // CHECK: sminv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 uint8_t test_vminv_u8(uint8x8_t a) {
-  // CHECK: test_vminv_u8
+  // CHECK-LABEL: test_vminv_u8
   return vminv_u8(a);
   // CHECK: uminv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 uint16_t test_vminv_u16(uint16x4_t a) {
-  // CHECK: test_vminv_u16
+  // CHECK-LABEL: test_vminv_u16
   return vminv_u16(a);
   // CHECK: uminv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 int8_t test_vminvq_s8(int8x16_t a) {
-  // CHECK: test_vminvq_s8
+  // CHECK-LABEL: test_vminvq_s8
   return vminvq_s8(a);
   // CHECK: sminv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 int16_t test_vminvq_s16(int16x8_t a) {
-  // CHECK: test_vminvq_s16
+  // CHECK-LABEL: test_vminvq_s16
   return vminvq_s16(a);
   // CHECK: sminv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 int32_t test_vminvq_s32(int32x4_t a) {
-  // CHECK: test_vminvq_s32
+  // CHECK-LABEL: test_vminvq_s32
   return vminvq_s32(a);
   // CHECK: sminv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 uint8_t test_vminvq_u8(uint8x16_t a) {
-  // CHECK: test_vminvq_u8
+  // CHECK-LABEL: test_vminvq_u8
   return vminvq_u8(a);
   // CHECK: uminv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 uint16_t test_vminvq_u16(uint16x8_t a) {
-  // CHECK: test_vminvq_u16
+  // CHECK-LABEL: test_vminvq_u16
   return vminvq_u16(a);
   // CHECK: uminv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 uint32_t test_vminvq_u32(uint32x4_t a) {
-  // CHECK: test_vminvq_u32
+  // CHECK-LABEL: test_vminvq_u32
   return vminvq_u32(a);
   // CHECK: uminv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 int8_t test_vaddv_s8(int8x8_t a) {
-  // CHECK: test_vaddv_s8
+  // CHECK-LABEL: test_vaddv_s8
   return vaddv_s8(a);
   // CHECK: addv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 int16_t test_vaddv_s16(int16x4_t a) {
-  // CHECK: test_vaddv_s16
+  // CHECK-LABEL: test_vaddv_s16
   return vaddv_s16(a);
   // CHECK: addv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 uint8_t test_vaddv_u8(uint8x8_t a) {
-  // CHECK: test_vaddv_u8
+  // CHECK-LABEL: test_vaddv_u8
   return vaddv_u8(a);
   // CHECK: addv {{b[0-9]+}}, {{v[0-9]+}}.8b
 }
 
 uint16_t test_vaddv_u16(uint16x4_t a) {
-  // CHECK: test_vaddv_u16
+  // CHECK-LABEL: test_vaddv_u16
   return vaddv_u16(a);
   // CHECK: addv {{h[0-9]+}}, {{v[0-9]+}}.4h
 }
 
 int8_t test_vaddvq_s8(int8x16_t a) {
-  // CHECK: test_vaddvq_s8
+  // CHECK-LABEL: test_vaddvq_s8
   return vaddvq_s8(a);
   // CHECK: addv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 int16_t test_vaddvq_s16(int16x8_t a) {
-  // CHECK: test_vaddvq_s16
+  // CHECK-LABEL: test_vaddvq_s16
   return vaddvq_s16(a);
   // CHECK: addv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 int32_t test_vaddvq_s32(int32x4_t a) {
-  // CHECK: test_vaddvq_s32
+  // CHECK-LABEL: test_vaddvq_s32
   return vaddvq_s32(a);
   // CHECK: addv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 uint8_t test_vaddvq_u8(uint8x16_t a) {
-  // CHECK: test_vaddvq_u8
+  // CHECK-LABEL: test_vaddvq_u8
   return vaddvq_u8(a);
   // CHECK: addv {{b[0-9]+}}, {{v[0-9]+}}.16b
 }
 
 uint16_t test_vaddvq_u16(uint16x8_t a) {
-  // CHECK: test_vaddvq_u16
+  // CHECK-LABEL: test_vaddvq_u16
   return vaddvq_u16(a);
   // CHECK: addv {{h[0-9]+}}, {{v[0-9]+}}.8h
 }
 
 uint32_t test_vaddvq_u32(uint32x4_t a) {
-  // CHECK: test_vaddvq_u32
+  // CHECK-LABEL: test_vaddvq_u32
   return vaddvq_u32(a);
   // CHECK: addv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 float32_t test_vmaxvq_f32(float32x4_t a) {
-  // CHECK: test_vmaxvq_f32
+  // CHECK-LABEL: test_vmaxvq_f32
   return vmaxvq_f32(a);
   // CHECK: fmaxv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 float32_t test_vminvq_f32(float32x4_t a) {
-  // CHECK: test_vminvq_f32
+  // CHECK-LABEL: test_vminvq_f32
   return vminvq_f32(a);
   // CHECK: fminv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 float32_t test_vmaxnmvq_f32(float32x4_t a) {
-  // CHECK: test_vmaxnmvq_f32
+  // CHECK-LABEL: test_vmaxnmvq_f32
   return vmaxnmvq_f32(a);
   // CHECK: fmaxnmv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
 
 float32_t test_vminnmvq_f32(float32x4_t a) {
-  // CHECK: test_vminnmvq_f32
+  // CHECK-LABEL: test_vminnmvq_f32
   return vminnmvq_f32(a);
   // CHECK: fminnmv {{s[0-9]+}}, {{v[0-9]+}}.4s
 }
diff --git a/test/CodeGen/aarch64-neon-copy.c b/test/CodeGen/aarch64-neon-copy.c
index eb91bf9..f8b3d90 100644
--- a/test/CodeGen/aarch64-neon-copy.c
+++ b/test/CodeGen/aarch64-neon-copy.c
@@ -7,1313 +7,1410 @@
 #include <arm_neon.h>
 
 uint8x8_t test_vset_lane_u8(uint8_t v1, uint8x8_t v2) {
-   // CHECK: test_vset_lane_u8
+   // CHECK-LABEL: test_vset_lane_u8
   return vset_lane_u8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 uint16x4_t test_vset_lane_u16(uint16_t v1, uint16x4_t v2) {
-   // CHECK: test_vset_lane_u16
+   // CHECK-LABEL: test_vset_lane_u16
   return vset_lane_u16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 uint32x2_t test_vset_lane_u32(uint32_t v1, uint32x2_t v2) {
-   // CHECK: test_vset_lane_u32
+   // CHECK-LABEL: test_vset_lane_u32
   return vset_lane_u32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{w[0-9]+}}
 }
 uint64x1_t test_vset_lane_u64(uint64_t v1, uint64x1_t v2) {
-   // CHECK: test_vset_lane_u64
+   // CHECK-LABEL: test_vset_lane_u64
   return vset_lane_u64(v1, v2, 0);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int8x8_t test_vset_lane_s8(int8_t v1, int8x8_t v2) {
-   // CHECK: test_vset_lane_s8
+   // CHECK-LABEL: test_vset_lane_s8
   return vset_lane_s8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 int16x4_t test_vset_lane_s16(int16_t v1, int16x4_t v2) {
-   // CHECK: test_vset_lane_s16
+   // CHECK-LABEL: test_vset_lane_s16
   return vset_lane_s16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 int32x2_t test_vset_lane_s32(int32_t v1, int32x2_t v2) {
-   // CHECK: test_vset_lane_s32
+   // CHECK-LABEL: test_vset_lane_s32
   return vset_lane_s32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{w[0-9]+}}
 }
 
   int64x1_t test_vset_lane_s64(int64_t v1, int64x1_t v2) {
-   // CHECK: test_vset_lane_s64
+   // CHECK-LABEL: test_vset_lane_s64
   return vset_lane_s64(v1, v2, 0);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint8x16_t test_vsetq_lane_u8(uint8_t v1, uint8x16_t v2) {
-   // CHECK: test_vsetq_lane_u8
+   // CHECK-LABEL: test_vsetq_lane_u8
   return vsetq_lane_u8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 uint16x8_t test_vsetq_lane_u16(uint16_t v1, uint16x8_t v2) {
-   // CHECK: test_vsetq_lane_u16
+   // CHECK-LABEL: test_vsetq_lane_u16
   return vsetq_lane_u16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 uint32x4_t test_vsetq_lane_u32(uint32_t v1, uint32x4_t v2) {
-   // CHECK: test_vsetq_lane_u32
+   // CHECK-LABEL: test_vsetq_lane_u32
   return vsetq_lane_u32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{w[0-9]+}}
 }
 
   uint64x2_t test_vsetq_lane_u64(uint64_t v1, uint64x2_t v2) {
-   // CHECK: test_vsetq_lane_u64
+   // CHECK-LABEL: test_vsetq_lane_u64
   return vsetq_lane_u64(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.d[1], {{x[0-9]+}}
 }
 
 int8x16_t test_vsetq_lane_s8(int8_t v1, int8x16_t v2) {
-   // CHECK: test_vsetq_lane_s8
+   // CHECK-LABEL: test_vsetq_lane_s8
   return vsetq_lane_s8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 int16x8_t test_vsetq_lane_s16(int16_t v1, int16x8_t v2) {
-   // CHECK: test_vsetq_lane_s16
+   // CHECK-LABEL: test_vsetq_lane_s16
   return vsetq_lane_s16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 int32x4_t test_vsetq_lane_s32(int32_t v1, int32x4_t v2) {
-   // CHECK: test_vsetq_lane_s32
+   // CHECK-LABEL: test_vsetq_lane_s32
   return vsetq_lane_s32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{w[0-9]+}}
 }
 
 int64x2_t test_vsetq_lane_s64(int64_t v1, int64x2_t v2) {
-   // CHECK: test_vsetq_lane_s64
+   // CHECK-LABEL: test_vsetq_lane_s64
   return vsetq_lane_s64(v1, v2, 0);
   // CHECK: ins {{v[0-9]+}}.d[0], {{x[0-9]+}}
 }
 
 poly8x8_t test_vset_lane_p8(poly8_t v1, poly8x8_t v2) {
-   // CHECK: test_vset_lane_p8
+   // CHECK-LABEL: test_vset_lane_p8
   return vset_lane_p8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 poly16x4_t test_vset_lane_p16(poly16_t v1, poly16x4_t v2) {
-   // CHECK: test_vset_lane_p16
+   // CHECK-LABEL: test_vset_lane_p16
   return vset_lane_p16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 poly8x16_t test_vsetq_lane_p8(poly8_t v1, poly8x16_t v2) {
-   // CHECK: test_vsetq_lane_p8
+   // CHECK-LABEL: test_vsetq_lane_p8
   return vsetq_lane_p8(v1, v2, 6);
   // CHECK: ins {{v[0-9]+}}.b[6], {{w[0-9]+}}
 }
 
 poly16x8_t test_vsetq_lane_p16(poly16_t v1, poly16x8_t v2) {
-   // CHECK: test_vsetq_lane_p16
+   // CHECK-LABEL: test_vsetq_lane_p16
   return vsetq_lane_p16(v1, v2, 2);
   // CHECK: ins {{v[0-9]+}}.h[2], {{w[0-9]+}}
 }
 
 float32x2_t test_vset_lane_f32(float32_t v1, float32x2_t v2) {
-   // CHECK: test_vset_lane_f32
+   // CHECK-LABEL: test_vset_lane_f32
   return vset_lane_f32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
 }
 
 float32x4_t test_vsetq_lane_f32(float32_t v1, float32x4_t v2) {
-   // CHECK: test_vsetq_lane_f32
+   // CHECK-LABEL: test_vsetq_lane_f32
   return vsetq_lane_f32(v1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
 }
 
 float64x1_t test_vset_lane_f64(float64_t v1, float64x1_t v2) {
-   // CHECK: test_vset_lane_f64
+   // CHECK-LABEL: test_vset_lane_f64
   return vset_lane_f64(v1, v2, 0);
   // CHECK: ret
 }
 
 float64x2_t test_vsetq_lane_f64(float64_t v1, float64x2_t v2) {
-   // CHECK: test_vsetq_lane_f64
+   // CHECK-LABEL: test_vsetq_lane_f64
   return vsetq_lane_f64(v1, v2, 0);
   // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[0]
 }
 
 uint8_t test_vget_lane_u8(uint8x8_t v1) {
-  // CHECK: test_vget_lane_u8
+  // CHECK-LABEL: test_vget_lane_u8
   return vget_lane_u8(v1, 7);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
 uint16_t test_vget_lane_u16(uint16x4_t v1) {
-  // CHECK: test_vget_lane_u16
+  // CHECK-LABEL: test_vget_lane_u16
   return vget_lane_u16(v1, 3);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
 uint32_t test_vget_lane_u32(uint32x2_t v1) {
-  // CHECK: test_vget_lane_u32
+  // CHECK-LABEL: test_vget_lane_u32
   return vget_lane_u32(v1, 1);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 uint64_t test_vget_lane_u64(uint64x1_t v1) {
-  // CHECK: test_vget_lane_u64
+  // CHECK-LABEL: test_vget_lane_u64
   return vget_lane_u64(v1, 0);
   // CHECK: fmov {{x[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8_t test_vgetq_lane_u8(uint8x16_t v1) {
-  // CHECK: test_vgetq_lane_u8
+  // CHECK-LABEL: test_vgetq_lane_u8
   return vgetq_lane_u8(v1, 15);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[15]
 }
 
 uint16_t test_vgetq_lane_u16(uint16x8_t v1) {
-  // CHECK: test_vgetq_lane_u16
+  // CHECK-LABEL: test_vgetq_lane_u16
   return vgetq_lane_u16(v1, 6);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[6]
 }
 
 uint32_t test_vgetq_lane_u32(uint32x4_t v1) {
-  // CHECK: test_vgetq_lane_u32
+  // CHECK-LABEL: test_vgetq_lane_u32
   return vgetq_lane_u32(v1, 2);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[2]
 }
 
 uint64_t test_vgetq_lane_u64(uint64x2_t v1) {
-  // CHECK: test_vgetq_lane_u64
+  // CHECK-LABEL: test_vgetq_lane_u64
   return vgetq_lane_u64(v1, 1);
   // CHECK: umov {{x[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 poly8_t test_vget_lane_p8(poly8x8_t v1) {
-  // CHECK: test_vget_lane_p8
+  // CHECK-LABEL: test_vget_lane_p8
   return vget_lane_p8(v1, 7);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
 poly16_t test_vget_lane_p16(poly16x4_t v1) {
-  // CHECK: test_vget_lane_p16
+  // CHECK-LABEL: test_vget_lane_p16
   return vget_lane_p16(v1, 3);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
 poly8_t test_vgetq_lane_p8(poly8x16_t v1) {
-  // CHECK: test_vgetq_lane_p8
+  // CHECK-LABEL: test_vgetq_lane_p8
   return vgetq_lane_p8(v1, 14);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[14]
 }
 
 poly16_t test_vgetq_lane_p16(poly16x8_t v1) {
-  // CHECK: test_vgetq_lane_p16
+  // CHECK-LABEL: test_vgetq_lane_p16
   return vgetq_lane_p16(v1, 6);
   // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[6]
 }
 
 int32_t test_vget_lane_s8(int8x8_t v1) {
-  // CHECK: test_vget_lane_s8
+  // CHECK-LABEL: test_vget_lane_s8
   return vget_lane_s8(v1, 7)+1;
   // CHECK: smov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
 int32_t test_vget_lane_s16(int16x4_t v1) {
-  // CHECK: test_vget_lane_s16
+  // CHECK-LABEL: test_vget_lane_s16
   return vget_lane_s16(v1, 3)+1;
   // CHECK: smov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
 int64_t test_vget_lane_s32(int32x2_t v1) {
-  // CHECK: test_vget_lane_s32
+  // CHECK-LABEL: test_vget_lane_s32
   return vget_lane_s32(v1, 1);
   // CHECK: smov {{x[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 int64_t test_vget_lane_s64(int64x1_t v1) {
-  // CHECK: test_vget_lane_s64
+  // CHECK-LABEL: test_vget_lane_s64
   return vget_lane_s64(v1, 0);
   // CHECK: fmov {{x[0-9]+}}, {{d[0-9]+}}
 }
 
 int32_t test_vgetq_lane_s8(int8x16_t v1) {
-  // CHECK: test_vgetq_lane_s8
+  // CHECK-LABEL: test_vgetq_lane_s8
   return vgetq_lane_s8(v1, 15)+1;
   // CHECK: smov {{w[0-9]+}}, {{v[0-9]+}}.b[15]
 }
 
 int32_t test_vgetq_lane_s16(int16x8_t v1) {
-  // CHECK: test_vgetq_lane_s16
+  // CHECK-LABEL: test_vgetq_lane_s16
   return vgetq_lane_s16(v1, 6)+1;
   // CHECK: smov {{w[0-9]+}}, {{v[0-9]+}}.h[6]
 }
 
 int64_t test_vgetq_lane_s32(int32x4_t v1) {
-  // CHECK: test_vgetq_lane_s32
+  // CHECK-LABEL: test_vgetq_lane_s32
   return vgetq_lane_s32(v1, 2);
   // CHECK: smov {{x[0-9]+}}, {{v[0-9]+}}.s[2]
 }
 
 int64_t test_vgetq_lane_s64(int64x2_t v1) {
-  // CHECK: test_vgetq_lane_s64
+  // CHECK-LABEL: test_vgetq_lane_s64
   return vgetq_lane_s64(v1, 1);
   // CHECK: umov {{x[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 int8x8_t test_vcopy_lane_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vcopy_lane_s8
+  // CHECK-LABEL: test_vcopy_lane_s8
   return vcopy_lane_s8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 int16x4_t test_vcopy_lane_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vcopy_lane_s16
+  // CHECK-LABEL: test_vcopy_lane_s16
   return vcopy_lane_s16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 poly8x8_t test_vcopy_lane_p8(poly8x8_t v1, poly8x8_t v2) {
-  // CHECK: test_vcopy_lane_p8
+  // CHECK-LABEL: test_vcopy_lane_p8
   return vcopy_lane_p8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 poly16x4_t test_vcopy_lane_p16(poly16x4_t v1, poly16x4_t v2) {
-  // CHECK: test_vcopy_lane_p16
+  // CHECK-LABEL: test_vcopy_lane_p16
   return vcopy_lane_p16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vcopy_lane_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vcopy_lane_s32
+  // CHECK-LABEL: test_vcopy_lane_s32
   return vcopy_lane_s32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 float32x2_t test_vcopy_lane_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcopy_lane_f32
+  // CHECK-LABEL: test_vcopy_lane_f32
   return vcopy_lane_f32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 uint8x8_t test_vcopy_lane_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vcopy_lane_u8
+  // CHECK-LABEL: test_vcopy_lane_u8
   return vcopy_lane_u8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 uint16x4_t test_vcopy_lane_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vcopy_lane_u16
+  // CHECK-LABEL: test_vcopy_lane_u16
   return vcopy_lane_u16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 uint32x2_t test_vcopy_lane_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vcopy_lane_u32
+  // CHECK-LABEL: test_vcopy_lane_u32
   return vcopy_lane_u32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 int8x8_t test_vcopy_laneq_s8(int8x8_t v1, int8x16_t v2) {
-  // CHECK: test_vcopy_laneq_s8
+  // CHECK-LABEL: test_vcopy_laneq_s8
   return vcopy_laneq_s8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 int16x4_t test_vcopy_laneq_s16(int16x4_t v1, int16x8_t v2) {
-  // CHECK: test_vcopy_laneq_s16
+  // CHECK-LABEL: test_vcopy_laneq_s16
   return vcopy_laneq_s16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 poly8x8_t test_vcopy_laneq_p8(poly8x8_t v1, poly8x16_t v2) {
-  // CHECK: test_vcopy_laneq_p8
+  // CHECK-LABEL: test_vcopy_laneq_p8
   return vcopy_laneq_p8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 poly16x4_t test_vcopy_laneq_p16(poly16x4_t v1, poly16x8_t v2) {
-  // CHECK: test_vcopy_laneq_p16
+  // CHECK-LABEL: test_vcopy_laneq_p16
   return vcopy_laneq_p16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 int32x2_t test_vcopy_laneq_s32(int32x2_t v1, int32x4_t v2) {
-  // CHECK: test_vcopy_laneq_s32
+  // CHECK-LABEL: test_vcopy_laneq_s32
   return vcopy_laneq_s32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 float32x2_t test_vcopy_laneq_f32(float32x2_t v1, float32x4_t v2) {
-  // CHECK: test_vcopy_laneq_f32
+  // CHECK-LABEL: test_vcopy_laneq_f32
   return vcopy_laneq_f32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 uint8x8_t test_vcopy_laneq_u8(uint8x8_t v1, uint8x16_t v2) {
-  // CHECK: test_vcopy_laneq_u8
+  // CHECK-LABEL: test_vcopy_laneq_u8
   return vcopy_laneq_u8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 uint16x4_t test_vcopy_laneq_u16(uint16x4_t v1, uint16x8_t v2) {
-  // CHECK: test_vcopy_laneq_u16
+  // CHECK-LABEL: test_vcopy_laneq_u16
   return vcopy_laneq_u16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 uint32x2_t test_vcopy_laneq_u32(uint32x2_t v1, uint32x4_t v2) {
-  // CHECK: test_vcopy_laneq_u32
+  // CHECK-LABEL: test_vcopy_laneq_u32
   return vcopy_laneq_u32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 int8x16_t test_vcopyq_lane_s8(int8x16_t v1, int8x8_t v2) {
-  // CHECK: test_vcopyq_lane_s8
+  // CHECK-LABEL: test_vcopyq_lane_s8
   return vcopyq_lane_s8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 int16x8_t test_vcopyq_lane_s16(int16x8_t v1, int16x4_t v2) {
-  // CHECK: test_vcopyq_lane_s16
+  // CHECK-LABEL: test_vcopyq_lane_s16
   return vcopyq_lane_s16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 poly8x16_t test_vcopyq_lane_p8(poly8x16_t v1, poly8x8_t v2) {
-  // CHECK: test_vcopyq_lane_p8
+  // CHECK-LABEL: test_vcopyq_lane_p8
   return vcopyq_lane_p8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 poly16x8_t test_vcopyq_lane_p16(poly16x8_t v1, poly16x4_t v2) {
-  // CHECK: test_vcopyq_lane_p16
+  // CHECK-LABEL: test_vcopyq_lane_p16
   return vcopyq_lane_p16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 int32x4_t test_vcopyq_lane_s32(int32x4_t v1, int32x2_t v2) {
-  // CHECK: test_vcopyq_lane_s32
+  // CHECK-LABEL: test_vcopyq_lane_s32
   return vcopyq_lane_s32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 int64x2_t test_vcopyq_lane_s64(int64x2_t v1, int64x1_t v2) {
-  // CHECK: test_vcopyq_lane_s64
+  // CHECK-LABEL: test_vcopyq_lane_s64
   return vcopyq_lane_s64(v1, 1, v2, 0);
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
 }
 
 float32x4_t test_vcopyq_lane_f32(float32x4_t v1, float32x2_t v2) {
-  // CHECK: test_vcopyq_lane_f32
+  // CHECK-LABEL: test_vcopyq_lane_f32
   return vcopyq_lane_f32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 float64x2_t test_vcopyq_lane_f64(float64x2_t v1, float64x1_t v2) {
-  // CHECK: test_vcopyq_lane_f64
+  // CHECK-LABEL: test_vcopyq_lane_f64
   return vcopyq_lane_f64(v1, 1, v2, 0);
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
 }
 
 uint8x16_t test_vcopyq_lane_u8(uint8x16_t v1, uint8x8_t v2) {
-  // CHECK: test_vcopyq_lane_u8
+  // CHECK-LABEL: test_vcopyq_lane_u8
   return vcopyq_lane_u8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 uint16x8_t test_vcopyq_lane_u16(uint16x8_t v1, uint16x4_t v2) {
-  // CHECK: test_vcopyq_lane_u16
+  // CHECK-LABEL: test_vcopyq_lane_u16
   return vcopyq_lane_u16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 uint32x4_t test_vcopyq_lane_u32(uint32x4_t v1, uint32x2_t v2) {
-  // CHECK: test_vcopyq_lane_u32
+  // CHECK-LABEL: test_vcopyq_lane_u32
   return vcopyq_lane_u32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 uint64x2_t test_vcopyq_lane_u64(uint64x2_t v1, uint64x1_t v2) {
-  // CHECK: test_vcopyq_lane_u64
+  // CHECK-LABEL: test_vcopyq_lane_u64
   return vcopyq_lane_u64(v1, 1, v2, 0);
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
 }
 
 int8x16_t test_vcopyq_laneq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vcopyq_laneq_s8
+  // CHECK-LABEL: test_vcopyq_laneq_s8
   return vcopyq_laneq_s8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 int16x8_t test_vcopyq_laneq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vcopyq_laneq_s16
+  // CHECK-LABEL: test_vcopyq_laneq_s16
   return vcopyq_laneq_s16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 poly8x16_t test_vcopyq_laneq_p8(poly8x16_t v1, poly8x16_t v2) {
-  // CHECK: test_vcopyq_laneq_p8
+  // CHECK-LABEL: test_vcopyq_laneq_p8
   return vcopyq_laneq_p8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 poly16x8_t test_vcopyq_laneq_p16(poly16x8_t v1, poly16x8_t v2) {
-  // CHECK: test_vcopyq_laneq_p16
+  // CHECK-LABEL: test_vcopyq_laneq_p16
   return vcopyq_laneq_p16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 int32x4_t test_vcopyq_laneq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vcopyq_laneq_s32
+  // CHECK-LABEL: test_vcopyq_laneq_s32
   return vcopyq_laneq_s32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 float32x4_t test_vcopyq_laneq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcopyq_laneq_f32
+  // CHECK-LABEL: test_vcopyq_laneq_f32
   return vcopyq_laneq_f32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 int64x2_t test_vcopyq_laneq_s64(int64x2_t v1, int64x2_t v2) {
-  // CHECK: test_vcopyq_laneq_s64
+  // CHECK-LABEL: test_vcopyq_laneq_s64
   return vcopyq_laneq_s64(v1, 1, v2, 1);
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[1]
 }
 
 uint8x16_t test_vcopyq_laneq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vcopyq_laneq_u8
+  // CHECK-LABEL: test_vcopyq_laneq_u8
   return vcopyq_laneq_u8(v1, 5, v2, 3);
   // CHECK: ins {{v[0-9]+}}.b[5], {{v[0-9]+}}.b[3]
 }
 
 uint16x8_t test_vcopyq_laneq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vcopyq_laneq_u16
+  // CHECK-LABEL: test_vcopyq_laneq_u16
   return vcopyq_laneq_u16(v1, 2, v2, 3);
   // CHECK: ins {{v[0-9]+}}.h[2], {{v[0-9]+}}.h[3]
 }
 
 uint32x4_t test_vcopyq_laneq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vcopyq_laneq_u32
+  // CHECK-LABEL: test_vcopyq_laneq_u32
   return vcopyq_laneq_u32(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
 }
 
 uint64x2_t test_vcopyq_laneq_u64(uint64x2_t v1, uint64x2_t v2) {
-  // CHECK: test_vcopyq_laneq_u64
+  // CHECK-LABEL: test_vcopyq_laneq_u64
   return vcopyq_laneq_u64(v1, 0, v2, 1);
   // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
 }
 
 int8x8_t test_vcreate_s8(uint64_t v1) {
-  // CHECK: test_vcreate_s8
+  // CHECK-LABEL: test_vcreate_s8
   return vcreate_s8(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int16x4_t test_vcreate_s16(uint64_t v1) {
-  // CHECK: test_vcreate_s16
+  // CHECK-LABEL: test_vcreate_s16
   return vcreate_s16(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int32x2_t test_vcreate_s32(uint64_t v1) {
-  // CHECK: test_vcreate_s32
+  // CHECK-LABEL: test_vcreate_s32
   return vcreate_s32(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int64x1_t test_vcreate_s64(uint64_t v1) {
-  // CHECK: test_vcreate_s64
+  // CHECK-LABEL: test_vcreate_s64
   return vcreate_s64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint8x8_t test_vcreate_u8(uint64_t v1) {
-  // CHECK: test_vcreate_u8
+  // CHECK-LABEL: test_vcreate_u8
   return vcreate_u8(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint16x4_t test_vcreate_u16(uint64_t v1) {
-  // CHECK: test_vcreate_u16
+  // CHECK-LABEL: test_vcreate_u16
   return vcreate_u16(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint32x2_t test_vcreate_u32(uint64_t v1) {
-  // CHECK: test_vcreate_u32
+  // CHECK-LABEL: test_vcreate_u32
   return vcreate_u32(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint64x1_t test_vcreate_u64(uint64_t v1) {
-  // CHECK: test_vcreate_u64
+  // CHECK-LABEL: test_vcreate_u64
   return vcreate_u64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 poly8x8_t test_vcreate_p8(uint64_t v1) {
-  // CHECK: test_vcreate_p8
+  // CHECK-LABEL: test_vcreate_p8
   return vcreate_p8(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 poly16x4_t test_vcreate_p16(uint64_t v1) {
-  // CHECK: test_vcreate_p16
+  // CHECK-LABEL: test_vcreate_p16
   return vcreate_p16(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 float16x4_t test_vcreate_f16(uint64_t v1) {
-  // CHECK: test_vcreate_f16
+  // CHECK-LABEL: test_vcreate_f16
   return vcreate_f16(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 float32x2_t test_vcreate_f32(uint64_t v1) {
-  // CHECK: test_vcreate_f32
+  // CHECK-LABEL: test_vcreate_f32
   return vcreate_f32(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 float64x1_t test_vcreate_f64(uint64_t v1) {
-  // CHECK: test_vcreate_f64
+  // CHECK-LABEL: test_vcreate_f64
   return vcreate_f64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint8x8_t test_vdup_n_u8(uint8_t v1) {
-  // CHECK: test_vdup_n_u8
+  // CHECK-LABEL: test_vdup_n_u8
   return vdup_n_u8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 uint16x4_t test_vdup_n_u16(uint16_t v1) {
-  // CHECK: test_vdup_n_u16
+  // CHECK-LABEL: test_vdup_n_u16
   return vdup_n_u16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 uint32x2_t test_vdup_n_u32(uint32_t v1) {
-  // CHECK: test_vdup_n_u32
+  // CHECK-LABEL: test_vdup_n_u32
   return vdup_n_u32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{w[0-9]+}}
 }
 
 uint64x1_t test_vdup_n_u64(uint64_t v1) {
-  // CHECK: test_vdup_n_u64
+  // CHECK-LABEL: test_vdup_n_u64
   return vdup_n_u64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint8x16_t test_vdupq_n_u8(uint8_t v1) {
-  // CHECK: test_vdupq_n_u8
+  // CHECK-LABEL: test_vdupq_n_u8
   return vdupq_n_u8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 uint16x8_t test_vdupq_n_u16(uint16_t v1) {
-  // CHECK: test_vdupq_n_u16
+  // CHECK-LABEL: test_vdupq_n_u16
   return vdupq_n_u16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 uint32x4_t test_vdupq_n_u32(uint32_t v1) {
-  // CHECK: test_vdupq_n_u32
+  // CHECK-LABEL: test_vdupq_n_u32
   return vdupq_n_u32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{w[0-9]+}}
 }
 
 uint64x2_t test_vdupq_n_u64(uint64_t v1) {
-  // CHECK: test_vdupq_n_u64
+  // CHECK-LABEL: test_vdupq_n_u64
   return vdupq_n_u64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{x[0-9]+}}
 }
 
 int8x8_t test_vdup_n_s8(int8_t v1) {
-  // CHECK: test_vdup_n_s8
+  // CHECK-LABEL: test_vdup_n_s8
   return vdup_n_s8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 int16x4_t test_vdup_n_s16(int16_t v1) {
-  // CHECK: test_vdup_n_s16
+  // CHECK-LABEL: test_vdup_n_s16
   return vdup_n_s16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 int32x2_t test_vdup_n_s32(int32_t v1) {
-  // CHECK: test_vdup_n_s32
+  // CHECK-LABEL: test_vdup_n_s32
   return vdup_n_s32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{w[0-9]+}}
 }
 
 int64x1_t test_vdup_n_s64(int64_t v1) {
-  // CHECK: test_vdup_n_s64
+  // CHECK-LABEL: test_vdup_n_s64
   return vdup_n_s64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int8x16_t test_vdupq_n_s8(int8_t v1) {
-  // CHECK: test_vdupq_n_s8
+  // CHECK-LABEL: test_vdupq_n_s8
   return vdupq_n_s8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 int16x8_t test_vdupq_n_s16(int16_t v1) {
-  // CHECK: test_vdupq_n_s16
+  // CHECK-LABEL: test_vdupq_n_s16
   return vdupq_n_s16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 int32x4_t test_vdupq_n_s32(int32_t v1) {
-  // CHECK: test_vdupq_n_s32
+  // CHECK-LABEL: test_vdupq_n_s32
   return vdupq_n_s32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{w[0-9]+}}
 }
 
 int64x2_t test_vdupq_n_s64(int64_t v1) {
-  // CHECK: test_vdupq_n_s64
+  // CHECK-LABEL: test_vdupq_n_s64
   return vdupq_n_s64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{x[0-9]+}}
 }
 
 poly8x8_t test_vdup_n_p8(poly8_t v1) {
-  // CHECK: test_vdup_n_p8
+  // CHECK-LABEL: test_vdup_n_p8
   return vdup_n_p8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 poly16x4_t test_vdup_n_p16(poly16_t v1) {
-  // CHECK: test_vdup_n_p16
+  // CHECK-LABEL: test_vdup_n_p16
   return vdup_n_p16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 poly8x16_t test_vdupq_n_p8(poly8_t v1) {
-  // CHECK: test_vdupq_n_p8
+  // CHECK-LABEL: test_vdupq_n_p8
   return vdupq_n_p8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 poly16x8_t test_vdupq_n_p16(poly16_t v1) {
-  // CHECK: test_vdupq_n_p16
+  // CHECK-LABEL: test_vdupq_n_p16
   return vdupq_n_p16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 float32x2_t test_vdup_n_f32(float32_t v1) {
-  // CHECK: test_vdup_n_f32
+  // CHECK-LABEL: test_vdup_n_f32
   return vdup_n_f32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float64x1_t test_vdup_n_f64(float64_t v1) {
-  // CHECK: test_vdup_n_f64
+  // CHECK-LABEL: test_vdup_n_f64
   return vdup_n_f64(v1);
   // CHECK: ret
 }
 
 float32x4_t test_vdupq_n_f32(float32_t v1) {
-  // CHECK: test_vdupq_n_f32
+  // CHECK-LABEL: test_vdupq_n_f32
   return vdupq_n_f32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vdupq_n_f64(float64_t v1) {
-  // CHECK: test_vdupq_n_f64
+  // CHECK-LABEL: test_vdupq_n_f64
   return vdupq_n_f64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 int8x8_t test_vdup_lane_s8(int8x8_t v1) {
-  // CHECK: test_vdup_lane_s8
+  // CHECK-LABEL: test_vdup_lane_s8
   return vdup_lane_s8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 int16x4_t test_vdup_lane_s16(int16x4_t v1) {
-  // CHECK: test_vdup_lane_s16
+  // CHECK-LABEL: test_vdup_lane_s16
   return vdup_lane_s16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 int32x2_t test_vdup_lane_s32(int32x2_t v1) {
-  // CHECK: test_vdup_lane_s32
+  // CHECK-LABEL: test_vdup_lane_s32
   return vdup_lane_s32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int64x1_t test_vdup_lane_s64(int64x1_t v1) {
-  // CHECK: test_vdup_lane_s64
+  // CHECK-LABEL: test_vdup_lane_s64
   return vdup_lane_s64(v1, 0);
   // CHECK: ret
 }
 
 int8x16_t test_vdupq_lane_s8(int8x8_t v1) {
-  // CHECK: test_vdupq_lane_s8
+  // CHECK-LABEL: test_vdupq_lane_s8
   return vdupq_lane_s8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 int16x8_t test_vdupq_lane_s16(int16x4_t v1) {
-  // CHECK: test_vdupq_lane_s16
+  // CHECK-LABEL: test_vdupq_lane_s16
   return vdupq_lane_s16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 int32x4_t test_vdupq_lane_s32(int32x2_t v1) {
-  // CHECK: test_vdupq_lane_s32
+  // CHECK-LABEL: test_vdupq_lane_s32
   return vdupq_lane_s32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int64x2_t test_vdupq_lane_s64(int64x1_t v1) {
-  // CHECK: test_vdupq_lane_s64
+  // CHECK-LABEL: test_vdupq_lane_s64
   return vdupq_lane_s64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 uint8x8_t test_vdup_lane_u8(uint8x8_t v1) {
-  // CHECK: test_vdup_lane_u8
+  // CHECK-LABEL: test_vdup_lane_u8
   return vdup_lane_u8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 uint16x4_t test_vdup_lane_u16(uint16x4_t v1) {
-  // CHECK: test_vdup_lane_u16
+  // CHECK-LABEL: test_vdup_lane_u16
   return vdup_lane_u16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 uint32x2_t test_vdup_lane_u32(uint32x2_t v1) {
-  // CHECK: test_vdup_lane_u32
+  // CHECK-LABEL: test_vdup_lane_u32
   return vdup_lane_u32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 uint64x1_t test_vdup_lane_u64(uint64x1_t v1) {
-  // CHECK: test_vdup_lane_u64
+  // CHECK-LABEL: test_vdup_lane_u64
   return vdup_lane_u64(v1, 0);
   // CHECK: ret
 }
 
 uint8x16_t test_vdupq_lane_u8(uint8x8_t v1) {
-  // CHECK: test_vdupq_lane_u8
+  // CHECK-LABEL: test_vdupq_lane_u8
   return vdupq_lane_u8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 uint16x8_t test_vdupq_lane_u16(uint16x4_t v1) {
-  // CHECK: test_vdupq_lane_u16
+  // CHECK-LABEL: test_vdupq_lane_u16
   return vdupq_lane_u16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 uint32x4_t test_vdupq_lane_u32(uint32x2_t v1) {
-  // CHECK: test_vdupq_lane_u32
+  // CHECK-LABEL: test_vdupq_lane_u32
   return vdupq_lane_u32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 uint64x2_t test_vdupq_lane_u64(uint64x1_t v1) {
-  // CHECK: test_vdupq_lane_u64
+  // CHECK-LABEL: test_vdupq_lane_u64
   return vdupq_lane_u64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 int8x8_t test_vdup_laneq_s8(int8x16_t v1) {
-  // CHECK: test_vdup_laneq_s8
+  // CHECK-LABEL: test_vdup_laneq_s8
   return vdup_laneq_s8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 int16x4_t test_vdup_laneq_s16(int16x8_t v1) {
-  // CHECK: test_vdup_laneq_s16
+  // CHECK-LABEL: test_vdup_laneq_s16
   return vdup_laneq_s16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 int32x2_t test_vdup_laneq_s32(int32x4_t v1) {
-  // CHECK: test_vdup_laneq_s32
+  // CHECK-LABEL: test_vdup_laneq_s32
   return vdup_laneq_s32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 int64x1_t test_vdup_laneq_s64(int64x2_t v1) {
-  // CHECK: test_vdup_laneq_s64
+  // CHECK-LABEL: test_vdup_laneq_s64
   return vdup_laneq_s64(v1, 0);
   // CHECK: ret
 }
 
 int8x16_t test_vdupq_laneq_s8(int8x16_t v1) {
-  // CHECK: test_vdupq_laneq_s8
+  // CHECK-LABEL: test_vdupq_laneq_s8
   return vdupq_laneq_s8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 int16x8_t test_vdupq_laneq_s16(int16x8_t v1) {
-  // CHECK: test_vdupq_laneq_s16
+  // CHECK-LABEL: test_vdupq_laneq_s16
   return vdupq_laneq_s16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 int32x4_t test_vdupq_laneq_s32(int32x4_t v1) {
-  // CHECK: test_vdupq_laneq_s32
+  // CHECK-LABEL: test_vdupq_laneq_s32
   return vdupq_laneq_s32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 int64x2_t test_vdupq_laneq_s64(int64x2_t v1) {
-  // CHECK: test_vdupq_laneq_s64
+  // CHECK-LABEL: test_vdupq_laneq_s64
   return vdupq_laneq_s64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 uint8x8_t test_vdup_laneq_u8(uint8x16_t v1) {
-  // CHECK: test_vdup_laneq_u8
+  // CHECK-LABEL: test_vdup_laneq_u8
   return vdup_laneq_u8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 uint16x4_t test_vdup_laneq_u16(uint16x8_t v1) {
-  // CHECK: test_vdup_laneq_u16
+  // CHECK-LABEL: test_vdup_laneq_u16
   return vdup_laneq_u16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 uint32x2_t test_vdup_laneq_u32(uint32x4_t v1) {
-  // CHECK: test_vdup_laneq_u32
+  // CHECK-LABEL: test_vdup_laneq_u32
   return vdup_laneq_u32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 uint64x1_t test_vdup_laneq_u64(uint64x2_t v1) {
-  // CHECK: test_vdup_laneq_u64
+  // CHECK-LABEL: test_vdup_laneq_u64
   return vdup_laneq_u64(v1, 0);
   // CHECK: ret
 }
 
 uint8x16_t test_vdupq_laneq_u8(uint8x16_t v1) {
-  // CHECK: test_vdupq_laneq_u8
+  // CHECK-LABEL: test_vdupq_laneq_u8
   return vdupq_laneq_u8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 uint16x8_t test_vdupq_laneq_u16(uint16x8_t v1) {
-  // CHECK: test_vdupq_laneq_u16
+  // CHECK-LABEL: test_vdupq_laneq_u16
   return vdupq_laneq_u16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 uint32x4_t test_vdupq_laneq_u32(uint32x4_t v1) {
-  // CHECK: test_vdupq_laneq_u32
+  // CHECK-LABEL: test_vdupq_laneq_u32
   return vdupq_laneq_u32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 uint64x2_t test_vdupq_laneq_u64(uint64x2_t v1) {
-  // CHECK: test_vdupq_laneq_u64
+  // CHECK-LABEL: test_vdupq_laneq_u64
   return vdupq_laneq_u64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 poly8x8_t test_vdup_lane_p8(poly8x8_t v1) {
-  // CHECK: test_vdup_lane_p8
+  // CHECK-LABEL: test_vdup_lane_p8
   return vdup_lane_p8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 poly16x4_t test_vdup_lane_p16(poly16x4_t v1) {
-  // CHECK: test_vdup_lane_p16
+  // CHECK-LABEL: test_vdup_lane_p16
   return vdup_lane_p16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 poly8x16_t test_vdupq_lane_p8(poly8x8_t v1) {
-  // CHECK: test_vdupq_lane_p8
+  // CHECK-LABEL: test_vdupq_lane_p8
   return vdupq_lane_p8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 poly16x8_t test_vdupq_lane_p16(poly16x4_t v1) {
-  // CHECK: test_vdupq_lane_p16
+  // CHECK-LABEL: test_vdupq_lane_p16
   return vdupq_lane_p16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 poly8x8_t test_vdup_laneq_p8(poly8x16_t v1) {
-  // CHECK: test_vdup_laneq_p8
+  // CHECK-LABEL: test_vdup_laneq_p8
   return vdup_laneq_p8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.8b, {{v[0-9]+}}.b[5]
 }
 
 poly16x4_t test_vdup_laneq_p16(poly16x8_t v1) {
-  // CHECK: test_vdup_laneq_p16
+  // CHECK-LABEL: test_vdup_laneq_p16
   return vdup_laneq_p16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 poly8x16_t test_vdupq_laneq_p8(poly8x16_t v1) {
-  // CHECK: test_vdupq_laneq_p8
+  // CHECK-LABEL: test_vdupq_laneq_p8
   return vdupq_laneq_p8(v1, 5);
   // CHECK: dup {{v[0-9]+}}.16b, {{v[0-9]+}}.b[5]
 }
 
 poly16x8_t test_vdupq_laneq_p16(poly16x8_t v1) {
-  // CHECK: test_vdupq_laneq_p16
+  // CHECK-LABEL: test_vdupq_laneq_p16
   return vdupq_laneq_p16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 float16x4_t test_vdup_lane_f16(float16x4_t v1) {
-  // CHECK: test_vdup_lane_f16
+  // CHECK-LABEL: test_vdup_lane_f16
   return vdup_lane_f16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 float32x2_t test_vdup_lane_f32(float32x2_t v1) {
-  // CHECK: test_vdup_lane_f32
+  // CHECK-LABEL: test_vdup_lane_f32
   return vdup_lane_f32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 float64x1_t test_vdup_lane_f64(float64x1_t v1) {
-  // CHECK: test_vdup_lane_f64
+  // CHECK-LABEL: test_vdup_lane_f64
   return vdup_lane_f64(v1, 0);
   // CHECK: ret
 }
 
 float16x4_t test_vdup_laneq_f16(float16x8_t v1) {
-  // CHECK: test_vdup_laneq_f16
+  // CHECK-LABEL: test_vdup_laneq_f16
   return vdup_laneq_f16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.4h, {{v[0-9]+}}.h[2]
 }
 
 float32x2_t test_vdup_laneq_f32(float32x4_t v1) {
-  // CHECK: test_vdup_laneq_f32
+  // CHECK-LABEL: test_vdup_laneq_f32
   return vdup_laneq_f32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
 }
 
 float64x1_t test_vdup_laneq_f64(float64x2_t v1) {
-  // CHECK: test_vdup_laneq_f64
+  // CHECK-LABEL: test_vdup_laneq_f64
   return vdup_laneq_f64(v1, 0);
   // CHECK: ret
 }
 
 float16x8_t test_vdupq_lane_f16(float16x4_t v1) {
-  // CHECK: test_vdupq_lane_f16
+  // CHECK-LABEL: test_vdupq_lane_f16
   return vdupq_lane_f16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 float32x4_t test_vdupq_lane_f32(float32x2_t v1) {
-  // CHECK: test_vdupq_lane_f32
+  // CHECK-LABEL: test_vdupq_lane_f32
   return vdupq_lane_f32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float64x2_t test_vdupq_lane_f64(float64x1_t v1) {
-  // CHECK: test_vdupq_lane_f64
+  // CHECK-LABEL: test_vdupq_lane_f64
   return vdupq_lane_f64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 float16x8_t test_vdupq_laneq_f16(float16x8_t v1) {
-  // CHECK: test_vdupq_laneq_f16
+  // CHECK-LABEL: test_vdupq_laneq_f16
   return vdupq_laneq_f16(v1, 2);
   // CHECK: dup {{v[0-9]+}}.8h, {{v[0-9]+}}.h[2]
 }
 
 float32x4_t test_vdupq_laneq_f32(float32x4_t v1) {
-  // CHECK: test_vdupq_laneq_f32
+  // CHECK-LABEL: test_vdupq_laneq_f32
   return vdupq_laneq_f32(v1, 1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
 }
 
 float64x2_t test_vdupq_laneq_f64(float64x2_t v1) {
-  // CHECK: test_vdupq_laneq_f64
+  // CHECK-LABEL: test_vdupq_laneq_f64
   return vdupq_laneq_f64(v1, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 int8x8_t test_vmov_n_s8(int8_t v1) {
-  // CHECK: test_vmov_n_s8
+  // CHECK-LABEL: test_vmov_n_s8
   return vmov_n_s8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 int16x4_t test_vmov_n_s16(int16_t v1) {
-  // CHECK: test_vmov_n_s16
+  // CHECK-LABEL: test_vmov_n_s16
   return vmov_n_s16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 int32x2_t test_vmov_n_s32(int32_t v1) {
-  // CHECK: test_vmov_n_s32
+  // CHECK-LABEL: test_vmov_n_s32
   return vmov_n_s32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{w[0-9]+}}
 }
 
 int64x1_t test_vmov_n_s64(int64_t v1) {
-  // CHECK: test_vmov_n_s64
+  // CHECK-LABEL: test_vmov_n_s64
   return vmov_n_s64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 int8x16_t test_vmovq_n_s8(int8_t v1) {
-  // CHECK: test_vmovq_n_s8
+  // CHECK-LABEL: test_vmovq_n_s8
   return vmovq_n_s8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 int16x8_t test_vmovq_n_s16(int16_t v1) {
-  // CHECK: test_vmovq_n_s16
+  // CHECK-LABEL: test_vmovq_n_s16
   return vmovq_n_s16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 int32x4_t test_vmovq_n_s32(int32_t v1) {
-  // CHECK: test_vmovq_n_s32
+  // CHECK-LABEL: test_vmovq_n_s32
   return vmovq_n_s32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{w[0-9]+}}
 }
 
 int64x2_t test_vmovq_n_s64(int64_t v1) {
-  // CHECK: test_vmovq_n_s64
+  // CHECK-LABEL: test_vmovq_n_s64
   return vmovq_n_s64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{x[0-9]+}}
 }
 
 uint8x8_t test_vmov_n_u8(uint8_t v1) {
-  // CHECK: test_vmov_n_u8
+  // CHECK-LABEL: test_vmov_n_u8
   return vmov_n_u8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 uint16x4_t test_vmov_n_u16(uint16_t v1) {
-  // CHECK: test_vmov_n_u16
+  // CHECK-LABEL: test_vmov_n_u16
   return vmov_n_u16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 uint32x2_t test_vmov_n_u32(uint32_t v1) {
-  // CHECK: test_vmov_n_u32
+  // CHECK-LABEL: test_vmov_n_u32
   return vmov_n_u32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{w[0-9]+}}
 }
 
 uint64x1_t test_vmov_n_u64(uint64_t v1) {
-  // CHECK: test_vmov_n_u64
+  // CHECK-LABEL: test_vmov_n_u64
   return vmov_n_u64(v1);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 uint8x16_t test_vmovq_n_u8(uint8_t v1) {
-  // CHECK: test_vmovq_n_u8
+  // CHECK-LABEL: test_vmovq_n_u8
   return vmovq_n_u8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 uint16x8_t test_vmovq_n_u16(uint16_t v1) {
-  // CHECK: test_vmovq_n_u16
+  // CHECK-LABEL: test_vmovq_n_u16
   return vmovq_n_u16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 uint32x4_t test_vmovq_n_u32(uint32_t v1) {
-  // CHECK: test_vmovq_n_u32
+  // CHECK-LABEL: test_vmovq_n_u32
   return vmovq_n_u32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{w[0-9]+}}
 }
 
 uint64x2_t test_vmovq_n_u64(uint64_t v1) {
-  // CHECK: test_vmovq_n_u64
+  // CHECK-LABEL: test_vmovq_n_u64
   return vmovq_n_u64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{x[0-9]+}}
 }
 
 poly8x8_t test_vmov_n_p8(poly8_t v1) {
-  // CHECK: test_vmov_n_p8
+  // CHECK-LABEL: test_vmov_n_p8
   return vmov_n_p8(v1);
   // CHECK: dup {{v[0-9]+}}.8b, {{w[0-9]+}}
 }
 
 poly16x4_t test_vmov_n_p16(poly16_t v1) {
-  // CHECK: test_vmov_n_p16
+  // CHECK-LABEL: test_vmov_n_p16
   return vmov_n_p16(v1);
   // CHECK: dup {{v[0-9]+}}.4h, {{w[0-9]+}}
 }
 
 poly8x16_t test_vmovq_n_p8(poly8_t v1) {
-  // CHECK: test_vmovq_n_p8
+  // CHECK-LABEL: test_vmovq_n_p8
   return vmovq_n_p8(v1);
   // CHECK: dup {{v[0-9]+}}.16b, {{w[0-9]+}}
 }
 
 poly16x8_t test_vmovq_n_p16(poly16_t v1) {
-  // CHECK: test_vmovq_n_p16
+  // CHECK-LABEL: test_vmovq_n_p16
   return vmovq_n_p16(v1);
   // CHECK: dup {{v[0-9]+}}.8h, {{w[0-9]+}}
 }
 
 float32x2_t test_vmov_n_f32(float32_t v1) {
-  // CHECK: test_vmov_n_f32
+  // CHECK-LABEL: test_vmov_n_f32
   return vmov_n_f32(v1);
   // CHECK: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
 }
 
 float64x1_t test_vmov_n_f64(float64_t v1) {
-  // CHECK: test_vmov_n_f64
+  // CHECK-LABEL: test_vmov_n_f64
   return vmov_n_f64(v1);
   // CHECK: ret
 }
 
 float32x4_t test_vmovq_n_f32(float32_t v1) {
-  // CHECK: test_vmovq_n_f32
+  // CHECK-LABEL: test_vmovq_n_f32
   return vmovq_n_f32(v1);
   // CHECK: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
 }
 
 float64x2_t test_vmovq_n_f64(float64_t v1) {
-  // CHECK: test_vmovq_n_f64
+  // CHECK-LABEL: test_vmovq_n_f64
   return vmovq_n_f64(v1);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vcopy_lane_s64
+// CHECK-LABEL: test_vcopy_lane_s64
 int64x1_t test_vcopy_lane_s64(int64x1_t a, int64x1_t c) {
   return vcopy_lane_s64(a, 0, c, 0);
 // CHECK: fmov {{d[0-9]+}}, {{d[0-9]+}}
 // CHECK-NOT: dup {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vcopy_lane_u64
+// CHECK-LABEL: test_vcopy_lane_u64
 uint64x1_t test_vcopy_lane_u64(uint64x1_t a, uint64x1_t c) {
   return vcopy_lane_u64(a, 0, c, 0);
 // CHECK: fmov {{d[0-9]+}}, {{d[0-9]+}}
 // CHECK-NOT: dup {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vcopy_lane_f64
+// CHECK-LABEL: test_vcopy_lane_f64
 float64x1_t test_vcopy_lane_f64(float64x1_t a, float64x1_t c) {
   return vcopy_lane_f64(a, 0, c, 0);
 // CHECK: fmov {{d[0-9]+}}, {{d[0-9]+}}
 // CHECK-NOT: dup {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vcopy_laneq_s64
+// CHECK-LABEL: test_vcopy_laneq_s64
 int64x1_t test_vcopy_laneq_s64(int64x1_t a, int64x2_t c) {
   return vcopy_laneq_s64(a, 0, c, 1);
 // CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vcopy_laneq_u64
+// CHECK-LABEL: test_vcopy_laneq_u64
 uint64x1_t test_vcopy_laneq_u64(uint64x1_t a, uint64x2_t c) {
   return vcopy_laneq_u64(a, 0, c, 1);
 // CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vcopy_laneq_f64
+// CHECK-LABEL: test_vcopy_laneq_f64
 float64x1_t test_vcopy_laneq_f64(float64x1_t a, float64x2_t c) {
   return vcopy_laneq_f64(a, 0, c, 1);
 // CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vcopy_laneq_p64
+// CHECK-LABEL: test_vcopy_laneq_p64
 poly64x1_t test_vcopy_laneq_p64(poly64x1_t a, poly64x2_t c) {
   return vcopy_laneq_p64(a, 0, c, 1);
 // CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vcopyq_laneq_f64
+// CHECK-LABEL: test_vcopyq_laneq_f64
 float64x2_t test_vcopyq_laneq_f64(float64x2_t a, float64x2_t c) {
 // CHECK: ins  {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[1]
   return vcopyq_laneq_f64(a, 1, c, 1);
 }
 
-// CHECK: test_vget_lane_f16
+// CHECK-LABEL: test_vget_lane_f16
 int test_vget_lane_f16(float16x4_t v1) {
   float16_t a = vget_lane_f16(v1, 3);
   return (int)a;
 // CHECK: dup {{h[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vgetq_lane_f16
+// CHECK-LABEL: test_vgetq_lane_f16
 int test_vgetq_lane_f16(float16x8_t v1) {
   float16_t a = vgetq_lane_f16(v1, 7);
   return (int)a;
 // CHECK: dup {{h[0-9]+}}, {{v[0-9]+}}.h[7]
 }
 
-// CHECK: test_vget_lane_f16_2
-float test_vget_lane_f16_2(float16x4_t v1) {
+// CHECK-LABEL: test2_vget_lane_f16
+float test2_vget_lane_f16(float16x4_t v1) {
   float16_t a = vget_lane_f16(v1, 3);
   return (float)a;
 // CHECK: dup {{h[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vgetq_lane_f16_2
-float test_vgetq_lane_f16_2(float16x8_t v1) {
+// CHECK-LABEL: test2_vgetq_lane_f16
+float test2_vgetq_lane_f16(float16x8_t v1) {
   float16_t a = vgetq_lane_f16(v1, 7);
   return (float)a;
 // CHECK: dup {{h[0-9]+}}, {{v[0-9]+}}.h[7]
 }
 
-// CHECK: test_vset_lane_f16
+// CHECK-LABEL: test_vset_lane_f16
 float16x4_t test_vset_lane_f16(float16x4_t v1) {
-  float16_t a;
+  float16_t a = 0.0;
   return vset_lane_f16(a, v1, 3);
-// CHECK: fmov  {{s[0-9]+}}, wzr
-// CHECK-NEXT: ins {{v[0-9]+}}.h[3],  {{v[0-9]+}}.h[0]
+// CHECK: ins {{v[0-9]+}}.h[3], wzr
 }
 
-// CHECK: test_vsetq_lane_f16
+// CHECK-LABEL: test_vsetq_lane_f16
 float16x8_t test_vsetq_lane_f16(float16x8_t v1) {
-  float16_t a;
+  float16_t a = 0.0;
   return vsetq_lane_f16(a, v1, 7);
-// CHECK: fmov  {{s[0-9]+}}, wzr
-// CHECK-NEXT: ins {{v[0-9]+}}.h[7],  {{v[0-9]+}}.h[0]
+// CHECK: ins {{v[0-9]+}}.h[7], wzr
 }
 
-// CHECK: test_vset_lane_f16_2
-float16x4_t test_vset_lane_f16_2(float16x4_t v1) {
+// CHECK-LABEL: test2_vset_lane_f16
+float16x4_t test2_vset_lane_f16(float16x4_t v1) {
+  float16_t a = 1.0;
+  return vset_lane_f16(a, v1, 3);
+// CHECK:  movz    {{w[0-9]+}}, #15360
+// CHECK-NEXT: ins {{v[0-9]+}}.h[3], {{w[0-9]+}}
+}
+
+// CHECK-LABEL: test2_vsetq_lane_f16
+float16x8_t test2_vsetq_lane_f16(float16x8_t v1) {
+  float16_t a = 1.0;
+  return vsetq_lane_f16(a, v1, 7);
+// CHECK:  movz    {{w[0-9]+}}, #15360
+// CHECK-NEXT: ins {{v[0-9]+}}.h[7],  {{w[0-9]+}}
+}
+
+// CHECK-LABEL: test_vget_vset_lane_f16
+float16x4_t test_vget_vset_lane_f16(float16x4_t v1) {
   float16_t a = vget_lane_f16(v1, 0);
   return vset_lane_f16(a, v1, 3);
 // CHECK: ins {{v[0-9]+}}.h[3],  {{v[0-9]+}}.h[0]
 }
 
-// CHECK: test_vsetq_lane_f16_2
-float16x8_t test_vsetq_lane_f16_2(float16x8_t v1) {
+// CHECK-LABEL: test_vgetq_vsetq_lane_f16
+float16x8_t test_vgetq_vsetq_lane_f16(float16x8_t v1) {
   float16_t a = vgetq_lane_f16(v1, 0);
   return vsetq_lane_f16(a, v1, 7);
 // CHECK: ins {{v[0-9]+}}.h[7],  {{v[0-9]+}}.h[0]
 }
 
+// CHECK-LABEL: test4_vset_lane_f16
+float16x4_t test4_vset_lane_f16(float16x4_t v1, float b, float c) {
+  float16_t a = (float16_t)b;
+  return vset_lane_f16(a, v1, 3);
+// CHECK: fmov {{w[0-9]+}},  {{s[0-9]+}}
+// CHECK: ins {{v[0-9]+}}.h[3],  {{w[0-9]+}}
+}
 
-// CHECK: test_vsetq_lane_f16_3
-float16x8_t test_vsetq_lane_f16_3(float16x8_t v1, float b, float c) {
+// CHECK-LABEL: test4_vsetq_lane_f16
+float16x8_t test4_vsetq_lane_f16(float16x8_t v1, float b, float c) {
   float16_t a = (float16_t)b;
   return vsetq_lane_f16(a, v1, 7);
+// CHECK: fmov {{w[0-9]+}},  {{s[0-9]+}}
 // CHECK: ins {{v[0-9]+}}.h[7],  {{w[0-9]+}}
 }
 
-// CHECK: test_vsetq_lane_f16_4
-float16x8_t test_vsetq_lane_f16_4(float16x8_t v1, float b, float c) {
+// CHECK-LABEL: test5_vset_lane_f16
+float16x4_t test5_vset_lane_f16(float16x4_t v1, float b, float c) {
+  float16_t a = (float16_t)b;
+  return vset_lane_f16(a, v1, 3);
+// CHECK: fmov {{w[0-9]+}},  {{s[0-9]+}}
+// CHECK: ins {{v[0-9]+}}.h[3],  {{w[0-9]+}}
+}
+
+// CHECK-LABEL: test5_vsetq_lane_f16
+float16x8_t test5_vsetq_lane_f16(float16x8_t v1, float b, float c) {
   float16_t a = (float16_t)b + 1.0;
   return vsetq_lane_f16(a, v1, 7);
+// CHECK: fmov {{w[0-9]+}},  {{s[0-9]+}}
 // CHECK: ins {{v[0-9]+}}.h[7],  {{w[0-9]+}}
 }
 
+// CHECK-LABEL: test_vset_vget_lane_f16
+int test_vset_vget_lane_f16(float16x4_t a) {
+  float16x4_t b;
+  b = vset_lane_f16(3.5, a, 3);
+  float16_t c = vget_lane_f16(b, 3);
+  return (int)c;
+// CHECK: movz x{{[0-9]+}}, #3
+}
+
+// CHECK-LABEL: test_vsetq_vgetq_lane_f16
+int test_vsetq_vgetq_lane_f16(float16x8_t a) {
+  float16x8_t b;
+  b = vsetq_lane_f16(3.5, a, 5);
+  float16_t c = vgetq_lane_f16(b, 5);
+  return (int)c;
+// CHECK: movz x{{[0-9]+}}, #3
+}
+
+// CHECK-LABEL: test_vdup_laneq_p64:
+poly64x1_t test_vdup_laneq_p64(poly64x2_t vec) {
+  return vdup_laneq_p64(vec, 0);
+// CHECK-NEXT: ret
+}
+
+// CHECK-LABEL: test_vdup_laneq_p64_1
+poly64x1_t test_vdup_laneq_p64_1(poly64x2_t vec) {
+  return vdup_laneq_p64(vec, 1);
+// CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
+}
+
+// CHECK-LABEL: test_vget_lane_f32
+float32_t test_vget_lane_f32_1(float32x2_t v) {
+  return vget_lane_f32(v, 1);
+// CHECK: dup {{s[0-9]+}}, {{v[0-9]+}}.s[1]
+}
+
+// CHECK-LABEL: test_vget_lane_f64:
+float64_t test_vget_lane_f64(float64x1_t v) {
+  return vget_lane_f64(v, 0);
+// CHECK-NEXT: ret
+}
+
+// CHECK-LABEL: test_vgetq_lane_f64_1
+float64_t test_vgetq_lane_f64_1(float64x2_t v) {
+  return vgetq_lane_f64(v, 1);
+// CHECK: dup {{d[0-9]+}}, {{v[0-9]+}}.d[1]
+}
+
+// CHECK-LABEL: test_vget_lane_f32:
+float32_t test_vget_lane_f32(float32x2_t v) {
+  return vget_lane_f32(v, 0);
+// CHECK-NEXT: ret
+}
+
+// CHECK-LABEL: test_vgetq_lane_f32:
+float32_t test_vgetq_lane_f32(float32x4_t v) {
+  return vgetq_lane_f32(v, 0);
+// CHECK-NEXT: ret
+}
+
+// CHECK-LABEL: test_vgetq_lane_f64:
+float64_t test_vgetq_lane_f64(float64x2_t v) {
+  return vgetq_lane_f64(v, 0);
+// CHECK-NEXT: ret
+}
+
diff --git a/test/CodeGen/aarch64-neon-crypto.c b/test/CodeGen/aarch64-neon-crypto.c
deleted file mode 100644
index 240f379..0000000
--- a/test/CodeGen/aarch64-neon-crypto.c
+++ /dev/null
@@ -1,94 +0,0 @@
-// REQUIRES: aarch64-registered-target
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
-// RUN:   -target-feature +crypto -S -O3 -o - %s | FileCheck %s
-// RUN: not %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
-// RUN:   -S -O3 -o - %s 2>&1 | FileCheck --check-prefix=CHECK-NO-CRYPTO %s
-
-// Test new aarch64 intrinsics and types
-
-#include <arm_neon.h>
-
-uint8x16_t test_vaeseq_u8(uint8x16_t data, uint8x16_t key) {
-  // CHECK: test_vaeseq_u8
-  // CHECK-NO-CRYPTO: warning: implicit declaration of function 'vaeseq_u8' is invalid in C99
-  return vaeseq_u8(data, key);
-  // CHECK: aese {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
-}
-
-uint8x16_t test_vaesdq_u8(uint8x16_t data, uint8x16_t key) {
-  // CHECK: test_vaesdq_u8
-  return vaesdq_u8(data, key);
-  // CHECK: aesd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
-}
-
-uint8x16_t test_vaesmcq_u8(uint8x16_t data) {
-  // CHECK: test_vaesmcq_u8
-  return vaesmcq_u8(data);
-  // CHECK: aesmc {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
-}
-
-uint8x16_t test_vaesimcq_u8(uint8x16_t data) {
-  // CHECK: test_vaesimcq_u8
-  return vaesimcq_u8(data);
-  // CHECK: aesimc {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
-}
-
-uint32_t test_vsha1h_u32(uint32_t hash_e) {
-  // CHECK: test_vsha1h_u32
-  return vsha1h_u32(hash_e);
-  // CHECK: sha1h {{s[0-9]+}}, {{s[0-9]+}}
-}
-
-uint32x4_t test_vsha1su1q_u32(uint32x4_t tw0_3, uint32x4_t w12_15) {
-  // CHECK: test_vsha1su1q_u32
-  return vsha1su1q_u32(tw0_3, w12_15);
-  // CHECK: sha1su1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha256su0q_u32(uint32x4_t w0_3, uint32x4_t w4_7) {
-  // CHECK: test_vsha256su0q_u32
-  return vsha256su0q_u32(w0_3, w4_7);
-  // CHECK: sha256su0 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha1cq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
-  // CHECK: test_vsha1cq_u32
-  return vsha1cq_u32(hash_abcd, hash_e, wk);
-  // CHECK: sha1c {{q[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha1pq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
-  // CHECK: test_vsha1pq_u32
-  return vsha1pq_u32(hash_abcd, hash_e, wk);
-  // CHECK: sha1p {{q[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha1mq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
-  // CHECK: test_vsha1mq_u32
-  return vsha1mq_u32(hash_abcd, hash_e, wk);
-  // CHECK: sha1m {{q[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha1su0q_u32(uint32x4_t w0_3, uint32x4_t w4_7, uint32x4_t w8_11) {
-  // CHECK: test_vsha1su0q_u32
-  return vsha1su0q_u32(w0_3, w4_7, w8_11);
-  // CHECK: sha1su0 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha256hq_u32(uint32x4_t hash_abcd, uint32x4_t hash_efgh, uint32x4_t wk) {
-  // CHECK: test_vsha256hq_u32
-  return vsha256hq_u32(hash_abcd, hash_efgh, wk);
-  // CHECK: sha256h {{q[0-9]+}}, {{q[0-9]+}}, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha256h2q_u32(uint32x4_t hash_efgh, uint32x4_t hash_abcd, uint32x4_t wk) {
-  // CHECK: test_vsha256h2q_u32
-  return vsha256h2q_u32(hash_efgh, hash_abcd, wk);
-  // CHECK: sha256h2 {{q[0-9]+}}, {{q[0-9]+}}, {{v[0-9]+}}.4s
-}
-
-uint32x4_t test_vsha256su1q_u32(uint32x4_t tw0_3, uint32x4_t w8_11, uint32x4_t w12_15) {
-  // CHECK: test_vsha256su1q_u32
-  return vsha256su1q_u32(tw0_3, w8_11, w12_15);
-  // CHECK: sha256su1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
-}
diff --git a/test/CodeGen/aarch64-neon-extract.c b/test/CodeGen/aarch64-neon-extract.c
index faf35af..77d574c 100644
--- a/test/CodeGen/aarch64-neon-extract.c
+++ b/test/CodeGen/aarch64-neon-extract.c
@@ -1,148 +1,151 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int8x8_t test_vext_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vext_s8
+  // CHECK-LABEL: test_vext_s8
   return vext_s8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x2
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?2}}
 }
 
 int16x4_t test_vext_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vext_s16
+  // CHECK-LABEL: test_vext_s16
   return vext_s16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x6
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?6}}
 }
 
 int32x2_t test_vext_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vext_s32
+  // CHECK-LABEL: test_vext_s32
   return vext_s32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x4
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?4}}
 }
 
 int64x1_t test_vext_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vext_s64
+  // CHECK-LABEL: test_vext_s64
   return vext_s64(a, b, 0);
 }
 
 int8x16_t test_vextq_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vextq_s8
+  // CHECK-LABEL: test_vextq_s8
   return vextq_s8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x2
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?2}}
 }
 
 int16x8_t test_vextq_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vextq_s16
+  // CHECK-LABEL: test_vextq_s16
   return vextq_s16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x6
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?6}}
 }
 
 int32x4_t test_vextq_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vextq_s32
+  // CHECK-LABEL: test_vextq_s32
   return vextq_s32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x4
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?4}}
 }
 
 int64x2_t test_vextq_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vextq_s64
+  // CHECK-LABEL: test_vextq_s64
   return vextq_s64(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x8
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?8}}
 }
 
 uint8x8_t test_vext_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vext_u8
+  // CHECK-LABEL: test_vext_u8
   return vext_u8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x2
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?2}}
 }
 
 uint16x4_t test_vext_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vext_u16
+  // CHECK-LABEL: test_vext_u16
   return vext_u16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x6
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?6}}
 }
 
 uint32x2_t test_vext_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vext_u32
+  // CHECK-LABEL: test_vext_u32
   return vext_u32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x4
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?4}}
 }
 
 uint64x1_t test_vext_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vext_u64
+  // CHECK-LABEL: test_vext_u64
   return vext_u64(a, b, 0);
 }
 
 uint8x16_t test_vextq_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vextq_u8
+  // CHECK-LABEL: test_vextq_u8
   return vextq_u8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x2
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?2}}
 }
 
 uint16x8_t test_vextq_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vextq_u16
+  // CHECK-LABEL: test_vextq_u16
   return vextq_u16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x6
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?6}}
 }
 
 uint32x4_t test_vextq_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vextq_u32
+  // CHECK-LABEL: test_vextq_u32
   return vextq_u32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x4
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?4}}
 }
 
 uint64x2_t test_vextq_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vextq_u64
+  // CHECK-LABEL: test_vextq_u64
   return vextq_u64(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x8
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?8}}
 }
 
 float32x2_t test_vext_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vext_f32
+  // CHECK-LABEL: test_vext_f32
   return vext_f32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x4
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?4}}
 }
 
 float64x1_t test_vext_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vext_f64
+  // CHECK-LABEL: test_vext_f64
   return vext_f64(a, b, 0);
 }
 
 float32x4_t test_vextq_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vextq_f32
+  // CHECK-LABEL: test_vextq_f32
   return vextq_f32(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x4
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?4}}
 }
 
 float64x2_t test_vextq_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vextq_f64
+  // CHECK-LABEL: test_vextq_f64
   return vextq_f64(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x8
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?8}}
 }
 
 poly8x8_t test_vext_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vext_p8
+  // CHECK-LABEL: test_vext_p8
   return vext_p8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x2
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?2}}
 }
 
 poly16x4_t test_vext_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vext_p16
+  // CHECK-LABEL: test_vext_p16
   return vext_p16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x6
+  // CHECK: ext {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{(0x)?6}}
 }
 
 poly8x16_t test_vextq_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vextq_p8
+  // CHECK-LABEL: test_vextq_p8
   return vextq_p8(a, b, 2);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x2
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?2}}
 }
 
 poly16x8_t test_vextq_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vextq_p16
+  // CHECK-LABEL: test_vextq_p16
   return vextq_p16(a, b, 3);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x6
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{(0x)?6}}
 }
diff --git a/test/CodeGen/aarch64-neon-fcvt-intrinsics.c b/test/CodeGen/aarch64-neon-fcvt-intrinsics.c
index 98f1389..77022f5 100644
--- a/test/CodeGen/aarch64-neon-fcvt-intrinsics.c
+++ b/test/CodeGen/aarch64-neon-fcvt-intrinsics.c
@@ -1,133 +1,136 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 float32_t test_vcvtxd_f32_f64(float64_t a) {
-// CHECK: test_vcvtxd_f32_f64
+// CHECK-LABEL: test_vcvtxd_f32_f64
 // CHECK: fcvtxn {{s[0-9]+}}, {{d[0-9]+}}
   return (float32_t)vcvtxd_f32_f64(a);
 }
 
 int32_t test_vcvtas_s32_f32(float32_t a) {
-// CHECK: test_vcvtas_s32_f32
-// CHECK: fcvtas {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtas_s32_f32
+// CHECK: fcvtas {{[ws][0-9]+}}, {{s[0-9]+}}
   return (int32_t)vcvtas_s32_f32(a);
 }
 
 int64_t test_test_vcvtad_s64_f64(float64_t a) {
-// CHECK: test_test_vcvtad_s64_f64
-// CHECK: fcvtas {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_test_vcvtad_s64_f64
+// CHECK: fcvtas {{[dx][0-9]+}}, {{d[0-9]+}}
   return (int64_t)vcvtad_s64_f64(a);
 }
 
 uint32_t test_vcvtas_u32_f32(float32_t a) {
-// CHECK: test_vcvtas_u32_f32
-// CHECK: fcvtau {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtas_u32_f32
+// CHECK: fcvtau {{[ws][0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vcvtas_u32_f32(a);
 }
 
 uint64_t test_vcvtad_u64_f64(float64_t a) {
-// CHECK: test_vcvtad_u64_f64
-// CHECK: fcvtau {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtad_u64_f64
+// CHECK: fcvtau {{[xd][0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vcvtad_u64_f64(a);
 }
 
 int32_t test_vcvtms_s32_f32(float32_t a) {
-// CHECK: test_vcvtms_s32_f32
-// CHECK: fcvtms {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtms_s32_f32
+// CHECK: fcvtms {{[sw][0-9]+}}, {{s[0-9]+}}
   return (int32_t)vcvtms_s32_f32(a);
 }
 
 int64_t test_vcvtmd_s64_f64(float64_t a) {
-// CHECK: test_vcvtmd_s64_f64
-// CHECK: fcvtms {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtmd_s64_f64
+// CHECK: fcvtms {{[dx][0-9]+}}, {{d[0-9]+}}
   return (int64_t)vcvtmd_s64_f64(a);
 }
 
 uint32_t test_vcvtms_u32_f32(float32_t a) {
-// CHECK: test_vcvtms_u32_f32
-// CHECK: fcvtmu {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtms_u32_f32
+// CHECK: fcvtmu {{[ws][0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vcvtms_u32_f32(a);
 }
 
 uint64_t test_vcvtmd_u64_f64(float64_t a) {
-// CHECK: test_vcvtmd_u64_f64
-// CHECK: fcvtmu {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtmd_u64_f64
+// CHECK: fcvtmu {{[xd][0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vcvtmd_u64_f64(a);
 }
 
 int32_t test_vcvtns_s32_f32(float32_t a) {
-// CHECK: test_vcvtns_s32_f32
-// CHECK: fcvtns {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtns_s32_f32
+// CHECK: fcvtns {{[sw][0-9]+}}, {{s[0-9]+}}
   return (int32_t)vcvtns_s32_f32(a);
 }
 
 int64_t test_vcvtnd_s64_f64(float64_t a) {
-// CHECK: test_vcvtnd_s64_f64
-// CHECK: fcvtns {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtnd_s64_f64
+// CHECK: fcvtns {{[dx][0-9]+}}, {{d[0-9]+}}
   return (int64_t)vcvtnd_s64_f64(a);
 }
 
 uint32_t test_vcvtns_u32_f32(float32_t a) {
-// CHECK: test_vcvtns_u32_f32
-// CHECK: fcvtnu {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtns_u32_f32
+// CHECK: fcvtnu {{[sw][0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vcvtns_u32_f32(a);
 }
 
 uint64_t test_vcvtnd_u64_f64(float64_t a) {
-// CHECK: test_vcvtnd_u64_f64
-// CHECK: fcvtnu {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtnd_u64_f64
+// CHECK: fcvtnu {{[dx][0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vcvtnd_u64_f64(a);
 }
 
 int32_t test_vcvtps_s32_f32(float32_t a) {
-// CHECK: test_vcvtps_s32_f32
-// CHECK: fcvtps {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtps_s32_f32
+// CHECK: fcvtps {{[sw][0-9]+}}, {{s[0-9]+}}
   return (int32_t)vcvtps_s32_f32(a);
 }
 
 int64_t test_vcvtpd_s64_f64(float64_t a) {
-// CHECK: test_vcvtpd_s64_f64
-// CHECK: fcvtps {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtpd_s64_f64
+// CHECK: fcvtps {{[dx][0-9]+}}, {{d[0-9]+}}
   return (int64_t)vcvtpd_s64_f64(a);
 }
 
 uint32_t test_vcvtps_u32_f32(float32_t a) {
-// CHECK: test_vcvtps_u32_f32
-// CHECK: fcvtpu {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvtps_u32_f32
+// CHECK: fcvtpu {{[sw][0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vcvtps_u32_f32(a);
 }
 
 uint64_t test_vcvtpd_u64_f64(float64_t a) {
-// CHECK: test_vcvtpd_u64_f64
-// CHECK: fcvtpu {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtpd_u64_f64
+// CHECK: fcvtpu {{[dx][0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vcvtpd_u64_f64(a);
 }
 
 int32_t test_vcvts_s32_f32(float32_t a) {
-// CHECK: test_vcvts_s32_f32
-// CHECK: fcvtzs {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvts_s32_f32
+// CHECK: fcvtzs {{[sw][0-9]+}}, {{s[0-9]+}}
   return (int32_t)vcvts_s32_f32(a);
 }
 
 int64_t test_vcvtd_s64_f64(float64_t a) {
-// CHECK: test_vcvtd_s64_f64
-// CHECK: fcvtzs {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtd_s64_f64
+// CHECK: fcvtzs {{[dx][0-9]+}}, {{d[0-9]+}}
   return (int64_t)vcvtd_s64_f64(a);
 }
 
 uint32_t test_vcvts_u32_f32(float32_t a) {
-// CHECK: test_vcvts_u32_f32
-// CHECK: fcvtzu {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvts_u32_f32
+// CHECK: fcvtzu {{[sw][0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vcvts_u32_f32(a);
 }
 
 uint64_t test_vcvtd_u64_f64(float64_t a) {
-// CHECK: test_vcvtd_u64_f64
-// CHECK: fcvtzu {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtd_u64_f64
+// CHECK: fcvtzu {{[dx][0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vcvtd_u64_f64(a);
 }
diff --git a/test/CodeGen/aarch64-neon-fma.c b/test/CodeGen/aarch64-neon-fma.c
new file mode 100644
index 0000000..85603c5
--- /dev/null
+++ b/test/CodeGen/aarch64-neon-fma.c
@@ -0,0 +1,204 @@
+// REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck -check-prefix=CHECK-FMA %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu -S -O3 -o - %s | FileCheck %s
+
+// Test new aarch64 intrinsics and types
+
+#include <arm_neon.h>
+
+float32x2_t test_vmla_n_f32(float32x2_t a, float32x2_t b, float32_t c) {
+  // CHECK-LABEL: test_vmla_n_f32
+  return vmla_n_f32(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK-FMA: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+float32x4_t test_vmlaq_n_f32(float32x4_t a, float32x4_t b, float32_t c) {
+  // CHECK-LABEL: test_vmlaq_n_f32
+  return vmlaq_n_f32(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK-FMA: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+float64x2_t test_vmlaq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
+  // CHECK-LABEL: test_vmlaq_n_f64
+  return vmlaq_n_f64(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
+  // CHECK: fadd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+  // CHECK-FMA: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
+  // CHECK-FMA: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+}
+
+float32x4_t test_vmlsq_n_f32(float32x4_t a, float32x4_t b, float32_t c) {
+  // CHECK-LABEL: test_vmlsq_n_f32
+  return vmlsq_n_f32(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: dup {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK-FMA: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+float32x2_t test_vmls_n_f32(float32x2_t a, float32x2_t b, float32_t c) {
+  // CHECK-LABEL: test_vmls_n_f32
+  return vmls_n_f32(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: dup {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK-FMA: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+float64x2_t test_vmlsq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
+  // CHECK-LABEL: test_vmlsq_n_f64
+  return vmlsq_n_f64(a, b, c);
+  // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
+  // CHECK: fsub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+  // CHECK-FMA: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
+  // CHECK-FMA: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+}
+
+float32x2_t test_vmla_lane_f32_0(float32x2_t a, float32x2_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmla_lane_f32_0
+  return vmla_lane_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+float32x4_t test_vmlaq_lane_f32_0(float32x4_t a, float32x4_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_f32_0
+  return vmlaq_lane_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+float32x2_t test_vmla_laneq_f32_0(float32x2_t a, float32x2_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmla_laneq_f32_0
+  return vmla_laneq_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+float32x4_t test_vmlaq_laneq_f32_0(float32x4_t a, float32x4_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_f32_0
+  return vmlaq_laneq_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+float32x2_t test_vmls_lane_f32_0(float32x2_t a, float32x2_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmls_lane_f32_0
+  return vmls_lane_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+float32x4_t test_vmlsq_lane_f32_0(float32x4_t a, float32x4_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_f32_0
+  return vmlsq_lane_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+float32x2_t test_vmls_laneq_f32_0(float32x2_t a, float32x2_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmls_laneq_f32_0
+  return vmls_laneq_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[0]
+}
+
+float32x4_t test_vmlsq_laneq_f32_0(float32x4_t a, float32x4_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_f32_0
+  return vmlsq_laneq_f32(a, b, v, 0);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+  // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[0]
+}
+
+float32x2_t test_vmla_lane_f32(float32x2_t a, float32x2_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmla_lane_f32
+  return vmla_lane_f32(a, b, v, 1);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+  // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+}
+
+float32x4_t test_vmlaq_lane_f32(float32x4_t a, float32x4_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmlaq_lane_f32
+  return vmlaq_lane_f32(a, b, v, 1);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+  // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+}
+
+float32x2_t test_vmla_laneq_f32(float32x2_t a, float32x2_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmla_laneq_f32
+  return vmla_laneq_f32(a, b, v, 3);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+  // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+float32x4_t test_vmlaq_laneq_f32(float32x4_t a, float32x4_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmlaq_laneq_f32
+  return vmlaq_laneq_f32(a, b, v, 3);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+  // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+float32x2_t test_vmls_lane_f32(float32x2_t a, float32x2_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmls_lane_f32
+  return vmls_lane_f32(a, b, v, 1);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+  // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[1]
+}
+
+float32x4_t test_vmlsq_lane_f32(float32x4_t a, float32x4_t b, float32x2_t v) {
+  // CHECK-LABEL: test_vmlsq_lane_f32
+  return vmlsq_lane_f32(a, b, v, 1);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+  // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[1]
+}
+float32x2_t test_vmls_laneq_f32(float32x2_t a, float32x2_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmls_laneq_f32
+  return vmls_laneq_f32(a, b, v, 3);
+  // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+  // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  // CHECK-FMA: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.s[3]
+}
+
+float32x4_t test_vmlsq_laneq_f32(float32x4_t a, float32x4_t b, float32x4_t v) {
+  // CHECK-LABEL: test_vmlsq_laneq_f32
+  return vmlsq_laneq_f32(a, b, v, 3);
+  // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+  // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  // CHECK-FMA: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.s[3]
+}
+
+float64x2_t test_vfmaq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
+  // CHECK-LABEL: test_vfmaq_n_f64:
+  return vfmaq_n_f64(a, b, c);
+  // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+\.2d|v[0-9]+\.d\[0\]}}
+}
+
+float64x2_t test_vfmsq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
+  // CHECK-LABEL: test_vfmsq_n_f64:
+  return vfmsq_n_f64(a, b, c);
+  // CHECK: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+\.2d|v[0-9]+\.d\[0\]}}
+}
diff --git a/test/CodeGen/aarch64-neon-intrinsics.c b/test/CodeGen/aarch64-neon-intrinsics.c
index 6e9b7f1..4e8c405 100644
--- a/test/CodeGen/aarch64-neon-intrinsics.c
+++ b/test/CodeGen/aarch64-neon-intrinsics.c
@@ -1,109 +1,112 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
-// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-AARCH64
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-ARM64
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int8x8_t test_vadd_s8(int8x8_t v1, int8x8_t v2) {
-   // CHECK: test_vadd_s8
+   // CHECK-LABEL: test_vadd_s8
   return vadd_s8(v1, v2);
   // CHECK: add {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vadd_s16(int16x4_t v1, int16x4_t v2) {
-   // CHECK: test_vadd_s16
+   // CHECK-LABEL: test_vadd_s16
   return vadd_s16(v1, v2);
   // CHECK: add {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vadd_s32(int32x2_t v1, int32x2_t v2) {
-   // CHECK: test_vadd_s32
+   // CHECK-LABEL: test_vadd_s32
   return vadd_s32(v1, v2);
   // CHECK: add {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vadd_s64(int64x1_t v1, int64x1_t v2) {
-  // CHECK: test_vadd_s64
+  // CHECK-LABEL: test_vadd_s64
   return vadd_s64(v1, v2);
   // CHECK: add {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 float32x2_t test_vadd_f32(float32x2_t v1, float32x2_t v2) {
-   // CHECK: test_vadd_f32
+   // CHECK-LABEL: test_vadd_f32
   return vadd_f32(v1, v2);
   // CHECK: fadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vadd_u8(uint8x8_t v1, uint8x8_t v2) {
-   // CHECK: test_vadd_u8
+   // CHECK-LABEL: test_vadd_u8
   return vadd_u8(v1, v2);
   // CHECK: add {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vadd_u16(uint16x4_t v1, uint16x4_t v2) {
-   // CHECK: test_vadd_u16
+   // CHECK-LABEL: test_vadd_u16
   return vadd_u16(v1, v2);
   // CHECK: add {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vadd_u32(uint32x2_t v1, uint32x2_t v2) {
-   // CHECK: test_vadd_u32
+   // CHECK-LABEL: test_vadd_u32
   return vadd_u32(v1, v2);
   // CHECK: add {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vadd_u64(uint64x1_t v1, uint64x1_t v2) {
-   // CHECK: test_vadd_u64
+   // CHECK-LABEL: test_vadd_u64
   return vadd_u64(v1, v2);
   // CHECK: add {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vaddq_s8(int8x16_t v1, int8x16_t v2) {
-   // CHECK: test_vaddq_s8
+   // CHECK-LABEL: test_vaddq_s8
   return vaddq_s8(v1, v2);
   // CHECK: add {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vaddq_s16(int16x8_t v1, int16x8_t v2) {
-   // CHECK: test_vaddq_s16
+   // CHECK-LABEL: test_vaddq_s16
   return vaddq_s16(v1, v2);
   // CHECK: add {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vaddq_s32(int32x4_t v1,int32x4_t  v2) {
-   // CHECK: test_vaddq_s32
+   // CHECK-LABEL: test_vaddq_s32
   return vaddq_s32(v1, v2);
   // CHECK: add {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vaddq_s64(int64x2_t v1, int64x2_t v2) {
-   // CHECK: test_vaddq_s64
+   // CHECK-LABEL: test_vaddq_s64
   return vaddq_s64(v1, v2);
   // CHECK: add {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x4_t test_vaddq_f32(float32x4_t v1, float32x4_t v2) {
-   // CHECK: test_vaddq_f32
+   // CHECK-LABEL: test_vaddq_f32
   return vaddq_f32(v1, v2);
   // CHECK: fadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vaddq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vaddq_f64
+  // CHECK-LABEL: test_vaddq_f64
   return vaddq_f64(v1, v2);
   // CHECK: fadd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vaddq_u8(uint8x16_t v1, uint8x16_t v2) {
-   // CHECK: test_vaddq_u8
+   // CHECK-LABEL: test_vaddq_u8
   return vaddq_u8(v1, v2);
   // CHECK: add {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vaddq_u16(uint16x8_t v1, uint16x8_t v2) {
-   // CHECK: test_vaddq_u16
+   // CHECK-LABEL: test_vaddq_u16
   return vaddq_u16(v1, v2);
   // CHECK: add {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
@@ -115,107 +118,107 @@
 }
 
 uint64x2_t test_vaddq_u64(uint64x2_t v1, uint64x2_t v2) {
-   // CHECK: test_vaddq_u64
+   // CHECK-LABEL: test_vaddq_u64
   return vaddq_u64(v1, v2);
   // CHECK: add {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vsub_s8(int8x8_t v1, int8x8_t v2) {
-   // CHECK: test_vsub_s8
+   // CHECK-LABEL: test_vsub_s8
   return vsub_s8(v1, v2);
   // CHECK: sub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int16x4_t test_vsub_s16(int16x4_t v1, int16x4_t v2) {
-   // CHECK: test_vsub_s16
+   // CHECK-LABEL: test_vsub_s16
   return vsub_s16(v1, v2);
   // CHECK: sub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int32x2_t test_vsub_s32(int32x2_t v1, int32x2_t v2) {
-   // CHECK: test_vsub_s32
+   // CHECK-LABEL: test_vsub_s32
   return vsub_s32(v1, v2);
   // CHECK: sub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vsub_s64(int64x1_t v1, int64x1_t v2) {
-   // CHECK: test_vsub_s64
+   // CHECK-LABEL: test_vsub_s64
   return vsub_s64(v1, v2);
   // CHECK: sub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 float32x2_t test_vsub_f32(float32x2_t v1, float32x2_t v2) {
-   // CHECK: test_vsub_f32
+   // CHECK-LABEL: test_vsub_f32
   return vsub_f32(v1, v2);
   // CHECK: fsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vsub_u8(uint8x8_t v1, uint8x8_t v2) {
-   // CHECK: test_vsub_u8
+   // CHECK-LABEL: test_vsub_u8
   return vsub_u8(v1, v2);
   // CHECK: sub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vsub_u16(uint16x4_t v1, uint16x4_t v2) {
-   // CHECK: test_vsub_u16
+   // CHECK-LABEL: test_vsub_u16
   return vsub_u16(v1, v2);
   // CHECK: sub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vsub_u32(uint32x2_t v1, uint32x2_t v2) {
-   // CHECK: test_vsub_u32
+   // CHECK-LABEL: test_vsub_u32
   return vsub_u32(v1, v2);
   // CHECK: sub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vsub_u64(uint64x1_t v1, uint64x1_t v2) {
-   // CHECK: test_vsub_u64
+   // CHECK-LABEL: test_vsub_u64
   return vsub_u64(v1, v2);
   // CHECK: sub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vsubq_s8(int8x16_t v1, int8x16_t v2) {
-   // CHECK: test_vsubq_s8
+   // CHECK-LABEL: test_vsubq_s8
   return vsubq_s8(v1, v2);
   // CHECK: sub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vsubq_s16(int16x8_t v1, int16x8_t v2) {
-   // CHECK: test_vsubq_s16
+   // CHECK-LABEL: test_vsubq_s16
   return vsubq_s16(v1, v2);
   // CHECK: sub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vsubq_s32(int32x4_t v1,int32x4_t  v2) {
-   // CHECK: test_vsubq_s32
+   // CHECK-LABEL: test_vsubq_s32
   return vsubq_s32(v1, v2);
   // CHECK: sub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vsubq_s64(int64x2_t v1, int64x2_t v2) {
-   // CHECK: test_vsubq_s64
+   // CHECK-LABEL: test_vsubq_s64
   return vsubq_s64(v1, v2);
   // CHECK: sub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x4_t test_vsubq_f32(float32x4_t v1, float32x4_t v2) {
-   // CHECK: test_vsubq_f32
+   // CHECK-LABEL: test_vsubq_f32
   return vsubq_f32(v1, v2);
   // CHECK: fsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vsubq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vsubq_f64
+  // CHECK-LABEL: test_vsubq_f64
   return vsubq_f64(v1, v2);
   // CHECK: fsub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vsubq_u8(uint8x16_t v1, uint8x16_t v2) {
-   // CHECK: test_vsubq_u8
+   // CHECK-LABEL: test_vsubq_u8
   return vsubq_u8(v1, v2);
   // CHECK: sub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vsubq_u16(uint16x8_t v1, uint16x8_t v2) {
-   // CHECK: test_vsubq_u16
+   // CHECK-LABEL: test_vsubq_u16
   return vsubq_u16(v1, v2);
   // CHECK: sub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
@@ -227,98 +230,98 @@
 }
 
 uint64x2_t test_vsubq_u64(uint64x2_t v1, uint64x2_t v2) {
-   // CHECK: test_vsubq_u64
+   // CHECK-LABEL: test_vsubq_u64
   return vsubq_u64(v1, v2);
   // CHECK: sub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vmul_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vmul_s8
+  // CHECK-LABEL: test_vmul_s8
   return vmul_s8(v1, v2);
   // CHECK: mul {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vmul_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vmul_s16
+  // CHECK-LABEL: test_vmul_s16
   return vmul_s16(v1, v2);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vmul_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vmul_s32
+  // CHECK-LABEL: test_vmul_s32
   return vmul_s32(v1, v2);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vmul_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vmul_f32
+  // CHECK-LABEL: test_vmul_f32
   return vmul_f32(v1, v2);
   // CHECK: fmul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 
 uint8x8_t test_vmul_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vmul_u8
+  // CHECK-LABEL: test_vmul_u8
   return vmul_u8(v1, v2);
   // CHECK: mul {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vmul_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vmul_u16
+  // CHECK-LABEL: test_vmul_u16
   return vmul_u16(v1, v2);
   // CHECK: mul {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vmul_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vmul_u32
+  // CHECK-LABEL: test_vmul_u32
   return vmul_u32(v1, v2);
   // CHECK: mul {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vmulq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vmulq_s8
+  // CHECK-LABEL: test_vmulq_s8
   return vmulq_s8(v1, v2);
   // CHECK: mul {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vmulq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vmulq_s16
+  // CHECK-LABEL: test_vmulq_s16
   return vmulq_s16(v1, v2);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vmulq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vmulq_s32
+  // CHECK-LABEL: test_vmulq_s32
   return vmulq_s32(v1, v2);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
     
 uint8x16_t test_vmulq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vmulq_u8
+  // CHECK-LABEL: test_vmulq_u8
   return vmulq_u8(v1, v2);
   // CHECK: mul {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vmulq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vmulq_u16
+  // CHECK-LABEL: test_vmulq_u16
   return vmulq_u16(v1, v2);
   // CHECK: mul {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vmulq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vmulq_u32
+  // CHECK-LABEL: test_vmulq_u32
   return vmulq_u32(v1, v2);
   // CHECK: mul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vmulq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vmulq_f32
+  // CHECK-LABEL: test_vmulq_f32
   return vmulq_f32(v1, v2);
   // CHECK: fmul {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmulq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vmulq_f64
+  // CHECK-LABEL: test_vmulq_f64
   return vmulq_f64(v1, v2);
   // CHECK: fmul {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
@@ -337,1037 +340,1048 @@
 
 
 int8x8_t test_vmla_s8(int8x8_t v1, int8x8_t v2, int8x8_t v3) {
-  // CHECK: test_vmla_s8
+  // CHECK-LABEL: test_vmla_s8
   return vmla_s8(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vmla_s16(int16x4_t v1, int16x4_t v2, int16x4_t v3) {
-  // CHECK: test_vmla_s16
+  // CHECK-LABEL: test_vmla_s16
   return vmla_s16(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vmla_s32(int32x2_t v1, int32x2_t v2, int32x2_t v3) {
-  // CHECK: test_vmla_s32
+  // CHECK-LABEL: test_vmla_s32
   return vmla_s32(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vmla_f32(float32x2_t v1, float32x2_t v2, float32x2_t v3) {
-  // CHECK: test_vmla_f32
+  // CHECK-LABEL: test_vmla_f32
   return vmla_f32(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vmla_u8(uint8x8_t v1, uint8x8_t v2, uint8x8_t v3) {
-  // CHECK: test_vmla_u8
+  // CHECK-LABEL: test_vmla_u8
   return vmla_u8(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vmla_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) {
-  // CHECK: test_vmla_u16
+  // CHECK-LABEL: test_vmla_u16
   return vmla_u16(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vmla_u32(uint32x2_t v1, uint32x2_t v2, uint32x2_t v3) {
-  // CHECK: test_vmla_u32
+  // CHECK-LABEL: test_vmla_u32
   return vmla_u32(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vmlaq_s8(int8x16_t v1, int8x16_t v2, int8x16_t v3) {
-  // CHECK: test_vmlaq_s8
+  // CHECK-LABEL: test_vmlaq_s8
   return vmlaq_s8(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vmlaq_s16(int16x8_t v1, int16x8_t v2, int16x8_t v3) {
-  // CHECK: test_vmlaq_s16
+  // CHECK-LABEL: test_vmlaq_s16
   return vmlaq_s16(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vmlaq_s32(int32x4_t v1, int32x4_t v2, int32x4_t v3) {
-  // CHECK: test_vmlaq_s32
+  // CHECK-LABEL: test_vmlaq_s32
   return vmlaq_s32(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 } 
 
 float32x4_t test_vmlaq_f32(float32x4_t v1, float32x4_t v2, float32x4_t v3) {
-  // CHECK: test_vmlaq_f32
+  // CHECK-LABEL: test_vmlaq_f32
   return vmlaq_f32(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vmlaq_u8(uint8x16_t v1, uint8x16_t v2, uint8x16_t v3) {
-   // CHECK: test_vmlaq_u8
+   // CHECK-LABEL: test_vmlaq_u8
   return vmlaq_u8(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vmlaq_u16(uint16x8_t v1, uint16x8_t v2, uint16x8_t v3) {
-  // CHECK: test_vmlaq_u16
+  // CHECK-LABEL: test_vmlaq_u16
   return vmlaq_u16(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vmlaq_u32(uint32x4_t v1, uint32x4_t v2, uint32x4_t v3) {
-  // CHECK: test_vmlaq_u32
+  // CHECK-LABEL: test_vmlaq_u32
   return vmlaq_u32(v1, v2, v3);
   // CHECK: mla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmlaq_f64(float64x2_t v1, float64x2_t v2, float64x2_t v3) {
-  // CHECK: test_vmlaq_f64
+  // CHECK-LABEL: test_vmlaq_f64
   return vmlaq_f64(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vmls_s8(int8x8_t v1, int8x8_t v2, int8x8_t v3) {
-  // CHECK: test_vmls_s8
+  // CHECK-LABEL: test_vmls_s8
   return vmls_s8(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vmls_s16(int16x4_t v1, int16x4_t v2, int16x4_t v3) {
-  // CHECK: test_vmls_s16
+  // CHECK-LABEL: test_vmls_s16
   return vmls_s16(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vmls_s32(int32x2_t v1, int32x2_t v2, int32x2_t v3) {
-  // CHECK: test_vmls_s32
+  // CHECK-LABEL: test_vmls_s32
   return vmls_s32(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vmls_f32(float32x2_t v1, float32x2_t v2, float32x2_t v3) {
-  // CHECK: test_vmls_f32
+  // CHECK-LABEL: test_vmls_f32
   return vmls_f32(v1, v2, v3);
   // CHECK: fmls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vmls_u8(uint8x8_t v1, uint8x8_t v2, uint8x8_t v3) {
-  // CHECK: test_vmls_u8
+  // CHECK-LABEL: test_vmls_u8
   return vmls_u8(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vmls_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) {
-  // CHECK: test_vmls_u16
+  // CHECK-LABEL: test_vmls_u16
   return vmls_u16(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vmls_u32(uint32x2_t v1, uint32x2_t v2, uint32x2_t v3) {
-  // CHECK: test_vmls_u32
+  // CHECK-LABEL: test_vmls_u32
   return vmls_u32(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 int8x16_t test_vmlsq_s8(int8x16_t v1, int8x16_t v2, int8x16_t v3) {
-  // CHECK: test_vmlsq_s8
+  // CHECK-LABEL: test_vmlsq_s8
   return vmlsq_s8(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vmlsq_s16(int16x8_t v1, int16x8_t v2, int16x8_t v3) {
-  // CHECK: test_vmlsq_s16
+  // CHECK-LABEL: test_vmlsq_s16
   return vmlsq_s16(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vmlsq_s32(int32x4_t v1, int32x4_t v2, int32x4_t v3) {
-  // CHECK: test_vmlsq_s32
+  // CHECK-LABEL: test_vmlsq_s32
   return vmlsq_s32(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vmlsq_f32(float32x4_t v1, float32x4_t v2, float32x4_t v3) {
-  // CHECK: test_vmlsq_f32
+  // CHECK-LABEL: test_vmlsq_f32
   return vmlsq_f32(v1, v2, v3);
   // CHECK: fmls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint8x16_t test_vmlsq_u8(uint8x16_t v1, uint8x16_t v2, uint8x16_t v3) {
-  // CHECK: test_vmlsq_u8
+  // CHECK-LABEL: test_vmlsq_u8
   return vmlsq_u8(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vmlsq_u16(uint16x8_t v1, uint16x8_t v2, uint16x8_t v3) {
-  // CHECK: test_vmlsq_u16
+  // CHECK-LABEL: test_vmlsq_u16
   return vmlsq_u16(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vmlsq_u32(uint32x4_t v1, uint32x4_t v2, uint32x4_t v3) {
-  // CHECK: test_vmlsq_u32
+  // CHECK-LABEL: test_vmlsq_u32
   return vmlsq_u32(v1, v2, v3);
   // CHECK: mls {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmlsq_f64(float64x2_t v1, float64x2_t v2, float64x2_t v3) {
-  // CHECK: test_vmlsq_f64
+  // CHECK-LABEL: test_vmlsq_f64
   return vmlsq_f64(v1, v2, v3);
   // CHECK: fmls {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 float32x2_t test_vfma_f32(float32x2_t v1, float32x2_t v2, float32x2_t v3) {
-  // CHECK: test_vfma_f32
+  // CHECK-LABEL: test_vfma_f32
   return vfma_f32(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vfmaq_f32(float32x4_t v1, float32x4_t v2, float32x4_t v3) {
-  // CHECK: test_vfmaq_f32
+  // CHECK-LABEL: test_vfmaq_f32
   return vfmaq_f32(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vfmaq_f64(float64x2_t v1, float64x2_t v2, float64x2_t v3) {
-  // CHECK: test_vfmaq_f64
+  // CHECK-LABEL: test_vfmaq_f64
   return vfmaq_f64(v1, v2, v3);
   // CHECK: fmla {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 float32x2_t test_vfms_f32(float32x2_t v1, float32x2_t v2, float32x2_t v3) {
-  // CHECK: test_vfms_f32
+  // CHECK-LABEL: test_vfms_f32
   return vfms_f32(v1, v2, v3);
-  // CHECK: fmls v0.2s, v1.2s, v2.2s
+  // CHECK: fmls v0.2s, {{v1.2s, v2.2s|v2.2s, v1.2s}}
 }
 
 float32x4_t test_vfmsq_f32(float32x4_t v1, float32x4_t v2, float32x4_t v3) {
-  // CHECK: test_vfmsq_f32
+  // CHECK-LABEL: test_vfmsq_f32
   return vfmsq_f32(v1, v2, v3);
-  // CHECK: fmls v0.4s, v1.4s, v2.4s
+  // CHECK: fmls v0.4s, {{v1.4s, v2.4s|v2.4s, v1.4s}}
 }
 
 float64x2_t test_vfmsq_f64(float64x2_t v1, float64x2_t v2, float64x2_t v3) {
   // CHECK: vfmsq_f64
   return vfmsq_f64(v1, v2, v3);
-  // CHECK: fmls v0.2d, v1.2d, v2.2d
+  // CHECK: fmls v0.2d, {{v1.2d, v2.2d|v2.2d, v1.2d}}
 }
 
 float64x2_t test_vdivq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vdivq_f64
+  // CHECK-LABEL: test_vdivq_f64
   return vdivq_f64(v1, v2);
   // CHECK: fdiv {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x4_t test_vdivq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vdivq_f32
+  // CHECK-LABEL: test_vdivq_f32
   return vdivq_f32(v1, v2);
   // CHECK: fdiv {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x2_t test_vdiv_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vdiv_f32
+  // CHECK-LABEL: test_vdiv_f32
   return vdiv_f32(v1, v2);
   // CHECK: fdiv {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x8_t test_vaba_s8(int8x8_t v1, int8x8_t v2, int8x8_t v3) {
-  // CHECK: test_vaba_s8
+  // CHECK-LABEL: test_vaba_s8
   return vaba_s8(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vaba_s16(int16x4_t v1, int16x4_t v2, int16x4_t v3) {
-  // CHECK: test_vaba_s16
+  // CHECK-LABEL: test_vaba_s16
   return vaba_s16(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vaba_s32(int32x2_t v1, int32x2_t v2, int32x2_t v3) {
-  // CHECK: test_vaba_s32
+  // CHECK-LABEL: test_vaba_s32
   return vaba_s32(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vaba_u8(uint8x8_t v1, uint8x8_t v2, uint8x8_t v3) {
-  // CHECK: test_vaba_u8
+  // CHECK-LABEL: test_vaba_u8
   return vaba_u8(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vaba_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) {
-  // CHECK: test_vaba_u16
+  // CHECK-LABEL: test_vaba_u16
   return vaba_u16(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vaba_u32(uint32x2_t v1, uint32x2_t v2, uint32x2_t v3) {
-  // CHECK: test_vaba_u32
+  // CHECK-LABEL: test_vaba_u32
   return vaba_u32(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vabaq_s8(int8x16_t v1, int8x16_t v2, int8x16_t v3) {
-  // CHECK: test_vabaq_s8
+  // CHECK-LABEL: test_vabaq_s8
   return vabaq_s8(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vabaq_s16(int16x8_t v1, int16x8_t v2, int16x8_t v3) {
-  // CHECK: test_vabaq_s16
+  // CHECK-LABEL: test_vabaq_s16
   return vabaq_s16(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vabaq_s32(int32x4_t v1, int32x4_t v2, int32x4_t v3) {
-  // CHECK: test_vabaq_s32
+  // CHECK-LABEL: test_vabaq_s32
   return vabaq_s32(v1, v2, v3);
   // CHECK: saba {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vabaq_u8(uint8x16_t v1, uint8x16_t v2, uint8x16_t v3) {
-  // CHECK: test_vabaq_u8
+  // CHECK-LABEL: test_vabaq_u8
   return vabaq_u8(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vabaq_u16(uint16x8_t v1, uint16x8_t v2, uint16x8_t v3) {
-  // CHECK: test_vabaq_u16
+  // CHECK-LABEL: test_vabaq_u16
   return vabaq_u16(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vabaq_u32(uint32x4_t v1, uint32x4_t v2, uint32x4_t v3) {
-  // CHECK: test_vabaq_u32
+  // CHECK-LABEL: test_vabaq_u32
   return vabaq_u32(v1, v2, v3);
   // CHECK: uaba {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int8x8_t test_vabd_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vabd_s8
+  // CHECK-LABEL: test_vabd_s8
   return vabd_s8(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vabd_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vabd_s16
+  // CHECK-LABEL: test_vabd_s16
   return vabd_s16(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vabd_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vabd_s32
+  // CHECK-LABEL: test_vabd_s32
   return vabd_s32(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vabd_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vabd_u8
+  // CHECK-LABEL: test_vabd_u8
   return vabd_u8(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vabd_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vabd_u16
+  // CHECK-LABEL: test_vabd_u16
   return vabd_u16(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vabd_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vabd_u32
+  // CHECK-LABEL: test_vabd_u32
   return vabd_u32(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vabd_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vabd_f32
+  // CHECK-LABEL: test_vabd_f32
   return vabd_f32(v1, v2);
   // CHECK: fabd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vabdq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vabdq_s8
+  // CHECK-LABEL: test_vabdq_s8
   return vabdq_s8(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vabdq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vabdq_s16
+  // CHECK-LABEL: test_vabdq_s16
   return vabdq_s16(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vabdq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vabdq_s32
+  // CHECK-LABEL: test_vabdq_s32
   return vabdq_s32(v1, v2);
   // CHECK: sabd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vabdq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vabdq_u8
+  // CHECK-LABEL: test_vabdq_u8
   return vabdq_u8(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vabdq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vabdq_u16
+  // CHECK-LABEL: test_vabdq_u16
   return vabdq_u16(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vabdq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vabdq_u32
+  // CHECK-LABEL: test_vabdq_u32
   return vabdq_u32(v1, v2);
   // CHECK: uabd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vabdq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vabdq_f32
+  // CHECK-LABEL: test_vabdq_f32
   return vabdq_f32(v1, v2);
   // CHECK: fabd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vabdq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vabdq_f64
+  // CHECK-LABEL: test_vabdq_f64
   return vabdq_f64(v1, v2);
   // CHECK: fabd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vbsl_s8(uint8x8_t v1, int8x8_t v2, int8x8_t v3) {
-  // CHECK: test_vbsl_s8
+  // CHECK-LABEL: test_vbsl_s8
   return vbsl_s8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vbsl_s16(uint16x4_t v1, int16x4_t v2, int16x4_t v3) {
-  // CHECK: test_vbsl_s16
+  // CHECK-LABEL: test_vbsl_s16
   return vbsl_s16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int32x2_t test_vbsl_s32(uint32x2_t v1, int32x2_t v2, int32x2_t v3) {
-  // CHECK: test_vbsl_s32
+  // CHECK-LABEL: test_vbsl_s32
   return vbsl_s32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint64x1_t test_vbsl_s64(uint64x1_t v1, uint64x1_t v2, uint64x1_t v3) {
-  // CHECK: test_vbsl_s64
+  // CHECK-LABEL: test_vbsl_s64
   return vbsl_s64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vbsl_u8(uint8x8_t v1, uint8x8_t v2, uint8x8_t v3) {
-  // CHECK: test_vbsl_u8
+  // CHECK-LABEL: test_vbsl_u8
   return vbsl_u8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vbsl_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) {
-  // CHECK: test_vbsl_u16
+  // CHECK-LABEL: test_vbsl_u16
   return vbsl_u16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint32x2_t test_vbsl_u32(uint32x2_t v1, uint32x2_t v2, uint32x2_t v3) {
-  // CHECK: test_vbsl_u32
+  // CHECK-LABEL: test_vbsl_u32
   return vbsl_u32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint64x1_t test_vbsl_u64(uint64x1_t v1, uint64x1_t v2, uint64x1_t v3) {
-  // CHECK: test_vbsl_u64
+  // CHECK-LABEL: test_vbsl_u64
   return vbsl_u64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 float32x2_t test_vbsl_f32(float32x2_t v1, float32x2_t v2, float32x2_t v3) {
-  // CHECK: test_vbsl_f32
+  // CHECK-LABEL: test_vbsl_f32
   return vbsl_f32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 float64x1_t test_vbsl_f64(uint64x1_t v1, float64x1_t v2, float64x1_t v3) {
-  // CHECK: test_vbsl_f64
+  // CHECK-LABEL: test_vbsl_f64
   return vbsl_f64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vbsl_p8(uint8x8_t v1, poly8x8_t v2, poly8x8_t v3) {
-  // CHECK: test_vbsl_p8
+  // CHECK-LABEL: test_vbsl_p8
   return vbsl_p8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly16x4_t test_vbsl_p16(uint16x4_t v1, poly16x4_t v2, poly16x4_t v3) {
-  // CHECK: test_vbsl_p16
+  // CHECK-LABEL: test_vbsl_p16
   return vbsl_p16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vbslq_s8(uint8x16_t v1, int8x16_t v2, int8x16_t v3) {
-  // CHECK: test_vbslq_s8
+  // CHECK-LABEL: test_vbslq_s8
   return vbslq_s8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vbslq_s16(uint16x8_t v1, int16x8_t v2, int16x8_t v3) {
-  // CHECK: test_vbslq_s16
+  // CHECK-LABEL: test_vbslq_s16
   return vbslq_s16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vbslq_s32(uint32x4_t v1, int32x4_t v2, int32x4_t v3) {
-  // CHECK: test_vbslq_s32
+  // CHECK-LABEL: test_vbslq_s32
   return vbslq_s32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int64x2_t test_vbslq_s64(uint64x2_t v1, int64x2_t v2, int64x2_t v3) {
-  // CHECK: test_vbslq_s64
+  // CHECK-LABEL: test_vbslq_s64
   return vbslq_s64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vbslq_u8(uint8x16_t v1, uint8x16_t v2, uint8x16_t v3) {
-  // CHECK: test_vbslq_u8
+  // CHECK-LABEL: test_vbslq_u8
   return vbslq_u8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vbslq_u16(uint16x8_t v1, uint16x8_t v2, uint16x8_t v3) {
-  // CHECK: test_vbslq_u16
+  // CHECK-LABEL: test_vbslq_u16
   return vbslq_u16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vbslq_u32(uint32x4_t v1, int32x4_t v2, int32x4_t v3) {
-  // CHECK: test_vbslq_u32
+  // CHECK-LABEL: test_vbslq_u32
   return vbslq_s32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint64x2_t test_vbslq_u64(uint64x2_t v1, uint64x2_t v2, uint64x2_t v3) {
-  // CHECK: test_vbslq_u64
+  // CHECK-LABEL: test_vbslq_u64
   return vbslq_u64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 float32x4_t test_vbslq_f32(uint32x4_t v1, float32x4_t v2, float32x4_t v3) {
-  // CHECK: test_vbslq_f32
+  // CHECK-LABEL: test_vbslq_f32
   return vbslq_f32(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vbslq_p8(uint8x16_t v1, poly8x16_t v2, poly8x16_t v3) {
-  // CHECK: test_vbslq_p8
+  // CHECK-LABEL: test_vbslq_p8
   return vbslq_p8(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x8_t test_vbslq_p16(uint16x8_t v1, poly16x8_t v2, poly16x8_t v3) {
-  // CHECK: test_vbslq_p16
+  // CHECK-LABEL: test_vbslq_p16
   return vbslq_p16(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 float64x2_t test_vbslq_f64(uint64x2_t v1, float64x2_t v2, float64x2_t v3) {
-  // CHECK: test_vbslq_f64
+  // CHECK-LABEL: test_vbslq_f64
   return vbslq_f64(v1, v2, v3);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 float32x2_t test_vrecps_f32(float32x2_t v1, float32x2_t v2) {
-   // CHECK: test_vrecps_f32
+   // CHECK-LABEL: test_vrecps_f32
    return vrecps_f32(v1, v2);
    // CHECK: frecps {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vrecpsq_f32(float32x4_t v1, float32x4_t v2) {
-   // CHECK: test_vrecpsq_f32
+   // CHECK-LABEL: test_vrecpsq_f32
    return vrecpsq_f32(v1, v2);
    // CHECK: frecps {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vrecpsq_f64(float64x2_t v1, float64x2_t v2) {
-   // CHECK: test_vrecpsq_f64
+   // CHECK-LABEL: test_vrecpsq_f64
   return vrecpsq_f64(v1, v2);
   // CHECK: frecps {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x2_t test_vrsqrts_f32(float32x2_t v1, float32x2_t v2) {
-   // CHECK: test_vrsqrts_f32
+   // CHECK-LABEL: test_vrsqrts_f32
   return vrsqrts_f32(v1, v2);
   // CHECK: frsqrts {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vrsqrtsq_f32(float32x4_t v1, float32x4_t v2) {
-   // CHECK: test_vrsqrtsq_f32
+   // CHECK-LABEL: test_vrsqrtsq_f32
   return vrsqrtsq_f32(v1, v2);
   // CHECK: frsqrts {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vrsqrtsq_f64(float64x2_t v1, float64x2_t v2) {
-   // CHECK: test_vrsqrtsq_f64
+   // CHECK-LABEL: test_vrsqrtsq_f64
   return vrsqrtsq_f64(v1, v2);
   // CHECK: frsqrts {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint32x2_t test_vcage_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcage_f32
+  // CHECK-LABEL: test_vcage_f32
   return vcage_f32(v1, v2);
   // CHECK: facge {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcage_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcage_f64
+  // CHECK-LABEL: test_vcage_f64
   return vcage_f64(a, b);
   // CHECK: facge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x4_t test_vcageq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcageq_f32
+  // CHECK-LABEL: test_vcageq_f32
   return vcageq_f32(v1, v2);
   // CHECK: facge {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vcageq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcageq_f64
+  // CHECK-LABEL: test_vcageq_f64
   return vcageq_f64(v1, v2);
   // CHECK: facge {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint32x2_t test_vcagt_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcagt_f32
+  // CHECK-LABEL: test_vcagt_f32
   return vcagt_f32(v1, v2);
   // CHECK: facgt {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcagt_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcagt_f64
+  // CHECK-LABEL: test_vcagt_f64
   return vcagt_f64(a, b);
   // CHECK: facgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x4_t test_vcagtq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcagtq_f32
+  // CHECK-LABEL: test_vcagtq_f32
   return vcagtq_f32(v1, v2);
   // CHECK: facgt {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vcagtq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcagtq_f64
+  // CHECK-LABEL: test_vcagtq_f64
   return vcagtq_f64(v1, v2);
   // CHECK: facgt {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint32x2_t test_vcale_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcale_f32
+  // CHECK-LABEL: test_vcale_f32
   return vcale_f32(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facge {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vcale_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcale_f64
+  // CHECK-LABEL: test_vcale_f64
   return vcale_f64(a, b);
   // CHECK: facge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x4_t test_vcaleq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcaleq_f32
+  // CHECK-LABEL: test_vcaleq_f32
   return vcaleq_f32(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facge {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint64x2_t test_vcaleq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcaleq_f64
+  // CHECK-LABEL: test_vcaleq_f64
   return vcaleq_f64(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facge {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint32x2_t test_vcalt_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcalt_f32
+  // CHECK-LABEL: test_vcalt_f32
   return vcalt_f32(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facgt {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vcalt_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcalt_f64
+  // CHECK-LABEL: test_vcalt_f64
   return vcalt_f64(a, b);
   // CHECK: facgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x4_t test_vcaltq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcaltq_f32
+  // CHECK-LABEL: test_vcaltq_f32
   return vcaltq_f32(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facgt {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint64x2_t test_vcaltq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcaltq_f64
+  // CHECK-LABEL: test_vcaltq_f64
   return vcaltq_f64(v1, v2);
   // Using registers other than v0, v1 are possible, but would be odd.
   // CHECK: facgt {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint8x8_t test_vtst_s8(int8x8_t v1, int8x8_t v2) {
-   // CHECK: test_vtst_s8
+   // CHECK-LABEL: test_vtst_s8
   return vtst_s8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vtst_s16(int16x4_t v1, int16x4_t v2) {
-   // CHECK: test_vtst_s16
+   // CHECK-LABEL: test_vtst_s16
   return vtst_s16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vtst_s32(int32x2_t v1, int32x2_t v2) {
-   // CHECK: test_vtst_s32
+   // CHECK-LABEL: test_vtst_s32
   return vtst_s32(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vtst_u8(uint8x8_t v1, uint8x8_t v2) {
-   // CHECK: test_vtst_u8
+   // CHECK-LABEL: test_vtst_u8
   return vtst_u8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vtst_u16(uint16x4_t v1, uint16x4_t v2) {
-   // CHECK: test_vtst_u16
+   // CHECK-LABEL: test_vtst_u16
   return vtst_u16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vtst_u32(uint32x2_t v1, uint32x2_t v2) {
-   // CHECK: test_vtst_u32
+   // CHECK-LABEL: test_vtst_u32
   return vtst_u32(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x16_t test_vtstq_s8(int8x16_t v1, int8x16_t v2) {
-   // CHECK: test_vtstq_s8
+   // CHECK-LABEL: test_vtstq_s8
   return vtstq_s8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vtstq_s16(int16x8_t v1, int16x8_t v2) {
-   // CHECK: test_vtstq_s16
+   // CHECK-LABEL: test_vtstq_s16
   return vtstq_s16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vtstq_s32(int32x4_t v1, int32x4_t v2) {
-   // CHECK: test_vtstq_s32
+   // CHECK-LABEL: test_vtstq_s32
   return vtstq_s32(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vtstq_u8(uint8x16_t v1, uint8x16_t v2) {
-   // CHECK: test_vtstq_u8
+   // CHECK-LABEL: test_vtstq_u8
   return vtstq_u8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vtstq_u16(uint16x8_t v1, uint16x8_t v2) {
-   // CHECK: test_vtstq_u16
+   // CHECK-LABEL: test_vtstq_u16
   return vtstq_u16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vtstq_u32(uint32x4_t v1, uint32x4_t v2) {
-   // CHECK: test_vtstq_u32
+   // CHECK-LABEL: test_vtstq_u32
   return vtstq_u32(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vtstq_s64(int64x2_t v1, int64x2_t v2) {
-   // CHECK: test_vtstq_s64
+   // CHECK-LABEL: test_vtstq_s64
   return vtstq_s64(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vtstq_u64(uint64x2_t v1, uint64x2_t v2) {
-   // CHECK: test_vtstq_u64
+   // CHECK-LABEL: test_vtstq_u64
   return vtstq_u64(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x8_t test_vtst_p8(poly8x8_t v1, poly8x8_t v2) {
-   // CHECK: test_vtst_p8
+   // CHECK-LABEL: test_vtst_p8
   return vtst_p8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vtst_p16(poly16x4_t v1, poly16x4_t v2) {
-   // CHECK: test_vtst_p16
+   // CHECK-LABEL: test_vtst_p16
   return vtst_p16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint8x16_t test_vtstq_p8(poly8x16_t v1, poly8x16_t v2) {
-   // CHECK: test_vtstq_p8
+   // CHECK-LABEL: test_vtstq_p8
   return vtstq_p8(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vtstq_p16(poly16x8_t v1, poly16x8_t v2) {
-   // CHECK: test_vtstq_p16
+   // CHECK-LABEL: test_vtstq_p16
   return vtstq_p16(v1, v2);
   // CHECK: cmtst {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
+uint64x1_t test_vtst_s64(int64x1_t a, int64x1_t b) {
+  // CHECK-LABEL: test_vtst_s64
+  return vtst_s64(a, b);
+  // CHECK: cmtst {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+uint64x1_t test_vtst_u64(uint64x1_t a, uint64x1_t b) {
+  // CHECK-LABEL: test_vtst_u64
+  return vtst_u64(a, b);
+  // CHECK: cmtst {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
 
 uint8x8_t test_vceq_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vceq_s8
+  // CHECK-LABEL: test_vceq_s8
   return vceq_s8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vceq_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vceq_s16
+  // CHECK-LABEL: test_vceq_s16
   return vceq_s16(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vceq_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vceq_s32
+  // CHECK-LABEL: test_vceq_s32
   return vceq_s32(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vceq_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vceq_s64
+  // CHECK-LABEL: test_vceq_s64
   return vceq_s64(a, b);
   // CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x1_t test_vceq_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vceq_u64
+  // CHECK-LABEL: test_vceq_u64
   return vceq_u64(a, b);
   // CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x2_t test_vceq_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vceq_f32
+  // CHECK-LABEL: test_vceq_f32
   return vceq_f32(v1, v2);
   // CHECK: fcmeq {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vceq_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vceq_f64
+  // CHECK-LABEL: test_vceq_f64
   return vceq_f64(a, b);
   // CHECK: fcmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vceq_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vceq_u8
+  // CHECK-LABEL: test_vceq_u8
   return vceq_u8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vceq_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vceq_u16
+  // CHECK-LABEL: test_vceq_u16
   return vceq_u16(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vceq_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vceq_u32
+  // CHECK-LABEL: test_vceq_u32
   return vceq_u32(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vceq_p8(poly8x8_t v1, poly8x8_t v2) {
-  // CHECK: test_vceq_p8
+  // CHECK-LABEL: test_vceq_p8
   return vceq_p8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vceqq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vceqq_s8
+  // CHECK-LABEL: test_vceqq_s8
   return vceqq_s8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vceqq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vceqq_s16
+  // CHECK-LABEL: test_vceqq_s16
   return vceqq_s16(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vceqq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vceqq_s32
+  // CHECK-LABEL: test_vceqq_s32
   return vceqq_s32(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vceqq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vceqq_f32
+  // CHECK-LABEL: test_vceqq_f32
   return vceqq_f32(v1, v2);
   // CHECK: fcmeq {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vceqq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vceqq_u8
+  // CHECK-LABEL: test_vceqq_u8
   return vceqq_u8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vceqq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vceqq_u16
+  // CHECK-LABEL: test_vceqq_u16
   return vceqq_u16(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vceqq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vceqq_u32
+  // CHECK-LABEL: test_vceqq_u32
   return vceqq_u32(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vceqq_p8(poly8x16_t v1, poly8x16_t v2) {
-  // CHECK: test_vceqq_p8
+  // CHECK-LABEL: test_vceqq_p8
   return vceqq_p8(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 
 uint64x2_t test_vceqq_s64(int64x2_t v1, int64x2_t v2) {
-  // CHECK: test_vceqq_s64
+  // CHECK-LABEL: test_vceqq_s64
   return vceqq_s64(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vceqq_u64(uint64x2_t v1, uint64x2_t v2) {
-  // CHECK: test_vceqq_u64
+  // CHECK-LABEL: test_vceqq_u64
   return vceqq_u64(v1, v2);
   // CHECK: cmeq {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vceqq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vceqq_f64
+  // CHECK-LABEL: test_vceqq_f64
   return vceqq_f64(v1, v2);
   // CHECK: fcmeq {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 uint8x8_t test_vcge_s8(int8x8_t v1, int8x8_t v2) {
-// CHECK: test_vcge_s8
+// CHECK-LABEL: test_vcge_s8
   return vcge_s8(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vcge_s16(int16x4_t v1, int16x4_t v2) {
-// CHECK: test_vcge_s16
+// CHECK-LABEL: test_vcge_s16
   return vcge_s16(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vcge_s32(int32x2_t v1, int32x2_t v2) {
-// CHECK: test_vcge_s32
+// CHECK-LABEL: test_vcge_s32
   return vcge_s32(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcge_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vcge_s64
+  // CHECK-LABEL: test_vcge_s64
   return vcge_s64(a, b);
   // CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x1_t test_vcge_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vcge_u64
+  // CHECK-LABEL: test_vcge_u64
   return vcge_u64(a, b);
   // CHECK: cmhs {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x2_t test_vcge_f32(float32x2_t v1, float32x2_t v2) {
-// CHECK: test_vcge_f32
+// CHECK-LABEL: test_vcge_f32
   return vcge_f32(v1, v2);
 // CHECK: fcmge {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcge_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcge_f64
+  // CHECK-LABEL: test_vcge_f64
   return vcge_f64(a, b);
   // CHECK: fcmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vcge_u8(uint8x8_t v1, uint8x8_t v2) {
-// CHECK: test_vcge_u8
+// CHECK-LABEL: test_vcge_u8
   return vcge_u8(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vcge_u16(uint16x4_t v1, uint16x4_t v2) {
-// CHECK: test_vcge_u16
+// CHECK-LABEL: test_vcge_u16
   return vcge_u16(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vcge_u32(uint32x2_t v1, uint32x2_t v2) {
-// CHECK: test_vcge_u32
+// CHECK-LABEL: test_vcge_u32
   return vcge_u32(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x16_t test_vcgeq_s8(int8x16_t v1, int8x16_t v2) {
-// CHECK: test_vcgeq_s8
+// CHECK-LABEL: test_vcgeq_s8
   return vcgeq_s8(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vcgeq_s16(int16x8_t v1, int16x8_t v2) {
-// CHECK: test_vcgeq_s16
+// CHECK-LABEL: test_vcgeq_s16
   return vcgeq_s16(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vcgeq_s32(int32x4_t v1, int32x4_t v2) {
-// CHECK: test_vcgeq_s32
+// CHECK-LABEL: test_vcgeq_s32
   return vcgeq_s32(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vcgeq_f32(float32x4_t v1, float32x4_t v2) {
-// CHECK: test_vcgeq_f32
+// CHECK-LABEL: test_vcgeq_f32
   return vcgeq_f32(v1, v2);
 // CHECK: fcmge {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vcgeq_u8(uint8x16_t v1, uint8x16_t v2) {
-// CHECK: test_vcgeq_u8
+// CHECK-LABEL: test_vcgeq_u8
   return vcgeq_u8(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vcgeq_u16(uint16x8_t v1, uint16x8_t v2) {
-// CHECK: test_vcgeq_u16
+// CHECK-LABEL: test_vcgeq_u16
   return vcgeq_u16(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vcgeq_u32(uint32x4_t v1, uint32x4_t v2) {
-// CHECK: test_vcgeq_u32
+// CHECK-LABEL: test_vcgeq_u32
   return vcgeq_u32(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vcgeq_s64(int64x2_t v1, int64x2_t v2) {
-// CHECK: test_vcgeq_s64
+// CHECK-LABEL: test_vcgeq_s64
   return vcgeq_s64(v1, v2);
 // CHECK: cmge {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vcgeq_u64(uint64x2_t v1, uint64x2_t v2) {
-// CHECK: test_vcgeq_u64
+// CHECK-LABEL: test_vcgeq_u64
   return vcgeq_u64(v1, v2);
 // CHECK: cmhs {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vcgeq_f64(float64x2_t v1, float64x2_t v2) {
-// CHECK: test_vcgeq_f64
+// CHECK-LABEL: test_vcgeq_f64
   return vcgeq_f64(v1, v2);
 // CHECK: fcmge {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
@@ -1376,242 +1390,242 @@
 // LE condition predicate implemented as GE, so check reversed operands.
 // Using registers other than v0, v1 are possible, but would be odd.
 uint8x8_t test_vcle_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vcle_s8
+  // CHECK-LABEL: test_vcle_s8
   return vcle_s8(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.8b, v1.8b, v0.8b
 }
 
 uint16x4_t test_vcle_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vcle_s16
+  // CHECK-LABEL: test_vcle_s16
   return vcle_s16(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.4h, v1.4h, v0.4h
 }
 
 uint32x2_t test_vcle_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vcle_s32
+  // CHECK-LABEL: test_vcle_s32
   return vcle_s32(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vcle_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vcle_s64
+  // CHECK-LABEL: test_vcle_s64
   return vcle_s64(a, b);
   // CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x1_t test_vcle_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vcle_u64
+  // CHECK-LABEL: test_vcle_u64
   return vcle_u64(a, b);
   // CHECK: cmhs {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x2_t test_vcle_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcle_f32
+  // CHECK-LABEL: test_vcle_f32
   return vcle_f32(v1, v2);
   // CHECK: fcmge {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vcle_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcle_f64
+  // CHECK-LABEL: test_vcle_f64
   return vcle_f64(a, b);
   // CHECK: fcmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vcle_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vcle_u8
+  // CHECK-LABEL: test_vcle_u8
   return vcle_u8(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.8b, v1.8b, v0.8b
 }
 
 uint16x4_t test_vcle_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vcle_u16
+  // CHECK-LABEL: test_vcle_u16
   return vcle_u16(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.4h, v1.4h, v0.4h
 }
 
 uint32x2_t test_vcle_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vcle_u32
+  // CHECK-LABEL: test_vcle_u32
   return vcle_u32(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint8x16_t test_vcleq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vcleq_s8
+  // CHECK-LABEL: test_vcleq_s8
   return vcleq_s8(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.16b, v1.16b, v0.16b
 }
 
 uint16x8_t test_vcleq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vcleq_s16
+  // CHECK-LABEL: test_vcleq_s16
   return vcleq_s16(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.8h, v1.8h, v0.8h
 }
 
 uint32x4_t test_vcleq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vcleq_s32
+  // CHECK-LABEL: test_vcleq_s32
   return vcleq_s32(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint32x4_t test_vcleq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcleq_f32
+  // CHECK-LABEL: test_vcleq_f32
   return vcleq_f32(v1, v2);
   // CHECK: fcmge {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint8x16_t test_vcleq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vcleq_u8
+  // CHECK-LABEL: test_vcleq_u8
   return vcleq_u8(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.16b, v1.16b, v0.16b
 }
 
 uint16x8_t test_vcleq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vcleq_u16
+  // CHECK-LABEL: test_vcleq_u16
   return vcleq_u16(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.8h, v1.8h, v0.8h
 }
 
 uint32x4_t test_vcleq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vcleq_u32
+  // CHECK-LABEL: test_vcleq_u32
   return vcleq_u32(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint64x2_t test_vcleq_s64(int64x2_t v1, int64x2_t v2) {
-  // CHECK: test_vcleq_s64
+  // CHECK-LABEL: test_vcleq_s64
   return vcleq_s64(v1, v2);
   // CHECK: cmge {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint64x2_t test_vcleq_u64(uint64x2_t v1, uint64x2_t v2) {
-  // CHECK: test_vcleq_u64
+  // CHECK-LABEL: test_vcleq_u64
   return vcleq_u64(v1, v2);
   // CHECK: cmhs {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint64x2_t test_vcleq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcleq_f64
+  // CHECK-LABEL: test_vcleq_f64
   return vcleq_f64(v1, v2);
   // CHECK: fcmge {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 
 uint8x8_t test_vcgt_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vcgt_s8
+  // CHECK-LABEL: test_vcgt_s8
   return vcgt_s8(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vcgt_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vcgt_s16
+  // CHECK-LABEL: test_vcgt_s16
   return vcgt_s16(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vcgt_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vcgt_s32
+  // CHECK-LABEL: test_vcgt_s32
   return vcgt_s32(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcgt_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vcgt_s64
+  // CHECK-LABEL: test_vcgt_s64
   return vcgt_s64(a, b);
   // CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x1_t test_vcgt_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vcgt_u64
+  // CHECK-LABEL: test_vcgt_u64
   return vcgt_u64(a, b);
   // CHECK: cmhi {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x2_t test_vcgt_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vcgt_f32
+  // CHECK-LABEL: test_vcgt_f32
   return vcgt_f32(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vcgt_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vcgt_f64
+  // CHECK-LABEL: test_vcgt_f64
   return vcgt_f64(a, b);
   // CHECK: fcmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vcgt_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vcgt_u8
+  // CHECK-LABEL: test_vcgt_u8
   return vcgt_u8(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vcgt_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vcgt_u16
+  // CHECK-LABEL: test_vcgt_u16
   return vcgt_u16(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vcgt_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vcgt_u32
+  // CHECK-LABEL: test_vcgt_u32
   return vcgt_u32(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x16_t test_vcgtq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vcgtq_s8
+  // CHECK-LABEL: test_vcgtq_s8
   return vcgtq_s8(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vcgtq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vcgtq_s16
+  // CHECK-LABEL: test_vcgtq_s16
   return vcgtq_s16(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vcgtq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vcgtq_s32
+  // CHECK-LABEL: test_vcgtq_s32
   return vcgtq_s32(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vcgtq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcgtq_f32
+  // CHECK-LABEL: test_vcgtq_f32
   return vcgtq_f32(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vcgtq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vcgtq_u8
+  // CHECK-LABEL: test_vcgtq_u8
   return vcgtq_u8(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vcgtq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vcgtq_u16
+  // CHECK-LABEL: test_vcgtq_u16
   return vcgtq_u16(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vcgtq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vcgtq_u32
+  // CHECK-LABEL: test_vcgtq_u32
   return vcgtq_u32(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vcgtq_s64(int64x2_t v1, int64x2_t v2) {
-  // CHECK: test_vcgtq_s64
+  // CHECK-LABEL: test_vcgtq_s64
   return vcgtq_s64(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vcgtq_u64(uint64x2_t v1, uint64x2_t v2) {
-  // CHECK: test_vcgtq_u64
+  // CHECK-LABEL: test_vcgtq_u64
   return vcgtq_u64(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x2_t test_vcgtq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcgtq_f64
+  // CHECK-LABEL: test_vcgtq_f64
   return vcgtq_f64(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
@@ -1622,381 +1636,381 @@
 // Using registers other than v0, v1 are possible, but would be odd.
 
 uint8x8_t test_vclt_s8(int8x8_t v1, int8x8_t v2) {
-  // CHECK: test_vclt_s8
+  // CHECK-LABEL: test_vclt_s8
   return vclt_s8(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.8b, v1.8b, v0.8b
 }
 
 uint16x4_t test_vclt_s16(int16x4_t v1, int16x4_t v2) {
-  // CHECK: test_vclt_s16
+  // CHECK-LABEL: test_vclt_s16
   return vclt_s16(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.4h, v1.4h, v0.4h
 }
 
 uint32x2_t test_vclt_s32(int32x2_t v1, int32x2_t v2) {
-  // CHECK: test_vclt_s32
+  // CHECK-LABEL: test_vclt_s32
   return vclt_s32(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vclt_s64(int64x1_t a, int64x1_t b) {
-  // CHECK: test_vclt_s64
+  // CHECK-LABEL: test_vclt_s64
   return vclt_s64(a, b);
   // CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x1_t test_vclt_u64(uint64x1_t a, uint64x1_t b) {
-  // CHECK: test_vclt_u64
+  // CHECK-LABEL: test_vclt_u64
   return vclt_u64(a, b);
   // CHECK: cmhi {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint32x2_t test_vclt_f32(float32x2_t v1, float32x2_t v2) {
-  // CHECK: test_vclt_f32
+  // CHECK-LABEL: test_vclt_f32
   return vclt_f32(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint64x1_t test_vclt_f64(float64x1_t a, float64x1_t b) {
-  // CHECK: test_vclt_f64
+  // CHECK-LABEL: test_vclt_f64
   return vclt_f64(a, b);
   // CHECK: fcmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vclt_u8(uint8x8_t v1, uint8x8_t v2) {
-  // CHECK: test_vclt_u8
+  // CHECK-LABEL: test_vclt_u8
   return vclt_u8(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.8b, v1.8b, v0.8b
 }
 
 uint16x4_t test_vclt_u16(uint16x4_t v1, uint16x4_t v2) {
-  // CHECK: test_vclt_u16
+  // CHECK-LABEL: test_vclt_u16
   return vclt_u16(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.4h, v1.4h, v0.4h
 }
 
 uint32x2_t test_vclt_u32(uint32x2_t v1, uint32x2_t v2) {
-  // CHECK: test_vclt_u32
+  // CHECK-LABEL: test_vclt_u32
   return vclt_u32(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.2s, v1.2s, v0.2s
 }
 
 uint8x16_t test_vcltq_s8(int8x16_t v1, int8x16_t v2) {
-  // CHECK: test_vcltq_s8
+  // CHECK-LABEL: test_vcltq_s8
   return vcltq_s8(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.16b, v1.16b, v0.16b
 }
 
 uint16x8_t test_vcltq_s16(int16x8_t v1, int16x8_t v2) {
-  // CHECK: test_vcltq_s16
+  // CHECK-LABEL: test_vcltq_s16
   return vcltq_s16(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.8h, v1.8h, v0.8h
 }
 
 uint32x4_t test_vcltq_s32(int32x4_t v1, int32x4_t v2) {
-  // CHECK: test_vcltq_s32
+  // CHECK-LABEL: test_vcltq_s32
   return vcltq_s32(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint32x4_t test_vcltq_f32(float32x4_t v1, float32x4_t v2) {
-  // CHECK: test_vcltq_f32
+  // CHECK-LABEL: test_vcltq_f32
   return vcltq_f32(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint8x16_t test_vcltq_u8(uint8x16_t v1, uint8x16_t v2) {
-  // CHECK: test_vcltq_u8
+  // CHECK-LABEL: test_vcltq_u8
   return vcltq_u8(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.16b, v1.16b, v0.16b
 }
 
 uint16x8_t test_vcltq_u16(uint16x8_t v1, uint16x8_t v2) {
-  // CHECK: test_vcltq_u16
+  // CHECK-LABEL: test_vcltq_u16
   return vcltq_u16(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.8h, v1.8h, v0.8h
 }
 
 uint32x4_t test_vcltq_u32(uint32x4_t v1, uint32x4_t v2) {
-  // CHECK: test_vcltq_u32
+  // CHECK-LABEL: test_vcltq_u32
   return vcltq_u32(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.4s, v1.4s, v0.4s
 }
 
 uint64x2_t test_vcltq_s64(int64x2_t v1, int64x2_t v2) {
-  // CHECK: test_vcltq_s64
+  // CHECK-LABEL: test_vcltq_s64
   return vcltq_s64(v1, v2);
   // CHECK: cmgt {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint64x2_t test_vcltq_u64(uint64x2_t v1, uint64x2_t v2) {
-  // CHECK: test_vcltq_u64
+  // CHECK-LABEL: test_vcltq_u64
   return vcltq_u64(v1, v2);
   // CHECK: cmhi {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 uint64x2_t test_vcltq_f64(float64x2_t v1, float64x2_t v2) {
-  // CHECK: test_vcltq_f64
+  // CHECK-LABEL: test_vcltq_f64
   return vcltq_f64(v1, v2);
   // CHECK: fcmgt {{v[0-9]+}}.2d, v1.2d, v0.2d
 }
 
 
 int8x8_t test_vhadd_s8(int8x8_t v1, int8x8_t v2) {
-// CHECK: test_vhadd_s8
+// CHECK-LABEL: test_vhadd_s8
   return vhadd_s8(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vhadd_s16(int16x4_t v1, int16x4_t v2) {
-// CHECK: test_vhadd_s16
+// CHECK-LABEL: test_vhadd_s16
   return vhadd_s16(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vhadd_s32(int32x2_t v1, int32x2_t v2) {
-// CHECK: test_vhadd_s32
+// CHECK-LABEL: test_vhadd_s32
   return vhadd_s32(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vhadd_u8(uint8x8_t v1, uint8x8_t v2) {
-// CHECK: test_vhadd_u8
+// CHECK-LABEL: test_vhadd_u8
   return vhadd_u8(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vhadd_u16(uint16x4_t v1, uint16x4_t v2) {
-// CHECK: test_vhadd_u16
+// CHECK-LABEL: test_vhadd_u16
   return vhadd_u16(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vhadd_u32(uint32x2_t v1, uint32x2_t v2) {
-// CHECK: test_vhadd_u32
+// CHECK-LABEL: test_vhadd_u32
   return vhadd_u32(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vhaddq_s8(int8x16_t v1, int8x16_t v2) {
-// CHECK: test_vhaddq_s8
+// CHECK-LABEL: test_vhaddq_s8
   return vhaddq_s8(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vhaddq_s16(int16x8_t v1, int16x8_t v2) {
-// CHECK: test_vhaddq_s16
+// CHECK-LABEL: test_vhaddq_s16
   return vhaddq_s16(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vhaddq_s32(int32x4_t v1, int32x4_t v2) {
-// CHECK: test_vhaddq_s32
+// CHECK-LABEL: test_vhaddq_s32
   return vhaddq_s32(v1, v2);
   // CHECK: shadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vhaddq_u8(uint8x16_t v1, uint8x16_t v2) {
-// CHECK: test_vhaddq_u8
+// CHECK-LABEL: test_vhaddq_u8
   return vhaddq_u8(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vhaddq_u16(uint16x8_t v1, uint16x8_t v2) {
-// CHECK: test_vhaddq_u16
+// CHECK-LABEL: test_vhaddq_u16
   return vhaddq_u16(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vhaddq_u32(uint32x4_t v1, uint32x4_t v2) {
-// CHECK: test_vhaddq_u32
+// CHECK-LABEL: test_vhaddq_u32
   return vhaddq_u32(v1, v2);
   // CHECK: uhadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 
 int8x8_t test_vhsub_s8(int8x8_t v1, int8x8_t v2) {
-// CHECK: test_vhsub_s8
+// CHECK-LABEL: test_vhsub_s8
   return vhsub_s8(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vhsub_s16(int16x4_t v1, int16x4_t v2) {
-// CHECK: test_vhsub_s16
+// CHECK-LABEL: test_vhsub_s16
   return vhsub_s16(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vhsub_s32(int32x2_t v1, int32x2_t v2) {
-// CHECK: test_vhsub_s32
+// CHECK-LABEL: test_vhsub_s32
   return vhsub_s32(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vhsub_u8(uint8x8_t v1, uint8x8_t v2) {
-// CHECK: test_vhsub_u8
+// CHECK-LABEL: test_vhsub_u8
   return vhsub_u8(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vhsub_u16(uint16x4_t v1, uint16x4_t v2) {
-// CHECK: test_vhsub_u16
+// CHECK-LABEL: test_vhsub_u16
   return vhsub_u16(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vhsub_u32(uint32x2_t v1, uint32x2_t v2) {
-// CHECK: test_vhsub_u32
+// CHECK-LABEL: test_vhsub_u32
   return vhsub_u32(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vhsubq_s8(int8x16_t v1, int8x16_t v2) {
-// CHECK: test_vhsubq_s8
+// CHECK-LABEL: test_vhsubq_s8
   return vhsubq_s8(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vhsubq_s16(int16x8_t v1, int16x8_t v2) {
-// CHECK: test_vhsubq_s16
+// CHECK-LABEL: test_vhsubq_s16
   return vhsubq_s16(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vhsubq_s32(int32x4_t v1, int32x4_t v2) {
-// CHECK: test_vhsubq_s32
+// CHECK-LABEL: test_vhsubq_s32
   return vhsubq_s32(v1, v2);
   // CHECK: shsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vhsubq_u8(uint8x16_t v1, uint8x16_t v2) {
-// CHECK: test_vhsubq_u8
+// CHECK-LABEL: test_vhsubq_u8
   return vhsubq_u8(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vhsubq_u16(uint16x8_t v1, uint16x8_t v2) {
-// CHECK: test_vhsubq_u16
+// CHECK-LABEL: test_vhsubq_u16
   return vhsubq_u16(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vhsubq_u32(uint32x4_t v1, uint32x4_t v2) {
-// CHECK: test_vhsubq_u32
+// CHECK-LABEL: test_vhsubq_u32
   return vhsubq_u32(v1, v2);
   // CHECK: uhsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 
 int8x8_t test_vrhadd_s8(int8x8_t v1, int8x8_t v2) {
-// CHECK: test_vrhadd_s8
+// CHECK-LABEL: test_vrhadd_s8
   return vrhadd_s8(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vrhadd_s16(int16x4_t v1, int16x4_t v2) {
-// CHECK: test_vrhadd_s16
+// CHECK-LABEL: test_vrhadd_s16
   return vrhadd_s16(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vrhadd_s32(int32x2_t v1, int32x2_t v2) {
-// CHECK: test_vrhadd_s32
+// CHECK-LABEL: test_vrhadd_s32
   return vrhadd_s32(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vrhadd_u8(uint8x8_t v1, uint8x8_t v2) {
-// CHECK: test_vrhadd_u8
+// CHECK-LABEL: test_vrhadd_u8
   return vrhadd_u8(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vrhadd_u16(uint16x4_t v1, uint16x4_t v2) {
-// CHECK: test_vrhadd_u16
+// CHECK-LABEL: test_vrhadd_u16
   return vrhadd_u16(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vrhadd_u32(uint32x2_t v1, uint32x2_t v2) {
-// CHECK: test_vrhadd_u32
+// CHECK-LABEL: test_vrhadd_u32
   return vrhadd_u32(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vrhaddq_s8(int8x16_t v1, int8x16_t v2) {
-// CHECK: test_vrhaddq_s8
+// CHECK-LABEL: test_vrhaddq_s8
   return vrhaddq_s8(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vrhaddq_s16(int16x8_t v1, int16x8_t v2) {
-// CHECK: test_vrhaddq_s16
+// CHECK-LABEL: test_vrhaddq_s16
   return vrhaddq_s16(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vrhaddq_s32(int32x4_t v1, int32x4_t v2) {
-// CHECK: test_vrhaddq_s32
+// CHECK-LABEL: test_vrhaddq_s32
   return vrhaddq_s32(v1, v2);
 // CHECK: srhadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vrhaddq_u8(uint8x16_t v1, uint8x16_t v2) {
-// CHECK: test_vrhaddq_u8
+// CHECK-LABEL: test_vrhaddq_u8
   return vrhaddq_u8(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vrhaddq_u16(uint16x8_t v1, uint16x8_t v2) {
-// CHECK: test_vrhaddq_u16
+// CHECK-LABEL: test_vrhaddq_u16
   return vrhaddq_u16(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vrhaddq_u32(uint32x4_t v1, uint32x4_t v2) {
-// CHECK: test_vrhaddq_u32
+// CHECK-LABEL: test_vrhaddq_u32
   return vrhaddq_u32(v1, v2);
 // CHECK: urhadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 int8x8_t test_vqadd_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vqadd_s8
+// CHECK-LABEL: test_vqadd_s8
   return vqadd_s8(a, b);
   // CHECK: sqadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vqadd_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqadd_s16
+// CHECK-LABEL: test_vqadd_s16
   return vqadd_s16(a, b);
   // CHECK: sqadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqadd_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqadd_s32
+// CHECK-LABEL: test_vqadd_s32
   return vqadd_s32(a, b);
   // CHECK: sqadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vqadd_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vqadd_s64
+// CHECK-LABEL: test_vqadd_s64
   return vqadd_s64(a, b);
 // CHECK:  sqadd {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vqadd_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vqadd_u8
+// CHECK-LABEL: test_vqadd_u8
   return vqadd_u8(a, b);
   // CHECK: uqadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vqadd_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vqadd_u16
+// CHECK-LABEL: test_vqadd_u16
   return vqadd_u16(a, b);
   // CHECK: uqadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vqadd_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vqadd_u32
+// CHECK-LABEL: test_vqadd_u32
   return vqadd_u32(a, b);
   // CHECK: uqadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
@@ -2008,3764 +2022,3793 @@
 }
 
 int8x16_t test_vqaddq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vqaddq_s8
+// CHECK-LABEL: test_vqaddq_s8
   return vqaddq_s8(a, b);
   // CHECK: sqadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vqaddq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqaddq_s16
+// CHECK-LABEL: test_vqaddq_s16
   return vqaddq_s16(a, b);
   // CHECK: sqadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqaddq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqaddq_s32
+// CHECK-LABEL: test_vqaddq_s32
   return vqaddq_s32(a, b);
   // CHECK: sqadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vqaddq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vqaddq_s64
+// CHECK-LABEL: test_vqaddq_s64
   return vqaddq_s64(a, b);
 // CHECK: sqadd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vqaddq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vqaddq_u8
+// CHECK-LABEL: test_vqaddq_u8
   return vqaddq_u8(a, b);
   // CHECK: uqadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vqaddq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vqaddq_u16
+// CHECK-LABEL: test_vqaddq_u16
   return vqaddq_u16(a, b);
   // CHECK: uqadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vqaddq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vqaddq_u32
+// CHECK-LABEL: test_vqaddq_u32
   return vqaddq_u32(a, b);
   // CHECK: uqadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vqaddq_u64(uint64x2_t a, uint64x2_t b) {
-// CHECK: test_vqaddq_u64
+// CHECK-LABEL: test_vqaddq_u64
   return vqaddq_u64(a, b);
 // CHECK: uqadd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vqsub_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vqsub_s8
+// CHECK-LABEL: test_vqsub_s8
   return vqsub_s8(a, b);
   // CHECK: sqsub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vqsub_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqsub_s16
+// CHECK-LABEL: test_vqsub_s16
   return vqsub_s16(a, b);
   // CHECK: sqsub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqsub_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqsub_s32
+// CHECK-LABEL: test_vqsub_s32
   return vqsub_s32(a, b);
   // CHECK: sqsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vqsub_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vqsub_s64
+// CHECK-LABEL: test_vqsub_s64
   return vqsub_s64(a, b);
 // CHECK: sqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vqsub_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vqsub_u8
+// CHECK-LABEL: test_vqsub_u8
   return vqsub_u8(a, b);
   // CHECK: uqsub {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vqsub_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vqsub_u16
+// CHECK-LABEL: test_vqsub_u16
   return vqsub_u16(a, b);
   // CHECK: uqsub {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vqsub_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vqsub_u32
+// CHECK-LABEL: test_vqsub_u32
   return vqsub_u32(a, b);
   // CHECK: uqsub {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vqsub_u64(uint64x1_t a, uint64x1_t b) {
-// CHECK: test_vqsub_u64
+// CHECK-LABEL: test_vqsub_u64
   return vqsub_u64(a, b);
 // CHECK:  uqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vqsubq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vqsubq_s8
+// CHECK-LABEL: test_vqsubq_s8
   return vqsubq_s8(a, b);
   // CHECK: sqsub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vqsubq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqsubq_s16
+// CHECK-LABEL: test_vqsubq_s16
   return vqsubq_s16(a, b);
   // CHECK: sqsub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqsubq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqsubq_s32
+// CHECK-LABEL: test_vqsubq_s32
   return vqsubq_s32(a, b);
   // CHECK: sqsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vqsubq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vqsubq_s64
+// CHECK-LABEL: test_vqsubq_s64
   return vqsubq_s64(a, b);
 // CHECK: sqsub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vqsubq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vqsubq_u8
+// CHECK-LABEL: test_vqsubq_u8
   return vqsubq_u8(a, b);
   // CHECK: uqsub {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vqsubq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vqsubq_u16
+// CHECK-LABEL: test_vqsubq_u16
   return vqsubq_u16(a, b);
   // CHECK: uqsub {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vqsubq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vqsubq_u32
+// CHECK-LABEL: test_vqsubq_u32
   return vqsubq_u32(a, b);
   // CHECK: uqsub {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vqsubq_u64(uint64x2_t a, uint64x2_t b) {
-// CHECK: test_vqsubq_u64
+// CHECK-LABEL: test_vqsubq_u64
   return vqsubq_u64(a, b);
   // CHECK: uqsub {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vshl_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vshl_s8
+// CHECK-LABEL: test_vshl_s8
   return vshl_s8(a, b);
 // CHECK: sshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vshl_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vshl_s16
+// CHECK-LABEL: test_vshl_s16
   return vshl_s16(a, b);
 // CHECK: sshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vshl_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vshl_s32
+// CHECK-LABEL: test_vshl_s32
   return vshl_s32(a, b);
 // CHECK: sshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vshl_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vshl_s64
+// CHECK-LABEL: test_vshl_s64
   return vshl_s64(a, b);
 // CHECK: sshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vshl_u8(uint8x8_t a, int8x8_t b) {
-// CHECK: test_vshl_u8
+// CHECK-LABEL: test_vshl_u8
   return vshl_u8(a, b);
 // CHECK: ushl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vshl_u16(uint16x4_t a, int16x4_t b) {
-// CHECK: test_vshl_u16
+// CHECK-LABEL: test_vshl_u16
   return vshl_u16(a, b);
 // CHECK: ushl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vshl_u32(uint32x2_t a, int32x2_t b) {
-// CHECK: test_vshl_u32
+// CHECK-LABEL: test_vshl_u32
   return vshl_u32(a, b);
 // CHECK: ushl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vshl_u64(uint64x1_t a, int64x1_t b) {
-// CHECK: test_vshl_u64
+// CHECK-LABEL: test_vshl_u64
   return vshl_u64(a, b);
 // CHECK: ushl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vshlq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vshlq_s8
+// CHECK-LABEL: test_vshlq_s8
   return vshlq_s8(a, b);
 // CHECK: sshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vshlq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vshlq_s16
+// CHECK-LABEL: test_vshlq_s16
   return vshlq_s16(a, b);
 // CHECK: sshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vshlq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vshlq_s32
+// CHECK-LABEL: test_vshlq_s32
   return vshlq_s32(a, b);
 // CHECK: sshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vshlq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vshlq_s64
+// CHECK-LABEL: test_vshlq_s64
   return vshlq_s64(a, b);
 // CHECK: sshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vshlq_u8(uint8x16_t a, int8x16_t b) {
-// CHECK: test_vshlq_u8
+// CHECK-LABEL: test_vshlq_u8
   return vshlq_u8(a, b);
 // CHECK: ushl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vshlq_u16(uint16x8_t a, int16x8_t b) {
-// CHECK: test_vshlq_u16
+// CHECK-LABEL: test_vshlq_u16
   return vshlq_u16(a, b);
 // CHECK: ushl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vshlq_u32(uint32x4_t a, int32x4_t b) {
-// CHECK: test_vshlq_u32
+// CHECK-LABEL: test_vshlq_u32
   return vshlq_u32(a, b);
 // CHECK: ushl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vshlq_u64(uint64x2_t a, int64x2_t b) {
-// CHECK: test_vshlq_u64
+// CHECK-LABEL: test_vshlq_u64
   return vshlq_u64(a, b);
 // CHECK: ushl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vqshl_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vqshl_s8
+// CHECK-LABEL: test_vqshl_s8
   return vqshl_s8(a, b);
 // CHECK: sqshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vqshl_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqshl_s16
+// CHECK-LABEL: test_vqshl_s16
   return vqshl_s16(a, b);
 // CHECK: sqshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqshl_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqshl_s32
+// CHECK-LABEL: test_vqshl_s32
   return vqshl_s32(a, b);
 // CHECK: sqshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vqshl_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vqshl_s64
+// CHECK-LABEL: test_vqshl_s64
   return vqshl_s64(a, b);
 // CHECK: sqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vqshl_u8(uint8x8_t a, int8x8_t b) {
-// CHECK: test_vqshl_u8
+// CHECK-LABEL: test_vqshl_u8
   return vqshl_u8(a, b);
 // CHECK: uqshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vqshl_u16(uint16x4_t a, int16x4_t b) {
-// CHECK: test_vqshl_u16
+// CHECK-LABEL: test_vqshl_u16
   return vqshl_u16(a, b);
 // CHECK: uqshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vqshl_u32(uint32x2_t a, int32x2_t b) {
-// CHECK: test_vqshl_u32
+// CHECK-LABEL: test_vqshl_u32
   return vqshl_u32(a, b);
 // CHECK: uqshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vqshl_u64(uint64x1_t a, int64x1_t b) {
-// CHECK: test_vqshl_u64
+// CHECK-LABEL: test_vqshl_u64
   return vqshl_u64(a, b);
 // CHECK: uqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vqshlq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vqshlq_s8
+// CHECK-LABEL: test_vqshlq_s8
   return vqshlq_s8(a, b);
 // CHECK: sqshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vqshlq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqshlq_s16
+// CHECK-LABEL: test_vqshlq_s16
   return vqshlq_s16(a, b);
 // CHECK: sqshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqshlq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqshlq_s32
+// CHECK-LABEL: test_vqshlq_s32
   return vqshlq_s32(a, b);
 // CHECK: sqshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vqshlq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vqshlq_s64
+// CHECK-LABEL: test_vqshlq_s64
   return vqshlq_s64(a, b);
 // CHECK: sqshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vqshlq_u8(uint8x16_t a, int8x16_t b) {
-// CHECK: test_vqshlq_u8
+// CHECK-LABEL: test_vqshlq_u8
   return vqshlq_u8(a, b);
 // CHECK: uqshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vqshlq_u16(uint16x8_t a, int16x8_t b) {
-// CHECK: test_vqshlq_u16
+// CHECK-LABEL: test_vqshlq_u16
   return vqshlq_u16(a, b);
 // CHECK: uqshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vqshlq_u32(uint32x4_t a, int32x4_t b) {
-// CHECK: test_vqshlq_u32
+// CHECK-LABEL: test_vqshlq_u32
   return vqshlq_u32(a, b);
 // CHECK: uqshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vqshlq_u64(uint64x2_t a, int64x2_t b) {
-// CHECK: test_vqshlq_u32
+// CHECK-LABEL: test_vqshlq_u64
   return vqshlq_u64(a, b);
 // CHECK: uqshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vrshl_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vrshl_s8
+// CHECK-LABEL: test_vrshl_s8
   return vrshl_s8(a, b);
 // CHECK: srshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vrshl_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vrshl_s16
+// CHECK-LABEL: test_vrshl_s16
   return vrshl_s16(a, b);
 // CHECK: srshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vrshl_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vrshl_s32
+// CHECK-LABEL: test_vrshl_s32
   return vrshl_s32(a, b);
 // CHECK: srshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vrshl_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vrshl_s64
+// CHECK-LABEL: test_vrshl_s64
   return vrshl_s64(a, b);
 // CHECK: srshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vrshl_u8(uint8x8_t a, int8x8_t b) {
-// CHECK: test_vrshl_u8
+// CHECK-LABEL: test_vrshl_u8
   return vrshl_u8(a, b);
 // CHECK: urshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vrshl_u16(uint16x4_t a, int16x4_t b) {
-// CHECK: test_vrshl_u16
+// CHECK-LABEL: test_vrshl_u16
   return vrshl_u16(a, b);
 // CHECK: urshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vrshl_u32(uint32x2_t a, int32x2_t b) {
-// CHECK: test_vrshl_u32
+// CHECK-LABEL: test_vrshl_u32
   return vrshl_u32(a, b);
 // CHECK: urshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vrshl_u64(uint64x1_t a, int64x1_t b) {
-// CHECK: test_vrshl_u64
+// CHECK-LABEL: test_vrshl_u64
   return vrshl_u64(a, b);
 // CHECK: urshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vrshlq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vrshlq_s8
+// CHECK-LABEL: test_vrshlq_s8
   return vrshlq_s8(a, b);
 // CHECK: srshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vrshlq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vrshlq_s16
+// CHECK-LABEL: test_vrshlq_s16
   return vrshlq_s16(a, b);
 // CHECK: srshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vrshlq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vrshlq_s32
+// CHECK-LABEL: test_vrshlq_s32
   return vrshlq_s32(a, b);
 // CHECK: srshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vrshlq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vrshlq_s64
+// CHECK-LABEL: test_vrshlq_s64
   return vrshlq_s64(a, b);
 // CHECK: srshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vrshlq_u8(uint8x16_t a, int8x16_t b) {
-// CHECK: test_vrshlq_u8
+// CHECK-LABEL: test_vrshlq_u8
   return vrshlq_u8(a, b);
 // CHECK: urshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vrshlq_u16(uint16x8_t a, int16x8_t b) {
-// CHECK: test_vrshlq_u16
+// CHECK-LABEL: test_vrshlq_u16
   return vrshlq_u16(a, b);
 // CHECK: urshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vrshlq_u32(uint32x4_t a, int32x4_t b) {
-// CHECK: test_vrshlq_u32
+// CHECK-LABEL: test_vrshlq_u32
   return vrshlq_u32(a, b);
 // CHECK: urshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vrshlq_u64(uint64x2_t a, int64x2_t b) {
-// CHECK: test_vrshlq_u64
+// CHECK-LABEL: test_vrshlq_u64
   return vrshlq_u64(a, b);
 // CHECK: urshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vqrshl_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vqrshl_s8
+// CHECK-LABEL: test_vqrshl_s8
   return vqrshl_s8(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vqrshl_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqrshl_s16
+// CHECK-LABEL: test_vqrshl_s16
   return vqrshl_s16(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqrshl_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqrshl_s32
+// CHECK-LABEL: test_vqrshl_s32
   return vqrshl_s32(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int64x1_t test_vqrshl_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vqrshl_s64
+// CHECK-LABEL: test_vqrshl_s64
   return vqrshl_s64(a, b);
 // CHECK: sqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8x8_t test_vqrshl_u8(uint8x8_t a, int8x8_t b) {
-// CHECK: test_vqrshl_u8
+// CHECK-LABEL: test_vqrshl_u8
   return vqrshl_u8(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vqrshl_u16(uint16x4_t a, int16x4_t b) {
-// CHECK: test_vqrshl_u16
+// CHECK-LABEL: test_vqrshl_u16
   return vqrshl_u16(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vqrshl_u32(uint32x2_t a, int32x2_t b) {
-// CHECK: test_vqrshl_u32
+// CHECK-LABEL: test_vqrshl_u32
   return vqrshl_u32(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint64x1_t test_vqrshl_u64(uint64x1_t a, int64x1_t b) {
-// CHECK: test_vqrshl_u64
+// CHECK-LABEL: test_vqrshl_u64
   return vqrshl_u64(a, b);
 // CHECK: uqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8x16_t test_vqrshlq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vqrshlq_s8
+// CHECK-LABEL: test_vqrshlq_s8
   return vqrshlq_s8(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vqrshlq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqrshlq_s16
+// CHECK-LABEL: test_vqrshlq_s16
   return vqrshlq_s16(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqrshlq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqrshlq_s32
+// CHECK-LABEL: test_vqrshlq_s32
   return vqrshlq_s32(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vqrshlq_s64(int64x2_t a, int64x2_t b) {
-// CHECK: test_vqrshlq_s64
+// CHECK-LABEL: test_vqrshlq_s64
   return vqrshlq_s64(a, b);
 // CHECK: sqrshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vqrshlq_u8
+// CHECK-LABEL: test_vqrshlq_u8
 uint8x16_t test_vqrshlq_u8(uint8x16_t a, int8x16_t b) {
   return vqrshlq_u8(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vqrshlq_u16(uint16x8_t a, int16x8_t b) {
-// CHECK: test_vqrshlq_u16
+// CHECK-LABEL: test_vqrshlq_u16
   return vqrshlq_u16(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vqrshlq_u32(uint32x4_t a, int32x4_t b) {
-// CHECK: test_vqrshlq_u32
+// CHECK-LABEL: test_vqrshlq_u32
   return vqrshlq_u32(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vqrshlq_u64(uint64x2_t a, int64x2_t b) {
-// CHECK: test_vqrshlq_u64
+// CHECK-LABEL: test_vqrshlq_u64
   return vqrshlq_u64(a, b);
 // CHECK: uqrshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
+poly64x1_t test_vsli_n_p64(poly64x1_t a, poly64x1_t b) {
+// CHECK-LABEL: test_vsli_n_p64
+  return vsli_n_p64(a, b, 0); 
+// CHECK: sli {{d[0-9]+}}, {{d[0-9]+}}, #0
+}
+
+poly64x2_t test_vsliq_n_p64(poly64x2_t a, poly64x2_t b) {
+// CHECK-LABEL: test_vsliq_n_p64
+  return vsliq_n_p64(a, b, 0); 
+// CHECK: sli {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
+}
+
 int8x8_t test_vmax_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vmax_s8
+// CHECK-LABEL: test_vmax_s8
   return vmax_s8(a, b);
 // CHECK: smax {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vmax_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vmax_s16
+// CHECK-LABEL: test_vmax_s16
   return vmax_s16(a, b);
 // CHECK: smax {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vmax_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vmax_s32
+// CHECK-LABEL: test_vmax_s32
   return vmax_s32(a, b);
 // CHECK: smax {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vmax_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vmax_u8
+// CHECK-LABEL: test_vmax_u8
   return vmax_u8(a, b);
 // CHECK: umax {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vmax_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vmax_u16
+// CHECK-LABEL: test_vmax_u16
   return vmax_u16(a, b);
 // CHECK: umax {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vmax_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vmax_u32
+// CHECK-LABEL: test_vmax_u32
   return vmax_u32(a, b);
 // CHECK: umax {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vmax_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vmax_f32
+// CHECK-LABEL: test_vmax_f32
   return vmax_f32(a, b);
 // CHECK: fmax {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vmaxq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vmaxq_s8
+// CHECK-LABEL: test_vmaxq_s8
   return vmaxq_s8(a, b);
 // CHECK: smax {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vmaxq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vmaxq_s16
+// CHECK-LABEL: test_vmaxq_s16
   return vmaxq_s16(a, b);
 // CHECK: smax {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vmaxq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vmaxq_s32
+// CHECK-LABEL: test_vmaxq_s32
   return vmaxq_s32(a, b);
 // CHECK: smax {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vmaxq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vmaxq_u8
+// CHECK-LABEL: test_vmaxq_u8
   return vmaxq_u8(a, b);
 // CHECK: umax {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vmaxq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vmaxq_u16
+// CHECK-LABEL: test_vmaxq_u16
   return vmaxq_u16(a, b);
 // CHECK: umax {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vmaxq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vmaxq_u32
+// CHECK-LABEL: test_vmaxq_u32
   return vmaxq_u32(a, b);
 // CHECK: umax {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vmaxq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vmaxq_f32
+// CHECK-LABEL: test_vmaxq_f32
   return vmaxq_f32(a, b);
 // CHECK: fmax {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmaxq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vmaxq_f64
+// CHECK-LABEL: test_vmaxq_f64
   return vmaxq_f64(a, b);
 // CHECK: fmax {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 
 int8x8_t test_vmin_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vmin_s8
+// CHECK-LABEL: test_vmin_s8
   return vmin_s8(a, b);
 // CHECK: smin {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vmin_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vmin_s16
+// CHECK-LABEL: test_vmin_s16
   return vmin_s16(a, b);
 // CHECK: smin {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vmin_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vmin_s32
+// CHECK-LABEL: test_vmin_s32
   return vmin_s32(a, b);
 // CHECK: smin {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vmin_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vmin_u8
+// CHECK-LABEL: test_vmin_u8
   return vmin_u8(a, b);
 // CHECK: umin {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vmin_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vmin_u16
+// CHECK-LABEL: test_vmin_u16
   return vmin_u16(a, b);
 // CHECK: umin {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vmin_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vmin_u32
+// CHECK-LABEL: test_vmin_u32
   return vmin_u32(a, b);
 // CHECK: umin {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vmin_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vmin_f32
+// CHECK-LABEL: test_vmin_f32
   return vmin_f32(a, b);
 // CHECK: fmin {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vminq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vminq_s8
+// CHECK-LABEL: test_vminq_s8
   return vminq_s8(a, b);
 // CHECK: smin {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vminq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vminq_s16
+// CHECK-LABEL: test_vminq_s16
   return vminq_s16(a, b);
 // CHECK: smin {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vminq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vminq_s32
+// CHECK-LABEL: test_vminq_s32
   return vminq_s32(a, b);
 // CHECK: smin {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vminq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vminq_u8
+// CHECK-LABEL: test_vminq_u8
   return vminq_u8(a, b);
 // CHECK: umin {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vminq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vminq_u16
+// CHECK-LABEL: test_vminq_u16
   return vminq_u16(a, b);
 // CHECK: umin {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vminq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vminq_u32
+// CHECK-LABEL: test_vminq_u32
   return vminq_u32(a, b);
 // CHECK: umin {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vminq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vminq_f32
+// CHECK-LABEL: test_vminq_f32
   return vminq_f32(a, b);
 // CHECK: fmin {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vminq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vminq_f64
+// CHECK-LABEL: test_vminq_f64
   return vminq_f64(a, b);
 // CHECK: fmin {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x2_t test_vmaxnm_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vmaxnm_f32
+// CHECK-LABEL: test_vmaxnm_f32
   return vmaxnm_f32(a, b);
 // CHECK: fmaxnm {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vmaxnmq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vmaxnmq_f32
+// CHECK-LABEL: test_vmaxnmq_f32
   return vmaxnmq_f32(a, b);
 // CHECK: fmaxnm {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmaxnmq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vmaxnmq_f64
+// CHECK-LABEL: test_vmaxnmq_f64
   return vmaxnmq_f64(a, b);
 // CHECK: fmaxnm {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x2_t test_vminnm_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vminnm_f32
+// CHECK-LABEL: test_vminnm_f32
   return vminnm_f32(a, b);
 // CHECK: fminnm {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vminnmq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vminnmq_f32
+// CHECK-LABEL: test_vminnmq_f32
   return vminnmq_f32(a, b);
 // CHECK: fminnm {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vminnmq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vminnmq_f64
+// CHECK-LABEL: test_vminnmq_f64
   return vminnmq_f64(a, b);
 // CHECK: fminnm {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vpmax_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vpmax_s8
+// CHECK-LABEL: test_vpmax_s8
   return vpmax_s8(a, b);
 // CHECK: smaxp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vpmax_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vpmax_s16
+// CHECK-LABEL: test_vpmax_s16
   return vpmax_s16(a, b);
 // CHECK: smaxp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vpmax_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vpmax_s32
+// CHECK-LABEL: test_vpmax_s32
   return vpmax_s32(a, b);
 // CHECK: smaxp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vpmax_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vpmax_u8
+// CHECK-LABEL: test_vpmax_u8
   return vpmax_u8(a, b);
 // CHECK: umaxp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vpmax_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vpmax_u16
+// CHECK-LABEL: test_vpmax_u16
   return vpmax_u16(a, b);
 // CHECK: umaxp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vpmax_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vpmax_u32
+// CHECK-LABEL: test_vpmax_u32
   return vpmax_u32(a, b);
 // CHECK: umaxp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vpmax_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vpmax_f32
+// CHECK-LABEL: test_vpmax_f32
   return vpmax_f32(a, b);
 // CHECK: fmaxp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vpmaxq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vpmaxq_s8
+// CHECK-LABEL: test_vpmaxq_s8
   return vpmaxq_s8(a, b);
 // CHECK: smaxp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vpmaxq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vpmaxq_s16
+// CHECK-LABEL: test_vpmaxq_s16
   return vpmaxq_s16(a, b);
 // CHECK: smaxp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vpmaxq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vpmaxq_s32
+// CHECK-LABEL: test_vpmaxq_s32
   return vpmaxq_s32(a, b);
 // CHECK: smaxp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vpmaxq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vpmaxq_u8
+// CHECK-LABEL: test_vpmaxq_u8
   return vpmaxq_u8(a, b);
 // CHECK: umaxp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vpmaxq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vpmaxq_u16
+// CHECK-LABEL: test_vpmaxq_u16
   return vpmaxq_u16(a, b);
 // CHECK: umaxp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vpmaxq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vpmaxq_u32
+// CHECK-LABEL: test_vpmaxq_u32
   return vpmaxq_u32(a, b);
 // CHECK: umaxp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vpmaxq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vpmaxq_f32
+// CHECK-LABEL: test_vpmaxq_f32
   return vpmaxq_f32(a, b);
 // CHECK: fmaxp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vpmaxq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vpmaxq_f64
+// CHECK-LABEL: test_vpmaxq_f64
   return vpmaxq_f64(a, b);
 // CHECK: fmaxp {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vpmin_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vpmin_s8
+// CHECK-LABEL: test_vpmin_s8
   return vpmin_s8(a, b);
 // CHECK: sminp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vpmin_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vpmin_s16
+// CHECK-LABEL: test_vpmin_s16
   return vpmin_s16(a, b);
 // CHECK: sminp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vpmin_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vpmin_s32
+// CHECK-LABEL: test_vpmin_s32
   return vpmin_s32(a, b);
 // CHECK: sminp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vpmin_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vpmin_u8
+// CHECK-LABEL: test_vpmin_u8
   return vpmin_u8(a, b);
 // CHECK: uminp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vpmin_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vpmin_u16
+// CHECK-LABEL: test_vpmin_u16
   return vpmin_u16(a, b);
 // CHECK: uminp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vpmin_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vpmin_u32
+// CHECK-LABEL: test_vpmin_u32
   return vpmin_u32(a, b);
 // CHECK: uminp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vpmin_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vpmin_f32
+// CHECK-LABEL: test_vpmin_f32
   return vpmin_f32(a, b);
 // CHECK: fminp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vpminq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vpminq_s8
+// CHECK-LABEL: test_vpminq_s8
   return vpminq_s8(a, b);
 // CHECK: sminp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vpminq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vpminq_s16
+// CHECK-LABEL: test_vpminq_s16
   return vpminq_s16(a, b);
 // CHECK: sminp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vpminq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vpminq_s32
+// CHECK-LABEL: test_vpminq_s32
   return vpminq_s32(a, b);
 // CHECK: sminp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vpminq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vpminq_u8
+// CHECK-LABEL: test_vpminq_u8
   return vpminq_u8(a, b);
 // CHECK: uminp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vpminq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vpminq_u16
+// CHECK-LABEL: test_vpminq_u16
   return vpminq_u16(a, b);
 // CHECK: uminp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vpminq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vpminq_u32
+// CHECK-LABEL: test_vpminq_u32
   return vpminq_u32(a, b);
 // CHECK: uminp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vpminq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vpminq_f32
+// CHECK-LABEL: test_vpminq_f32
   return vpminq_f32(a, b);
 // CHECK: fminp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vpminq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vpminq_f64
+// CHECK-LABEL: test_vpminq_f64
   return vpminq_f64(a, b);
 // CHECK: fminp {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x2_t test_vpmaxnm_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vpmaxnm_f32
+// CHECK-LABEL: test_vpmaxnm_f32
   return vpmaxnm_f32(a, b);
 // CHECK: fmaxnmp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vpmaxnmq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vpmaxnmq_f32
+// CHECK-LABEL: test_vpmaxnmq_f32
   return vpmaxnmq_f32(a, b);
 // CHECK: fmaxnmp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vpmaxnmq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vpmaxnmq_f64
+// CHECK-LABEL: test_vpmaxnmq_f64
   return vpmaxnmq_f64(a, b);
 // CHECK: fmaxnmp {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 float32x2_t test_vpminnm_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vpminnm_f32
+// CHECK-LABEL: test_vpminnm_f32
   return vpminnm_f32(a, b);
 // CHECK: fminnmp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vpminnmq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vpminnmq_f32
+// CHECK-LABEL: test_vpminnmq_f32
   return vpminnmq_f32(a, b);
 // CHECK: fminnmp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vpminnmq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vpminnmq_f64
+// CHECK-LABEL: test_vpminnmq_f64
   return vpminnmq_f64(a, b);
 // CHECK: fminnmp {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vpadd_s8(int8x8_t a, int8x8_t b) {
-// CHECK: test_vpadd_s8
+// CHECK-LABEL: test_vpadd_s8
   return vpadd_s8(a, b);
 // CHECK: addp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4_t test_vpadd_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vpadd_s16
+// CHECK-LABEL: test_vpadd_s16
   return vpadd_s16(a, b);
 // CHECK: addp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vpadd_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vpadd_s32
+// CHECK-LABEL: test_vpadd_s32
   return vpadd_s32(a, b);
 // CHECK: addp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint8x8_t test_vpadd_u8(uint8x8_t a, uint8x8_t b) {
-// CHECK: test_vpadd_u8
+// CHECK-LABEL: test_vpadd_u8
   return vpadd_u8(a, b);
 // CHECK: addp {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint16x4_t test_vpadd_u16(uint16x4_t a, uint16x4_t b) {
-// CHECK: test_vpadd_u16
+// CHECK-LABEL: test_vpadd_u16
   return vpadd_u16(a, b);
 // CHECK: addp {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint32x2_t test_vpadd_u32(uint32x2_t a, uint32x2_t b) {
-// CHECK: test_vpadd_u32
+// CHECK-LABEL: test_vpadd_u32
   return vpadd_u32(a, b);
 // CHECK: addp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x2_t test_vpadd_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vpadd_f32
+// CHECK-LABEL: test_vpadd_f32
   return vpadd_f32(a, b);
 // CHECK: faddp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int8x16_t test_vpaddq_s8(int8x16_t a, int8x16_t b) {
-// CHECK: test_vpaddq_s8
+// CHECK-LABEL: test_vpaddq_s8
   return vpaddq_s8(a, b);
 // CHECK: addp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x8_t test_vpaddq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vpaddq_s16
+// CHECK-LABEL: test_vpaddq_s16
   return vpaddq_s16(a, b);
 // CHECK: addp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vpaddq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vpaddq_s32
+// CHECK-LABEL: test_vpaddq_s32
   return vpaddq_s32(a, b);
 // CHECK: addp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint8x16_t test_vpaddq_u8(uint8x16_t a, uint8x16_t b) {
-// CHECK: test_vpaddq_u8
+// CHECK-LABEL: test_vpaddq_u8
   return vpaddq_u8(a, b);
 // CHECK: addp {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x8_t test_vpaddq_u16(uint16x8_t a, uint16x8_t b) {
-// CHECK: test_vpaddq_u16
+// CHECK-LABEL: test_vpaddq_u16
   return vpaddq_u16(a, b);
 // CHECK: addp {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x4_t test_vpaddq_u32(uint32x4_t a, uint32x4_t b) {
-// CHECK: test_vpaddq_u32
+// CHECK-LABEL: test_vpaddq_u32
   return vpaddq_u32(a, b);
 // CHECK: addp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float32x4_t test_vpaddq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vpaddq_f32
+// CHECK-LABEL: test_vpaddq_f32
   return vpaddq_f32(a, b);
 // CHECK: faddp {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vpaddq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vpaddq_f64
+// CHECK-LABEL: test_vpaddq_f64
   return vpaddq_f64(a, b);
 // CHECK: faddp {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int16x4_t test_vqdmulh_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqdmulh_s16
+// CHECK-LABEL: test_vqdmulh_s16
   return vqdmulh_s16(a, b);
 // CHECK: sqdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqdmulh_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqdmulh_s32
+// CHECK-LABEL: test_vqdmulh_s32
   return vqdmulh_s32(a, b);
 // CHECK: sqdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vqdmulhq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqdmulhq_s16
+// CHECK-LABEL: test_vqdmulhq_s16
   return vqdmulhq_s16(a, b);
 // CHECK: sqdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqdmulhq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqdmulhq_s32
+// CHECK-LABEL: test_vqdmulhq_s32
   return vqdmulhq_s32(a, b);
 // CHECK: sqdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x4_t test_vqrdmulh_s16(int16x4_t a, int16x4_t b) {
-// CHECK: test_vqrdmulh_s16
+// CHECK-LABEL: test_vqrdmulh_s16
   return vqrdmulh_s16(a, b);
 // CHECK: sqrdmulh {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int32x2_t test_vqrdmulh_s32(int32x2_t a, int32x2_t b) {
-// CHECK: test_vqrdmulh_s32
+// CHECK-LABEL: test_vqrdmulh_s32
   return vqrdmulh_s32(a, b);
 // CHECK: sqrdmulh {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vqrdmulhq_s16(int16x8_t a, int16x8_t b) {
-// CHECK: test_vqrdmulhq_s16
+// CHECK-LABEL: test_vqrdmulhq_s16
   return vqrdmulhq_s16(a, b);
 // CHECK: sqrdmulh {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x4_t test_vqrdmulhq_s32(int32x4_t a, int32x4_t b) {
-// CHECK: test_vqrdmulhq_s32
+// CHECK-LABEL: test_vqrdmulhq_s32
   return vqrdmulhq_s32(a, b);
 // CHECK: sqrdmulh {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
-
 float32x2_t test_vmulx_f32(float32x2_t a, float32x2_t b) {
-// CHECK: test_vmulx_f32
+// CHECK-LABEL: test_vmulx_f32
   return vmulx_f32(a, b);
 // CHECK: fmulx {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 float32x4_t test_vmulxq_f32(float32x4_t a, float32x4_t b) {
-// CHECK: test_vmulxq_f32
+// CHECK-LABEL: test_vmulxq_f32
   return vmulxq_f32(a, b);
 // CHECK: fmulx {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vmulxq_f64(float64x2_t a, float64x2_t b) {
-// CHECK: test_vmulxq_f64
+// CHECK-LABEL: test_vmulxq_f64
   return vmulxq_f64(a, b);
 // CHECK: fmulx {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vshl_n_s8(int8x8_t a) {
-// CHECK: test_vshl_n_s8
+// CHECK-LABEL: test_vshl_n_s8
   return vshl_n_s8(a, 3);
 // CHECK: shl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vshl_n_s16(int16x4_t a) {
-// CHECK: test_vshl_n_s16
+// CHECK-LABEL: test_vshl_n_s16
   return vshl_n_s16(a, 3);
 // CHECK: shl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vshl_n_s32(int32x2_t a) {
-// CHECK: test_vshl_n_s32
+// CHECK-LABEL: test_vshl_n_s32
   return vshl_n_s32(a, 3);
 // CHECK: shl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vshlq_n_s8(int8x16_t a) {
-// CHECK: test_vshlq_n_s8
+// CHECK-LABEL: test_vshlq_n_s8
   return vshlq_n_s8(a, 3);
 // CHECK: shl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vshlq_n_s16(int16x8_t a) {
-// CHECK: test_vshlq_n_s16
+// CHECK-LABEL: test_vshlq_n_s16
   return vshlq_n_s16(a, 3);
 // CHECK: shl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vshlq_n_s32(int32x4_t a) {
-// CHECK: test_vshlq_n_s32
+// CHECK-LABEL: test_vshlq_n_s32
   return vshlq_n_s32(a, 3);
 // CHECK: shl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vshlq_n_s64(int64x2_t a) {
-// CHECK: test_vshlq_n_s64
+// CHECK-LABEL: test_vshlq_n_s64
   return vshlq_n_s64(a, 3);
 // CHECK: shl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vshl_n_u8(int8x8_t a) {
-// CHECK: test_vshl_n_u8
+// CHECK-LABEL: test_vshl_n_u8
   return vshl_n_u8(a, 3);
 // CHECK: shl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vshl_n_u16(int16x4_t a) {
-// CHECK: test_vshl_n_u16
+// CHECK-LABEL: test_vshl_n_u16
   return vshl_n_u16(a, 3);
 // CHECK: shl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vshl_n_u32(int32x2_t a) {
-// CHECK: test_vshl_n_u32
+// CHECK-LABEL: test_vshl_n_u32
   return vshl_n_u32(a, 3);
 // CHECK: shl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vshlq_n_u8(int8x16_t a) {
-// CHECK: test_vshlq_n_u8
+// CHECK-LABEL: test_vshlq_n_u8
   return vshlq_n_u8(a, 3);
 // CHECK: shl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vshlq_n_u16(int16x8_t a) {
-// CHECK: test_vshlq_n_u16
+// CHECK-LABEL: test_vshlq_n_u16
   return vshlq_n_u16(a, 3);
 // CHECK: shl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vshlq_n_u32(int32x4_t a) {
-// CHECK: test_vshlq_n_u32
+// CHECK-LABEL: test_vshlq_n_u32
   return vshlq_n_u32(a, 3);
 // CHECK: shl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vshlq_n_u64(int64x2_t a) {
-// CHECK: test_vshlq_n_u64
+// CHECK-LABEL: test_vshlq_n_u64
   return vshlq_n_u64(a, 3);
 // CHECK: shl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vshr_n_s8(int8x8_t a) {
-  // CHECK: test_vshr_n_s8
+  // CHECK-LABEL: test_vshr_n_s8
   return vshr_n_s8(a, 3);
   // CHECK: sshr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vshr_n_s16(int16x4_t a) {
-  // CHECK: test_vshr_n_s16
+  // CHECK-LABEL: test_vshr_n_s16
   return vshr_n_s16(a, 3);
   // CHECK: sshr {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vshr_n_s32(int32x2_t a) {
-  // CHECK: test_vshr_n_s32
+  // CHECK-LABEL: test_vshr_n_s32
   return vshr_n_s32(a, 3);
   // CHECK: sshr {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vshrq_n_s8(int8x16_t a) {
-  // CHECK: test_vshrq_n_s8
+  // CHECK-LABEL: test_vshrq_n_s8
   return vshrq_n_s8(a, 3);
   // CHECK: sshr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vshrq_n_s16(int16x8_t a) {
-  // CHECK: test_vshrq_n_s16
+  // CHECK-LABEL: test_vshrq_n_s16
   return vshrq_n_s16(a, 3);
   // CHECK: sshr {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vshrq_n_s32(int32x4_t a) {
-  // CHECK: test_vshrq_n_s32
+  // CHECK-LABEL: test_vshrq_n_s32
   return vshrq_n_s32(a, 3);
   // CHECK: sshr {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vshrq_n_s64(int64x2_t a) {
-  // CHECK: test_vshrq_n_s64
+  // CHECK-LABEL: test_vshrq_n_s64
   return vshrq_n_s64(a, 3);
   // CHECK: sshr {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vshr_n_u8(int8x8_t a) {
-  // CHECK: test_vshr_n_u8
+  // CHECK-LABEL: test_vshr_n_u8
   return vshr_n_u8(a, 3);
   // CHECK: ushr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vshr_n_u16(int16x4_t a) {
-  // CHECK: test_vshr_n_u16
+  // CHECK-LABEL: test_vshr_n_u16
   return vshr_n_u16(a, 3);
   // CHECK: ushr {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vshr_n_u32(int32x2_t a) {
-  // CHECK: test_vshr_n_u32
+  // CHECK-LABEL: test_vshr_n_u32
   return vshr_n_u32(a, 3);
   // CHECK: ushr {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vshrq_n_u8(int8x16_t a) {
-  // CHECK: test_vshrq_n_u8
+  // CHECK-LABEL: test_vshrq_n_u8
   return vshrq_n_u8(a, 3);
   // CHECK: ushr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vshrq_n_u16(int16x8_t a) {
-  // CHECK: test_vshrq_n_u16
+  // CHECK-LABEL: test_vshrq_n_u16
   return vshrq_n_u16(a, 3);
   // CHECK: ushr {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vshrq_n_u32(int32x4_t a) {
-  // CHECK: test_vshrq_n_u32
+  // CHECK-LABEL: test_vshrq_n_u32
   return vshrq_n_u32(a, 3);
   // CHECK: ushr {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vshrq_n_u64(int64x2_t a) {
-  // CHECK: test_vshrq_n_u64
+  // CHECK-LABEL: test_vshrq_n_u64
   return vshrq_n_u64(a, 3);
   // CHECK: ushr {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vsra_n_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsra_n_s8
+  // CHECK-LABEL: test_vsra_n_s8
   return vsra_n_s8(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vsra_n_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsra_n_s16
+  // CHECK-LABEL: test_vsra_n_s16
   return vsra_n_s16(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vsra_n_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsra_n_s32
+  // CHECK-LABEL: test_vsra_n_s32
   return vsra_n_s32(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vsraq_n_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsraq_n_s8
+  // CHECK-LABEL: test_vsraq_n_s8
   return vsraq_n_s8(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vsraq_n_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsraq_n_s16
+  // CHECK-LABEL: test_vsraq_n_s16
   return vsraq_n_s16(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vsraq_n_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsraq_n_s32
+  // CHECK-LABEL: test_vsraq_n_s32
   return vsraq_n_s32(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vsraq_n_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsraq_n_s64
+  // CHECK-LABEL: test_vsraq_n_s64
   return vsraq_n_s64(a, b, 3);
   // CHECK: ssra {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vsra_n_u8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsra_n_u8
+  // CHECK-LABEL: test_vsra_n_u8
   return vsra_n_u8(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vsra_n_u16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsra_n_u16
+  // CHECK-LABEL: test_vsra_n_u16
   return vsra_n_u16(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vsra_n_u32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsra_n_u32
+  // CHECK-LABEL: test_vsra_n_u32
   return vsra_n_u32(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vsraq_n_u8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsraq_n_u8
+  // CHECK-LABEL: test_vsraq_n_u8
   return vsraq_n_u8(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vsraq_n_u16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsraq_n_u16
+  // CHECK-LABEL: test_vsraq_n_u16
   return vsraq_n_u16(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vsraq_n_u32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsraq_n_u32
+  // CHECK-LABEL: test_vsraq_n_u32
   return vsraq_n_u32(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vsraq_n_u64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsraq_n_u64
+  // CHECK-LABEL: test_vsraq_n_u64
   return vsraq_n_u64(a, b, 3);
   // CHECK: usra {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vrshr_n_s8(int8x8_t a) {
-  // CHECK: test_vrshr_n_s8
+  // CHECK-LABEL: test_vrshr_n_s8
   return vrshr_n_s8(a, 3);
   // CHECK: srshr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vrshr_n_s16(int16x4_t a) {
-  // CHECK: test_vrshr_n_s16
+  // CHECK-LABEL: test_vrshr_n_s16
   return vrshr_n_s16(a, 3);
   // CHECK: srshr {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vrshr_n_s32(int32x2_t a) {
-  // CHECK: test_vrshr_n_s32
+  // CHECK-LABEL: test_vrshr_n_s32
   return vrshr_n_s32(a, 3);
   // CHECK: srshr {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vrshrq_n_s8(int8x16_t a) {
-  // CHECK: test_vrshrq_n_s8
+  // CHECK-LABEL: test_vrshrq_n_s8
   return vrshrq_n_s8(a, 3);
   // CHECK: srshr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vrshrq_n_s16(int16x8_t a) {
-  // CHECK: test_vrshrq_n_s16
+  // CHECK-LABEL: test_vrshrq_n_s16
   return vrshrq_n_s16(a, 3);
   // CHECK: srshr {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vrshrq_n_s32(int32x4_t a) {
-  // CHECK: test_vrshrq_n_s32
+  // CHECK-LABEL: test_vrshrq_n_s32
   return vrshrq_n_s32(a, 3);
   // CHECK: srshr {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vrshrq_n_s64(int64x2_t a) {
-  // CHECK: test_vrshrq_n_s64
+  // CHECK-LABEL: test_vrshrq_n_s64
   return vrshrq_n_s64(a, 3);
   // CHECK: srshr {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vrshr_n_u8(int8x8_t a) {
-  // CHECK: test_vrshr_n_u8
+  // CHECK-LABEL: test_vrshr_n_u8
   return vrshr_n_u8(a, 3);
   // CHECK: urshr {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vrshr_n_u16(int16x4_t a) {
-  // CHECK: test_vrshr_n_u16
+  // CHECK-LABEL: test_vrshr_n_u16
   return vrshr_n_u16(a, 3);
   // CHECK: urshr {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vrshr_n_u32(int32x2_t a) {
-  // CHECK: test_vrshr_n_u32
+  // CHECK-LABEL: test_vrshr_n_u32
   return vrshr_n_u32(a, 3);
   // CHECK: urshr {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vrshrq_n_u8(int8x16_t a) {
-  // CHECK: test_vrshrq_n_u8
+  // CHECK-LABEL: test_vrshrq_n_u8
   return vrshrq_n_u8(a, 3);
   // CHECK: urshr {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vrshrq_n_u16(int16x8_t a) {
-  // CHECK: test_vrshrq_n_u16
+  // CHECK-LABEL: test_vrshrq_n_u16
   return vrshrq_n_u16(a, 3);
   // CHECK: urshr {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vrshrq_n_u32(int32x4_t a) {
-  // CHECK: test_vrshrq_n_u32
+  // CHECK-LABEL: test_vrshrq_n_u32
   return vrshrq_n_u32(a, 3);
   // CHECK: urshr {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vrshrq_n_u64(int64x2_t a) {
-  // CHECK: test_vrshrq_n_u64
+  // CHECK-LABEL: test_vrshrq_n_u64
   return vrshrq_n_u64(a, 3);
   // CHECK: urshr {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vrsra_n_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vrsra_n_s8
+  // CHECK-LABEL: test_vrsra_n_s8
   return vrsra_n_s8(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vrsra_n_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vrsra_n_s16
+  // CHECK-LABEL: test_vrsra_n_s16
   return vrsra_n_s16(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vrsra_n_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vrsra_n_s32
+  // CHECK-LABEL: test_vrsra_n_s32
   return vrsra_n_s32(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vrsraq_n_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vrsraq_n_s8
+  // CHECK-LABEL: test_vrsraq_n_s8
   return vrsraq_n_s8(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vrsraq_n_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vrsraq_n_s16
+  // CHECK-LABEL: test_vrsraq_n_s16
   return vrsraq_n_s16(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vrsraq_n_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vrsraq_n_s32
+  // CHECK-LABEL: test_vrsraq_n_s32
   return vrsraq_n_s32(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vrsraq_n_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vrsraq_n_s64
+  // CHECK-LABEL: test_vrsraq_n_s64
   return vrsraq_n_s64(a, b, 3);
   // CHECK: srsra {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vrsra_n_u8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vrsra_n_u8
+  // CHECK-LABEL: test_vrsra_n_u8
   return vrsra_n_u8(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vrsra_n_u16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vrsra_n_u16
+  // CHECK-LABEL: test_vrsra_n_u16
   return vrsra_n_u16(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vrsra_n_u32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vrsra_n_u32
+  // CHECK-LABEL: test_vrsra_n_u32
   return vrsra_n_u32(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vrsraq_n_u8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vrsraq_n_u8
+  // CHECK-LABEL: test_vrsraq_n_u8
   return vrsraq_n_u8(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vrsraq_n_u16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vrsraq_n_u16
+  // CHECK-LABEL: test_vrsraq_n_u16
   return vrsraq_n_u16(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vrsraq_n_u32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vrsraq_n_u32
+  // CHECK-LABEL: test_vrsraq_n_u32
   return vrsraq_n_u32(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vrsraq_n_u64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vrsraq_n_u64
+  // CHECK-LABEL: test_vrsraq_n_u64
   return vrsraq_n_u64(a, b, 3);
   // CHECK: ursra {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vsri_n_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsri_n_s8
+  // CHECK-LABEL: test_vsri_n_s8
   return vsri_n_s8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vsri_n_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsri_n_s16
+  // CHECK-LABEL: test_vsri_n_s16
   return vsri_n_s16(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vsri_n_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsri_n_s32
+  // CHECK-LABEL: test_vsri_n_s32
   return vsri_n_s32(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vsriq_n_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsriq_n_s8
+  // CHECK-LABEL: test_vsriq_n_s8
   return vsriq_n_s8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vsriq_n_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsriq_n_s16
+  // CHECK-LABEL: test_vsriq_n_s16
   return vsriq_n_s16(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vsriq_n_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsriq_n_s32
+  // CHECK-LABEL: test_vsriq_n_s32
   return vsriq_n_s32(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vsriq_n_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsriq_n_s64
+  // CHECK-LABEL: test_vsriq_n_s64
   return vsriq_n_s64(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vsri_n_u8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsri_n_u8
+  // CHECK-LABEL: test_vsri_n_u8
   return vsri_n_u8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vsri_n_u16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsri_n_u16
+  // CHECK-LABEL: test_vsri_n_u16
   return vsri_n_u16(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vsri_n_u32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsri_n_u32
+  // CHECK-LABEL: test_vsri_n_u32
   return vsri_n_u32(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vsriq_n_u8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsriq_n_u8
+  // CHECK-LABEL: test_vsriq_n_u8
   return vsriq_n_u8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vsriq_n_u16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsriq_n_u16
+  // CHECK-LABEL: test_vsriq_n_u16
   return vsriq_n_u16(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vsriq_n_u32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsriq_n_u32
+  // CHECK-LABEL: test_vsriq_n_u32
   return vsriq_n_u32(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vsriq_n_u64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsriq_n_u64
+  // CHECK-LABEL: test_vsriq_n_u64
   return vsriq_n_u64(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 poly8x8_t test_vsri_n_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vsri_n_p8
+  // CHECK-LABEL: test_vsri_n_p8
   return vsri_n_p8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 poly16x4_t test_vsri_n_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vsri_n_p16
+  // CHECK-LABEL: test_vsri_n_p16
   return vsri_n_p16(a, b, 15);
   // CHECK: sri {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #15
 }
 
 poly8x16_t test_vsriq_n_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vsriq_n_p8
+  // CHECK-LABEL: test_vsriq_n_p8
   return vsriq_n_p8(a, b, 3);
   // CHECK: sri {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 poly16x8_t test_vsriq_n_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vsriq_n_p16
+  // CHECK-LABEL: test_vsriq_n_p16
   return vsriq_n_p16(a, b, 15);
   // CHECK: sri {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #15
 }
 
 int8x8_t test_vsli_n_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsli_n_s8
+  // CHECK-LABEL: test_vsli_n_s8
   return vsli_n_s8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vsli_n_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsli_n_s16
+  // CHECK-LABEL: test_vsli_n_s16
   return vsli_n_s16(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vsli_n_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsli_n_s32
+  // CHECK-LABEL: test_vsli_n_s32
   return vsli_n_s32(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vsliq_n_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsliq_n_s8
+  // CHECK-LABEL: test_vsliq_n_s8
   return vsliq_n_s8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vsliq_n_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsliq_n_s16
+  // CHECK-LABEL: test_vsliq_n_s16
   return vsliq_n_s16(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vsliq_n_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsliq_n_s32
+  // CHECK-LABEL: test_vsliq_n_s32
   return vsliq_n_s32(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vsliq_n_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsliq_n_s64
+  // CHECK-LABEL: test_vsliq_n_s64
   return vsliq_n_s64(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 uint8x8_t test_vsli_n_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vsli_n_u8
+  // CHECK-LABEL: test_vsli_n_u8
   return vsli_n_u8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 uint16x4_t test_vsli_n_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vsli_n_u16
+  // CHECK-LABEL: test_vsli_n_u16
   return vsli_n_u16(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 uint32x2_t test_vsli_n_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vsli_n_u32
+  // CHECK-LABEL: test_vsli_n_u32
   return vsli_n_u32(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 uint8x16_t test_vsliq_n_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vsliq_n_u8
+  // CHECK-LABEL: test_vsliq_n_u8
   return vsliq_n_u8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 uint16x8_t test_vsliq_n_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vsliq_n_u16
+  // CHECK-LABEL: test_vsliq_n_u16
   return vsliq_n_u16(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 uint32x4_t test_vsliq_n_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vsliq_n_u32
+  // CHECK-LABEL: test_vsliq_n_u32
   return vsliq_n_u32(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 uint64x2_t test_vsliq_n_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vsliq_n_u64
+  // CHECK-LABEL: test_vsliq_n_u64
   return vsliq_n_u64(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 poly8x8_t test_vsli_n_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vsli_n_p8
+  // CHECK-LABEL: test_vsli_n_p8
   return vsli_n_p8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 poly16x4_t test_vsli_n_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vsli_n_p16
+  // CHECK-LABEL: test_vsli_n_p16
   return vsli_n_p16(a, b, 15);
   // CHECK: sli {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #15
 }
 
 poly8x16_t test_vsliq_n_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vsliq_n_p8
+  // CHECK-LABEL: test_vsliq_n_p8
   return vsliq_n_p8(a, b, 3);
   // CHECK: sli {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 poly16x8_t test_vsliq_n_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vsliq_n_p16
+  // CHECK-LABEL: test_vsliq_n_p16
   return vsliq_n_p16(a, b, 15);
   // CHECK: sli {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #15
 }
 
 int8x8_t test_vqshlu_n_s8(int8x8_t a) {
-  // CHECK: test_vqshlu_n_s8
+  // CHECK-LABEL: test_vqshlu_n_s8
   return vqshlu_n_s8(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #3
 }
 
 int16x4_t test_vqshlu_n_s16(int16x4_t a) {
-  // CHECK: test_vqshlu_n_s16
+  // CHECK-LABEL: test_vqshlu_n_s16
   return vqshlu_n_s16(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #3
 }
 
 int32x2_t test_vqshlu_n_s32(int32x2_t a) {
-  // CHECK: test_vqshlu_n_s32
+  // CHECK-LABEL: test_vqshlu_n_s32
   return vqshlu_n_s32(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #3
 }
 
 int8x16_t test_vqshluq_n_s8(int8x16_t a) {
-  // CHECK: test_vqshluq_n_s8
+  // CHECK-LABEL: test_vqshluq_n_s8
   return vqshluq_n_s8(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #3
 }
 
 int16x8_t test_vqshluq_n_s16(int16x8_t a) {
-  // CHECK: test_vqshluq_n_s16
+  // CHECK-LABEL: test_vqshluq_n_s16
   return vqshluq_n_s16(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #3
 }
 
 int32x4_t test_vqshluq_n_s32(int32x4_t a) {
-  // CHECK: test_vqshluq_n_s32
+  // CHECK-LABEL: test_vqshluq_n_s32
   return vqshluq_n_s32(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #3
 }
 
 int64x2_t test_vqshluq_n_s64(int64x2_t a) {
-  // CHECK: test_vqshluq_n_s64
+  // CHECK-LABEL: test_vqshluq_n_s64
   return vqshluq_n_s64(a, 3);
   // CHECK: sqshlu {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #3
 }
 
 int8x8_t test_vshrn_n_s16(int16x8_t a) {
-  // CHECK: test_vshrn_n_s16
+  // CHECK-LABEL: test_vshrn_n_s16
   return vshrn_n_s16(a, 3);
   // CHECK: shrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vshrn_n_s32(int32x4_t a) {
-  // CHECK: test_vshrn_n_s32
+  // CHECK-LABEL: test_vshrn_n_s32
   return vshrn_n_s32(a, 9);
   // CHECK: shrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vshrn_n_s64(int64x2_t a) {
-  // CHECK: test_vshrn_n_s64
+  // CHECK-LABEL: test_vshrn_n_s64
   return vshrn_n_s64(a, 19);
   // CHECK: shrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x8_t test_vshrn_n_u16(uint16x8_t a) {
-  // CHECK: test_vshrn_n_u16
+  // CHECK-LABEL: test_vshrn_n_u16
   return vshrn_n_u16(a, 3);
   // CHECK: shrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x4_t test_vshrn_n_u32(uint32x4_t a) {
-  // CHECK: test_vshrn_n_u32
+  // CHECK-LABEL: test_vshrn_n_u32
   return vshrn_n_u32(a, 9);
   // CHECK: shrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x2_t test_vshrn_n_u64(uint64x2_t a) {
-  // CHECK: test_vshrn_n_u64
+  // CHECK-LABEL: test_vshrn_n_u64
   return vshrn_n_u64(a, 19);
   // CHECK: shrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vshrn_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vshrn_high_n_s16
+  // CHECK-LABEL: test_vshrn_high_n_s16
   return vshrn_high_n_s16(a, b, 3);
   // CHECK: shrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vshrn_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vshrn_high_n_s32
+  // CHECK-LABEL: test_vshrn_high_n_s32
   return vshrn_high_n_s32(a, b, 9);
   // CHECK: shrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vshrn_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vshrn_high_n_s64
+  // CHECK-LABEL: test_vshrn_high_n_s64
   return vshrn_high_n_s64(a, b, 19);
   // CHECK: shrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x16_t test_vshrn_high_n_u16(uint8x8_t a, uint16x8_t b) {
-  // CHECK: test_vshrn_high_n_u16
+  // CHECK-LABEL: test_vshrn_high_n_u16
   return vshrn_high_n_u16(a, b, 3);
   // CHECK: shrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x8_t test_vshrn_high_n_u32(uint16x4_t a, uint32x4_t b) {
-  // CHECK: test_vshrn_high_n_u32
+  // CHECK-LABEL: test_vshrn_high_n_u32
   return vshrn_high_n_u32(a, b, 9);
   // CHECK: shrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x4_t test_vshrn_high_n_u64(uint32x2_t a, uint64x2_t b) {
-  // CHECK: test_vshrn_high_n_u64
+  // CHECK-LABEL: test_vshrn_high_n_u64
   return vshrn_high_n_u64(a, b, 19);
   // CHECK: shrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int8x8_t test_vqshrun_n_s16(int16x8_t a) {
-  // CHECK: test_vqshrun_n_s16
+  // CHECK-LABEL: test_vqshrun_n_s16
   return vqshrun_n_s16(a, 3);
   // CHECK: sqshrun {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vqshrun_n_s32(int32x4_t a) {
-  // CHECK: test_vqshrun_n_s32
+  // CHECK-LABEL: test_vqshrun_n_s32
   return vqshrun_n_s32(a, 9);
   // CHECK: sqshrun {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vqshrun_n_s64(int64x2_t a) {
-  // CHECK: test_vqshrun_n_s64
+  // CHECK-LABEL: test_vqshrun_n_s64
   return vqshrun_n_s64(a, 19);
   // CHECK: sqshrun {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vqshrun_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqshrun_high_n_s16
+  // CHECK-LABEL: test_vqshrun_high_n_s16
   return vqshrun_high_n_s16(a, b, 3);
   // CHECK: sqshrun2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vqshrun_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqshrun_high_n_s32
+  // CHECK-LABEL: test_vqshrun_high_n_s32
   return vqshrun_high_n_s32(a, b, 9);
   // CHECK: sqshrun2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vqshrun_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqshrun_high_n_s64
+  // CHECK-LABEL: test_vqshrun_high_n_s64
   return vqshrun_high_n_s64(a, b, 19);
   // CHECK: sqshrun2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int8x8_t test_vrshrn_n_s16(int16x8_t a) {
-  // CHECK: test_vrshrn_n_s16
+  // CHECK-LABEL: test_vrshrn_n_s16
   return vrshrn_n_s16(a, 3);
   // CHECK: rshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vrshrn_n_s32(int32x4_t a) {
-  // CHECK: test_vrshrn_n_s32
+  // CHECK-LABEL: test_vrshrn_n_s32
   return vrshrn_n_s32(a, 9);
   // CHECK: rshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vrshrn_n_s64(int64x2_t a) {
-  // CHECK: test_vrshrn_n_s64
+  // CHECK-LABEL: test_vrshrn_n_s64
   return vrshrn_n_s64(a, 19);
   // CHECK: rshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x8_t test_vrshrn_n_u16(uint16x8_t a) {
-  // CHECK: test_vrshrn_n_u16
+  // CHECK-LABEL: test_vrshrn_n_u16
   return vrshrn_n_u16(a, 3);
   // CHECK: rshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x4_t test_vrshrn_n_u32(uint32x4_t a) {
-  // CHECK: test_vrshrn_n_u32
+  // CHECK-LABEL: test_vrshrn_n_u32
   return vrshrn_n_u32(a, 9);
   // CHECK: rshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x2_t test_vrshrn_n_u64(uint64x2_t a) {
-  // CHECK: test_vrshrn_n_u64
+  // CHECK-LABEL: test_vrshrn_n_u64
   return vrshrn_n_u64(a, 19);
   // CHECK: rshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vrshrn_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vrshrn_high_n_s16
+  // CHECK-LABEL: test_vrshrn_high_n_s16
   return vrshrn_high_n_s16(a, b, 3);
   // CHECK: rshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vrshrn_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vrshrn_high_n_s32
+  // CHECK-LABEL: test_vrshrn_high_n_s32
   return vrshrn_high_n_s32(a, b, 9);
   // CHECK: rshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vrshrn_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vrshrn_high_n_s64
+  // CHECK-LABEL: test_vrshrn_high_n_s64
   return vrshrn_high_n_s64(a, b, 19);
   // CHECK: rshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x16_t test_vrshrn_high_n_u16(uint8x8_t a, uint16x8_t b) {
-  // CHECK: test_vrshrn_high_n_u16
+  // CHECK-LABEL: test_vrshrn_high_n_u16
   return vrshrn_high_n_u16(a, b, 3);
   // CHECK: rshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x8_t test_vrshrn_high_n_u32(uint16x4_t a, uint32x4_t b) {
-  // CHECK: test_vrshrn_high_n_u32
+  // CHECK-LABEL: test_vrshrn_high_n_u32
   return vrshrn_high_n_u32(a, b, 9);
   // CHECK: rshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x4_t test_vrshrn_high_n_u64(uint32x2_t a, uint64x2_t b) {
-  // CHECK: test_vrshrn_high_n_u64
+  // CHECK-LABEL: test_vrshrn_high_n_u64
   return vrshrn_high_n_u64(a, b, 19);
   // CHECK: rshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int8x8_t test_vqrshrun_n_s16(int16x8_t a) {
-  // CHECK: test_vqrshrun_n_s16
+  // CHECK-LABEL: test_vqrshrun_n_s16
   return vqrshrun_n_s16(a, 3);
   // CHECK: sqrshrun {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vqrshrun_n_s32(int32x4_t a) {
-  // CHECK: test_vqrshrun_n_s32
+  // CHECK-LABEL: test_vqrshrun_n_s32
   return vqrshrun_n_s32(a, 9);
   // CHECK: sqrshrun {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vqrshrun_n_s64(int64x2_t a) {
-  // CHECK: test_vqrshrun_n_s64
+  // CHECK-LABEL: test_vqrshrun_n_s64
   return vqrshrun_n_s64(a, 19);
   // CHECK: sqrshrun {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vqrshrun_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqrshrun_high_n_s16
+  // CHECK-LABEL: test_vqrshrun_high_n_s16
   return vqrshrun_high_n_s16(a, b, 3);
   // CHECK: sqrshrun2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vqrshrun_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqrshrun_high_n_s32
+  // CHECK-LABEL: test_vqrshrun_high_n_s32
   return vqrshrun_high_n_s32(a, b, 9);
   // CHECK: sqrshrun2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vqrshrun_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqrshrun_high_n_s64
+  // CHECK-LABEL: test_vqrshrun_high_n_s64
   return vqrshrun_high_n_s64(a, b, 19);
   // CHECK: sqrshrun2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int8x8_t test_vqshrn_n_s16(int16x8_t a) {
-  // CHECK: test_vqshrn_n_s16
+  // CHECK-LABEL: test_vqshrn_n_s16
   return vqshrn_n_s16(a, 3);
   // CHECK: sqshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vqshrn_n_s32(int32x4_t a) {
-  // CHECK: test_vqshrn_n_s32
+  // CHECK-LABEL: test_vqshrn_n_s32
   return vqshrn_n_s32(a, 9);
   // CHECK: sqshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vqshrn_n_s64(int64x2_t a) {
-  // CHECK: test_vqshrn_n_s64
+  // CHECK-LABEL: test_vqshrn_n_s64
   return vqshrn_n_s64(a, 19);
   // CHECK: sqshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x8_t test_vqshrn_n_u16(uint16x8_t a) {
-  // CHECK: test_vqshrn_n_u16
+  // CHECK-LABEL: test_vqshrn_n_u16
   return vqshrn_n_u16(a, 3);
   // CHECK: uqshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x4_t test_vqshrn_n_u32(uint32x4_t a) {
-  // CHECK: test_vqshrn_n_u32
+  // CHECK-LABEL: test_vqshrn_n_u32
   return vqshrn_n_u32(a, 9);
   // CHECK: uqshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x2_t test_vqshrn_n_u64(uint64x2_t a) {
-  // CHECK: test_vqshrn_n_u64
+  // CHECK-LABEL: test_vqshrn_n_u64
   return vqshrn_n_u64(a, 19);
   // CHECK: uqshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vqshrn_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqshrn_high_n_s16
+  // CHECK-LABEL: test_vqshrn_high_n_s16
   return vqshrn_high_n_s16(a, b, 3);
   // CHECK: sqshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vqshrn_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqshrn_high_n_s32
+  // CHECK-LABEL: test_vqshrn_high_n_s32
   return vqshrn_high_n_s32(a, b, 9);
   // CHECK: sqshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vqshrn_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqshrn_high_n_s64
+  // CHECK-LABEL: test_vqshrn_high_n_s64
   return vqshrn_high_n_s64(a, b, 19);
   // CHECK: sqshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x16_t test_vqshrn_high_n_u16(uint8x8_t a, uint16x8_t b) {
-  // CHECK: test_vqshrn_high_n_u16
+  // CHECK-LABEL: test_vqshrn_high_n_u16
   return vqshrn_high_n_u16(a, b, 3);
   // CHECK: uqshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x8_t test_vqshrn_high_n_u32(uint16x4_t a, uint32x4_t b) {
-  // CHECK: test_vqshrn_high_n_u32
+  // CHECK-LABEL: test_vqshrn_high_n_u32
   return vqshrn_high_n_u32(a, b, 9);
   // CHECK: uqshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x4_t test_vqshrn_high_n_u64(uint32x2_t a, uint64x2_t b) {
-  // CHECK: test_vqshrn_high_n_u64
+  // CHECK-LABEL: test_vqshrn_high_n_u64
   return vqshrn_high_n_u64(a, b, 19);
   // CHECK: uqshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int8x8_t test_vqrshrn_n_s16(int16x8_t a) {
-  // CHECK: test_vqrshrn_n_s16
+  // CHECK-LABEL: test_vqrshrn_n_s16
   return vqrshrn_n_s16(a, 3);
   // CHECK: sqrshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 int16x4_t test_vqrshrn_n_s32(int32x4_t a) {
-  // CHECK: test_vqrshrn_n_s32
+  // CHECK-LABEL: test_vqrshrn_n_s32
   return vqrshrn_n_s32(a, 9);
   // CHECK: sqrshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 int32x2_t test_vqrshrn_n_s64(int64x2_t a) {
-  // CHECK: test_vqrshrn_n_s64
+  // CHECK-LABEL: test_vqrshrn_n_s64
   return vqrshrn_n_s64(a, 19);
   // CHECK: sqrshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x8_t test_vqrshrn_n_u16(uint16x8_t a) {
-  // CHECK: test_vqrshrn_n_u16
+  // CHECK-LABEL: test_vqrshrn_n_u16
   return vqrshrn_n_u16(a, 3);
   // CHECK: uqrshrn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x4_t test_vqrshrn_n_u32(uint32x4_t a) {
-  // CHECK: test_vqrshrn_n_u32
+  // CHECK-LABEL: test_vqrshrn_n_u32
   return vqrshrn_n_u32(a, 9);
   // CHECK: uqrshrn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x2_t test_vqrshrn_n_u64(uint64x2_t a) {
-  // CHECK: test_vqrshrn_n_u64
+  // CHECK-LABEL: test_vqrshrn_n_u64
   return vqrshrn_n_u64(a, 19);
   // CHECK: uqrshrn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, #19
 }
 
 int8x16_t test_vqrshrn_high_n_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqrshrn_high_n_s16
+  // CHECK-LABEL: test_vqrshrn_high_n_s16
   return vqrshrn_high_n_s16(a, b, 3);
   // CHECK: sqrshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 int16x8_t test_vqrshrn_high_n_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqrshrn_high_n_s32
+  // CHECK-LABEL: test_vqrshrn_high_n_s32
   return vqrshrn_high_n_s32(a, b, 9);
   // CHECK: sqrshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 int32x4_t test_vqrshrn_high_n_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqrshrn_high_n_s64
+  // CHECK-LABEL: test_vqrshrn_high_n_s64
   return vqrshrn_high_n_s64(a, b, 19);
   // CHECK: sqrshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 uint8x16_t test_vqrshrn_high_n_u16(uint8x8_t a, uint16x8_t b) {
-  // CHECK: test_vqrshrn_high_n_u16
+  // CHECK-LABEL: test_vqrshrn_high_n_u16
   return vqrshrn_high_n_u16(a, b, 3);
   // CHECK: uqrshrn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, #3
 }
 
 uint16x8_t test_vqrshrn_high_n_u32(uint16x4_t a, uint32x4_t b) {
-  // CHECK: test_vqrshrn_high_n_u32
+  // CHECK-LABEL: test_vqrshrn_high_n_u32
   return vqrshrn_high_n_u32(a, b, 9);
   // CHECK: uqrshrn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, #9
 }
 
 uint32x4_t test_vqrshrn_high_n_u64(uint32x2_t a, uint64x2_t b) {
-  // CHECK: test_vqrshrn_high_n_u64
+  // CHECK-LABEL: test_vqrshrn_high_n_u64
   return vqrshrn_high_n_u64(a, b, 19);
   // CHECK: uqrshrn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, #19
 }
 
 int16x8_t test_vshll_n_s8(int8x8_t a) {
-// CHECK: test_vshll_n_s8
+// CHECK-LABEL: test_vshll_n_s8
   return vshll_n_s8(a, 3);
 // CHECK: sshll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #3
 }
 
 int32x4_t test_vshll_n_s16(int16x4_t a) {
-// CHECK: test_vshll_n_s16
+// CHECK-LABEL: test_vshll_n_s16
   return vshll_n_s16(a, 9);
 // CHECK: sshll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #9
 }
 
 int64x2_t test_vshll_n_s32(int32x2_t a) {
-// CHECK: test_vshll_n_s32
+// CHECK-LABEL: test_vshll_n_s32
   return vshll_n_s32(a, 19);
 // CHECK: sshll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #19
 }
 
 uint16x8_t test_vshll_n_u8(uint8x8_t a) {
-// CHECK: test_vshll_n_u8
+// CHECK-LABEL: test_vshll_n_u8
   return vshll_n_u8(a, 3);
 // CHECK: ushll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #3
 }
 
 uint32x4_t test_vshll_n_u16(uint16x4_t a) {
-// CHECK: test_vshll_n_u16
+// CHECK-LABEL: test_vshll_n_u16
   return vshll_n_u16(a, 9);
 // CHECK: ushll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #9
 }
 
 uint64x2_t test_vshll_n_u32(uint32x2_t a) {
-// CHECK: test_vshll_n_u32
+// CHECK-LABEL: test_vshll_n_u32
   return vshll_n_u32(a, 19);
 // CHECK: ushll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #19
 }
 
 int16x8_t test_vshll_high_n_s8(int8x16_t a) {
-// CHECK: test_vshll_high_n_s8
+// CHECK-LABEL: test_vshll_high_n_s8
   return vshll_high_n_s8(a, 3);
 // CHECK: sshll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #3
 }
 
 int32x4_t test_vshll_high_n_s16(int16x8_t a) {
-// CHECK: test_vshll_high_n_s16
+// CHECK-LABEL: test_vshll_high_n_s16
   return vshll_high_n_s16(a, 9);
 // CHECK: sshll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #9
 }
 
 int64x2_t test_vshll_high_n_s32(int32x4_t a) {
-// CHECK: test_vshll_high_n_s32
+// CHECK-LABEL: test_vshll_high_n_s32
   return vshll_high_n_s32(a, 19);
 // CHECK: sshll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #19
 }
 
 uint16x8_t test_vshll_high_n_u8(uint8x16_t a) {
-// CHECK: test_vshll_high_n_u8
+// CHECK-LABEL: test_vshll_high_n_u8
   return vshll_high_n_u8(a, 3);
 // CHECK: ushll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #3
 }
 
 uint32x4_t test_vshll_high_n_u16(uint16x8_t a) {
-// CHECK: test_vshll_high_n_u16
+// CHECK-LABEL: test_vshll_high_n_u16
   return vshll_high_n_u16(a, 9);
 // CHECK: ushll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #9
 }
 
 uint64x2_t test_vshll_high_n_u32(uint32x4_t a) {
-// CHECK: test_vshll_high_n_u32
+// CHECK-LABEL: test_vshll_high_n_u32
   return vshll_high_n_u32(a, 19);
 // CHECK: ushll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #19
 }
 
 int16x8_t test_vmovl_s8(int8x8_t a) {
-// CHECK: test_vmovl_s8
+// CHECK-LABEL: test_vmovl_s8
   return vmovl_s8(a);
 // CHECK: sshll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #0
 }
 
 int32x4_t test_vmovl_s16(int16x4_t a) {
-// CHECK: test_vmovl_s16
+// CHECK-LABEL: test_vmovl_s16
   return vmovl_s16(a);
 // CHECK: sshll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #0
 }
 
 int64x2_t test_vmovl_s32(int32x2_t a) {
-// CHECK: test_vmovl_s32
+// CHECK-LABEL: test_vmovl_s32
   return vmovl_s32(a);
 // CHECK: sshll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #0
 }
 
 uint16x8_t test_vmovl_u8(uint8x8_t a) {
-// CHECK: test_vmovl_u8
+// CHECK-LABEL: test_vmovl_u8
   return vmovl_u8(a);
 // CHECK: ushll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #0
 }
 
 uint32x4_t test_vmovl_u16(uint16x4_t a) {
-// CHECK: test_vmovl_u16
+// CHECK-LABEL: test_vmovl_u16
   return vmovl_u16(a);
 // CHECK: ushll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #0
 }
 
 uint64x2_t test_vmovl_u32(uint32x2_t a) {
-// CHECK: test_vmovl_u32
+// CHECK-LABEL: test_vmovl_u32
   return vmovl_u32(a);
 // CHECK: ushll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #0
 }
 
 int16x8_t test_vmovl_high_s8(int8x16_t a) {
-// CHECK: test_vmovl_high_s8
+// CHECK-LABEL: test_vmovl_high_s8
   return vmovl_high_s8(a);
 // CHECK: sshll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #0
 }
 
 int32x4_t test_vmovl_high_s16(int16x8_t a) {
-// CHECK: test_vmovl_high_s16
+// CHECK-LABEL: test_vmovl_high_s16
   return vmovl_high_s16(a);
 // CHECK: sshll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #0
 }
 
 int64x2_t test_vmovl_high_s32(int32x4_t a) {
-// CHECK: test_vmovl_high_s32
+// CHECK-LABEL: test_vmovl_high_s32
   return vmovl_high_s32(a);
 // CHECK: sshll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #0
 }
 
 uint16x8_t test_vmovl_high_u8(uint8x16_t a) {
-// CHECK: test_vmovl_high_u8
+// CHECK-LABEL: test_vmovl_high_u8
   return vmovl_high_u8(a);
 // CHECK: ushll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #0
 }
 
 uint32x4_t test_vmovl_high_u16(uint16x8_t a) {
-// CHECK: test_vmovl_high_u16
+// CHECK-LABEL: test_vmovl_high_u16
   return vmovl_high_u16(a);
 // CHECK: ushll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #0
 }
 
 uint64x2_t test_vmovl_high_u32(uint32x4_t a) {
-// CHECK: test_vmovl_high_u32
+// CHECK-LABEL: test_vmovl_high_u32
   return vmovl_high_u32(a);
 // CHECK: ushll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #0
 }
 
 float32x2_t test_vcvt_n_f32_s32(int32x2_t a) {
-  // CHECK: test_vcvt_n_f32_s32
+  // CHECK-LABEL: test_vcvt_n_f32_s32
   return vcvt_n_f32_s32(a, 31);
   // CHECK: scvtf {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #31
 }
 
 float32x4_t test_vcvtq_n_f32_s32(int32x4_t a) {
-  // CHECK: test_vcvtq_n_f32_s32
+  // CHECK-LABEL: test_vcvtq_n_f32_s32
   return vcvtq_n_f32_s32(a, 31);
   // CHECK: scvtf {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #31
 }
 
 float64x2_t test_vcvtq_n_f64_s64(int64x2_t a) {
-  // CHECK: test_vcvtq_n_f64_s64
+  // CHECK-LABEL: test_vcvtq_n_f64_s64
   return vcvtq_n_f64_s64(a, 50);
   // CHECK: scvtf {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #50
 }
 
 float32x2_t test_vcvt_n_f32_u32(uint32x2_t a) {
-  // CHECK: test_vcvt_n_f32_u32
+  // CHECK-LABEL: test_vcvt_n_f32_u32
   return vcvt_n_f32_u32(a, 31);
   // CHECK: ucvtf {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #31
 }
 
 float32x4_t test_vcvtq_n_f32_u32(uint32x4_t a) {
-  // CHECK: test_vcvtq_n_f32_u32
+  // CHECK-LABEL: test_vcvtq_n_f32_u32
   return vcvtq_n_f32_u32(a, 31);
   // CHECK: ucvtf {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #31
 }
 
 float64x2_t test_vcvtq_n_f64_u64(uint64x2_t a) {
-  // CHECK: test_vcvtq_n_f64_u64
+  // CHECK-LABEL: test_vcvtq_n_f64_u64
   return vcvtq_n_f64_u64(a, 50);
   // CHECK: ucvtf {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #50
 }
 
 int32x2_t test_vcvt_n_s32_f32(float32x2_t a) {
-  // CHECK: test_vcvt_n_s32_f32
+  // CHECK-LABEL: test_vcvt_n_s32_f32
   return vcvt_n_s32_f32(a, 31);
   // CHECK: fcvtzs {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #31
 }
 
 int32x4_t test_vcvtq_n_s32_f32(float32x4_t a) {
-  // CHECK: test_vcvtq_n_s32_f32
+  // CHECK-LABEL: test_vcvtq_n_s32_f32
   return vcvtq_n_s32_f32(a, 31);
   // CHECK: fcvtzs {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #31
 }
 
 int64x2_t test_vcvtq_n_s64_f64(float64x2_t a) {
-  // CHECK: test_vcvtq_n_s64_f64
+  // CHECK-LABEL: test_vcvtq_n_s64_f64
   return vcvtq_n_s64_f64(a, 50);
   // CHECK: fcvtzs {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #50
 }
 
 uint32x2_t test_vcvt_n_u32_f32(float32x2_t a) {
-  // CHECK: test_vcvt_n_u32_f32
+  // CHECK-LABEL: test_vcvt_n_u32_f32
   return vcvt_n_u32_f32(a, 31);
   // CHECK: fcvtzu {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #31
 }
 
 uint32x4_t test_vcvtq_n_u32_f32(float32x4_t a) {
-  // CHECK: test_vcvt_n_u32_f32
+  // CHECK-LABEL: test_vcvtq_n_u32_f32
   return vcvtq_n_u32_f32(a, 31);
   // CHECK: fcvtzu {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #31
 }
 
 uint64x2_t test_vcvtq_n_u64_f64(float64x2_t a) {
-  // CHECK: test_vcvtq_n_u64_f64
+  // CHECK-LABEL: test_vcvtq_n_u64_f64
   return vcvtq_n_u64_f64(a, 50);
   // CHECK: fcvtzu {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #50
 }
 
 int16x8_t test_vaddl_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vaddl_s8
+  // CHECK-LABEL: test_vaddl_s8
   return vaddl_s8(a, b);
-  // CHECK: saddl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: saddl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int32x4_t test_vaddl_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vaddl_s16
+  // CHECK-LABEL: test_vaddl_s16
   return vaddl_s16(a, b);
-  // CHECK: saddl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: saddl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vaddl_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vaddl_s32
+  // CHECK-LABEL: test_vaddl_s32
   return vaddl_s32(a, b);
-  // CHECK: saddl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: saddl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint16x8_t test_vaddl_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vaddl_u8
+  // CHECK-LABEL: test_vaddl_u8
   return vaddl_u8(a, b);
-  // CHECK: uaddl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: uaddl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint32x4_t test_vaddl_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vaddl_u16
+  // CHECK-LABEL: test_vaddl_u16
   return vaddl_u16(a, b);
-  // CHECK: uaddl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: uaddl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint64x2_t test_vaddl_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vaddl_u32
+  // CHECK-LABEL: test_vaddl_u32
   return vaddl_u32(a, b);
-  // CHECK: uaddl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: uaddl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vaddl_high_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vaddl_high_s8
+  // CHECK-LABEL: test_vaddl_high_s8
   return vaddl_high_s8(a, b);
-  // CHECK: saddl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: saddl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vaddl_high_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vaddl_high_s16
+  // CHECK-LABEL: test_vaddl_high_s16
   return vaddl_high_s16(a, b);
-  // CHECK: saddl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: saddl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vaddl_high_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vaddl_high_s32
+  // CHECK-LABEL: test_vaddl_high_s32
   return vaddl_high_s32(a, b);
-  // CHECK: saddl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: saddl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint16x8_t test_vaddl_high_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vaddl_high_u8
+  // CHECK-LABEL: test_vaddl_high_u8
   return vaddl_high_u8(a, b);
-  // CHECK: uaddl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: uaddl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint32x4_t test_vaddl_high_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vaddl_high_u16
+  // CHECK-LABEL: test_vaddl_high_u16
   return vaddl_high_u16(a, b);
-  // CHECK: uaddl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: uaddl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint64x2_t test_vaddl_high_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vaddl_high_u32
+  // CHECK-LABEL: test_vaddl_high_u32
   return vaddl_high_u32(a, b);
-  // CHECK: uaddl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: uaddl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vaddw_s8(int16x8_t a, int8x8_t b) {
-  // CHECK: test_vaddw_s8
+  // CHECK-LABEL: test_vaddw_s8
   return vaddw_s8(a, b);
-  // CHECK: saddw {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.8b
+  // CHECK: saddw {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8b
 }
 
 int32x4_t test_vaddw_s16(int32x4_t a, int16x4_t b) {
-  // CHECK: test_vaddw_s16
+  // CHECK-LABEL: test_vaddw_s16
   return vaddw_s16(a, b);
-  // CHECK: saddw {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.4h
+  // CHECK: saddw {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vaddw_s32(int64x2_t a, int32x2_t b) {
-  // CHECK: test_vaddw_s32
+  // CHECK-LABEL: test_vaddw_s32
   return vaddw_s32(a, b);
-  // CHECK: saddw {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.2s
+  // CHECK: saddw {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2s
 }
 
 uint16x8_t test_vaddw_u8(uint16x8_t a, uint8x8_t b) {
-  // CHECK: test_vaddw_u8
+  // CHECK-LABEL: test_vaddw_u8
   return vaddw_u8(a, b);
-  // CHECK: uaddw {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.8b
+  // CHECK: uaddw {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8b
 }
 
 uint32x4_t test_vaddw_u16(uint32x4_t a, uint16x4_t b) {
-  // CHECK: test_vaddw_u16
+  // CHECK-LABEL: test_vaddw_u16
   return vaddw_u16(a, b);
-  // CHECK: uaddw {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.4h
+  // CHECK: uaddw {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4h
 }
 
 uint64x2_t test_vaddw_u32(uint64x2_t a, uint32x2_t b) {
-  // CHECK: test_vaddw_u32
+  // CHECK-LABEL: test_vaddw_u32
   return vaddw_u32(a, b);
-  // CHECK: uaddw {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.2s
+  // CHECK: uaddw {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vaddw_high_s8(int16x8_t a, int8x16_t b) {
-  // CHECK: test_vaddw_high_s8
+  // CHECK-LABEL: test_vaddw_high_s8
   return vaddw_high_s8(a, b);
-  // CHECK: saddw2 {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.16b
+  // CHECK: saddw2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vaddw_high_s16(int32x4_t a, int16x8_t b) {
-  // CHECK: test_vaddw_high_s16
+  // CHECK-LABEL: test_vaddw_high_s16
   return vaddw_high_s16(a, b);
-  // CHECK: saddw2 {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.8h
+  // CHECK: saddw2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vaddw_high_s32(int64x2_t a, int32x4_t b) {
-  // CHECK: test_vaddw_high_s32
+  // CHECK-LABEL: test_vaddw_high_s32
   return vaddw_high_s32(a, b);
-  // CHECK: saddw2 {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.4s
+  // CHECK: saddw2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.4s
 }
 
 uint16x8_t test_vaddw_high_u8(uint16x8_t a, uint8x16_t b) {
-  // CHECK: test_vaddw_high_u8
+  // CHECK-LABEL: test_vaddw_high_u8
   return vaddw_high_u8(a, b);
-  // CHECK: uaddw2 {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.16b
+  // CHECK: uaddw2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.16b
 }
 
 uint32x4_t test_vaddw_high_u16(uint32x4_t a, uint16x8_t b) {
-  // CHECK: test_vaddw_high_u16
+  // CHECK-LABEL: test_vaddw_high_u16
   return vaddw_high_u16(a, b);
-  // CHECK: uaddw2 {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.8h
+  // CHECK: uaddw2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.8h
 }
 
 uint64x2_t test_vaddw_high_u32(uint64x2_t a, uint32x4_t b) {
-  // CHECK: test_vaddw_high_u32
+  // CHECK-LABEL: test_vaddw_high_u32
   return vaddw_high_u32(a, b);
-  // CHECK: uaddw2 {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.4s
+  // CHECK: uaddw2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vsubl_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vsubl_s8
+  // CHECK-LABEL: test_vsubl_s8
   return vsubl_s8(a, b);
-  // CHECK: ssubl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: ssubl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int32x4_t test_vsubl_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vsubl_s16
+  // CHECK-LABEL: test_vsubl_s16
   return vsubl_s16(a, b);
-  // CHECK: ssubl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: ssubl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vsubl_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vsubl_s32
+  // CHECK-LABEL: test_vsubl_s32
   return vsubl_s32(a, b);
-  // CHECK: ssubl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: ssubl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 uint16x8_t test_vsubl_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vsubl_u8
+  // CHECK-LABEL: test_vsubl_u8
   return vsubl_u8(a, b);
-  // CHECK: usubl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: usubl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint32x4_t test_vsubl_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vsubl_u16
+  // CHECK-LABEL: test_vsubl_u16
   return vsubl_u16(a, b);
-  // CHECK: usubl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: usubl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint64x2_t test_vsubl_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vsubl_u32
+  // CHECK-LABEL: test_vsubl_u32
   return vsubl_u32(a, b);
-  // CHECK: usubl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: usubl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vsubl_high_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vsubl_high_s8
+  // CHECK-LABEL: test_vsubl_high_s8
   return vsubl_high_s8(a, b);
-  // CHECK: ssubl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: ssubl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vsubl_high_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsubl_high_s16
+  // CHECK-LABEL: test_vsubl_high_s16
   return vsubl_high_s16(a, b);
-  // CHECK: ssubl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: ssubl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vsubl_high_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsubl_high_s32
+  // CHECK-LABEL: test_vsubl_high_s32
   return vsubl_high_s32(a, b);
-  // CHECK: ssubl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: ssubl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint16x8_t test_vsubl_high_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vsubl_high_u8
+  // CHECK-LABEL: test_vsubl_high_u8
   return vsubl_high_u8(a, b);
-  // CHECK: usubl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: usubl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint32x4_t test_vsubl_high_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vsubl_high_u16
+  // CHECK-LABEL: test_vsubl_high_u16
   return vsubl_high_u16(a, b);
-  // CHECK: usubl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: usubl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint64x2_t test_vsubl_high_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vsubl_high_u32
+  // CHECK-LABEL: test_vsubl_high_u32
   return vsubl_high_u32(a, b);
-  // CHECK: usubl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: usubl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vsubw_s8(int16x8_t a, int8x8_t b) {
-  // CHECK: test_vsubw_s8
+  // CHECK-LABEL: test_vsubw_s8
   return vsubw_s8(a, b);
-  // CHECK: ssubw {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.8b
+  // CHECK: ssubw {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8b
 }
 
 int32x4_t test_vsubw_s16(int32x4_t a, int16x4_t b) {
-  // CHECK: test_vsubw_s16
+  // CHECK-LABEL: test_vsubw_s16
   return vsubw_s16(a, b);
-  // CHECK: ssubw {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.4h
+  // CHECK: ssubw {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vsubw_s32(int64x2_t a, int32x2_t b) {
-  // CHECK: test_vsubw_s32
+  // CHECK-LABEL: test_vsubw_s32
   return vsubw_s32(a, b);
-  // CHECK: ssubw {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.2s
+  // CHECK: ssubw {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2s
 }
 
 uint16x8_t test_vsubw_u8(uint16x8_t a, uint8x8_t b) {
-  // CHECK: test_vsubw_u8
+  // CHECK-LABEL: test_vsubw_u8
   return vsubw_u8(a, b);
-  // CHECK: usubw {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.8b
+  // CHECK: usubw {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.8b
 }
 
 uint32x4_t test_vsubw_u16(uint32x4_t a, uint16x4_t b) {
-  // CHECK: test_vsubw_u16
+  // CHECK-LABEL: test_vsubw_u16
   return vsubw_u16(a, b);
-  // CHECK: usubw {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.4h
+  // CHECK: usubw {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.4h
 }
 
 uint64x2_t test_vsubw_u32(uint64x2_t a, uint32x2_t b) {
-  // CHECK: test_vsubw_u32
+  // CHECK-LABEL: test_vsubw_u32
   return vsubw_u32(a, b);
-  // CHECK: usubw {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.2s
+  // CHECK: usubw {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vsubw_high_s8(int16x8_t a, int8x16_t b) {
-  // CHECK: test_vsubw_high_s8
+  // CHECK-LABEL: test_vsubw_high_s8
   return vsubw_high_s8(a, b);
-  // CHECK: ssubw2 {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.16b
+  // CHECK: ssubw2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.16b
 }
 
 int32x4_t test_vsubw_high_s16(int32x4_t a, int16x8_t b) {
-  // CHECK: test_vsubw_high_s16
+  // CHECK-LABEL: test_vsubw_high_s16
   return vsubw_high_s16(a, b);
-  // CHECK: ssubw2 {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.8h
+  // CHECK: ssubw2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vsubw_high_s32(int64x2_t a, int32x4_t b) {
-  // CHECK: test_vsubw_high_s32
+  // CHECK-LABEL: test_vsubw_high_s32
   return vsubw_high_s32(a, b);
-  // CHECK: ssubw2 {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.4s
+  // CHECK: ssubw2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.4s
 }
 
 uint16x8_t test_vsubw_high_u8(uint16x8_t a, uint8x16_t b) {
-  // CHECK: test_vsubw_high_u8
+  // CHECK-LABEL: test_vsubw_high_u8
   return vsubw_high_u8(a, b);
-  // CHECK: usubw2 {{v[0-31]+}}.8h, {{v[0-31]+}}.8h, {{v[0-31]+}}.16b
+  // CHECK: usubw2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, {{v[0-9]+}}.16b
 }
 
 uint32x4_t test_vsubw_high_u16(uint32x4_t a, uint16x8_t b) {
-  // CHECK: test_vsubw_high_u16
+  // CHECK-LABEL: test_vsubw_high_u16
   return vsubw_high_u16(a, b);
-  // CHECK: usubw2 {{v[0-31]+}}.4s, {{v[0-31]+}}.4s, {{v[0-31]+}}.8h
+  // CHECK: usubw2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, {{v[0-9]+}}.8h
 }
 
 uint64x2_t test_vsubw_high_u32(uint64x2_t a, uint32x4_t b) {
-  // CHECK: test_vsubw_high_u32
+  // CHECK-LABEL: test_vsubw_high_u32
   return vsubw_high_u32(a, b);
-  // CHECK: usubw2 {{v[0-31]+}}.2d, {{v[0-31]+}}.2d, {{v[0-31]+}}.4s
+  // CHECK: usubw2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.4s
 }
 
 int8x8_t test_vaddhn_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vaddhn_s16
+  // CHECK-LABEL: test_vaddhn_s16
   return vaddhn_s16(a, b);
-  // CHECK: addhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: addhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x4_t test_vaddhn_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vaddhn_s32
+  // CHECK-LABEL: test_vaddhn_s32
   return vaddhn_s32(a, b);
-  // CHECK: addhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: addhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x2_t test_vaddhn_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vaddhn_s64
+  // CHECK-LABEL: test_vaddhn_s64
   return vaddhn_s64(a, b);
-  // CHECK: addhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: addhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x8_t test_vaddhn_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vaddhn_u16
+  // CHECK-LABEL: test_vaddhn_u16
   return vaddhn_u16(a, b);
-  // CHECK: addhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: addhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x4_t test_vaddhn_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vaddhn_u32
+  // CHECK-LABEL: test_vaddhn_u32
   return vaddhn_u32(a, b);
-  // CHECK: addhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: addhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x2_t test_vaddhn_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vaddhn_u64
+  // CHECK-LABEL: test_vaddhn_u64
   return vaddhn_u64(a, b);
-  // CHECK: addhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: addhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x16_t test_vaddhn_high_s16(int8x8_t r, int16x8_t a, int16x8_t b) {
-  // CHECK: test_vaddhn_high_s16
+  // CHECK-LABEL: test_vaddhn_high_s16
   return vaddhn_high_s16(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: addhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x8_t test_vaddhn_high_s32(int16x4_t r, int32x4_t a, int32x4_t b) {
-  // CHECK: test_vaddhn_high_s32
+  // CHECK-LABEL: test_vaddhn_high_s32
   return vaddhn_high_s32(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: addhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vaddhn_high_s64(int32x2_t r, int64x2_t a, int64x2_t b) {
-  // CHECK: test_vaddhn_high_s64
+  // CHECK-LABEL: test_vaddhn_high_s64
   return vaddhn_high_s64(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: addhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vaddhn_high_u16(uint8x8_t r, uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vaddhn_high_u16
+  // CHECK-LABEL: test_vaddhn_high_u16
   return vaddhn_high_u16(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: addhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x8_t test_vaddhn_high_u32(uint16x4_t r, uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vaddhn_high_u32
+  // CHECK-LABEL: test_vaddhn_high_u32
   return vaddhn_high_u32(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: addhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vaddhn_high_u64(uint32x2_t r, uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vaddhn_high_u64
+  // CHECK-LABEL: test_vaddhn_high_u64
   return vaddhn_high_u64(r, a, b);
-  // CHECK: addhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: addhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vraddhn_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vraddhn_s16
+  // CHECK-LABEL: test_vraddhn_s16
   return vraddhn_s16(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: raddhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x4_t test_vraddhn_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vraddhn_s32
+  // CHECK-LABEL: test_vraddhn_s32
   return vraddhn_s32(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: raddhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x2_t test_vraddhn_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vraddhn_s64
+  // CHECK-LABEL: test_vraddhn_s64
   return vraddhn_s64(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: raddhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x8_t test_vraddhn_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vraddhn_u16
+  // CHECK-LABEL: test_vraddhn_u16
   return vraddhn_u16(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: raddhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x4_t test_vraddhn_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vraddhn_u32
+  // CHECK-LABEL: test_vraddhn_u32
   return vraddhn_u32(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: raddhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x2_t test_vraddhn_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vraddhn_u64
+  // CHECK-LABEL: test_vraddhn_u64
   return vraddhn_u64(a, b);
-  // CHECK: raddhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: raddhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x16_t test_vraddhn_high_s16(int8x8_t r, int16x8_t a, int16x8_t b) {
-  // CHECK: test_vraddhn_high_s16
+  // CHECK-LABEL: test_vraddhn_high_s16
   return vraddhn_high_s16(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: raddhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x8_t test_vraddhn_high_s32(int16x4_t r, int32x4_t a, int32x4_t b) {
-  // CHECK: test_vraddhn_high_s32
+  // CHECK-LABEL: test_vraddhn_high_s32
   return vraddhn_high_s32(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: raddhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vraddhn_high_s64(int32x2_t r, int64x2_t a, int64x2_t b) {
-  // CHECK: test_vraddhn_high_s64
+  // CHECK-LABEL: test_vraddhn_high_s64
   return vraddhn_high_s64(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: raddhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vraddhn_high_u16(uint8x8_t r, uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vraddhn_high_u16
+  // CHECK-LABEL: test_vraddhn_high_u16
   return vraddhn_high_u16(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: raddhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x8_t test_vraddhn_high_u32(uint16x4_t r, uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vraddhn_high_u32
+  // CHECK-LABEL: test_vraddhn_high_u32
   return vraddhn_high_u32(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: raddhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vraddhn_high_u64(uint32x2_t r, uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vraddhn_high_u64
+  // CHECK-LABEL: test_vraddhn_high_u64
   return vraddhn_high_u64(r, a, b);
-  // CHECK: raddhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: raddhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vsubhn_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsubhn_s16
+  // CHECK-LABEL: test_vsubhn_s16
   return vsubhn_s16(a, b);
-  // CHECK: subhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: subhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x4_t test_vsubhn_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsubhn_s32
+  // CHECK-LABEL: test_vsubhn_s32
   return vsubhn_s32(a, b);
-  // CHECK: subhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: subhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x2_t test_vsubhn_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsubhn_s64
+  // CHECK-LABEL: test_vsubhn_s64
   return vsubhn_s64(a, b);
-  // CHECK: subhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: subhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x8_t test_vsubhn_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vsubhn_u16
+  // CHECK-LABEL: test_vsubhn_u16
   return vsubhn_u16(a, b);
-  // CHECK: subhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: subhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x4_t test_vsubhn_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vsubhn_u32
+  // CHECK-LABEL: test_vsubhn_u32
   return vsubhn_u32(a, b);
-  // CHECK: subhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: subhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x2_t test_vsubhn_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vsubhn_u64
+  // CHECK-LABEL: test_vsubhn_u64
   return vsubhn_u64(a, b);
-  // CHECK: subhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: subhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x16_t test_vsubhn_high_s16(int8x8_t r, int16x8_t a, int16x8_t b) {
-  // CHECK: test_vsubhn_high_s16
+  // CHECK-LABEL: test_vsubhn_high_s16
   return vsubhn_high_s16(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: subhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x8_t test_vsubhn_high_s32(int16x4_t r, int32x4_t a, int32x4_t b) {
-  // CHECK: test_vsubhn_high_s32
+  // CHECK-LABEL: test_vsubhn_high_s32
   return vsubhn_high_s32(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: subhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vsubhn_high_s64(int32x2_t r, int64x2_t a, int64x2_t b) {
-  // CHECK: test_vsubhn_high_s64
+  // CHECK-LABEL: test_vsubhn_high_s64
   return vsubhn_high_s64(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: subhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vsubhn_high_u16(uint8x8_t r, uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vsubhn_high_u16
+  // CHECK-LABEL: test_vsubhn_high_u16
   return vsubhn_high_u16(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: subhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x8_t test_vsubhn_high_u32(uint16x4_t r, uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vsubhn_high_u32
+  // CHECK-LABEL: test_vsubhn_high_u32
   return vsubhn_high_u32(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: subhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vsubhn_high_u64(uint32x2_t r, uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vsubhn_high_u64
+  // CHECK-LABEL: test_vsubhn_high_u64
   return vsubhn_high_u64(r, a, b);
-  // CHECK: subhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: subhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x8_t test_vrsubhn_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vrsubhn_s16
+  // CHECK-LABEL: test_vrsubhn_s16
   return vrsubhn_s16(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: rsubhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x4_t test_vrsubhn_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vrsubhn_s32
+  // CHECK-LABEL: test_vrsubhn_s32
   return vrsubhn_s32(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: rsubhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x2_t test_vrsubhn_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vrsubhn_s64
+  // CHECK-LABEL: test_vrsubhn_s64
   return vrsubhn_s64(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: rsubhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x8_t test_vrsubhn_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vrsubhn_u16
+  // CHECK-LABEL: test_vrsubhn_u16
   return vrsubhn_u16(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.8b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: rsubhn {{v[0-9]+}}.8b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x4_t test_vrsubhn_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vrsubhn_u32
+  // CHECK-LABEL: test_vrsubhn_u32
   return vrsubhn_u32(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.4h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: rsubhn {{v[0-9]+}}.4h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x2_t test_vrsubhn_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vrsubhn_u64
+  // CHECK-LABEL: test_vrsubhn_u64
   return vrsubhn_u64(a, b);
-  // CHECK: rsubhn {{v[0-31]+}}.2s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: rsubhn {{v[0-9]+}}.2s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int8x16_t test_vrsubhn_high_s16(int8x8_t r, int16x8_t a, int16x8_t b) {
-  // CHECK: test_vrsubhn_high_s16
+  // CHECK-LABEL: test_vrsubhn_high_s16
   return vrsubhn_high_s16(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: rsubhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int16x8_t test_vrsubhn_high_s32(int16x4_t r, int32x4_t a, int32x4_t b) {
-  // CHECK: test_vrsubhn_high_s32
+  // CHECK-LABEL: test_vrsubhn_high_s32
   return vrsubhn_high_s32(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: rsubhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vrsubhn_high_s64(int32x2_t r, int64x2_t a, int64x2_t b) {
-  // CHECK: test_vrsubhn_high_s64
+  // CHECK-LABEL: test_vrsubhn_high_s64
   return vrsubhn_high_s64(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: rsubhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint8x16_t test_vrsubhn_high_u16(uint8x8_t r, uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vrsubhn_high_u16
+  // CHECK-LABEL: test_vrsubhn_high_u16
   return vrsubhn_high_u16(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.16b, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: rsubhn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint16x8_t test_vrsubhn_high_u32(uint16x4_t r, uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vrsubhn_high_u32
+  // CHECK-LABEL: test_vrsubhn_high_u32
   return vrsubhn_high_u32(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.8h, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: rsubhn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint32x4_t test_vrsubhn_high_u64(uint32x2_t r, uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vrsubhn_high_u64
+  // CHECK-LABEL: test_vrsubhn_high_u64
   return vrsubhn_high_u64(r, a, b);
-  // CHECK: rsubhn2 {{v[0-31]+}}.4s, {{v[0-31]+}}.2d, {{v[0-31]+}}.2d
+  // CHECK: rsubhn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 int16x8_t test_vabdl_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vabdl_s8
+  // CHECK-LABEL: test_vabdl_s8
   return vabdl_s8(a, b);
-  // CHECK: sabdl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: sabdl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int32x4_t test_vabdl_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vabdl_s16
+  // CHECK-LABEL: test_vabdl_s16
   return vabdl_s16(a, b);
-  // CHECK: sabdl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: sabdl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vabdl_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vabdl_s32
+  // CHECK-LABEL: test_vabdl_s32
   return vabdl_s32(a, b);
-  // CHECK: sabdl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: sabdl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 uint16x8_t test_vabdl_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vabdl_u8
+  // CHECK-LABEL: test_vabdl_u8
   return vabdl_u8(a, b);
-  // CHECK: uabdl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: uabdl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint32x4_t test_vabdl_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vabdl_u16
+  // CHECK-LABEL: test_vabdl_u16
   return vabdl_u16(a, b);
-  // CHECK: uabdl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: uabdl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint64x2_t test_vabdl_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vabdl_u32
+  // CHECK-LABEL: test_vabdl_u32
   return vabdl_u32(a, b);
-  // CHECK: uabdl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: uabdl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vabal_s8(int16x8_t a, int8x8_t b, int8x8_t c) {
-  // CHECK: test_vabal_s8
+  // CHECK-LABEL: test_vabal_s8
   return vabal_s8(a, b, c);
-  // CHECK: sabal {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: sabal {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int32x4_t test_vabal_s16(int32x4_t a, int16x4_t b, int16x4_t c) {
-  // CHECK: test_vabal_s16
+  // CHECK-LABEL: test_vabal_s16
   return vabal_s16(a, b, c);
-  // CHECK: sabal {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: sabal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vabal_s32(int64x2_t a, int32x2_t b, int32x2_t c) {
-  // CHECK: test_vabal_s32
+  // CHECK-LABEL: test_vabal_s32
   return vabal_s32(a, b, c);
-  // CHECK: sabal {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: sabal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 uint16x8_t test_vabal_u8(uint16x8_t a, uint8x8_t b, uint8x8_t c) {
-  // CHECK: test_vabal_u8
+  // CHECK-LABEL: test_vabal_u8
   return vabal_u8(a, b, c);
-  // CHECK: uabal {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: uabal {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint32x4_t test_vabal_u16(uint32x4_t a, uint16x4_t b, uint16x4_t c) {
-  // CHECK: test_vabal_u16
+  // CHECK-LABEL: test_vabal_u16
   return vabal_u16(a, b, c);
-  // CHECK: uabal {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: uabal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint64x2_t test_vabal_u32(uint64x2_t a, uint32x2_t b, uint32x2_t c) {
-  // CHECK: test_vabal_u32
+  // CHECK-LABEL: test_vabal_u32
   return vabal_u32(a, b, c);
-  // CHECK: uabal {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: uabal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vabdl_high_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vabdl_high_s8
+  // CHECK-LABEL: test_vabdl_high_s8
   return vabdl_high_s8(a, b);
-  // CHECK: sabdl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: sabdl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int32x4_t test_vabdl_high_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vabdl_high_s16
+  // CHECK-LABEL: test_vabdl_high_s16
   return vabdl_high_s16(a, b);
-  // CHECK: sabdl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: sabdl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vabdl_high_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vabdl_high_s32
+  // CHECK-LABEL: test_vabdl_high_s32
   return vabdl_high_s32(a, b);
-  // CHECK: sabdl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: sabdl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint16x8_t test_vabdl_high_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vabdl_high_u8
+  // CHECK-LABEL: test_vabdl_high_u8
   return vabdl_high_u8(a, b);
-  // CHECK: uabdl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: uabdl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint32x4_t test_vabdl_high_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vabdl_high_u16
+  // CHECK-LABEL: test_vabdl_high_u16
   return vabdl_high_u16(a, b);
-  // CHECK: uabdl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: uabdl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint64x2_t test_vabdl_high_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vabdl_high_u32
+  // CHECK-LABEL: test_vabdl_high_u32
   return vabdl_high_u32(a, b);
-  // CHECK: uabdl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: uabdl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vabal_high_s8(int16x8_t a, int8x16_t b, int8x16_t c) {
-  // CHECK: test_vabal_high_s8
+  // CHECK-LABEL: test_vabal_high_s8
   return vabal_high_s8(a, b, c);
-  // CHECK: sabal2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: sabal2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int32x4_t test_vabal_high_s16(int32x4_t a, int16x8_t b, int16x8_t c) {
-  // CHECK: test_vabal_high_s16
+  // CHECK-LABEL: test_vabal_high_s16
   return vabal_high_s16(a, b, c);
-  // CHECK: sabal2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: sabal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vabal_high_s32(int64x2_t a, int32x4_t b, int32x4_t c) {
-  // CHECK: test_vabal_high_s32
+  // CHECK-LABEL: test_vabal_high_s32
   return vabal_high_s32(a, b, c);
-  // CHECK: sabal2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: sabal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint16x8_t test_vabal_high_u8(uint16x8_t a, uint8x16_t b, uint8x16_t c) {
-  // CHECK: test_vabal_high_u8
+  // CHECK-LABEL: test_vabal_high_u8
   return vabal_high_u8(a, b, c);
-  // CHECK: uabal2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: uabal2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint32x4_t test_vabal_high_u16(uint32x4_t a, uint16x8_t b, uint16x8_t c) {
-  // CHECK: test_vabal_high_u16
+  // CHECK-LABEL: test_vabal_high_u16
   return vabal_high_u16(a, b, c);
-  // CHECK: uabal2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: uabal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint64x2_t test_vabal_high_u32(uint64x2_t a, uint32x4_t b, uint32x4_t c) {
-  // CHECK: test_vabal_high_u32
+  // CHECK-LABEL: test_vabal_high_u32
   return vabal_high_u32(a, b, c);
-  // CHECK: uabal2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: uabal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vmull_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vmull_s8
+  // CHECK-LABEL: test_vmull_s8
   return vmull_s8(a, b);
-  // CHECK: smull {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: smull {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int32x4_t test_vmull_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vmull_s16
+  // CHECK-LABEL: test_vmull_s16
   return vmull_s16(a, b);
-  // CHECK: smull {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: smull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vmull_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vmull_s32
+  // CHECK-LABEL: test_vmull_s32
   return vmull_s32(a, b);
-  // CHECK: smull {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: smull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 uint16x8_t test_vmull_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vmull_u8
+  // CHECK-LABEL: test_vmull_u8
   return vmull_u8(a, b);
-  // CHECK: umull {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: umull {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint32x4_t test_vmull_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vmull_u16
+  // CHECK-LABEL: test_vmull_u16
   return vmull_u16(a, b);
-  // CHECK: umull {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: umull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint64x2_t test_vmull_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vmull_u32
+  // CHECK-LABEL: test_vmull_u32
   return vmull_u32(a, b);
-  // CHECK: umull {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: umull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vmull_high_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vmull_high_s8
+  // CHECK-LABEL: test_vmull_high_s8
   return vmull_high_s8(a, b);
-  // CHECK: smull2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: smull2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int32x4_t test_vmull_high_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vmull_high_s16
+  // CHECK-LABEL: test_vmull_high_s16
   return vmull_high_s16(a, b);
-  // CHECK: smull2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: smull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vmull_high_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vmull_high_s32
+  // CHECK-LABEL: test_vmull_high_s32
   return vmull_high_s32(a, b);
-  // CHECK: smull2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: smull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint16x8_t test_vmull_high_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vmull_high_u8
+  // CHECK-LABEL: test_vmull_high_u8
   return vmull_high_u8(a, b);
-  // CHECK: umull2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: umull2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint32x4_t test_vmull_high_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vmull_high_u16
+  // CHECK-LABEL: test_vmull_high_u16
   return vmull_high_u16(a, b);
-  // CHECK: umull2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: umull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint64x2_t test_vmull_high_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vmull_high_u32
+  // CHECK-LABEL: test_vmull_high_u32
   return vmull_high_u32(a, b);
-  // CHECK: umull2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: umull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vmlal_s8(int16x8_t a, int8x8_t b, int8x8_t c) {
-  // CHECK: test_vmlal_s8
+  // CHECK-LABEL: test_vmlal_s8
   return vmlal_s8(a, b, c);
-  // CHECK: smlal {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: smlal {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int32x4_t test_vmlal_s16(int32x4_t a, int16x4_t b, int16x4_t c) {
-  // CHECK: test_vmlal_s16
+  // CHECK-LABEL: test_vmlal_s16
   return vmlal_s16(a, b, c);
-  // CHECK: smlal {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: smlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vmlal_s32(int64x2_t a, int32x2_t b, int32x2_t c) {
-  // CHECK: test_vmlal_s32
+  // CHECK-LABEL: test_vmlal_s32
   return vmlal_s32(a, b, c);
-  // CHECK: smlal {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: smlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 uint16x8_t test_vmlal_u8(uint16x8_t a, uint8x8_t b, uint8x8_t c) {
-  // CHECK: test_vmlal_u8
+  // CHECK-LABEL: test_vmlal_u8
   return vmlal_u8(a, b, c);
-  // CHECK: umlal {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: umlal {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint32x4_t test_vmlal_u16(uint32x4_t a, uint16x4_t b, uint16x4_t c) {
-  // CHECK: test_vmlal_u16
+  // CHECK-LABEL: test_vmlal_u16
   return vmlal_u16(a, b, c);
-  // CHECK: umlal {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: umlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint64x2_t test_vmlal_u32(uint64x2_t a, uint32x2_t b, uint32x2_t c) {
-  // CHECK: test_vmlal_u32
+  // CHECK-LABEL: test_vmlal_u32
   return vmlal_u32(a, b, c);
-  // CHECK: umlal {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: umlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vmlal_high_s8(int16x8_t a, int8x16_t b, int8x16_t c) {
-  // CHECK: test_vmlal_high_s8
+  // CHECK-LABEL: test_vmlal_high_s8
   return vmlal_high_s8(a, b, c);
-  // CHECK: smlal2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: smlal2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int32x4_t test_vmlal_high_s16(int32x4_t a, int16x8_t b, int16x8_t c) {
-  // CHECK: test_vmlal_high_s16
+  // CHECK-LABEL: test_vmlal_high_s16
   return vmlal_high_s16(a, b, c);
-  // CHECK: smlal2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: smlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vmlal_high_s32(int64x2_t a, int32x4_t b, int32x4_t c) {
-  // CHECK: test_vmlal_high_s32
+  // CHECK-LABEL: test_vmlal_high_s32
   return vmlal_high_s32(a, b, c);
-  // CHECK: smlal2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: smlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint16x8_t test_vmlal_high_u8(uint16x8_t a, uint8x16_t b, uint8x16_t c) {
-  // CHECK: test_vmlal_high_u8
+  // CHECK-LABEL: test_vmlal_high_u8
   return vmlal_high_u8(a, b, c);
-  // CHECK: umlal2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: umlal2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint32x4_t test_vmlal_high_u16(uint32x4_t a, uint16x8_t b, uint16x8_t c) {
-  // CHECK: test_vmlal_high_u16
+  // CHECK-LABEL: test_vmlal_high_u16
   return vmlal_high_u16(a, b, c);
-  // CHECK: umlal2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: umlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint64x2_t test_vmlal_high_u32(uint64x2_t a, uint32x4_t b, uint32x4_t c) {
-  // CHECK: test_vmlal_high_u32
+  // CHECK-LABEL: test_vmlal_high_u32
   return vmlal_high_u32(a, b, c);
-  // CHECK: umlal2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: umlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int16x8_t test_vmlsl_s8(int16x8_t a, int8x8_t b, int8x8_t c) {
-  // CHECK: test_vmlsl_s8
+  // CHECK-LABEL: test_vmlsl_s8
   return vmlsl_s8(a, b, c);
-  // CHECK: smlsl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: smlsl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 int32x4_t test_vmlsl_s16(int32x4_t a, int16x4_t b, int16x4_t c) {
-  // CHECK: test_vmlsl_s16
+  // CHECK-LABEL: test_vmlsl_s16
   return vmlsl_s16(a, b, c);
-  // CHECK: smlsl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: smlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vmlsl_s32(int64x2_t a, int32x2_t b, int32x2_t c) {
-  // CHECK: test_vmlsl_s32
+  // CHECK-LABEL: test_vmlsl_s32
   return vmlsl_s32(a, b, c);
-  // CHECK: smlsl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: smlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 uint16x8_t test_vmlsl_u8(uint16x8_t a, uint8x8_t b, uint8x8_t c) {
-  // CHECK: test_vmlsl_u8
+  // CHECK-LABEL: test_vmlsl_u8
   return vmlsl_u8(a, b, c);
-  // CHECK: umlsl {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: umlsl {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint32x4_t test_vmlsl_u16(uint32x4_t a, uint16x4_t b, uint16x4_t c) {
-  // CHECK: test_vmlsl_u16
+  // CHECK-LABEL: test_vmlsl_u16
   return vmlsl_u16(a, b, c);
-  // CHECK: umlsl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: umlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint64x2_t test_vmlsl_u32(uint64x2_t a, uint32x2_t b, uint32x2_t c) {
-  // CHECK: test_vmlsl_u32
+  // CHECK-LABEL: test_vmlsl_u32
   return vmlsl_u32(a, b, c);
-  // CHECK: umlsl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: umlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int16x8_t test_vmlsl_high_s8(int16x8_t a, int8x16_t b, int8x16_t c) {
-  // CHECK: test_vmlsl_high_s8
+  // CHECK-LABEL: test_vmlsl_high_s8
   return vmlsl_high_s8(a, b, c);
-  // CHECK: smlsl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: smlsl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int32x4_t test_vmlsl_high_s16(int32x4_t a, int16x8_t b, int16x8_t c) {
-  // CHECK: test_vmlsl_high_s16
+  // CHECK-LABEL: test_vmlsl_high_s16
   return vmlsl_high_s16(a, b, c);
-  // CHECK: smlsl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: smlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vmlsl_high_s32(int64x2_t a, int32x4_t b, int32x4_t c) {
-  // CHECK: test_vmlsl_high_s32
+  // CHECK-LABEL: test_vmlsl_high_s32
   return vmlsl_high_s32(a, b, c);
-  // CHECK: smlsl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: smlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint16x8_t test_vmlsl_high_u8(uint16x8_t a, uint8x16_t b, uint8x16_t c) {
-  // CHECK: test_vmlsl_high_u8
+  // CHECK-LABEL: test_vmlsl_high_u8
   return vmlsl_high_u8(a, b, c);
-  // CHECK: umlsl2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: umlsl2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint32x4_t test_vmlsl_high_u16(uint32x4_t a, uint16x8_t b, uint16x8_t c) {
-  // CHECK: test_vmlsl_high_u16
+  // CHECK-LABEL: test_vmlsl_high_u16
   return vmlsl_high_u16(a, b, c);
-  // CHECK: umlsl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: umlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint64x2_t test_vmlsl_high_u32(uint64x2_t a, uint32x4_t b, uint32x4_t c) {
-  // CHECK: test_vmlsl_high_u32
+  // CHECK-LABEL: test_vmlsl_high_u32
   return vmlsl_high_u32(a, b, c);
-  // CHECK: umlsl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: umlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vqdmull_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vqdmull_s16
+  // CHECK-LABEL: test_vqdmull_s16
   return vqdmull_s16(a, b);
-  // CHECK: sqdmull {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: sqdmull {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int64x2_t test_vqdmull_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vqdmull_s32
+  // CHECK-LABEL: test_vqdmull_s32
   return vqdmull_s32(a, b);
-  // CHECK: sqdmull {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: sqdmull {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int32x4_t test_vqdmlal_s16(int32x4_t a, int16x4_t b, int16x4_t c) {
-  // CHECK: test_vqdmlal_s16
+  // CHECK-LABEL: test_vqdmlal_s16
   return vqdmlal_s16(a, b, c);
-  // CHECK: sqdmlal {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: sqdmlal {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vqdmlal_s32(int64x2_t a, int32x2_t b, int32x2_t c) {
-  // CHECK: test_vqdmlal_s32
+  // CHECK-LABEL: test_vqdmlal_s32
   return vqdmlal_s32(a, b, c);
-  // CHECK: sqdmlal {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: sqdmlal {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int32x4_t test_vqdmlsl_s16(int32x4_t a, int16x4_t b, int16x4_t c) {
-  // CHECK: test_vqdmlsl_s16
+  // CHECK-LABEL: test_vqdmlsl_s16
   return vqdmlsl_s16(a, b, c);
-  // CHECK: sqdmlsl {{v[0-31]+}}.4s, {{v[0-31]+}}.4h, {{v[0-31]+}}.4h
+  // CHECK: sqdmlsl {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int64x2_t test_vqdmlsl_s32(int64x2_t a, int32x2_t b, int32x2_t c) {
-  // CHECK: test_vqdmlsl_s32
+  // CHECK-LABEL: test_vqdmlsl_s32
   return vqdmlsl_s32(a, b, c);
-  // CHECK: sqdmlsl {{v[0-31]+}}.2d, {{v[0-31]+}}.2s, {{v[0-31]+}}.2s
+  // CHECK: sqdmlsl {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
 }
 
 int32x4_t test_vqdmull_high_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vqdmull_high_s16
+  // CHECK-LABEL: test_vqdmull_high_s16
   return vqdmull_high_s16(a, b);
-  // CHECK: sqdmull2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: sqdmull2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int64x2_t test_vqdmull_high_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vqdmull_high_s32
+  // CHECK-LABEL: test_vqdmull_high_s32
   return vqdmull_high_s32(a, b);
-  // CHECK: sqdmull2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: sqdmull2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vqdmlal_high_s16(int32x4_t a, int16x8_t b, int16x8_t c) {
-  // CHECK: test_vqdmlal_high_s16
+  // CHECK-LABEL: test_vqdmlal_high_s16
   return vqdmlal_high_s16(a, b, c);
-  // CHECK: sqdmlal2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: sqdmlal2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vqdmlal_high_s32(int64x2_t a, int32x4_t b, int32x4_t c) {
-  // CHECK: test_vqdmlal_high_s32
+  // CHECK-LABEL: test_vqdmlal_high_s32
   return vqdmlal_high_s32(a, b, c);
-  // CHECK: sqdmlal2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: sqdmlal2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int32x4_t test_vqdmlsl_high_s16(int32x4_t a, int16x8_t b, int16x8_t c) {
-  // CHECK: test_vqdmlsl_high_s16
+  // CHECK-LABEL: test_vqdmlsl_high_s16
   return vqdmlsl_high_s16(a, b, c);
-  // CHECK: sqdmlsl2 {{v[0-31]+}}.4s, {{v[0-31]+}}.8h, {{v[0-31]+}}.8h
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int64x2_t test_vqdmlsl_high_s32(int64x2_t a, int32x4_t b, int32x4_t c) {
-  // CHECK: test_vqdmlsl_high_s32
+  // CHECK-LABEL: test_vqdmlsl_high_s32
   return vqdmlsl_high_s32(a, b, c);
-  // CHECK: sqdmlsl2 {{v[0-31]+}}.2d, {{v[0-31]+}}.4s, {{v[0-31]+}}.4s
+  // CHECK: sqdmlsl2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 poly16x8_t test_vmull_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vmull_p8
+  // CHECK-LABEL: test_vmull_p8
   return vmull_p8(a, b);
-  // CHECK: pmull {{v[0-31]+}}.8h, {{v[0-31]+}}.8b, {{v[0-31]+}}.8b
+  // CHECK: pmull {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly16x8_t test_vmull_high_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vmull_high_p8
+  // CHECK-LABEL: test_vmull_high_p8
   return vmull_high_p8(a, b);
-  // CHECK: pmull2 {{v[0-31]+}}.8h, {{v[0-31]+}}.16b, {{v[0-31]+}}.16b
+  // CHECK: pmull2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int64_t test_vaddd_s64(int64_t a, int64_t b) {
-// CHECK: test_vaddd_s64
+// CHECK-LABEL: test_vaddd_s64
   return vaddd_s64(a, b);
-// CHECK: add {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK: add {{[xd][0-9]+}}, {{[xd][0-9]+}}, {{[xd][0-9]+}}
 }
 
 uint64_t test_vaddd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vaddd_u64
+// CHECK-LABEL: test_vaddd_u64
   return vaddd_u64(a, b);
-// CHECK: add {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK: add {{[xd][0-9]+}}, {{[xd][0-9]+}}, {{[xd][0-9]+}}
 }
 
 int64_t test_vsubd_s64(int64_t a, int64_t b) {
-// CHECK: test_vsubd_s64
+// CHECK-LABEL: test_vsubd_s64
   return vsubd_s64(a, b);
-// CHECK: sub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK: sub {{[xd][0-9]+}}, {{[xd][0-9]+}}, {{[xd][0-9]+}}
 }
 
 uint64_t test_vsubd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vsubd_u64
+// CHECK-LABEL: test_vsubd_u64
   return vsubd_u64(a, b);
-// CHECK: sub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK: sub {{[xd][0-9]+}}, {{[xd][0-9]+}}, {{[xd][0-9]+}}
 }
 
 int8_t test_vqaddb_s8(int8_t a, int8_t b) {
-// CHECK: test_vqaddb_s8
+// CHECK-LABEL: test_vqaddb_s8
   return vqaddb_s8(a, b);
-// CHECK: sqadd {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: sqadd {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
 int16_t test_vqaddh_s16(int16_t a, int16_t b) {
-// CHECK: test_vqaddh_s16
+// CHECK-LABEL: test_vqaddh_s16
   return vqaddh_s16(a, b);
-// CHECK: sqadd {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqadd {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 int32_t test_vqadds_s32(int32_t a, int32_t b) {
-// CHECK: test_vqadds_s32
+// CHECK-LABEL: test_vqadds_s32
   return vqadds_s32(a, b);
 // CHECK: sqadd {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 int64_t test_vqaddd_s64(int64_t a, int64_t b) {
-// CHECK: test_vqaddd_s64
+// CHECK-LABEL: test_vqaddd_s64
   return vqaddd_s64(a, b);
 // CHECK: sqadd {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8_t test_vqaddb_u8(uint8_t a, uint8_t b) {
-// CHECK: test_vqaddb_u8
+// CHECK-LABEL: test_vqaddb_u8
   return vqaddb_u8(a, b);
-// CHECK: uqadd {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: uqadd {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
 uint16_t test_vqaddh_u16(uint16_t a, uint16_t b) {
-// CHECK: test_vqaddh_u16
+// CHECK-LABEL: test_vqaddh_u16
   return vqaddh_u16(a, b);
-// CHECK: uqadd {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: uqadd {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 uint32_t test_vqadds_u32(uint32_t a, uint32_t b) {
-// CHECK: test_vqadds_u32
+// CHECK-LABEL: test_vqadds_u32
   return vqadds_u32(a, b);
 // CHECK: uqadd {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 uint64_t test_vqaddd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vqaddd_u64
+// CHECK-LABEL: test_vqaddd_u64
   return vqaddd_u64(a, b);
 // CHECK: uqadd {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int8_t test_vqsubb_s8(int8_t a, int8_t b) {
-// CHECK: test_vqsubb_s8
+// CHECK-LABEL: test_vqsubb_s8
   return vqsubb_s8(a, b);
-// CHECK: sqsub {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: sqsub {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
 int16_t test_vqsubh_s16(int16_t a, int16_t b) {
-// CHECK: test_vqsubh_s16
+// CHECK-LABEL: test_vqsubh_s16
   return vqsubh_s16(a, b);
-// CHECK: sqsub {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqsub {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 int32_t test_vqsubs_s32(int32_t a, int32_t b) {
-  // CHECK: test_vqsubs_s32
+  // CHECK-LABEL: test_vqsubs_s32
   return vqsubs_s32(a, b);
 // CHECK: sqsub {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 int64_t test_vqsubd_s64(int64_t a, int64_t b) {
-// CHECK: test_vqsubd_s64
+// CHECK-LABEL: test_vqsubd_s64
   return vqsubd_s64(a, b);
 // CHECK: sqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint8_t test_vqsubb_u8(uint8_t a, uint8_t b) {
-// CHECK: test_vqsubb_u8
+// CHECK-LABEL: test_vqsubb_u8
   return vqsubb_u8(a, b);
-// CHECK: uqsub {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: uqsub {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
 uint16_t test_vqsubh_u16(uint16_t a, uint16_t b) {
-// CHECK: test_vqsubh_u16
+// CHECK-LABEL: test_vqsubh_u16
   return vqsubh_u16(a, b);
-// CHECK: uqsub {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: uqsub {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 uint32_t test_vqsubs_u32(uint32_t a, uint32_t b) {
-// CHECK: test_vqsubs_u32
+// CHECK-LABEL: test_vqsubs_u32
   return vqsubs_u32(a, b);
 // CHECK: uqsub {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 uint64_t test_vqsubd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vqsubd_u64
+// CHECK-LABEL: test_vqsubd_u64
   return vqsubd_u64(a, b);
 // CHECK: uqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 int64_t test_vshld_s64(int64_t a, int64_t b) {
-// CHECK: test_vshld_s64
+// CHECK-LABEL: test_vshld_s64
   return vshld_s64(a, b);
 // CHECK: sshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64_t test_vshld_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vshld_u64
+// CHECK-LABEL: test_vshld_u64
   return vshld_u64(a, b);
 // CHECK: ushl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vqshlb_s8
+// CHECK-LABEL: test_vqshlb_s8
 int8_t test_vqshlb_s8(int8_t a, int8_t b) {
   return vqshlb_s8(a, b);
-// CHECK: sqshl {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: sqshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
-// CHECK: test_vqshlh_s16
+// CHECK-LABEL: test_vqshlh_s16
 int16_t test_vqshlh_s16(int16_t a, int16_t b) {
   return vqshlh_s16(a, b);
-// CHECK: sqshl {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
-// CHECK: test_vqshls_s32
+// CHECK-LABEL: test_vqshls_s32
 int32_t test_vqshls_s32(int32_t a, int32_t b) {
   return vqshls_s32(a, b);
 // CHECK: sqshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
-// CHECK: test_vqshld_s64
+// CHECK-LABEL: test_vqshld_s64
 int64_t test_vqshld_s64(int64_t a, int64_t b) {
   return vqshld_s64(a, b);
 // CHECK: sqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vqshlb_u8
+// CHECK-LABEL: test_vqshlb_u8
 uint8_t test_vqshlb_u8(uint8_t a, uint8_t b) {
   return vqshlb_u8(a, b);
-// CHECK: uqshl {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: uqshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
-// CHECK: test_vqshlh_u16
+// CHECK-LABEL: test_vqshlh_u16
 uint16_t test_vqshlh_u16(uint16_t a, uint16_t b) {
   return vqshlh_u16(a, b);
-// CHECK: uqshl {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: uqshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
-// CHECK: test_vqshls_u32
+// CHECK-LABEL: test_vqshls_u32
 uint32_t test_vqshls_u32(uint32_t a, uint32_t b) {
   return vqshls_u32(a, b);
 // CHECK: uqshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
-// CHECK: test_vqshld_u64
+// CHECK-LABEL: test_vqshld_u64
 uint64_t test_vqshld_u64(uint64_t a, uint64_t b) {
   return vqshld_u64(a, b);
 // CHECK: uqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vrshld_s64
+// CHECK-LABEL: test_vrshld_s64
 int64_t test_vrshld_s64(int64_t a, int64_t b) {
   return vrshld_s64(a, b);
 // CHECK: srshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 
-// CHECK: test_vrshld_u64
+// CHECK-LABEL: test_vrshld_u64
 uint64_t test_vrshld_u64(uint64_t a, uint64_t b) {
   return vrshld_u64(a, b);
 // CHECK: urshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vqrshlb_s8
+// CHECK-LABEL: test_vqrshlb_s8
 int8_t test_vqrshlb_s8(int8_t a, int8_t b) {
   return vqrshlb_s8(a, b);
-// CHECK: sqrshl {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: sqrshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
-// CHECK: test_vqrshlh_s16
+// CHECK-LABEL: test_vqrshlh_s16
 int16_t test_vqrshlh_s16(int16_t a, int16_t b) {
   return vqrshlh_s16(a, b);
-// CHECK: sqrshl {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqrshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
-// CHECK: test_vqrshls_s32
+// CHECK-LABEL: test_vqrshls_s32
 int32_t test_vqrshls_s32(int32_t a, int32_t b) {
   return vqrshls_s32(a, b);
 // CHECK: sqrshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
-// CHECK: test_vqrshld_s64
+// CHECK-LABEL: test_vqrshld_s64
 int64_t test_vqrshld_s64(int64_t a, int64_t b) {
   return vqrshld_s64(a, b);
 // CHECK: sqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vqrshlb_u8
+// CHECK-LABEL: test_vqrshlb_u8
 uint8_t test_vqrshlb_u8(uint8_t a, uint8_t b) {
   return vqrshlb_u8(a, b);
-// CHECK: uqrshl {{b[0-9]+}}, {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK: uqrshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
 }
 
-// CHECK: test_vqrshlh_u16
+// CHECK-LABEL: test_vqrshlh_u16
 uint16_t test_vqrshlh_u16(uint16_t a, uint16_t b) {
   return vqrshlh_u16(a, b);
-// CHECK: uqrshl {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: uqrshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
-// CHECK: test_vqrshls_u32
+// CHECK-LABEL: test_vqrshls_u32
 uint32_t test_vqrshls_u32(uint32_t a, uint32_t b) {
   return vqrshls_u32(a, b);
 // CHECK: uqrshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
-// CHECK: test_vqrshld_u64
+// CHECK-LABEL: test_vqrshld_u64
 uint64_t test_vqrshld_u64(uint64_t a, uint64_t b) {
   return vqrshld_u64(a, b);
 // CHECK: uqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vpaddd_s64
+// CHECK-LABEL: test_vpaddd_s64
 int64_t test_vpaddd_s64(int64x2_t a) {
   return vpaddd_s64(a);
 // CHECK: addp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vpadds_f32
+// CHECK-LABEL: test_vpadds_f32
 float32_t test_vpadds_f32(float32x2_t a) {
   return vpadds_f32(a);
 // CHECK: faddp {{s[0-9]+}}, {{v[0-9]+}}.2s
 }
 
-// CHECK: test_vpaddd_f64
+// CHECK-LABEL: test_vpaddd_f64
 float64_t test_vpaddd_f64(float64x2_t a) {
   return vpaddd_f64(a);
 // CHECK: faddp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vpmaxnms_f32
+// CHECK-LABEL: test_vpmaxnms_f32
 float32_t test_vpmaxnms_f32(float32x2_t a) {
   return vpmaxnms_f32(a);
 // CHECK: fmaxnmp {{s[0-9]+}}, {{v[0-9]+}}.2s
 }
 
-// CHECK: test_vpmaxnmqd_f64
+// CHECK-LABEL: test_vpmaxnmqd_f64
 float64_t test_vpmaxnmqd_f64(float64x2_t a) {
   return vpmaxnmqd_f64(a);
 // CHECK: fmaxnmp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vpmaxs_f32
+// CHECK-LABEL: test_vpmaxs_f32
 float32_t test_vpmaxs_f32(float32x2_t a) {
   return vpmaxs_f32(a);
 // CHECK: fmaxp {{s[0-9]+}}, {{v[0-9]+}}.2s
 }
 
-// CHECK: test_vpmaxqd_f64
+// CHECK-LABEL: test_vpmaxqd_f64
 float64_t test_vpmaxqd_f64(float64x2_t a) {
   return vpmaxqd_f64(a);
 // CHECK: fmaxp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vpminnms_f32
+// CHECK-LABEL: test_vpminnms_f32
 float32_t test_vpminnms_f32(float32x2_t a) {
   return vpminnms_f32(a);
 // CHECK: fminnmp {{s[0-9]+}}, {{v[0-9]+}}.2s
 }
 
-// CHECK: test_vpminnmqd_f64
+// CHECK-LABEL: test_vpminnmqd_f64
 float64_t test_vpminnmqd_f64(float64x2_t a) {
   return vpminnmqd_f64(a);
 // CHECK: fminnmp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
-// CHECK: test_vpmins_f32
+// CHECK-LABEL: test_vpmins_f32
 float32_t test_vpmins_f32(float32x2_t a) {
   return vpmins_f32(a);
 // CHECK: fminp {{s[0-9]+}}, {{v[0-9]+}}.2s
 }
 
-// CHECK: test_vpminqd_f64
+// CHECK-LABEL: test_vpminqd_f64
 float64_t test_vpminqd_f64(float64x2_t a) {
   return vpminqd_f64(a);
 // CHECK: fminp {{d[0-9]+}}, {{v[0-9]+}}.2d
 }
 
 int16_t test_vqdmulhh_s16(int16_t a, int16_t b) {
-// CHECK: test_vqdmulhh_s16
+// CHECK-LABEL: test_vqdmulhh_s16
   return vqdmulhh_s16(a, b);
-// CHECK: sqdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 int32_t test_vqdmulhs_s32(int32_t a, int32_t b) {
-// CHECK: test_vqdmulhs_s32
+// CHECK-LABEL: test_vqdmulhs_s32
   return vqdmulhs_s32(a, b);
 // CHECK: sqdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 int16_t test_vqrdmulhh_s16(int16_t a, int16_t b) {
-// CHECK: test_vqrdmulhh_s16
+// CHECK-LABEL: test_vqrdmulhh_s16
   return vqrdmulhh_s16(a, b);
-// CHECK: sqrdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK: sqrdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
 }
 
 int32_t test_vqrdmulhs_s32(int32_t a, int32_t b) {
-// CHECK: test_vqrdmulhs_s32
+// CHECK-LABEL: test_vqrdmulhs_s32
   return vqrdmulhs_s32(a, b);
 // CHECK: sqrdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 float32_t test_vmulxs_f32(float32_t a, float32_t b) {
-// CHECK: test_vmulxs_f32
+// CHECK-LABEL: test_vmulxs_f32
   return vmulxs_f32(a, b);
 // CHECK: fmulx {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 float64_t test_vmulxd_f64(float64_t a, float64_t b) {
-// CHECK: test_vmulxd_f64
+// CHECK-LABEL: test_vmulxd_f64
   return vmulxd_f64(a, b);
 // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
+float64x1_t test_vmulx_f64(float64x1_t a, float64x1_t b) {
+// CHECK-LABEL: test_vmulx_f64
+  return vmulx_f64(a, b);
+// CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
 float32_t test_vrecpss_f32(float32_t a, float32_t b) {
-// CHECK: test_vrecpss_f32
+// CHECK-LABEL: test_vrecpss_f32
   return vrecpss_f32(a, b);
 // CHECK: frecps {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 float64_t test_vrecpsd_f64(float64_t a, float64_t b) {
-// CHECK: test_vrecpsd_f64
+// CHECK-LABEL: test_vrecpsd_f64
   return vrecpsd_f64(a, b);
 // CHECK: frecps {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 float32_t test_vrsqrtss_f32(float32_t a, float32_t b) {
-// CHECK: test_vrsqrtss_f32
+// CHECK-LABEL: test_vrsqrtss_f32
   return vrsqrtss_f32(a, b);
 // CHECK: frsqrts {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
 }
 
 float64_t test_vrsqrtsd_f64(float64_t a, float64_t b) {
-// CHECK: test_vrsqrtsd_f64
+// CHECK-LABEL: test_vrsqrtsd_f64
   return vrsqrtsd_f64(a, b);
 // CHECK: frsqrts {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 float32_t test_vcvts_f32_s32(int32_t a) {
-// CHECK: test_vcvts_f32_s32
-// CHECK: scvtf {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvts_f32_s32
+// CHECK: scvtf {{s[0-9]+}}, {{[ws][0-9]+}}
   return vcvts_f32_s32(a);
 }
 
 float64_t test_vcvtd_f64_s64(int64_t a) {
-// CHECK: test_vcvtd_f64_s64
-// CHECK: scvtf {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtd_f64_s64
+// CHECK: scvtf {{d[0-9]+}}, {{[dx][0-9]+}}
   return vcvtd_f64_s64(a);
 }
 
 float32_t test_vcvts_f32_u32(uint32_t a) {
-// CHECK: test_vcvts_f32_u32
-// CHECK: ucvtf {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcvts_f32_u32
+// CHECK: ucvtf {{s[0-9]+}}, {{[ws][0-9]+}}
   return vcvts_f32_u32(a);
 }
 
 float64_t test_vcvtd_f64_u64(uint64_t a) {
-// CHECK: test_vcvtd_f64_u64
-// CHECK: ucvtf {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcvtd_f64_u64
+// CHECK: ucvtf {{d[0-9]+}}, {{[xd][0-9]+}}
   return vcvtd_f64_u64(a);
 }
 
 float32_t test_vrecpes_f32(float32_t a) {
-// CHECK: test_vrecpes_f32
+// CHECK-LABEL: test_vrecpes_f32
 // CHECK: frecpe {{s[0-9]+}}, {{s[0-9]+}}
   return vrecpes_f32(a);
 }
  
 float64_t test_vrecped_f64(float64_t a) {
-// CHECK: test_vrecped_f64
+// CHECK-LABEL: test_vrecped_f64
 // CHECK: frecpe {{d[0-9]+}}, {{d[0-9]+}}
   return vrecped_f64(a);
 }
  
 float32_t test_vrecpxs_f32(float32_t a) {
-// CHECK: test_vrecpxs_f32
+// CHECK-LABEL: test_vrecpxs_f32
 // CHECK: frecpx {{s[0-9]+}}, {{s[0-9]+}}
   return vrecpxs_f32(a);
  }
  
 float64_t test_vrecpxd_f64(float64_t a) {
-// CHECK: test_vrecpxd_f64
+// CHECK-LABEL: test_vrecpxd_f64
 // CHECK: frecpx {{d[0-9]+}}, {{d[0-9]+}}
   return vrecpxd_f64(a);
 }
 
+uint32x2_t test_vrsqrte_u32(uint32x2_t a) {
+// CHECK-LABEL: test_vrsqrte_u32
+// CHECK: ursqrte {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+  return vrsqrte_u32(a);
+}
+
+uint32x4_t test_vrsqrteq_u32(uint32x4_t a) {
+// CHECK-LABEL: test_vrsqrteq_u32
+// CHECK: ursqrte {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+  return vrsqrteq_u32(a);
+}
+
 float32_t test_vrsqrtes_f32(float32_t a) {
 // CHECK: vrsqrtes_f32
 // CHECK: frsqrte {{s[0-9]+}}, {{s[0-9]+}}
@@ -5779,2954 +5822,2746 @@
 }
 
 uint8x16_t test_vld1q_u8(uint8_t const *a) {
-  // CHECK: test_vld1q_u8
+  // CHECK-LABEL: test_vld1q_u8
   return vld1q_u8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.16b}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8_t test_vld1q_u16(uint16_t const *a) {
-  // CHECK: test_vld1q_u16
+  // CHECK-LABEL: test_vld1q_u16
   return vld1q_u16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8h}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4_t test_vld1q_u32(uint32_t const *a) {
-  // CHECK: test_vld1q_u32
+  // CHECK-LABEL: test_vld1q_u32
   return vld1q_u32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4s}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2_t test_vld1q_u64(uint64_t const *a) {
-  // CHECK: test_vld1q_u64
+  // CHECK-LABEL: test_vld1q_u64
   return vld1q_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2d}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16_t test_vld1q_s8(int8_t const *a) {
-  // CHECK: test_vld1q_s8
+  // CHECK-LABEL: test_vld1q_s8
   return vld1q_s8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.16b}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8_t test_vld1q_s16(int16_t const *a) {
-  // CHECK: test_vld1q_s16
+  // CHECK-LABEL: test_vld1q_s16
   return vld1q_s16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8h}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4_t test_vld1q_s32(int32_t const *a) {
-  // CHECK: test_vld1q_s32
+  // CHECK-LABEL: test_vld1q_s32
   return vld1q_s32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4s}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2_t test_vld1q_s64(int64_t const *a) {
-  // CHECK: test_vld1q_s64
+  // CHECK-LABEL: test_vld1q_s64
   return vld1q_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2d}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8_t test_vld1q_f16(float16_t const *a) {
-  // CHECK: test_vld1q_f16
+  // CHECK-LABEL: test_vld1q_f16
   return vld1q_f16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8h}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4_t test_vld1q_f32(float32_t const *a) {
-  // CHECK: test_vld1q_f32
+  // CHECK-LABEL: test_vld1q_f32
   return vld1q_f32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4s}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2_t test_vld1q_f64(float64_t const *a) {
-  // CHECK: test_vld1q_f64
+  // CHECK-LABEL: test_vld1q_f64
   return vld1q_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2d}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16_t test_vld1q_p8(poly8_t const *a) {
-  // CHECK: test_vld1q_p8
+  // CHECK-LABEL: test_vld1q_p8
   return vld1q_p8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.16b}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8_t test_vld1q_p16(poly16_t const *a) {
-  // CHECK: test_vld1q_p16
+  // CHECK-LABEL: test_vld1q_p16
   return vld1q_p16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8h}|ldr q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8_t test_vld1_u8(uint8_t const *a) {
-  // CHECK: test_vld1_u8
+  // CHECK-LABEL: test_vld1_u8
   return vld1_u8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8b}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4_t test_vld1_u16(uint16_t const *a) {
-  // CHECK: test_vld1_u16
+  // CHECK-LABEL: test_vld1_u16
   return vld1_u16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4h}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2_t test_vld1_u32(uint32_t const *a) {
-  // CHECK: test_vld1_u32
+  // CHECK-LABEL: test_vld1_u32
   return vld1_u32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2s}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1_t test_vld1_u64(uint64_t const *a) {
-  // CHECK: test_vld1_u64
+  // CHECK-LABEL: test_vld1_u64
   return vld1_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8_t test_vld1_s8(int8_t const *a) {
-  // CHECK: test_vld1_s8
+  // CHECK-LABEL: test_vld1_s8
   return vld1_s8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8b}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4_t test_vld1_s16(int16_t const *a) {
-  // CHECK: test_vld1_s16
+  // CHECK-LABEL: test_vld1_s16
   return vld1_s16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4h}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2_t test_vld1_s32(int32_t const *a) {
-  // CHECK: test_vld1_s32
+  // CHECK-LABEL: test_vld1_s32
   return vld1_s32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2s}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1_t test_vld1_s64(int64_t const *a) {
-  // CHECK: test_vld1_s64
+  // CHECK-LABEL: test_vld1_s64
   return vld1_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4_t test_vld1_f16(float16_t const *a) {
-  // CHECK: test_vld1_f16
+  // CHECK-LABEL: test_vld1_f16
   return vld1_f16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4h}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2_t test_vld1_f32(float32_t const *a) {
-  // CHECK: test_vld1_f32
+  // CHECK-LABEL: test_vld1_f32
   return vld1_f32(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.2s}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1_t test_vld1_f64(float64_t const *a) {
-  // CHECK: test_vld1_f64
+  // CHECK-LABEL: test_vld1_f64
   return vld1_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8_t test_vld1_p8(poly8_t const *a) {
-  // CHECK: test_vld1_p8
+  // CHECK-LABEL: test_vld1_p8
   return vld1_p8(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.8b}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4_t test_vld1_p16(poly16_t const *a) {
-  // CHECK: test_vld1_p16
+  // CHECK-LABEL: test_vld1_p16
   return vld1_p16(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1 {v[0-9]+.4h}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x2_t test_vld2q_u8(uint8_t const *a) {
-  // CHECK: test_vld2q_u8
+  // CHECK-LABEL: test_vld2q_u8
   return vld2q_u8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x2_t test_vld2q_u16(uint16_t const *a) {
-  // CHECK: test_vld2q_u16
+  // CHECK-LABEL: test_vld2q_u16
   return vld2q_u16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x2_t test_vld2q_u32(uint32_t const *a) {
-  // CHECK: test_vld2q_u32
+  // CHECK-LABEL: test_vld2q_u32
   return vld2q_u32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x2_t test_vld2q_u64(uint64_t const *a) {
-  // CHECK: test_vld2q_u64
+  // CHECK-LABEL: test_vld2q_u64
   return vld2q_u64(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x2_t test_vld2q_s8(int8_t const *a) {
-  // CHECK: test_vld2q_s8
+  // CHECK-LABEL: test_vld2q_s8
   return vld2q_s8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x2_t test_vld2q_s16(int16_t const *a) {
-  // CHECK: test_vld2q_s16
+  // CHECK-LABEL: test_vld2q_s16
   return vld2q_s16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x2_t test_vld2q_s32(int32_t const *a) {
-  // CHECK: test_vld2q_s32
+  // CHECK-LABEL: test_vld2q_s32
   return vld2q_s32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x2_t test_vld2q_s64(int64_t const *a) {
-  // CHECK: test_vld2q_s64
+  // CHECK-LABEL: test_vld2q_s64
   return vld2q_s64(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x2_t test_vld2q_f16(float16_t const *a) {
-  // CHECK: test_vld2q_f16
+  // CHECK-LABEL: test_vld2q_f16
   return vld2q_f16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x2_t test_vld2q_f32(float32_t const *a) {
-  // CHECK: test_vld2q_f32
+  // CHECK-LABEL: test_vld2q_f32
   return vld2q_f32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x2_t test_vld2q_f64(float64_t const *a) {
-  // CHECK: test_vld2q_f64
+  // CHECK-LABEL: test_vld2q_f64
   return vld2q_f64(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x2_t test_vld2q_p8(poly8_t const *a) {
-  // CHECK: test_vld2q_p8
+  // CHECK-LABEL: test_vld2q_p8
   return vld2q_p8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x2_t test_vld2q_p16(poly16_t const *a) {
-  // CHECK: test_vld2q_p16
+  // CHECK-LABEL: test_vld2q_p16
   return vld2q_p16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x2_t test_vld2_u8(uint8_t const *a) {
-  // CHECK: test_vld2_u8
+  // CHECK-LABEL: test_vld2_u8
   return vld2_u8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x2_t test_vld2_u16(uint16_t const *a) {
-  // CHECK: test_vld2_u16
+  // CHECK-LABEL: test_vld2_u16
   return vld2_u16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x2_t test_vld2_u32(uint32_t const *a) {
-  // CHECK: test_vld2_u32
+  // CHECK-LABEL: test_vld2_u32
   return vld2_u32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x2_t test_vld2_u64(uint64_t const *a) {
-  // CHECK: test_vld2_u64
+  // CHECK-LABEL: test_vld2_u64
   return vld2_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x2_t test_vld2_s8(int8_t const *a) {
-  // CHECK: test_vld2_s8
+  // CHECK-LABEL: test_vld2_s8
   return vld2_s8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x2_t test_vld2_s16(int16_t const *a) {
-  // CHECK: test_vld2_s16
+  // CHECK-LABEL: test_vld2_s16
   return vld2_s16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x2_t test_vld2_s32(int32_t const *a) {
-  // CHECK: test_vld2_s32
+  // CHECK-LABEL: test_vld2_s32
   return vld2_s32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x2_t test_vld2_s64(int64_t const *a) {
-  // CHECK: test_vld2_s64
+  // CHECK-LABEL: test_vld2_s64
   return vld2_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x2_t test_vld2_f16(float16_t const *a) {
-  // CHECK: test_vld2_f16
+  // CHECK-LABEL: test_vld2_f16
   return vld2_f16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x2_t test_vld2_f32(float32_t const *a) {
-  // CHECK: test_vld2_f32
+  // CHECK-LABEL: test_vld2_f32
   return vld2_f32(a);
-  // CHECK: ld2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x2_t test_vld2_f64(float64_t const *a) {
-  // CHECK: test_vld2_f64
+  // CHECK-LABEL: test_vld2_f64
   return vld2_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x2_t test_vld2_p8(poly8_t const *a) {
-  // CHECK: test_vld2_p8
+  // CHECK-LABEL: test_vld2_p8
   return vld2_p8(a);
-  // CHECK: ld2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x2_t test_vld2_p16(poly16_t const *a) {
-  // CHECK: test_vld2_p16
+  // CHECK-LABEL: test_vld2_p16
   return vld2_p16(a);
-  // CHECK: ld2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x3_t test_vld3q_u8(uint8_t const *a) {
-  // CHECK: test_vld3q_u8
+  // CHECK-LABEL: test_vld3q_u8
   return vld3q_u8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x3_t test_vld3q_u16(uint16_t const *a) {
-  // CHECK: test_vld3q_u16
+  // CHECK-LABEL: test_vld3q_u16
   return vld3q_u16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x3_t test_vld3q_u32(uint32_t const *a) {
-  // CHECK: test_vld3q_u32
+  // CHECK-LABEL: test_vld3q_u32
   return vld3q_u32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x3_t test_vld3q_u64(uint64_t const *a) {
-  // CHECK: test_vld3q_u64
+  // CHECK-LABEL: test_vld3q_u64
   return vld3q_u64(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x3_t test_vld3q_s8(int8_t const *a) {
-  // CHECK: test_vld3q_s8
+  // CHECK-LABEL: test_vld3q_s8
   return vld3q_s8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x3_t test_vld3q_s16(int16_t const *a) {
-  // CHECK: test_vld3q_s16
+  // CHECK-LABEL: test_vld3q_s16
   return vld3q_s16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x3_t test_vld3q_s32(int32_t const *a) {
-  // CHECK: test_vld3q_s32
+  // CHECK-LABEL: test_vld3q_s32
   return vld3q_s32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x3_t test_vld3q_s64(int64_t const *a) {
-  // CHECK: test_vld3q_s64
+  // CHECK-LABEL: test_vld3q_s64
   return vld3q_s64(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x3_t test_vld3q_f16(float16_t const *a) {
-  // CHECK: test_vld3q_f16
+  // CHECK-LABEL: test_vld3q_f16
   return vld3q_f16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x3_t test_vld3q_f32(float32_t const *a) {
-  // CHECK: test_vld3q_f32
+  // CHECK-LABEL: test_vld3q_f32
   return vld3q_f32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x3_t test_vld3q_f64(float64_t const *a) {
-  // CHECK: test_vld3q_f64
+  // CHECK-LABEL: test_vld3q_f64
   return vld3q_f64(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x3_t test_vld3q_p8(poly8_t const *a) {
-  // CHECK: test_vld3q_p8
+  // CHECK-LABEL: test_vld3q_p8
   return vld3q_p8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x3_t test_vld3q_p16(poly16_t const *a) {
-  // CHECK: test_vld3q_p16
+  // CHECK-LABEL: test_vld3q_p16
   return vld3q_p16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x3_t test_vld3_u8(uint8_t const *a) {
-  // CHECK: test_vld3_u8
+  // CHECK-LABEL: test_vld3_u8
   return vld3_u8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x3_t test_vld3_u16(uint16_t const *a) {
-  // CHECK: test_vld3_u16
+  // CHECK-LABEL: test_vld3_u16
   return vld3_u16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x3_t test_vld3_u32(uint32_t const *a) {
-  // CHECK: test_vld3_u32
+  // CHECK-LABEL: test_vld3_u32
   return vld3_u32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x3_t test_vld3_u64(uint64_t const *a) {
-  // CHECK: test_vld3_u64
+  // CHECK-LABEL: test_vld3_u64
   return vld3_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x3_t test_vld3_s8(int8_t const *a) {
-  // CHECK: test_vld3_s8
+  // CHECK-LABEL: test_vld3_s8
   return vld3_s8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x3_t test_vld3_s16(int16_t const *a) {
-  // CHECK: test_vld3_s16
+  // CHECK-LABEL: test_vld3_s16
   return vld3_s16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x3_t test_vld3_s32(int32_t const *a) {
-  // CHECK: test_vld3_s32
+  // CHECK-LABEL: test_vld3_s32
   return vld3_s32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x3_t test_vld3_s64(int64_t const *a) {
-  // CHECK: test_vld3_s64
+  // CHECK-LABEL: test_vld3_s64
   return vld3_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x3_t test_vld3_f16(float16_t const *a) {
-  // CHECK: test_vld3_f16
+  // CHECK-LABEL: test_vld3_f16
   return vld3_f16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x3_t test_vld3_f32(float32_t const *a) {
-  // CHECK: test_vld3_f32
+  // CHECK-LABEL: test_vld3_f32
   return vld3_f32(a);
-  // CHECK: ld3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x3_t test_vld3_f64(float64_t const *a) {
-  // CHECK: test_vld3_f64
+  // CHECK-LABEL: test_vld3_f64
   return vld3_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x3_t test_vld3_p8(poly8_t const *a) {
-  // CHECK: test_vld3_p8
+  // CHECK-LABEL: test_vld3_p8
   return vld3_p8(a);
-  // CHECK: ld3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x3_t test_vld3_p16(poly16_t const *a) {
-  // CHECK: test_vld3_p16
+  // CHECK-LABEL: test_vld3_p16
   return vld3_p16(a);
-  // CHECK: ld3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x4_t test_vld4q_u8(uint8_t const *a) {
-  // CHECK: test_vld4q_u8
+  // CHECK-LABEL: test_vld4q_u8
   return vld4q_u8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x4_t test_vld4q_u16(uint16_t const *a) {
-  // CHECK: test_vld4q_u16
+  // CHECK-LABEL: test_vld4q_u16
   return vld4q_u16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x4_t test_vld4q_u32(uint32_t const *a) {
-  // CHECK: test_vld4q_u32
+  // CHECK-LABEL: test_vld4q_u32
   return vld4q_u32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x4_t test_vld4q_u64(uint64_t const *a) {
-  // CHECK: test_vld4q_u64
+  // CHECK-LABEL: test_vld4q_u64
   return vld4q_u64(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x4_t test_vld4q_s8(int8_t const *a) {
-  // CHECK: test_vld4q_s8
+  // CHECK-LABEL: test_vld4q_s8
   return vld4q_s8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x4_t test_vld4q_s16(int16_t const *a) {
-  // CHECK: test_vld4q_s16
+  // CHECK-LABEL: test_vld4q_s16
   return vld4q_s16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x4_t test_vld4q_s32(int32_t const *a) {
-  // CHECK: test_vld4q_s32
+  // CHECK-LABEL: test_vld4q_s32
   return vld4q_s32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x4_t test_vld4q_s64(int64_t const *a) {
-  // CHECK: test_vld4q_s64
+  // CHECK-LABEL: test_vld4q_s64
   return vld4q_s64(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x4_t test_vld4q_f16(float16_t const *a) {
-  // CHECK: test_vld4q_f16
+  // CHECK-LABEL: test_vld4q_f16
   return vld4q_f16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x4_t test_vld4q_f32(float32_t const *a) {
-  // CHECK: test_vld4q_f32
+  // CHECK-LABEL: test_vld4q_f32
   return vld4q_f32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x4_t test_vld4q_f64(float64_t const *a) {
-  // CHECK: test_vld4q_f64
+  // CHECK-LABEL: test_vld4q_f64
   return vld4q_f64(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x4_t test_vld4q_p8(poly8_t const *a) {
-  // CHECK: test_vld4q_p8
+  // CHECK-LABEL: test_vld4q_p8
   return vld4q_p8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x4_t test_vld4q_p16(poly16_t const *a) {
-  // CHECK: test_vld4q_p16
+  // CHECK-LABEL: test_vld4q_p16
   return vld4q_p16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x4_t test_vld4_u8(uint8_t const *a) {
-  // CHECK: test_vld4_u8
+  // CHECK-LABEL: test_vld4_u8
   return vld4_u8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x4_t test_vld4_u16(uint16_t const *a) {
-  // CHECK: test_vld4_u16
+  // CHECK-LABEL: test_vld4_u16
   return vld4_u16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x4_t test_vld4_u32(uint32_t const *a) {
-  // CHECK: test_vld4_u32
+  // CHECK-LABEL: test_vld4_u32
   return vld4_u32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x4_t test_vld4_u64(uint64_t const *a) {
-  // CHECK: test_vld4_u64
+  // CHECK-LABEL: test_vld4_u64
   return vld4_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x4_t test_vld4_s8(int8_t const *a) {
-  // CHECK: test_vld4_s8
+  // CHECK-LABEL: test_vld4_s8
   return vld4_s8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x4_t test_vld4_s16(int16_t const *a) {
-  // CHECK: test_vld4_s16
+  // CHECK-LABEL: test_vld4_s16
   return vld4_s16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x4_t test_vld4_s32(int32_t const *a) {
-  // CHECK: test_vld4_s32
+  // CHECK-LABEL: test_vld4_s32
   return vld4_s32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x4_t test_vld4_s64(int64_t const *a) {
-  // CHECK: test_vld4_s64
+  // CHECK-LABEL: test_vld4_s64
   return vld4_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x4_t test_vld4_f16(float16_t const *a) {
-  // CHECK: test_vld4_f16
+  // CHECK-LABEL: test_vld4_f16
   return vld4_f16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x4_t test_vld4_f32(float32_t const *a) {
-  // CHECK: test_vld4_f32
+  // CHECK-LABEL: test_vld4_f32
   return vld4_f32(a);
-  // CHECK: ld4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x4_t test_vld4_f64(float64_t const *a) {
-  // CHECK: test_vld4_f64
+  // CHECK-LABEL: test_vld4_f64
   return vld4_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x4_t test_vld4_p8(poly8_t const *a) {
-  // CHECK: test_vld4_p8
+  // CHECK-LABEL: test_vld4_p8
   return vld4_p8(a);
-  // CHECK: ld4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x4_t test_vld4_p16(poly16_t const *a) {
-  // CHECK: test_vld4_p16
+  // CHECK-LABEL: test_vld4_p16
   return vld4_p16(a);
-  // CHECK: ld4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u8(uint8_t *a, uint8x16_t b) {
-  // CHECK: test_vst1q_u8
+  // CHECK-LABEL: test_vst1q_u8
   vst1q_u8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.16b}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u16(uint16_t *a, uint16x8_t b) {
-  // CHECK: test_vst1q_u16
+  // CHECK-LABEL: test_vst1q_u16
   vst1q_u16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8h}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u32(uint32_t *a, uint32x4_t b) {
-  // CHECK: test_vst1q_u32
+  // CHECK-LABEL: test_vst1q_u32
   vst1q_u32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4s}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u64(uint64_t *a, uint64x2_t b) {
-  // CHECK: test_vst1q_u64
+  // CHECK-LABEL: test_vst1q_u64
   vst1q_u64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2d}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s8(int8_t *a, int8x16_t b) {
-  // CHECK: test_vst1q_s8
+  // CHECK-LABEL: test_vst1q_s8
   vst1q_s8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.16b}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s16(int16_t *a, int16x8_t b) {
-  // CHECK: test_vst1q_s16
+  // CHECK-LABEL: test_vst1q_s16
   vst1q_s16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8h}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s32(int32_t *a, int32x4_t b) {
-  // CHECK: test_vst1q_s32
+  // CHECK-LABEL: test_vst1q_s32
   vst1q_s32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4s}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s64(int64_t *a, int64x2_t b) {
-  // CHECK: test_vst1q_s64
+  // CHECK-LABEL: test_vst1q_s64
   vst1q_s64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2d}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f16(float16_t *a, float16x8_t b) {
-  // CHECK: test_vst1q_f16
+  // CHECK-LABEL: test_vst1q_f16
   vst1q_f16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8h}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f32(float32_t *a, float32x4_t b) {
-  // CHECK: test_vst1q_f32
+  // CHECK-LABEL: test_vst1q_f32
   vst1q_f32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4s}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f64(float64_t *a, float64x2_t b) {
-  // CHECK: test_vst1q_f64
+  // CHECK-LABEL: test_vst1q_f64
   vst1q_f64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2d}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p8(poly8_t *a, poly8x16_t b) {
-  // CHECK: test_vst1q_p8
+  // CHECK-LABEL: test_vst1q_p8
   vst1q_p8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.16b}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p16(poly16_t *a, poly16x8_t b) {
-  // CHECK: test_vst1q_p16
+  // CHECK-LABEL: test_vst1q_p16
   vst1q_p16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8h}|str q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u8(uint8_t *a, uint8x8_t b) {
-  // CHECK: test_vst1_u8
+  // CHECK-LABEL: test_vst1_u8
   vst1_u8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8b}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u16(uint16_t *a, uint16x4_t b) {
-  // CHECK: test_vst1_u16
+  // CHECK-LABEL: test_vst1_u16
   vst1_u16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4h}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u32(uint32_t *a, uint32x2_t b) {
-  // CHECK: test_vst1_u32
+  // CHECK-LABEL: test_vst1_u32
   vst1_u32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2s}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u64(uint64_t *a, uint64x1_t b) {
-  // CHECK: test_vst1_u64
+  // CHECK-LABEL: test_vst1_u64
   vst1_u64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.1d}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s8(int8_t *a, int8x8_t b) {
-  // CHECK: test_vst1_s8
+  // CHECK-LABEL: test_vst1_s8
   vst1_s8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8b}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s16(int16_t *a, int16x4_t b) {
-  // CHECK: test_vst1_s16
+  // CHECK-LABEL: test_vst1_s16
   vst1_s16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4h}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s32(int32_t *a, int32x2_t b) {
-  // CHECK: test_vst1_s32
+  // CHECK-LABEL: test_vst1_s32
   vst1_s32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2s}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s64(int64_t *a, int64x1_t b) {
-  // CHECK: test_vst1_s64
+  // CHECK-LABEL: test_vst1_s64
   vst1_s64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.1d}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f16(float16_t *a, float16x4_t b) {
-  // CHECK: test_vst1_f16
+  // CHECK-LABEL: test_vst1_f16
   vst1_f16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4h}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f32(float32_t *a, float32x2_t b) {
-  // CHECK: test_vst1_f32
+  // CHECK-LABEL: test_vst1_f32
   vst1_f32(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.2s}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f64(float64_t *a, float64x1_t b) {
-  // CHECK: test_vst1_f64
+  // CHECK-LABEL: test_vst1_f64
   vst1_f64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.1d}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p8(poly8_t *a, poly8x8_t b) {
-  // CHECK: test_vst1_p8
+  // CHECK-LABEL: test_vst1_p8
   vst1_p8(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.8b}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p16(poly16_t *a, poly16x4_t b) {
-  // CHECK: test_vst1_p16
+  // CHECK-LABEL: test_vst1_p16
   vst1_p16(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.4h}|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_u8(uint8_t *a, uint8x16x2_t b) {
-  // CHECK: test_vst2q_u8
+  // CHECK-LABEL: test_vst2q_u8
   vst2q_u8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_u16(uint16_t *a, uint16x8x2_t b) {
-  // CHECK: test_vst2q_u16
+  // CHECK-LABEL: test_vst2q_u16
   vst2q_u16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_u32(uint32_t *a, uint32x4x2_t b) {
-  // CHECK: test_vst2q_u32
+  // CHECK-LABEL: test_vst2q_u32
   vst2q_u32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_u64(uint64_t *a, uint64x2x2_t b) {
-  // CHECK: test_vst2q_u64
+  // CHECK-LABEL: test_vst2q_u64
   vst2q_u64(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_s8(int8_t *a, int8x16x2_t b) {
-  // CHECK: test_vst2q_s8
+  // CHECK-LABEL: test_vst2q_s8
   vst2q_s8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_s16(int16_t *a, int16x8x2_t b) {
-  // CHECK: test_vst2q_s16
+  // CHECK-LABEL: test_vst2q_s16
   vst2q_s16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_s32(int32_t *a, int32x4x2_t b) {
-  // CHECK: test_vst2q_s32
+  // CHECK-LABEL: test_vst2q_s32
   vst2q_s32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_s64(int64_t *a, int64x2x2_t b) {
-  // CHECK: test_vst2q_s64
+  // CHECK-LABEL: test_vst2q_s64
   vst2q_s64(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_f16(float16_t *a, float16x8x2_t b) {
-  // CHECK: test_vst2q_f16
+  // CHECK-LABEL: test_vst2q_f16
   vst2q_f16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_f32(float32_t *a, float32x4x2_t b) {
-  // CHECK: test_vst2q_f32
+  // CHECK-LABEL: test_vst2q_f32
   vst2q_f32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_f64(float64_t *a, float64x2x2_t b) {
-  // CHECK: test_vst2q_f64
+  // CHECK-LABEL: test_vst2q_f64
   vst2q_f64(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_p8(poly8_t *a, poly8x16x2_t b) {
-  // CHECK: test_vst2q_p8
+  // CHECK-LABEL: test_vst2q_p8
   vst2q_p8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_p16(poly16_t *a, poly16x8x2_t b) {
-  // CHECK: test_vst2q_p16
+  // CHECK-LABEL: test_vst2q_p16
   vst2q_p16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_u8(uint8_t *a, uint8x8x2_t b) {
-  // CHECK: test_vst2_u8
+  // CHECK-LABEL: test_vst2_u8
   vst2_u8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_u16(uint16_t *a, uint16x4x2_t b) {
-  // CHECK: test_vst2_u16
+  // CHECK-LABEL: test_vst2_u16
   vst2_u16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_u32(uint32_t *a, uint32x2x2_t b) {
-  // CHECK: test_vst2_u32
+  // CHECK-LABEL: test_vst2_u32
   vst2_u32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_u64(uint64_t *a, uint64x1x2_t b) {
-  // CHECK: test_vst2_u64
+  // CHECK-LABEL: test_vst2_u64
   vst2_u64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_s8(int8_t *a, int8x8x2_t b) {
-  // CHECK: test_vst2_s8
+  // CHECK-LABEL: test_vst2_s8
   vst2_s8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_s16(int16_t *a, int16x4x2_t b) {
-  // CHECK: test_vst2_s16
+  // CHECK-LABEL: test_vst2_s16
   vst2_s16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_s32(int32_t *a, int32x2x2_t b) {
-  // CHECK: test_vst2_s32
+  // CHECK-LABEL: test_vst2_s32
   vst2_s32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_s64(int64_t *a, int64x1x2_t b) {
-  // CHECK: test_vst2_s64
+  // CHECK-LABEL: test_vst2_s64
   vst2_s64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_f16(float16_t *a, float16x4x2_t b) {
-  // CHECK: test_vst2_f16
+  // CHECK-LABEL: test_vst2_f16
   vst2_f16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_f32(float32_t *a, float32x2x2_t b) {
-  // CHECK: test_vst2_f32
+  // CHECK-LABEL: test_vst2_f32
   vst2_f32(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_f64(float64_t *a, float64x1x2_t b) {
-  // CHECK: test_vst2_f64
+  // CHECK-LABEL: test_vst2_f64
   vst2_f64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st2}} {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_p8(poly8_t *a, poly8x8x2_t b) {
-  // CHECK: test_vst2_p8
+  // CHECK-LABEL: test_vst2_p8
   vst2_p8(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_p16(poly16_t *a, poly16x4x2_t b) {
-  // CHECK: test_vst2_p16
+  // CHECK-LABEL: test_vst2_p16
   vst2_p16(a, b);
-  // CHECK: st2 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_u8(uint8_t *a, uint8x16x3_t b) {
-  // CHECK: test_vst3q_u8
+  // CHECK-LABEL: test_vst3q_u8
   vst3q_u8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_u16(uint16_t *a, uint16x8x3_t b) {
-  // CHECK: test_vst3q_u16
+  // CHECK-LABEL: test_vst3q_u16
   vst3q_u16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_u32(uint32_t *a, uint32x4x3_t b) {
-  // CHECK: test_vst3q_u32
+  // CHECK-LABEL: test_vst3q_u32
   vst3q_u32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_u64(uint64_t *a, uint64x2x3_t b) {
-  // CHECK: test_vst3q_u64
+  // CHECK-LABEL: test_vst3q_u64
   vst3q_u64(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_s8(int8_t *a, int8x16x3_t b) {
-  // CHECK: test_vst3q_s8
+  // CHECK-LABEL: test_vst3q_s8
   vst3q_s8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_s16(int16_t *a, int16x8x3_t b) {
-  // CHECK: test_vst3q_s16
+  // CHECK-LABEL: test_vst3q_s16
   vst3q_s16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_s32(int32_t *a, int32x4x3_t b) {
-  // CHECK: test_vst3q_s32
+  // CHECK-LABEL: test_vst3q_s32
   vst3q_s32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_s64(int64_t *a, int64x2x3_t b) {
-  // CHECK: test_vst3q_s64
+  // CHECK-LABEL: test_vst3q_s64
   vst3q_s64(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_f16(float16_t *a, float16x8x3_t b) {
-  // CHECK: test_vst3q_f16
+  // CHECK-LABEL: test_vst3q_f16
   vst3q_f16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_f32(float32_t *a, float32x4x3_t b) {
-  // CHECK: test_vst3q_f32
+  // CHECK-LABEL: test_vst3q_f32
   vst3q_f32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_f64(float64_t *a, float64x2x3_t b) {
-  // CHECK: test_vst3q_f64
+  // CHECK-LABEL: test_vst3q_f64
   vst3q_f64(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_p8(poly8_t *a, poly8x16x3_t b) {
-  // CHECK: test_vst3q_p8
+  // CHECK-LABEL: test_vst3q_p8
   vst3q_p8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_p16(poly16_t *a, poly16x8x3_t b) {
-  // CHECK: test_vst3q_p16
+  // CHECK-LABEL: test_vst3q_p16
   vst3q_p16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_u8(uint8_t *a, uint8x8x3_t b) {
-  // CHECK: test_vst3_u8
+  // CHECK-LABEL: test_vst3_u8
   vst3_u8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_u16(uint16_t *a, uint16x4x3_t b) {
-  // CHECK: test_vst3_u16
+  // CHECK-LABEL: test_vst3_u16
   vst3_u16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_u32(uint32_t *a, uint32x2x3_t b) {
-  // CHECK: test_vst3_u32
+  // CHECK-LABEL: test_vst3_u32
   vst3_u32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_u64(uint64_t *a, uint64x1x3_t b) {
-  // CHECK: test_vst3_u64
+  // CHECK-LABEL: test_vst3_u64
   vst3_u64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_s8(int8_t *a, int8x8x3_t b) {
-  // CHECK: test_vst3_s8
+  // CHECK-LABEL: test_vst3_s8
   vst3_s8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_s16(int16_t *a, int16x4x3_t b) {
-  // CHECK: test_vst3_s16
+  // CHECK-LABEL: test_vst3_s16
   vst3_s16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_s32(int32_t *a, int32x2x3_t b) {
-  // CHECK: test_vst3_s32
+  // CHECK-LABEL: test_vst3_s32
   vst3_s32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_s64(int64_t *a, int64x1x3_t b) {
-  // CHECK: test_vst3_s64
+  // CHECK-LABEL: test_vst3_s64
   vst3_s64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_f16(float16_t *a, float16x4x3_t b) {
-  // CHECK: test_vst3_f16
+  // CHECK-LABEL: test_vst3_f16
   vst3_f16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_f32(float32_t *a, float32x2x3_t b) {
-  // CHECK: test_vst3_f32
+  // CHECK-LABEL: test_vst3_f32
   vst3_f32(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_f64(float64_t *a, float64x1x3_t b) {
-  // CHECK: test_vst3_f64
+  // CHECK-LABEL: test_vst3_f64
   vst3_f64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st3}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_p8(poly8_t *a, poly8x8x3_t b) {
-  // CHECK: test_vst3_p8
+  // CHECK-LABEL: test_vst3_p8
   vst3_p8(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_p16(poly16_t *a, poly16x4x3_t b) {
-  // CHECK: test_vst3_p16
+  // CHECK-LABEL: test_vst3_p16
   vst3_p16(a, b);
-  // CHECK: st3 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_u8(uint8_t *a, uint8x16x4_t b) {
-  // CHECK: test_vst4q_u8
+  // CHECK-LABEL: test_vst4q_u8
   vst4q_u8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_u16(uint16_t *a, uint16x8x4_t b) {
-  // CHECK: test_vst4q_u16
+  // CHECK-LABEL: test_vst4q_u16
   vst4q_u16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_u32(uint32_t *a, uint32x4x4_t b) {
-  // CHECK: test_vst4q_u32
+  // CHECK-LABEL: test_vst4q_u32
   vst4q_u32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_u64(uint64_t *a, uint64x2x4_t b) {
-  // CHECK: test_vst4q_u64
+  // CHECK-LABEL: test_vst4q_u64
   vst4q_u64(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_s8(int8_t *a, int8x16x4_t b) {
-  // CHECK: test_vst4q_s8
+  // CHECK-LABEL: test_vst4q_s8
   vst4q_s8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_s16(int16_t *a, int16x8x4_t b) {
-  // CHECK: test_vst4q_s16
+  // CHECK-LABEL: test_vst4q_s16
   vst4q_s16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_s32(int32_t *a, int32x4x4_t b) {
-  // CHECK: test_vst4q_s32
+  // CHECK-LABEL: test_vst4q_s32
   vst4q_s32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_s64(int64_t *a, int64x2x4_t b) {
-  // CHECK: test_vst4q_s64
+  // CHECK-LABEL: test_vst4q_s64
   vst4q_s64(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_f16(float16_t *a, float16x8x4_t b) {
-  // CHECK: test_vst4q_f16
+  // CHECK-LABEL: test_vst4q_f16
   vst4q_f16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_f32(float32_t *a, float32x4x4_t b) {
-  // CHECK: test_vst4q_f32
+  // CHECK-LABEL: test_vst4q_f32
   vst4q_f32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_f64(float64_t *a, float64x2x4_t b) {
-  // CHECK: test_vst4q_f64
+  // CHECK-LABEL: test_vst4q_f64
   vst4q_f64(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_p8(poly8_t *a, poly8x16x4_t b) {
-  // CHECK: test_vst4q_p8
+  // CHECK-LABEL: test_vst4q_p8
   vst4q_p8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_p16(poly16_t *a, poly16x8x4_t b) {
-  // CHECK: test_vst4q_p16
+  // CHECK-LABEL: test_vst4q_p16
   vst4q_p16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_u8(uint8_t *a, uint8x8x4_t b) {
-  // CHECK: test_vst4_u8
+  // CHECK-LABEL: test_vst4_u8
   vst4_u8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_u16(uint16_t *a, uint16x4x4_t b) {
-  // CHECK: test_vst4_u16
+  // CHECK-LABEL: test_vst4_u16
   vst4_u16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_u32(uint32_t *a, uint32x2x4_t b) {
-  // CHECK: test_vst4_u32
+  // CHECK-LABEL: test_vst4_u32
   vst4_u32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_u64(uint64_t *a, uint64x1x4_t b) {
-  // CHECK: test_vst4_u64
+  // CHECK-LABEL: test_vst4_u64
   vst4_u64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_s8(int8_t *a, int8x8x4_t b) {
-  // CHECK: test_vst4_s8
+  // CHECK-LABEL: test_vst4_s8
   vst4_s8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+// CHECK: st4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_s16(int16_t *a, int16x4x4_t b) {
-  // CHECK: test_vst4_s16
+  // CHECK-LABEL: test_vst4_s16
   vst4_s16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_s32(int32_t *a, int32x2x4_t b) {
-  // CHECK: test_vst4_s32
+  // CHECK-LABEL: test_vst4_s32
   vst4_s32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_s64(int64_t *a, int64x1x4_t b) {
-  // CHECK: test_vst4_s64
+  // CHECK-LABEL: test_vst4_s64
   vst4_s64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_f16(float16_t *a, float16x4x4_t b) {
-  // CHECK: test_vst4_f16
+  // CHECK-LABEL: test_vst4_f16
   vst4_f16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_f32(float32_t *a, float32x2x4_t b) {
-  // CHECK: test_vst4_f32
+  // CHECK-LABEL: test_vst4_f32
   vst4_f32(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_f64(float64_t *a, float64x1x4_t b) {
-  // CHECK: test_vst4_f64
+  // CHECK-LABEL: test_vst4_f64
   vst4_f64(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{st1|st4}} {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_p8(poly8_t *a, poly8x8x4_t b) {
-  // CHECK: test_vst4_p8
+  // CHECK-LABEL: test_vst4_p8
   vst4_p8(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_p16(poly16_t *a, poly16x4x4_t b) {
-  // CHECK: test_vst4_p16
+  // CHECK-LABEL: test_vst4_p16
   vst4_p16(a, b);
-  // CHECK: st4 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x2_t test_vld1q_u8_x2(uint8_t const *a) {
   // CHECK-LABEL: test_vld1q_u8_x2
   return vld1q_u8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x2_t test_vld1q_u16_x2(uint16_t const *a) {
   // CHECK-LABEL: test_vld1q_u16_x2
   return vld1q_u16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x2_t test_vld1q_u32_x2(uint32_t const *a) {
   // CHECK-LABEL: test_vld1q_u32_x2
   return vld1q_u32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x2_t test_vld1q_u64_x2(uint64_t const *a) {
   // CHECK-LABEL: test_vld1q_u64_x2
   return vld1q_u64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x2_t test_vld1q_s8_x2(int8_t const *a) {
   // CHECK-LABEL: test_vld1q_s8_x2
   return vld1q_s8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x2_t test_vld1q_s16_x2(int16_t const *a) {
   // CHECK-LABEL: test_vld1q_s16_x2
   return vld1q_s16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x2_t test_vld1q_s32_x2(int32_t const *a) {
   // CHECK-LABEL: test_vld1q_s32_x2
   return vld1q_s32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x2_t test_vld1q_s64_x2(int64_t const *a) {
   // CHECK-LABEL: test_vld1q_s64_x2
   return vld1q_s64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x2_t test_vld1q_f16_x2(float16_t const *a) {
   // CHECK-LABEL: test_vld1q_f16_x2
   return vld1q_f16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x2_t test_vld1q_f32_x2(float32_t const *a) {
   // CHECK-LABEL: test_vld1q_f32_x2
   return vld1q_f32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x2_t test_vld1q_f64_x2(float64_t const *a) {
   // CHECK-LABEL: test_vld1q_f64_x2
   return vld1q_f64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x2_t test_vld1q_p8_x2(poly8_t const *a) {
   // CHECK-LABEL: test_vld1q_p8_x2
   return vld1q_p8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x2_t test_vld1q_p16_x2(poly16_t const *a) {
   // CHECK-LABEL: test_vld1q_p16_x2
   return vld1q_p16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x2_t test_vld1q_p64_x2(poly64_t const *a) {
   // CHECK-LABEL: test_vld1q_p64_x2
   return vld1q_p64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x2_t test_vld1_u8_x2(uint8_t const *a) {
   // CHECK-LABEL: test_vld1_u8_x2
   return vld1_u8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x2_t test_vld1_u16_x2(uint16_t const *a) {
   // CHECK-LABEL: test_vld1_u16_x2
   return vld1_u16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x2_t test_vld1_u32_x2(uint32_t const *a) {
   // CHECK-LABEL: test_vld1_u32_x2
   return vld1_u32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x2_t test_vld1_u64_x2(uint64_t const *a) {
   // CHECK-LABEL: test_vld1_u64_x2
   return vld1_u64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x2_t test_vld1_s8_x2(int8_t const *a) {
   // CHECK-LABEL: test_vld1_s8_x2
   return vld1_s8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x2_t test_vld1_s16_x2(int16_t const *a) {
   // CHECK-LABEL: test_vld1_s16_x2
   return vld1_s16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x2_t test_vld1_s32_x2(int32_t const *a) {
   // CHECK-LABEL: test_vld1_s32_x2
   return vld1_s32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x2_t test_vld1_s64_x2(int64_t const *a) {
   // CHECK-LABEL: test_vld1_s64_x2
   return vld1_s64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x2_t test_vld1_f16_x2(float16_t const *a) {
   // CHECK-LABEL: test_vld1_f16_x2
   return vld1_f16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x2_t test_vld1_f32_x2(float32_t const *a) {
   // CHECK-LABEL: test_vld1_f32_x2
   return vld1_f32_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x2_t test_vld1_f64_x2(float64_t const *a) {
   // CHECK-LABEL: test_vld1_f64_x2
   return vld1_f64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x2_t test_vld1_p8_x2(poly8_t const *a) {
   // CHECK-LABEL: test_vld1_p8_x2
   return vld1_p8_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x2_t test_vld1_p16_x2(poly16_t const *a) {
   // CHECK-LABEL: test_vld1_p16_x2
   return vld1_p16_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x2_t test_vld1_p64_x2(poly64_t const *a) {
   // CHECK-LABEL: test_vld1_p64_x2
   return vld1_p64_x2(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x3_t test_vld1q_u8_x3(uint8_t const *a) {
   // CHECK-LABEL: test_vld1q_u8_x3
   return vld1q_u8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x3_t test_vld1q_u16_x3(uint16_t const *a) {
   // CHECK-LABEL: test_vld1q_u16_x3
   return vld1q_u16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x3_t test_vld1q_u32_x3(uint32_t const *a) {
   // CHECK-LABEL: test_vld1q_u32_x3
   return vld1q_u32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x3_t test_vld1q_u64_x3(uint64_t const *a) {
   // CHECK-LABEL: test_vld1q_u64_x3
   return vld1q_u64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x3_t test_vld1q_s8_x3(int8_t const *a) {
   // CHECK-LABEL: test_vld1q_s8_x3
   return vld1q_s8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x3_t test_vld1q_s16_x3(int16_t const *a) {
   // CHECK-LABEL: test_vld1q_s16_x3
   return vld1q_s16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x3_t test_vld1q_s32_x3(int32_t const *a) {
   // CHECK-LABEL: test_vld1q_s32_x3
   return vld1q_s32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x3_t test_vld1q_s64_x3(int64_t const *a) {
   // CHECK-LABEL: test_vld1q_s64_x3
   return vld1q_s64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x3_t test_vld1q_f16_x3(float16_t const *a) {
   // CHECK-LABEL: test_vld1q_f16_x3
   return vld1q_f16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x3_t test_vld1q_f32_x3(float32_t const *a) {
   // CHECK-LABEL: test_vld1q_f32_x3
   return vld1q_f32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x3_t test_vld1q_f64_x3(float64_t const *a) {
   // CHECK-LABEL: test_vld1q_f64_x3
   return vld1q_f64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x3_t test_vld1q_p8_x3(poly8_t const *a) {
   // CHECK-LABEL: test_vld1q_p8_x3
   return vld1q_p8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x3_t test_vld1q_p16_x3(poly16_t const *a) {
   // CHECK-LABEL: test_vld1q_p16_x3
   return vld1q_p16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x3_t test_vld1q_p64_x3(poly64_t const *a) {
   // CHECK-LABEL: test_vld1q_p64_x3
   return vld1q_p64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x3_t test_vld1_u8_x3(uint8_t const *a) {
   // CHECK-LABEL: test_vld1_u8_x3
   return vld1_u8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x3_t test_vld1_u16_x3(uint16_t const *a) {
   // CHECK-LABEL: test_vld1_u16_x3
   return vld1_u16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x3_t test_vld1_u32_x3(uint32_t const *a) {
   // CHECK-LABEL: test_vld1_u32_x3
   return vld1_u32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x3_t test_vld1_u64_x3(uint64_t const *a) {
   // CHECK-LABEL: test_vld1_u64_x3
   return vld1_u64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x3_t test_vld1_s8_x3(int8_t const *a) {
   // CHECK-LABEL: test_vld1_s8_x3
   return vld1_s8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x3_t test_vld1_s16_x3(int16_t const *a) {
   // CHECK-LABEL: test_vld1_s16_x3
   return vld1_s16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x3_t test_vld1_s32_x3(int32_t const *a) {
   // CHECK-LABEL: test_vld1_s32_x3
   return vld1_s32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x3_t test_vld1_s64_x3(int64_t const *a) {
   // CHECK-LABEL: test_vld1_s64_x3
   return vld1_s64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x3_t test_vld1_f16_x3(float16_t const *a) {
   // CHECK-LABEL: test_vld1_f16_x3
   return vld1_f16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x3_t test_vld1_f32_x3(float32_t const *a) {
   // CHECK-LABEL: test_vld1_f32_x3
   return vld1_f32_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x3_t test_vld1_f64_x3(float64_t const *a) {
   // CHECK-LABEL: test_vld1_f64_x3
   return vld1_f64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x3_t test_vld1_p8_x3(poly8_t const *a) {
   // CHECK-LABEL: test_vld1_p8_x3
   return vld1_p8_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x3_t test_vld1_p16_x3(poly16_t const *a) {
   // CHECK-LABEL: test_vld1_p16_x3
   return vld1_p16_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x3_t test_vld1_p64_x3(poly64_t const *a) {
   // CHECK-LABEL: test_vld1_p64_x3
   return vld1_p64_x3(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x16x4_t test_vld1q_u8_x4(uint8_t const *a) {
   // CHECK-LABEL: test_vld1q_u8_x4
   return vld1q_u8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x8x4_t test_vld1q_u16_x4(uint16_t const *a) {
   // CHECK-LABEL: test_vld1q_u16_x4
   return vld1q_u16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x4x4_t test_vld1q_u32_x4(uint32_t const *a) {
   // CHECK-LABEL: test_vld1q_u32_x4
   return vld1q_u32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x2x4_t test_vld1q_u64_x4(uint64_t const *a) {
   // CHECK-LABEL: test_vld1q_u64_x4
   return vld1q_u64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x16x4_t test_vld1q_s8_x4(int8_t const *a) {
   // CHECK-LABEL: test_vld1q_s8_x4
   return vld1q_s8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x8x4_t test_vld1q_s16_x4(int16_t const *a) {
   // CHECK-LABEL: test_vld1q_s16_x4
   return vld1q_s16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x4x4_t test_vld1q_s32_x4(int32_t const *a) {
   // CHECK-LABEL: test_vld1q_s32_x4
   return vld1q_s32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x2x4_t test_vld1q_s64_x4(int64_t const *a) {
   // CHECK-LABEL: test_vld1q_s64_x4
   return vld1q_s64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x8x4_t test_vld1q_f16_x4(float16_t const *a) {
   // CHECK-LABEL: test_vld1q_f16_x4
   return vld1q_f16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x4x4_t test_vld1q_f32_x4(float32_t const *a) {
   // CHECK-LABEL: test_vld1q_f32_x4
   return vld1q_f32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x2x4_t test_vld1q_f64_x4(float64_t const *a) {
   // CHECK-LABEL: test_vld1q_f64_x4
   return vld1q_f64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x16x4_t test_vld1q_p8_x4(poly8_t const *a) {
   // CHECK-LABEL: test_vld1q_p8_x4
   return vld1q_p8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x8x4_t test_vld1q_p16_x4(poly16_t const *a) {
   // CHECK-LABEL: test_vld1q_p16_x4
   return vld1q_p16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x4_t test_vld1q_p64_x4(poly64_t const *a) {
   // CHECK-LABEL: test_vld1q_p64_x4
   return vld1q_p64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint8x8x4_t test_vld1_u8_x4(uint8_t const *a) {
   // CHECK-LABEL: test_vld1_u8_x4
   return vld1_u8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint16x4x4_t test_vld1_u16_x4(uint16_t const *a) {
   // CHECK-LABEL: test_vld1_u16_x4
   return vld1_u16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint32x2x4_t test_vld1_u32_x4(uint32_t const *a) {
   // CHECK-LABEL: test_vld1_u32_x4
   return vld1_u32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 uint64x1x4_t test_vld1_u64_x4(uint64_t const *a) {
   // CHECK-LABEL: test_vld1_u64_x4
   return vld1_u64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int8x8x4_t test_vld1_s8_x4(int8_t const *a) {
   // CHECK-LABEL: test_vld1_s8_x4
   return vld1_s8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int16x4x4_t test_vld1_s16_x4(int16_t const *a) {
   // CHECK-LABEL: test_vld1_s16_x4
   return vld1_s16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int32x2x4_t test_vld1_s32_x4(int32_t const *a) {
   // CHECK-LABEL: test_vld1_s32_x4
   return vld1_s32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64x1x4_t test_vld1_s64_x4(int64_t const *a) {
   // CHECK-LABEL: test_vld1_s64_x4
   return vld1_s64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float16x4x4_t test_vld1_f16_x4(float16_t const *a) {
   // CHECK-LABEL: test_vld1_f16_x4
   return vld1_f16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float32x2x4_t test_vld1_f32_x4(float32_t const *a) {
   // CHECK-LABEL: test_vld1_f32_x4
   return vld1_f32_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 float64x1x4_t test_vld1_f64_x4(float64_t const *a) {
   // CHECK-LABEL: test_vld1_f64_x4
   return vld1_f64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly8x8x4_t test_vld1_p8_x4(poly8_t const *a) {
   // CHECK-LABEL: test_vld1_p8_x4
   return vld1_p8_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly16x4x4_t test_vld1_p16_x4(poly16_t const *a) {
   // CHECK-LABEL: test_vld1_p16_x4
   return vld1_p16_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x4_t test_vld1_p64_x4(poly64_t const *a) {
   // CHECK-LABEL: test_vld1_p64_x4
   return vld1_p64_x4(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u8_x2(uint8_t *a, uint8x16x2_t b) {
-  // CHECK: test_vst1q_u8_x2
+  // CHECK-LABEL: test_vst1q_u8_x2
   vst1q_u8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u16_x2(uint16_t *a, uint16x8x2_t b) {
-  // CHECK: test_vst1q_u16_x2
+  // CHECK-LABEL: test_vst1q_u16_x2
   vst1q_u16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u32_x2(uint32_t *a, uint32x4x2_t b) {
-  // CHECK: test_vst1q_u32_x2
+  // CHECK-LABEL: test_vst1q_u32_x2
   vst1q_u32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u64_x2(uint64_t *a, uint64x2x2_t b) {
-  // CHECK: test_vst1q_u64_x2
+  // CHECK-LABEL: test_vst1q_u64_x2
   vst1q_u64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s8_x2(int8_t *a, int8x16x2_t b) {
-  // CHECK: test_vst1q_s8_x2
+  // CHECK-LABEL: test_vst1q_s8_x2
   vst1q_s8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s16_x2(int16_t *a, int16x8x2_t b) {
-  // CHECK: test_vst1q_s16_x2
+  // CHECK-LABEL: test_vst1q_s16_x2
   vst1q_s16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s32_x2(int32_t *a, int32x4x2_t b) {
-  // CHECK: test_vst1q_s32_x2
+  // CHECK-LABEL: test_vst1q_s32_x2
   vst1q_s32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s64_x2(int64_t *a, int64x2x2_t b) {
-  // CHECK: test_vst1q_s64_x2
+  // CHECK-LABEL: test_vst1q_s64_x2
   vst1q_s64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f16_x2(float16_t *a, float16x8x2_t b) {
-  // CHECK: test_vst1q_f16_x2
+  // CHECK-LABEL: test_vst1q_f16_x2
   vst1q_f16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f32_x2(float32_t *a, float32x4x2_t b) {
-  // CHECK: test_vst1q_f32_x2
+  // CHECK-LABEL: test_vst1q_f32_x2
   vst1q_f32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f64_x2(float64_t *a, float64x2x2_t b) {
-  // CHECK: test_vst1q_f64_x2
+  // CHECK-LABEL: test_vst1q_f64_x2
   vst1q_f64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p8_x2(poly8_t *a, poly8x16x2_t b) {
-  // CHECK: test_vst1q_p8_x2
+  // CHECK-LABEL: test_vst1q_p8_x2
   vst1q_p8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p16_x2(poly16_t *a, poly16x8x2_t b) {
-  // CHECK: test_vst1q_p16_x2
+  // CHECK-LABEL: test_vst1q_p16_x2
   vst1q_p16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p64_x2(poly64_t *a, poly64x2x2_t b) {
-  // CHECK: test_vst1q_p64_x2
+  // CHECK-LABEL: test_vst1q_p64_x2
   vst1q_p64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u8_x2(uint8_t *a, uint8x8x2_t b) {
-  // CHECK: test_vst1_u8_x2
+  // CHECK-LABEL: test_vst1_u8_x2
   vst1_u8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u16_x2(uint16_t *a, uint16x4x2_t b) {
-  // CHECK: test_vst1_u16_x2
+  // CHECK-LABEL: test_vst1_u16_x2
   vst1_u16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u32_x2(uint32_t *a, uint32x2x2_t b) {
-  // CHECK: test_vst1_u32_x2
+  // CHECK-LABEL: test_vst1_u32_x2
   vst1_u32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u64_x2(uint64_t *a, uint64x1x2_t b) {
-  // CHECK: test_vst1_u64_x2
+  // CHECK-LABEL: test_vst1_u64_x2
   vst1_u64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s8_x2(int8_t *a, int8x8x2_t b) {
-  // CHECK: test_vst1_s8_x2
+  // CHECK-LABEL: test_vst1_s8_x2
   vst1_s8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s16_x2(int16_t *a, int16x4x2_t b) {
-  // CHECK: test_vst1_s16_x2
+  // CHECK-LABEL: test_vst1_s16_x2
   vst1_s16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s32_x2(int32_t *a, int32x2x2_t b) {
-  // CHECK: test_vst1_s32_x2
+  // CHECK-LABEL: test_vst1_s32_x2
   vst1_s32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s64_x2(int64_t *a, int64x1x2_t b) {
-  // CHECK: test_vst1_s64_x2
+  // CHECK-LABEL: test_vst1_s64_x2
   vst1_s64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f16_x2(float16_t *a, float16x4x2_t b) {
-  // CHECK: test_vst1_f16_x2
+  // CHECK-LABEL: test_vst1_f16_x2
   vst1_f16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f32_x2(float32_t *a, float32x2x2_t b) {
-  // CHECK: test_vst1_f32_x2
+  // CHECK-LABEL: test_vst1_f32_x2
   vst1_f32_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f64_x2(float64_t *a, float64x1x2_t b) {
-  // CHECK: test_vst1_f64_x2
+  // CHECK-LABEL: test_vst1_f64_x2
   vst1_f64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p8_x2(poly8_t *a, poly8x8x2_t b) {
-  // CHECK: test_vst1_p8_x2
+  // CHECK-LABEL: test_vst1_p8_x2
   vst1_p8_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p16_x2(poly16_t *a, poly16x4x2_t b) {
-  // CHECK: test_vst1_p16_x2
+  // CHECK-LABEL: test_vst1_p16_x2
   vst1_p16_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p64_x2(poly64_t *a, poly64x1x2_t b) {
-  // CHECK: test_vst1_p64_x2
+  // CHECK-LABEL: test_vst1_p64_x2
   vst1_p64_x2(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u8_x3(uint8_t *a, uint8x16x3_t b) {
-  // CHECK: test_vst1q_u8_x3
+  // CHECK-LABEL: test_vst1q_u8_x3
   vst1q_u8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u16_x3(uint16_t *a, uint16x8x3_t b) {
-  // CHECK: test_vst1q_u16_x3
+  // CHECK-LABEL: test_vst1q_u16_x3
   vst1q_u16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u32_x3(uint32_t *a, uint32x4x3_t b) {
-  // CHECK: test_vst1q_u32_x3
+  // CHECK-LABEL: test_vst1q_u32_x3
   vst1q_u32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u64_x3(uint64_t *a, uint64x2x3_t b) {
-  // CHECK: test_vst1q_u64_x3
+  // CHECK-LABEL: test_vst1q_u64_x3
   vst1q_u64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s8_x3(int8_t *a, int8x16x3_t b) {
-  // CHECK: test_vst1q_s8_x3
+  // CHECK-LABEL: test_vst1q_s8_x3
   vst1q_s8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s16_x3(int16_t *a, int16x8x3_t b) {
-  // CHECK: test_vst1q_s16_x3
+  // CHECK-LABEL: test_vst1q_s16_x3
   vst1q_s16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s32_x3(int32_t *a, int32x4x3_t b) {
-  // CHECK: test_vst1q_s32_x3
+  // CHECK-LABEL: test_vst1q_s32_x3
   vst1q_s32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s64_x3(int64_t *a, int64x2x3_t b) {
-  // CHECK: test_vst1q_s64_x3
+  // CHECK-LABEL: test_vst1q_s64_x3
   vst1q_s64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f16_x3(float16_t *a, float16x8x3_t b) {
-  // CHECK: test_vst1q_f16_x3
+  // CHECK-LABEL: test_vst1q_f16_x3
   vst1q_f16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f32_x3(float32_t *a, float32x4x3_t b) {
-  // CHECK: test_vst1q_f32_x3
+  // CHECK-LABEL: test_vst1q_f32_x3
   vst1q_f32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f64_x3(float64_t *a, float64x2x3_t b) {
-  // CHECK: test_vst1q_f64_x3
+  // CHECK-LABEL: test_vst1q_f64_x3
   vst1q_f64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p8_x3(poly8_t *a, poly8x16x3_t b) {
-  // CHECK: test_vst1q_p8_x3
+  // CHECK-LABEL: test_vst1q_p8_x3
   vst1q_p8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p16_x3(poly16_t *a, poly16x8x3_t b) {
-  // CHECK: test_vst1q_p16_x3
+  // CHECK-LABEL: test_vst1q_p16_x3
   vst1q_p16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p64_x3(poly64_t *a, poly64x2x3_t b) {
-  // CHECK: test_vst1q_p64_x3
+  // CHECK-LABEL: test_vst1q_p64_x3
   vst1q_p64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u8_x3(uint8_t *a, uint8x8x3_t b) {
-  // CHECK: test_vst1_u8_x3
+  // CHECK-LABEL: test_vst1_u8_x3
   vst1_u8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u16_x3(uint16_t *a, uint16x4x3_t b) {
-  // CHECK: test_vst1_u16_x3
+  // CHECK-LABEL: test_vst1_u16_x3
   vst1_u16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u32_x3(uint32_t *a, uint32x2x3_t b) {
-  // CHECK: test_vst1_u32_x3
+  // CHECK-LABEL: test_vst1_u32_x3
   vst1_u32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u64_x3(uint64_t *a, uint64x1x3_t b) {
-  // CHECK: test_vst1_u64_x3
+  // CHECK-LABEL: test_vst1_u64_x3
   vst1_u64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s8_x3(int8_t *a, int8x8x3_t b) {
-  // CHECK: test_vst1_s8_x3
+  // CHECK-LABEL: test_vst1_s8_x3
   vst1_s8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s16_x3(int16_t *a, int16x4x3_t b) {
-  // CHECK: test_vst1_s16_x3
+  // CHECK-LABEL: test_vst1_s16_x3
   vst1_s16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s32_x3(int32_t *a, int32x2x3_t b) {
-  // CHECK: test_vst1_s32_x3
+  // CHECK-LABEL: test_vst1_s32_x3
   vst1_s32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s64_x3(int64_t *a, int64x1x3_t b) {
-  // CHECK: test_vst1_s64_x3
+  // CHECK-LABEL: test_vst1_s64_x3
   vst1_s64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f16_x3(float16_t *a, float16x4x3_t b) {
-  // CHECK: test_vst1_f16_x3
+  // CHECK-LABEL: test_vst1_f16_x3
   vst1_f16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f32_x3(float32_t *a, float32x2x3_t b) {
-  // CHECK: test_vst1_f32_x3
+  // CHECK-LABEL: test_vst1_f32_x3
   vst1_f32_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f64_x3(float64_t *a, float64x1x3_t b) {
-  // CHECK: test_vst1_f64_x3
+  // CHECK-LABEL: test_vst1_f64_x3
   vst1_f64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p8_x3(poly8_t *a, poly8x8x3_t b) {
-  // CHECK: test_vst1_p8_x3
+  // CHECK-LABEL: test_vst1_p8_x3
   vst1_p8_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p16_x3(poly16_t *a, poly16x4x3_t b) {
-  // CHECK: test_vst1_p16_x3
+  // CHECK-LABEL: test_vst1_p16_x3
   vst1_p16_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p64_x3(poly64_t *a, poly64x1x3_t b) {
-  // CHECK: test_vst1_p64_x3
+  // CHECK-LABEL: test_vst1_p64_x3
   vst1_p64_x3(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
-  // [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u8_x4(uint8_t *a, uint8x16x4_t b) {
-  // CHECK: test_vst1q_u8_x4
+  // CHECK-LABEL: test_vst1q_u8_x4
   vst1q_u8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u16_x4(uint16_t *a, uint16x8x4_t b) {
-  // CHECK: test_vst1q_u16_x4
+  // CHECK-LABEL: test_vst1q_u16_x4
   vst1q_u16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u32_x4(uint32_t *a, uint32x4x4_t b) {
-  // CHECK: test_vst1q_u32_x4
+  // CHECK-LABEL: test_vst1q_u32_x4
   vst1q_u32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_u64_x4(uint64_t *a, uint64x2x4_t b) {
-  // CHECK: test_vst1q_u64_x4
+  // CHECK-LABEL: test_vst1q_u64_x4
   vst1q_u64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s8_x4(int8_t *a, int8x16x4_t b) {
-  // CHECK: test_vst1q_s8_x4
+  // CHECK-LABEL: test_vst1q_s8_x4
   vst1q_s8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s16_x4(int16_t *a, int16x8x4_t b) {
-  // CHECK: test_vst1q_s16_x4
+  // CHECK-LABEL: test_vst1q_s16_x4
   vst1q_s16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s32_x4(int32_t *a, int32x4x4_t b) {
-  // CHECK: test_vst1q_s32_x4
+  // CHECK-LABEL: test_vst1q_s32_x4
   vst1q_s32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_s64_x4(int64_t *a, int64x2x4_t b) {
-  // CHECK: test_vst1q_s64_x4
+  // CHECK-LABEL: test_vst1q_s64_x4
   vst1q_s64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f16_x4(float16_t *a, float16x8x4_t b) {
-  // CHECK: test_vst1q_f16_x4
+  // CHECK-LABEL: test_vst1q_f16_x4
   vst1q_f16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f32_x4(float32_t *a, float32x4x4_t b) {
-  // CHECK: test_vst1q_f32_x4
+  // CHECK-LABEL: test_vst1q_f32_x4
   vst1q_f32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_f64_x4(float64_t *a, float64x2x4_t b) {
-  // CHECK: test_vst1q_f64_x4
+  // CHECK-LABEL: test_vst1q_f64_x4
   vst1q_f64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p8_x4(poly8_t *a, poly8x16x4_t b) {
-  // CHECK: test_vst1q_p8_x4
+  // CHECK-LABEL: test_vst1q_p8_x4
   vst1q_p8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p16_x4(poly16_t *a, poly16x8x4_t b) {
-  // CHECK: test_vst1q_p16_x4
+  // CHECK-LABEL: test_vst1q_p16_x4
   vst1q_p16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p64_x4(poly64_t *a, poly64x2x4_t b) {
-  // CHECK: test_vst1q_p64_x4
+  // CHECK-LABEL: test_vst1q_p64_x4
   vst1q_p64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u8_x4(uint8_t *a, uint8x8x4_t b) {
-  // CHECK: test_vst1_u8_x4
+  // CHECK-LABEL: test_vst1_u8_x4
   vst1_u8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u16_x4(uint16_t *a, uint16x4x4_t b) {
-  // CHECK: test_vst1_u16_x4
+  // CHECK-LABEL: test_vst1_u16_x4
   vst1_u16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u32_x4(uint32_t *a, uint32x2x4_t b) {
-  // CHECK: test_vst1_u32_x4
+  // CHECK-LABEL: test_vst1_u32_x4
   vst1_u32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_u64_x4(uint64_t *a, uint64x1x4_t b) {
-  // CHECK: test_vst1_u64_x4
+  // CHECK-LABEL: test_vst1_u64_x4
   vst1_u64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s8_x4(int8_t *a, int8x8x4_t b) {
-  // CHECK: test_vst1_s8_x4
+  // CHECK-LABEL: test_vst1_s8_x4
   vst1_s8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s16_x4(int16_t *a, int16x4x4_t b) {
-  // CHECK: test_vst1_s16_x4
+  // CHECK-LABEL: test_vst1_s16_x4
   vst1_s16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s32_x4(int32_t *a, int32x2x4_t b) {
-  // CHECK: test_vst1_s32_x4
+  // CHECK-LABEL: test_vst1_s32_x4
   vst1_s32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_s64_x4(int64_t *a, int64x1x4_t b) {
-  // CHECK: test_vst1_s64_x4
+  // CHECK-LABEL: test_vst1_s64_x4
   vst1_s64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f16_x4(float16_t *a, float16x4x4_t b) {
-  // CHECK: test_vst1_f16_x4
+  // CHECK-LABEL: test_vst1_f16_x4
   vst1_f16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f32_x4(float32_t *a, float32x2x4_t b) {
-  // CHECK: test_vst1_f32_x4
+  // CHECK-LABEL: test_vst1_f32_x4
   vst1_f32_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_f64_x4(float64_t *a, float64x1x4_t b) {
-  // CHECK: test_vst1_f64_x4
+  // CHECK-LABEL: test_vst1_f64_x4
   vst1_f64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p8_x4(poly8_t *a, poly8x8x4_t b) {
-  // CHECK: test_vst1_p8_x4
+  // CHECK-LABEL: test_vst1_p8_x4
   vst1_p8_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p16_x4(poly16_t *a, poly16x4x4_t b) {
-  // CHECK: test_vst1_p16_x4
+  // CHECK-LABEL: test_vst1_p16_x4
   vst1_p16_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h ?}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p64_x4(poly64_t *a, poly64x1x4_t b) {
-  // CHECK: test_vst1_p64_x4
+  // CHECK-LABEL: test_vst1_p64_x4
   vst1_p64_x4(a, b);
-  // CHECK: st1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ ?v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d ?}}}, [{{x[0-9]+|sp}}]
 }
 
 int64_t test_vceqd_s64(int64_t a, int64_t b) {
-// CHECK: test_vceqd_s64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vceqd_s64
+// CHECK: {{cmeq d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vceqd_s64(a, b);
 }
 
 uint64_t test_vceqd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vceqd_u64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vceqd_u64
+// CHECK: {{cmeq d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vceqd_u64(a, b);
-} 
+}
 
 int64_t test_vceqzd_s64(int64_t a) {
-// CHECK: test_vceqzd_s64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vceqzd_s64
+// CHECK: {{cmeq d[0-9]+, d[0-9]+, #0x0|cmp x0, #0}}
   return (int64_t)vceqzd_s64(a);
 }
 
 int64_t test_vceqzd_u64(int64_t a) {
-// CHECK: test_vceqzd_u64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vceqzd_u64
+// CHECK: {{cmeq d[0-9]+, d[0-9]+, #0x0|cmp x0, #0}}
   return (int64_t)vceqzd_u64(a);
 }
 
 int64_t test_vcged_s64(int64_t a, int64_t b) {
-// CHECK: test_vcged_s64
-// CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcged_s64
+// CHECK: {{cmge d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vcged_s64(a, b);
 }
 
 uint64_t test_vcged_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vcged_u64
-// CHECK: cmhs {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcged_u64
+// CHECK: {{cmhs d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
     return (uint64_t)vcged_u64(a, b);
 }
 
 int64_t test_vcgezd_s64(int64_t a) {
-// CHECK: test_vcgezd_s64
-// CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vcgezd_s64
+// CHECK: {{cmge d[0-9]+, d[0-9]+, #0x0|eor x0, x[0-9]+, x0, asr #63}}
   return (int64_t)vcgezd_s64(a);
 }
 
 int64_t test_vcgtd_s64(int64_t a, int64_t b) {
-// CHECK: test_vcgtd_s64
-// CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcgtd_s64
+// CHECK: {{cmgt d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vcgtd_s64(a, b);
 }
 
 uint64_t test_vcgtd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vcgtd_u64
-// CHECK: cmhi {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcgtd_u64
+// CHECK: {{cmhi d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (uint64_t)vcgtd_u64(a, b);
 }
 
 int64_t test_vcgtzd_s64(int64_t a) {
-// CHECK: test_vcgtzd_s64
-// CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vcgtzd_s64
+// CHECK: {{cmgt d[0-9]+, d[0-9]+, #0x0|cmp x0, #0}}
   return (int64_t)vcgtzd_s64(a);
 }
 
 int64_t test_vcled_s64(int64_t a, int64_t b) {
-// CHECK: test_vcled_s64
-// CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcled_s64
+// CHECK: {{cmge d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vcled_s64(a, b);
 }
 
 uint64_t test_vcled_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vcled_u64
-// CHECK: cmhs {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
-  return (uint64_t)vcled_u64(a, 0);
+// CHECK-LABEL: test_vcled_u64
+// CHECK: {{cmhs d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
+  return (uint64_t)vcled_u64(a, b);
 }
 
 int64_t test_vclezd_s64(int64_t a) {
-// CHECK: test_vclezd_s64
-// CHECK: cmle {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vclezd_s64
+// CHECK: {{cmle d[0-9]+, d[0-9]+, #0x0|cmp x0, #1}}
   return (int64_t)vclezd_s64(a);
 }
 
 int64_t test_vcltd_s64(int64_t a, int64_t b) {
-// CHECK: test_vcltd_s64
-// CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcltd_s64
+// CHECK: {{cmgt d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (int64_t)vcltd_s64(a, b);
 }
 
 uint64_t test_vcltd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vcltd_u64
-// CHECK: cmhi {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcltd_u64
+// CHECK: {{cmhi d[0-9]+, d[0-9]+, d[0-9]+|cmp x0, x1}}
   return (uint64_t)vcltd_u64(a, b);
 }
 
 int64_t test_vcltzd_s64(int64_t a) {
-// CHECK: test_vcltzd_s64
-// CHECK: cmlt {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vcltzd_s64
+// CHECK: {{cmlt d[0-9]+, d[0-9]+, #0x0|asr x0, x0, #63}}
   return (int64_t)vcltzd_s64(a);
 }
 
 int64_t test_vtstd_s64(int64_t a, int64_t b) {
-// CHECK: test_vtstd_s64
-// CHECK: cmtst {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vtstd_s64
+// CHECK: {{cmtst d[0-9]+, d[0-9]+, d[0-9]+|and x[0-9]+, x1, x0}}
   return (int64_t)vtstd_s64(a, b);
 }
 
 uint64_t test_vtstd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vtstd_u64
-// CHECK: cmtst {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vtstd_u64
+// CHECK: {{cmtst d[0-9]+, d[0-9]+, d[0-9]+|and x[0-9]+, x1, x0}}
   return (uint64_t)vtstd_u64(a, b);
 }
 
 int64_t test_vabsd_s64(int64_t a) {
-// CHECK: test_vabsd_s64
+// CHECK-LABEL: test_vabsd_s64
 // CHECK: abs {{d[0-9]+}}, {{d[0-9]+}}
   return (int64_t)vabsd_s64(a);
 }
 
 int8_t test_vqabsb_s8(int8_t a) {
-// CHECK: test_vqabsb_s8
-// CHECK: sqabs {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK-LABEL: test_vqabsb_s8
+// CHECK: sqabs {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
   return (int8_t)vqabsb_s8(a);
 }
 
 int16_t test_vqabsh_s16(int16_t a) {
-// CHECK: test_vqabsh_s16
-// CHECK: sqabs {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqabsh_s16
+// CHECK: sqabs {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
   return (int16_t)vqabsh_s16(a);
 }
 
 int32_t test_vqabss_s32(int32_t a) {
-// CHECK: test_vqabss_s32
+// CHECK-LABEL: test_vqabss_s32
 // CHECK: sqabs {{s[0-9]+}}, {{s[0-9]+}}
   return (int32_t)vqabss_s32(a);
 }
 
 int64_t test_vqabsd_s64(int64_t a) {
-// CHECK: test_vqabsd_s64
+// CHECK-LABEL: test_vqabsd_s64
 // CHECK: sqabs {{d[0-9]+}}, {{d[0-9]+}}
   return (int64_t)vqabsd_s64(a);
 }
 
 int64_t test_vnegd_s64(int64_t a) {
-// CHECK: test_vnegd_s64
-// CHECK: neg {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vnegd_s64
+// CHECK: neg {{[xd][0-9]+}}, {{[xd][0-9]+}}
   return (int64_t)vnegd_s64(a);
 }
 
 int8_t test_vqnegb_s8(int8_t a) {
-// CHECK: test_vqnegb_s8
-// CHECK: sqneg {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK-LABEL: test_vqnegb_s8
+// CHECK: sqneg {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
   return (int8_t)vqnegb_s8(a);
 }
 
 int16_t test_vqnegh_s16(int16_t a) {
-// CHECK: test_vqnegh_s16
-// CHECK: sqneg {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqnegh_s16
+// CHECK: sqneg {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
   return (int16_t)vqnegh_s16(a);
 }
 
 int32_t test_vqnegs_s32(int32_t a) {
-// CHECK: test_vqnegs_s32
+// CHECK-LABEL: test_vqnegs_s32
 // CHECK: sqneg {{s[0-9]+}}, {{s[0-9]+}}
   return (int32_t)vqnegs_s32(a);
 }
 
 int64_t test_vqnegd_s64(int64_t a) {
-// CHECK: test_vqnegd_s64
+// CHECK-LABEL: test_vqnegd_s64
 // CHECK: sqneg {{d[0-9]+}}, {{d[0-9]+}}
   return (int64_t)vqnegd_s64(a);
 }
 
 int8_t test_vuqaddb_s8(int8_t a, int8_t b) {
-// CHECK: test_vuqaddb_s8
-// CHECK: suqadd {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK-LABEL: test_vuqaddb_s8
+// CHECK: suqadd {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
   return (int8_t)vuqaddb_s8(a, b);
 }
 
 int16_t test_vuqaddh_s16(int16_t a, int16_t b) {
-// CHECK: test_vuqaddh_s16
-// CHECK: suqadd {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vuqaddh_s16
+// CHECK: suqadd {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
   return (int16_t)vuqaddh_s16(a, b);
 }
 
 int32_t test_vuqadds_s32(int32_t a, int32_t b) {
-// CHECK: test_vuqadds_s32
+// CHECK-LABEL: test_vuqadds_s32
 // CHECK: suqadd {{s[0-9]+}}, {{s[0-9]+}}
   return (int32_t)vuqadds_s32(a, b);
 }
 
 int64_t test_vuqaddd_s64(int64_t a, int64_t b) {
-// CHECK: test_vuqaddd_s64
+// CHECK-LABEL: test_vuqaddd_s64
 // CHECK: suqadd {{d[0-9]+}}, {{d[0-9]+}}
   return (int64_t)vuqaddd_s64(a, b);
 }
 
 uint8_t test_vsqaddb_u8(uint8_t a, uint8_t b) {
-// CHECK: test_vsqaddb_u8
-// CHECK: usqadd {{b[0-9]+}}, {{b[0-9]+}}
+// CHECK-LABEL: test_vsqaddb_u8
+// CHECK: usqadd {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}
   return (uint8_t)vsqaddb_u8(a, b);
 }
 
 uint16_t test_vsqaddh_u16(uint16_t a, uint16_t b) {
-// CHECK: test_vsqaddh_u16
-// CHECK: usqadd {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vsqaddh_u16
+// CHECK: usqadd {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
   return (uint16_t)vsqaddh_u16(a, b);
 }
 
 uint32_t test_vsqadds_u32(uint32_t a, uint32_t b) {
-// CHECK: test_vsqadds_u32
+// CHECK-LABEL: test_vsqadds_u32
 // CHECK: usqadd {{s[0-9]+}}, {{s[0-9]+}}
   return (uint32_t)vsqadds_u32(a, b);
 }
 
 uint64_t test_vsqaddd_u64(uint64_t a, uint64_t b) {
-// CHECK: test_vsqaddd_u64
+// CHECK-LABEL: test_vsqaddd_u64
 // CHECK: usqadd {{d[0-9]+}}, {{d[0-9]+}}
   return (uint64_t)vsqaddd_u64(a, b);
 }
 
 int32_t test_vqdmlalh_s16(int32_t a, int16_t b, int16_t c) {
-// CHECK: test_vqdmlalh_s16
-// CHECK: sqdmlal {{s[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-AARCH64-LABEL: test_vqdmlalh_s16
+// CHECK-AARCH64: sqdmlal {{s[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+
+// CHECK-ARM64-LABEL: test_vqdmlalh_s16
+// CHECK-ARM64: sqdmull v[[PROD:[0-9]+]].4s, {{v[0-9]+.4h}}, {{v[0-9]+.4h}}
+// CHECK-ARM64: sqadd {{s[0-9]+}}, {{s[0-9]+}}, s[[PROD]]
   return (int32_t)vqdmlalh_s16(a, b, c);
 }
 
 int64_t test_vqdmlals_s32(int64_t a, int32_t b, int32_t c) {
-// CHECK: test_vqdmlals_s32
+// CHECK-LABEL: test_vqdmlals_s32
 // CHECK: sqdmlal {{d[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
   return (int64_t)vqdmlals_s32(a, b, c);
 }
 
 int32_t test_vqdmlslh_s16(int32_t a, int16_t b, int16_t c) {
-// CHECK: test_vqdmlslh_s16
-// CHECK: sqdmlsl {{s[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-AARCH64-LABEL: test_vqdmlslh_s16
+// CHECK-AARCH64: sqdmlsl {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
+
+// CHECK-ARM64-LABEL: test_vqdmlslh_s16
+// CHECK-ARM64: sqdmull v[[PROD:[0-9]+]].4s, {{v[0-9]+.4h}}, {{v[0-9]+.4h}}
+// CHECK-ARM64: sqsub {{s[0-9]+}}, {{s[0-9]+}}, s[[PROD]]
   return (int32_t)vqdmlslh_s16(a, b, c);
 }
 
 int64_t test_vqdmlsls_s32(int64_t a, int32_t b, int32_t c) {
-// CHECK: test_vqdmlsls_s32
+// CHECK-LABEL: test_vqdmlsls_s32
 // CHECK: sqdmlsl {{d[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
   return (int64_t)vqdmlsls_s32(a, b, c);
 }
 
 int32_t test_vqdmullh_s16(int16_t a, int16_t b) {
-// CHECK: test_vqdmullh_s16
-// CHECK: sqdmull {{s[0-9]+}}, {{h[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqdmullh_s16
+// CHECK: sqdmull {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}
   return (int32_t)vqdmullh_s16(a, b);
 }
 
 int64_t test_vqdmulls_s32(int32_t a, int32_t b) {
-// CHECK: test_vqdmulls_s32
+// CHECK-LABEL: test_vqdmulls_s32
 // CHECK: sqdmull {{d[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
   return (int64_t)vqdmulls_s32(a, b);
 }
 
 int8_t test_vqmovunh_s16(int16_t a) {
-// CHECK: test_vqmovunh_s16
-// CHECK: sqxtun {{b[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqmovunh_s16
+// CHECK: sqxtun {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}
   return (int8_t)vqmovunh_s16(a);
 }
 
 int16_t test_vqmovuns_s32(int32_t a) {
-// CHECK: test_vqmovuns_s32
-// CHECK: sqxtun {{h[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vqmovuns_s32
+// CHECK: sqxtun {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}
   return (int16_t)vqmovuns_s32(a);
 }
 
 int32_t test_vqmovund_s64(int64_t a) {
-// CHECK: test_vqmovund_s64
+// CHECK-LABEL: test_vqmovund_s64
 // CHECK: sqxtun {{s[0-9]+}}, {{d[0-9]+}}
   return (int32_t)vqmovund_s64(a);
 }
 
 int8_t test_vqmovnh_s16(int16_t a) {
-// CHECK: test_vqmovnh_s16
-// CHECK: sqxtn {{b[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqmovnh_s16
+// CHECK: sqxtn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}
   return (int8_t)vqmovnh_s16(a);
 }
 
 int16_t test_vqmovns_s32(int32_t a) {
-// CHECK: test_vqmovns_s32
-// CHECK: sqxtn {{h[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vqmovns_s32
+// CHECK: sqxtn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}
   return (int16_t)vqmovns_s32(a);
 }
 
 int32_t test_vqmovnd_s64(int64_t a) {
-// CHECK: test_vqmovnd_s64
+// CHECK-LABEL: test_vqmovnd_s64
 // CHECK: sqxtn {{s[0-9]+}}, {{d[0-9]+}}
   return (int32_t)vqmovnd_s64(a);
 }
 
 int8_t test_vqmovnh_u16(int16_t a) {
-// CHECK: test_vqmovnh_u16
-// CHECK: uqxtn {{b[0-9]+}}, {{h[0-9]+}}
+// CHECK-LABEL: test_vqmovnh_u16
+// CHECK: uqxtn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}
   return (int8_t)vqmovnh_u16(a);
 }
 
 int16_t test_vqmovns_u32(int32_t a) {
-// CHECK: test_vqmovns_u32
-// CHECK: uqxtn {{h[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vqmovns_u32
+// CHECK: uqxtn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}
   return (int16_t)vqmovns_u32(a);
 }
 
 int32_t test_vqmovnd_u64(int64_t a) {
-// CHECK: test_vqmovnd_u64
+// CHECK-LABEL: test_vqmovnd_u64
 // CHECK: uqxtn {{s[0-9]+}}, {{d[0-9]+}}
   return (int32_t)vqmovnd_u64(a);
 }
 
 uint32_t test_vceqs_f32(float32_t a, float32_t b) {
-// CHECK: test_vceqs_f32
-// CHECK: fcmeq {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vceqs_f32
+// CHECK: {{fcmeq s0, s0, s1|fcmp s0, s1}}
   return (uint32_t)vceqs_f32(a, b);
 }
 
 uint64_t test_vceqd_f64(float64_t a, float64_t b) {
-// CHECK: test_vceqd_f64
-// CHECK: fcmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vceqd_f64
+// CHECK: {{fcmeq d0, d0, d1|fcmp d0, d1}}
   return (uint64_t)vceqd_f64(a, b);
 }
 
 uint32_t test_vceqzs_f32(float32_t a) {
-// CHECK: test_vceqzs_f32
-// CHECK: fcmeq {{s[0-9]+}}, {{s[0-9]+}}, #0.0
+// CHECK-LABEL: test_vceqzs_f32
+// CHECK: {{fcmeq s0, s0, #0.0|fcmp s0, #0.0}}
   return (uint32_t)vceqzs_f32(a);
 }
 
 uint64_t test_vceqzd_f64(float64_t a) {
-// CHECK: test_vceqzd_f64
-// CHECK: fcmeq {{d[0-9]+}}, {{d[0-9]+}}, #0.0
+// CHECK-LABEL: test_vceqzd_f64
+// CHECK: {{fcmeq d0, d0, #0.0|fcmp d0, #0.0}}
   return (uint64_t)vceqzd_f64(a);
 }
 
 uint32_t test_vcges_f32(float32_t a, float32_t b) {
-// CHECK: test_vcges_f32
-// CHECK: fcmge {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcges_f32
+// CHECK: {{fcmge s0, s0, s1|fcmp s0, s1}}
   return (uint32_t)vcges_f32(a, b);
 }
 
 uint64_t test_vcged_f64(float64_t a, float64_t b) {
-// CHECK: test_vcged_f64
-// CHECK: fcmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcged_f64
+// CHECK: {{fcmge d0, d0, d1|fcmp d0, d1}}
   return (uint64_t)vcged_f64(a, b);
 }
 
 uint32_t test_vcgezs_f32(float32_t a) {
-// CHECK: test_vcgezs_f32
-// CHECK: fcmge {{s[0-9]+}}, {{s[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcgezs_f32
+// CHECK: {{fcmge s0, s0, #0.0|fcmp s0, #0.0}}
   return (uint32_t)vcgezs_f32(a);
 }
 
 uint64_t test_vcgezd_f64(float64_t a) {
-// CHECK: test_vcgezd_f64
-// CHECK: fcmge {{d[0-9]+}}, {{d[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcgezd_f64
+// CHECK: {{fcmge d0, d0, #0.0|fcmp d0, #0.0}}
   return (uint64_t)vcgezd_f64(a);
 }
 
 uint32_t test_vcgts_f32(float32_t a, float32_t b) {
-// CHECK: test_vcgts_f32
-// CHECK: fcmgt {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcgts_f32
+// CHECK: {{fcmgt s0, s0, s1|fcmp s0, s1}}
   return (uint32_t)vcgts_f32(a, b);
 }
 
 uint64_t test_vcgtd_f64(float64_t a, float64_t b) {
-// CHECK: test_vcgtd_f64
-// CHECK: fcmgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcgtd_f64
+// CHECK: {{fcmgt d0, d0, d1|fcmp d0, d1}}
   return (uint64_t)vcgtd_f64(a, b);
 }
 
 uint32_t test_vcgtzs_f32(float32_t a) {
-// CHECK: test_vcgtzs_f32
-// CHECK: fcmgt {{s[0-9]+}}, {{s[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcgtzs_f32
+// CHECK: {{fcmgt s0, s0, #0.0|fcmp s0, #0.0}}
   return (uint32_t)vcgtzs_f32(a);
 }
 
 uint64_t test_vcgtzd_f64(float64_t a) {
-// CHECK: test_vcgtzd_f64
-// CHECK: fcmgt {{d[0-9]+}}, {{d[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcgtzd_f64
+// CHECK: {{fcmgt d0, d0, #0.0|fcmp d0, #0.0}}
   return (uint64_t)vcgtzd_f64(a);
 }
 
 uint32_t test_vcles_f32(float32_t a, float32_t b) {
-// CHECK: test_vcles_f32
-// CHECK: fcmge {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcles_f32
+// CHECK: {{fcmge s0, s1, s0|fcmp s0, s1}}
   return (uint32_t)vcles_f32(a, b);
 }
 
 uint64_t test_vcled_f64(float64_t a, float64_t b) {
-// CHECK: test_vcled_f64
-// CHECK: fcmge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcled_f64
+// CHECK: {{fcmge d0, d1, d0|fcmp d0, d1}}
   return (uint64_t)vcled_f64(a, b);
 }
 
 uint32_t test_vclezs_f32(float32_t a) {
-// CHECK: test_vclezs_f32
-// CHECK: fcmle {{s[0-9]+}}, {{s[0-9]+}}, #0.0
+// CHECK-LABEL: test_vclezs_f32
+// CHECK: {{fcmle s0, s0, #0.0|fcmp s0, #0.0}}
   return (uint32_t)vclezs_f32(a);
 }
 
 uint64_t test_vclezd_f64(float64_t a) {
-// CHECK: test_vclezd_f64
-// CHECK: fcmle {{d[0-9]+}}, {{d[0-9]+}}, #0.0
+// CHECK-LABEL: test_vclezd_f64
+// CHECK: {{fcmle d0, d0, #0.0|fcmp d0, #0.0}}
   return (uint64_t)vclezd_f64(a);
 }
 
 uint32_t test_vclts_f32(float32_t a, float32_t b) {
-// CHECK: test_vclts_f32
-// CHECK: fcmgt {{s[0-9]+}}, s1, s0
+// CHECK-LABEL: test_vclts_f32
+// CHECK: {{fcmgt s0, s1, s0|fcmp s0, s1}}
   return (uint32_t)vclts_f32(a, b);
 }
 
 uint64_t test_vcltd_f64(float64_t a, float64_t b) {
-// CHECK: test_vcltd_f64
-// CHECK: fcmgt {{d[0-9]+}}, d1, d0
+// CHECK-LABEL: test_vcltd_f64
+// CHECK: {{fcmgt d0, d1, d0|fcmp d0, d1}}
   return (uint64_t)vcltd_f64(a, b);
 }
 
 uint32_t test_vcltzs_f32(float32_t a) {
-// CHECK: test_vcltzs_f32
-// CHECK: fcmlt {{s[0-9]+}}, {{s[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcltzs_f32
+// CHECK: {{fcmlt s0, s0, #0.0|fcmp s0, #0.0}}
   return (uint32_t)vcltzs_f32(a);
 }
 
 uint64_t test_vcltzd_f64(float64_t a) {
-// CHECK: test_vcltzd_f64
-// CHECK: fcmlt {{d[0-9]+}}, {{d[0-9]+}}, #0.0
+// CHECK-LABEL: test_vcltzd_f64
+// CHECK: {{fcmlt d0, d0, #0.0|fcmp d0, #0.0}}
   return (uint64_t)vcltzd_f64(a);
 }
 
 uint32_t test_vcages_f32(float32_t a, float32_t b) {
-// CHECK: test_vcages_f32
-// CHECK: facge {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcages_f32
+// CHECK: facge s0, s0, s1
   return (uint32_t)vcages_f32(a, b);
 }
 
 uint64_t test_vcaged_f64(float64_t a, float64_t b) {
-// CHECK: test_vcaged_f64
-// CHECK: facge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcaged_f64
+// CHECK: facge d0, d0, d1
   return (uint64_t)vcaged_f64(a, b);
 }
 
 uint32_t test_vcagts_f32(float32_t a, float32_t b) {
-// CHECK: test_vcagts_f32
-// CHECK: facgt {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcagts_f32
+// CHECK: facgt s0, s0, s1
   return (uint32_t)vcagts_f32(a, b);
 }
 
 uint64_t test_vcagtd_f64(float64_t a, float64_t b) {
-// CHECK: test_vcagtd_f64
-// CHECK: facgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcagtd_f64
+// CHECK: facgt d0, d0, d1
   return (uint64_t)vcagtd_f64(a, b);
 }
 
 uint32_t test_vcales_f32(float32_t a, float32_t b) {
-// CHECK: test_vcales_f32
-// CHECK: facge {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcales_f32
+// CHECK: facge s0, s1, s0
   return (uint32_t)vcales_f32(a, b);
 }
 
 uint64_t test_vcaled_f64(float64_t a, float64_t b) {
-// CHECK: test_vcaled_f64
-// CHECK: facge {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcaled_f64
+// CHECK: facge d0, d1, d0
   return (uint64_t)vcaled_f64(a, b);
 }
 
 uint32_t test_vcalts_f32(float32_t a, float32_t b) {
-// CHECK: test_vcalts_f32
-// CHECK: facgt {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+// CHECK-LABEL: test_vcalts_f32
+// CHECK: facgt s0, s1, s0
   return (uint32_t)vcalts_f32(a, b);
 }
 
 uint64_t test_vcaltd_f64(float64_t a, float64_t b) {
-// CHECK: test_vcaltd_f64
-// CHECK: facgt {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+// CHECK-LABEL: test_vcaltd_f64
+// CHECK: facgt d0, d1, d0
   return (uint64_t)vcaltd_f64(a, b);
 }
 
 int64_t test_vshrd_n_s64(int64_t a) {
 // CHECK-LABEL: test_vshrd_n_s64
-// CHECK: sshr {{d[0-9]+}}, {{d[0-9]+}}, #1
+// CHECK: {{sshr d[0-9]+, d[0-9]+, #1|asr x0, x0, #1}}
   return (int64_t)vshrd_n_s64(a, 1);
 }
 
@@ -8738,7 +8573,7 @@
 
 uint64_t test_vshrd_n_u64(uint64_t a) {
 // CHECK-LABEL: test_vshrd_n_u64
-// CHECK: ushr {{d[0-9]+}}, {{d[0-9]+}}, #64
+// CHECK: {{ushr d[0-9]+, d[0-9]+, #64|lsr x0, x0, #63}}
   return (uint64_t)vshrd_n_u64(a, 64);
 }
 
@@ -8755,7 +8590,7 @@
 }
 
 int64x1_t test_vrshr_n_s64(int64x1_t a) {
-// CHECK: test_vrshr_n_s64
+// CHECK-LABEL: test_vrshr_n_s64
 // CHECK: srshr d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vrshr_n_s64(a, 1);
 }
@@ -8767,43 +8602,43 @@
 }
 
 uint64x1_t test_vrshr_n_u64(uint64x1_t a) {
-// CHECK: test_vrshr_n_u64
+// CHECK-LABEL: test_vrshr_n_u64
 // CHECK: urshr d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vrshr_n_u64(a, 1);
 }
 
 int64_t test_vsrad_n_s64(int64_t a, int64_t b) {
 // CHECK-LABEL: test_vsrad_n_s64
-// CHECK: ssra {{d[0-9]+}}, {{d[0-9]+}}, #63
+// CHECK: {{ssra d[0-9]+, d[0-9]+, #63|add x0, x0, x1, asr #63}}
   return (int64_t)vsrad_n_s64(a, b, 63);
 }
 
 int64x1_t test_vsra_n_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vsra_n_s64
+// CHECK-LABEL: test_vsra_n_s64
 // CHECK: ssra d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsra_n_s64(a, b, 1);
 }
 
 uint64_t test_vsrad_n_u64(uint64_t a, uint64_t b) {
 // CHECK-LABEL: test_vsrad_n_u64
-// CHECK: usra {{d[0-9]+}}, {{d[0-9]+}}, #63
+// CHECK: {{usra d[0-9]+, d[0-9]+, #63|add x0, x0, x1, lsr #63}}
   return (uint64_t)vsrad_n_u64(a, b, 63);
 }
 
 uint64x1_t test_vsra_n_u64(uint64x1_t a, uint64x1_t b) {
-// CHECK: test_vsra_n_u64
+// CHECK-LABEL: test_vsra_n_u64
 // CHECK: usra d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsra_n_u64(a, b, 1);
 }
 
 int64_t test_vrsrad_n_s64(int64_t a, int64_t b) {
 // CHECK-LABEL: test_vrsrad_n_s64
-// CHECK: srsra {{d[0-9]+}}, {{d[0-9]+}}, #63
+// CHECK: {{srsra d[0-9]+, d[0-9]+, #63}}
   return (int64_t)vrsrad_n_s64(a, b, 63);
 }
 
 int64x1_t test_vrsra_n_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vrsra_n_s64
+// CHECK-LABEL: test_vrsra_n_s64
 // CHECK: srsra d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vrsra_n_s64(a, b, 1);
 }
@@ -8815,43 +8650,43 @@
 }
 
 uint64x1_t test_vrsra_n_u64(uint64x1_t a, uint64x1_t b) {
-// CHECK: test_vrsra_n_u64
+// CHECK-LABEL: test_vrsra_n_u64
 // CHECK: ursra d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vrsra_n_u64(a, b, 1);
 }
 
 int64_t test_vshld_n_s64(int64_t a) {
 // CHECK-LABEL: test_vshld_n_s64
-// CHECK: shl {{d[0-9]+}}, {{d[0-9]+}}, #0
-  return (int64_t)vshld_n_s64(a, 0);
+// CHECK: {{shl d[0-9]+, d[0-9]+, #1|lsl x0, x0, #1}}
+  return (int64_t)vshld_n_s64(a, 1);
 }
 int64x1_t test_vshl_n_s64(int64x1_t a) {
-// CHECK: test_vshl_n_s64
+// CHECK-LABEL: test_vshl_n_s64
 // CHECK: shl d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vshl_n_s64(a, 1);
 }
 
 uint64_t test_vshld_n_u64(uint64_t a) {
 // CHECK-LABEL: test_vshld_n_u64
-// CHECK: shl {{d[0-9]+}}, {{d[0-9]+}}, #63
+// CHECK: {{shl d[0-9]+, d[0-9]+, #63|lsl x0, x0, #63}}
   return (uint64_t)vshld_n_u64(a, 63);
 }
 
 uint64x1_t test_vshl_n_u64(uint64x1_t a) {
-// CHECK: test_vshl_n_u64
+// CHECK-LABEL: test_vshl_n_u64
 // CHECK: shl d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vshl_n_u64(a, 1);
 }
 
 int8_t test_vqshlb_n_s8(int8_t a) {
 // CHECK-LABEL: test_vqshlb_n_s8
-// CHECK: sqshl {{b[0-9]+}}, {{b[0-9]+}}, #7
+// CHECK: sqshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, #7
   return (int8_t)vqshlb_n_s8(a, 7);
 }
 
 int16_t test_vqshlh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqshlh_n_s16
-// CHECK: sqshl {{h[0-9]+}}, {{h[0-9]+}}, #15
+// CHECK: sqshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, #15
   return (int16_t)vqshlh_n_s16(a, 15);
 }
 
@@ -8867,21 +8702,105 @@
   return (int64_t)vqshld_n_s64(a, 63);
 }
 
+int8x8_t test_vqshl_n_s8(int8x8_t a) {
+  // CHECK-LABEL: test_vqshl_n_s8
+  return vqshl_n_s8(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0
+}
+
+int8x16_t test_vqshlq_n_s8(int8x16_t a) {
+  // CHECK-LABEL: test_vqshlq_n_s8
+  return vqshlq_n_s8(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0
+}
+
+int16x4_t test_vqshl_n_s16(int16x4_t a) {
+  // CHECK-LABEL: test_vqshl_n_s16
+  return vqshl_n_s16(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0
+}
+
+int16x8_t test_vqshlq_n_s16(int16x8_t a) {
+  // CHECK-LABEL: test_vqshlq_n_s16
+  return vqshlq_n_s16(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0
+}
+
+int32x2_t test_vqshl_n_s32(int32x2_t a) {
+  // CHECK-LABEL: test_vqshl_n_s32
+  return vqshl_n_s32(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
+}
+
+int32x4_t test_vqshlq_n_s32(int32x4_t a) {
+  // CHECK-LABEL: test_vqshlq_n_s32
+  return vqshlq_n_s32(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
+}
+
+int64x2_t test_vqshlq_n_s64(int64x2_t a) {
+  // CHECK-LABEL: test_vqshlq_n_s64
+  return vqshlq_n_s64(a, 0);
+  // CHECK: sqshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
+}
+
+uint8x8_t test_vqshl_n_u8(uint8x8_t a) {
+  // CHECK-LABEL: test_vqshl_n_u8
+  return vqshl_n_u8(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0
+}
+
+uint8x16_t test_vqshlq_n_u8(uint8x16_t a) {
+  // CHECK-LABEL: test_vqshlq_n_u8
+  return vqshlq_n_u8(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0
+}
+
+uint16x4_t test_vqshl_n_u16(uint16x4_t a) {
+  // CHECK-LABEL: test_vqshl_n_u16
+  return vqshl_n_u16(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0
+}
+
+uint16x8_t test_vqshlq_n_u16(uint16x8_t a) {
+  // CHECK-LABEL: test_vqshlq_n_u16
+  return vqshlq_n_u16(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0
+}
+
+uint32x2_t test_vqshl_n_u32(uint32x2_t a) {
+  // CHECK-LABEL: test_vqshl_n_u32
+  return vqshl_n_u32(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
+}
+
+uint32x4_t test_vqshlq_n_u32(uint32x4_t a) {
+  // CHECK-LABEL: test_vqshlq_n_u32
+  return vqshlq_n_u32(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
+}
+
+uint64x2_t test_vqshlq_n_u64(uint64x2_t a) {
+  // CHECK-LABEL: test_vqshlq_n_u64
+  return vqshlq_n_u64(a, 0);
+  // CHECK: uqshl {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
+}
+
 int64x1_t test_vqshl_n_s64(int64x1_t a) {
-// CHECK: test_vqshl_n_s64
+// CHECK-LABEL: test_vqshl_n_s64
 // CHECK: sqshl d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vqshl_n_s64(a, 1);
 }
 
 uint8_t test_vqshlb_n_u8(uint8_t a) {
 // CHECK-LABEL: test_vqshlb_n_u8
-// CHECK: uqshl {{b[0-9]+}}, {{b[0-9]+}}, #7
+// CHECK: uqshl {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, #7
   return (uint8_t)vqshlb_n_u8(a, 7);
 }
 
 uint16_t test_vqshlh_n_u16(uint16_t a) {
 // CHECK-LABEL: test_vqshlh_n_u16
-// CHECK: uqshl {{h[0-9]+}}, {{h[0-9]+}}, #15
+// CHECK: uqshl {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, #15
   return (uint16_t)vqshlh_n_u16(a, 15);
 }
 
@@ -8898,20 +8817,20 @@
 }
 
 uint64x1_t test_vqshl_n_u64(uint64x1_t a) {
-// CHECK: test_vqshl_n_u64
+// CHECK-LABEL: test_vqshl_n_u64
 // CHECK: uqshl d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vqshl_n_u64(a, 1);
 }
 
 int8_t test_vqshlub_n_s8(int8_t a) {
 // CHECK-LABEL: test_vqshlub_n_s8
-// CHECK: sqshlu {{b[0-9]+}}, {{b[0-9]+}}, #7
+// CHECK: sqshlu {{b[0-9]+|v[0-9]+.8b}}, {{b[0-9]+|v[0-9]+.8b}}, #7
   return (int8_t)vqshlub_n_s8(a, 7);
 }
 
 int16_t test_vqshluh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqshluh_n_s16
-// CHECK: sqshlu {{h[0-9]+}}, {{h[0-9]+}}, #15
+// CHECK: sqshlu {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, #15
   return (int16_t)vqshluh_n_s16(a, 15);
 }
 
@@ -8928,7 +8847,7 @@
 }
 
 uint64x1_t test_vqshlu_n_s64(int64x1_t a) {
-// CHECK: test_vqshlu_n_s64
+// CHECK-LABEL: test_vqshlu_n_s64
 // CHECK: sqshlu d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vqshlu_n_s64(a, 1);
 }
@@ -8940,7 +8859,7 @@
 }
 
 int64x1_t test_vsri_n_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vsri_n_s64
+// CHECK-LABEL: test_vsri_n_s64
 // CHECK: sri d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsri_n_s64(a, b, 1);
 }
@@ -8952,7 +8871,7 @@
 }
 
 uint64x1_t test_vsri_n_u64(uint64x1_t a, uint64x1_t b) {
-// CHECK: test_vsri_n_u64
+// CHECK-LABEL: test_vsri_n_u64
 // CHECK: sri d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsri_n_u64(a, b, 1);
 }
@@ -8964,7 +8883,7 @@
 }
 
 int64x1_t test_vsli_n_s64(int64x1_t a, int64x1_t b) {
-// CHECK: test_vsli_n_s64
+// CHECK-LABEL: test_vsli_n_s64
 // CHECK: sli d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsli_n_s64(a, b, 1);
 }
@@ -8976,20 +8895,20 @@
 }
 
 uint64x1_t test_vsli_n_u64(uint64x1_t a, uint64x1_t b) {
-// CHECK: test_vsli_n_u64
+// CHECK-LABEL: test_vsli_n_u64
 // CHECK: sli d{{[0-9]+}}, d{{[0-9]+}}, #1
   return vsli_n_u64(a, b, 1);
 }
 
 int8_t test_vqshrnh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqshrnh_n_s16
-// CHECK: sqshrn {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: sqshrn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (int8_t)vqshrnh_n_s16(a, 8);
 }
 
 int16_t test_vqshrns_n_s32(int32_t a) {
 // CHECK-LABEL: test_vqshrns_n_s32
-// CHECK: sqshrn {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: sqshrn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (int16_t)vqshrns_n_s32(a, 16);
 }
 
@@ -9001,13 +8920,13 @@
 
 uint8_t test_vqshrnh_n_u16(uint16_t a) {
 // CHECK-LABEL: test_vqshrnh_n_u16
-// CHECK: uqshrn {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: uqshrn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (uint8_t)vqshrnh_n_u16(a, 8);
 }
 
 uint16_t test_vqshrns_n_u32(uint32_t a) {
 // CHECK-LABEL: test_vqshrns_n_u32
-// CHECK: uqshrn {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: uqshrn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (uint16_t)vqshrns_n_u32(a, 16);
 }
 
@@ -9019,13 +8938,13 @@
 
 int8_t test_vqrshrnh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqrshrnh_n_s16
-// CHECK: sqrshrn {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: sqrshrn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (int8_t)vqrshrnh_n_s16(a, 8);
 }
 
 int16_t test_vqrshrns_n_s32(int32_t a) {
 // CHECK-LABEL: test_vqrshrns_n_s32
-// CHECK: sqrshrn {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: sqrshrn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (int16_t)vqrshrns_n_s32(a, 16);
 }
 
@@ -9037,13 +8956,13 @@
 
 uint8_t test_vqrshrnh_n_u16(uint16_t a) {
 // CHECK-LABEL: test_vqrshrnh_n_u16
-// CHECK: uqrshrn {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: uqrshrn {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (uint8_t)vqrshrnh_n_u16(a, 8);
 }
 
 uint16_t test_vqrshrns_n_u32(uint32_t a) {
 // CHECK-LABEL: test_vqrshrns_n_u32
-// CHECK: uqrshrn {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: uqrshrn {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (uint16_t)vqrshrns_n_u32(a, 16);
 }
 
@@ -9055,13 +8974,13 @@
 
 int8_t test_vqshrunh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqshrunh_n_s16
-// CHECK: sqshrun {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: sqshrun {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (int8_t)vqshrunh_n_s16(a, 8);
 }
 
 int16_t test_vqshruns_n_s32(int32_t a) {
 // CHECK-LABEL: test_vqshruns_n_s32
-// CHECK: sqshrun {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: sqshrun {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (int16_t)vqshruns_n_s32(a, 16);
 }
 
@@ -9073,13 +8992,13 @@
 
 int8_t test_vqrshrunh_n_s16(int16_t a) {
 // CHECK-LABEL: test_vqrshrunh_n_s16
-// CHECK: sqrshrun {{b[0-9]+}}, {{h[0-9]+}}, #8
+// CHECK: sqrshrun {{b[0-9]+|v[0-9]+.8b}}, {{h[0-9]+|v[0-9]+.8h}}, #8
   return (int8_t)vqrshrunh_n_s16(a, 8);
 }
 
 int16_t test_vqrshruns_n_s32(int32_t a) {
 // CHECK-LABEL: test_vqrshruns_n_s32
-// CHECK: sqrshrun {{h[0-9]+}}, {{s[0-9]+}}, #16
+// CHECK: sqrshrun {{h[0-9]+|v[0-9]+.4h}}, {{s[0-9]+|v[0-9]+.4s}}, #16
   return (int16_t)vqrshruns_n_s32(a, 16);
 }
 
@@ -9090,2232 +9009,2232 @@
 }
 
 float32_t test_vcvts_n_f32_s32(int32_t a) {
-// CHECK: test_vcvts_n_f32_s32
+// CHECK-LABEL: test_vcvts_n_f32_s32
 // CHECK: scvtf {{s[0-9]+}}, {{s[0-9]+}}, #1
   return vcvts_n_f32_s32(a, 1);
 }
 
 float64_t test_vcvtd_n_f64_s64(int64_t a) {
-// CHECK: test_vcvtd_n_f64_s64
+// CHECK-LABEL: test_vcvtd_n_f64_s64
 // CHECK: scvtf {{d[0-9]+}}, {{d[0-9]+}}, #1
   return vcvtd_n_f64_s64(a, 1);
 }
 
 float32_t test_vcvts_n_f32_u32(uint32_t a) {
-// CHECK: test_vcvts_n_f32_u32
+// CHECK-LABEL: test_vcvts_n_f32_u32
 // CHECK: ucvtf {{s[0-9]+}}, {{s[0-9]+}}, #32
   return vcvts_n_f32_u32(a, 32);
 }
 
 float64_t test_vcvtd_n_f64_u64(uint64_t a) {
-// CHECK: test_vcvtd_n_f64_u64
+// CHECK-LABEL: test_vcvtd_n_f64_u64
 // CHECK: ucvtf {{d[0-9]+}}, {{d[0-9]+}}, #64
   return vcvtd_n_f64_u64(a, 64);
 }
 
 int32_t test_vcvts_n_s32_f32(float32_t a) {
-// CHECK: test_vcvts_n_s32_f32
+// CHECK-LABEL: test_vcvts_n_s32_f32
 // CHECK: fcvtzs {{s[0-9]+}}, {{s[0-9]+}}, #1
   return (int32_t)vcvts_n_s32_f32(a, 1);
 }
 
 int64_t test_vcvtd_n_s64_f64(float64_t a) {
-// CHECK: test_vcvtd_n_s64_f64
+// CHECK-LABEL: test_vcvtd_n_s64_f64
 // CHECK: fcvtzs {{d[0-9]+}}, {{d[0-9]+}}, #1
   return (int64_t)vcvtd_n_s64_f64(a, 1);
 }
 
 uint32_t test_vcvts_n_u32_f32(float32_t a) {
-// CHECK: test_vcvts_n_u32_f32
+// CHECK-LABEL: test_vcvts_n_u32_f32
 // CHECK: fcvtzu {{s[0-9]+}}, {{s[0-9]+}}, #32
   return (uint32_t)vcvts_n_u32_f32(a, 32);
 }
 
 uint64_t test_vcvtd_n_u64_f64(float64_t a) {
-// CHECK: test_vcvtd_n_u64_f64
+// CHECK-LABEL: test_vcvtd_n_u64_f64
 // CHECK: fcvtzu {{d[0-9]+}}, {{d[0-9]+}}, #64
   return (uint64_t)vcvtd_n_u64_f64(a, 64);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_s16
+// CHECK-LABEL: test_vreinterpret_s8_s16:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_s16(int16x4_t a) {
   return vreinterpret_s8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_s32
+// CHECK-LABEL: test_vreinterpret_s8_s32:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_s32(int32x2_t a) {
   return vreinterpret_s8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_s64
+// CHECK-LABEL: test_vreinterpret_s8_s64:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_s64(int64x1_t a) {
   return vreinterpret_s8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_u8
+// CHECK-LABEL: test_vreinterpret_s8_u8:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_u8(uint8x8_t a) {
   return vreinterpret_s8_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_u16
+// CHECK-LABEL: test_vreinterpret_s8_u16:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_u16(uint16x4_t a) {
   return vreinterpret_s8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_u32
+// CHECK-LABEL: test_vreinterpret_s8_u32:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_u32(uint32x2_t a) {
   return vreinterpret_s8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_u64
+// CHECK-LABEL: test_vreinterpret_s8_u64:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_u64(uint64x1_t a) {
   return vreinterpret_s8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_f16
+// CHECK-LABEL: test_vreinterpret_s8_f16:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_f16(float16x4_t a) {
   return vreinterpret_s8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_f32
+// CHECK-LABEL: test_vreinterpret_s8_f32:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_f32(float32x2_t a) {
   return vreinterpret_s8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_f64
+// CHECK-LABEL: test_vreinterpret_s8_f64:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_f64(float64x1_t a) {
   return vreinterpret_s8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_p8
+// CHECK-LABEL: test_vreinterpret_s8_p8:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_p8(poly8x8_t a) {
   return vreinterpret_s8_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_p16
+// CHECK-LABEL: test_vreinterpret_s8_p16:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_p16(poly16x4_t a) {
   return vreinterpret_s8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s8_p64
+// CHECK-LABEL: test_vreinterpret_s8_p64:
 // CHECK-NEXT: ret
 int8x8_t test_vreinterpret_s8_p64(poly64x1_t a) {
   return vreinterpret_s8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_s8
+// CHECK-LABEL: test_vreinterpret_s16_s8:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_s8(int8x8_t a) {
   return vreinterpret_s16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_s32
+// CHECK-LABEL: test_vreinterpret_s16_s32:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_s32(int32x2_t a) {
   return vreinterpret_s16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_s64
+// CHECK-LABEL: test_vreinterpret_s16_s64:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_s64(int64x1_t a) {
   return vreinterpret_s16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_u8
+// CHECK-LABEL: test_vreinterpret_s16_u8:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_u8(uint8x8_t a) {
   return vreinterpret_s16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_u16
+// CHECK-LABEL: test_vreinterpret_s16_u16:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_u16(uint16x4_t a) {
   return vreinterpret_s16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_u32
+// CHECK-LABEL: test_vreinterpret_s16_u32:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_u32(uint32x2_t a) {
   return vreinterpret_s16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_u64
+// CHECK-LABEL: test_vreinterpret_s16_u64:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_u64(uint64x1_t a) {
   return vreinterpret_s16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_f16
+// CHECK-LABEL: test_vreinterpret_s16_f16:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_f16(float16x4_t a) {
   return vreinterpret_s16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_f32
+// CHECK-LABEL: test_vreinterpret_s16_f32:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_f32(float32x2_t a) {
   return vreinterpret_s16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_f64
+// CHECK-LABEL: test_vreinterpret_s16_f64:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_f64(float64x1_t a) {
   return vreinterpret_s16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_p8
+// CHECK-LABEL: test_vreinterpret_s16_p8:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_p8(poly8x8_t a) {
   return vreinterpret_s16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_p16
+// CHECK-LABEL: test_vreinterpret_s16_p16:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_p16(poly16x4_t a) {
   return vreinterpret_s16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s16_p64
+// CHECK-LABEL: test_vreinterpret_s16_p64:
 // CHECK-NEXT: ret
 int16x4_t test_vreinterpret_s16_p64(poly64x1_t a) {
   return vreinterpret_s16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_s8
+// CHECK-LABEL: test_vreinterpret_s32_s8:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_s8(int8x8_t a) {
   return vreinterpret_s32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_s16
+// CHECK-LABEL: test_vreinterpret_s32_s16:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_s16(int16x4_t a) {
   return vreinterpret_s32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_s64
+// CHECK-LABEL: test_vreinterpret_s32_s64:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_s64(int64x1_t a) {
   return vreinterpret_s32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_u8
+// CHECK-LABEL: test_vreinterpret_s32_u8:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_u8(uint8x8_t a) {
   return vreinterpret_s32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_u16
+// CHECK-LABEL: test_vreinterpret_s32_u16:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_u16(uint16x4_t a) {
   return vreinterpret_s32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_u32
+// CHECK-LABEL: test_vreinterpret_s32_u32:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_u32(uint32x2_t a) {
   return vreinterpret_s32_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_u64
+// CHECK-LABEL: test_vreinterpret_s32_u64:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_u64(uint64x1_t a) {
   return vreinterpret_s32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_f16
+// CHECK-LABEL: test_vreinterpret_s32_f16:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_f16(float16x4_t a) {
   return vreinterpret_s32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_f32
+// CHECK-LABEL: test_vreinterpret_s32_f32:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_f32(float32x2_t a) {
   return vreinterpret_s32_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_f64
+// CHECK-LABEL: test_vreinterpret_s32_f64:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_f64(float64x1_t a) {
   return vreinterpret_s32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_p8
+// CHECK-LABEL: test_vreinterpret_s32_p8:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_p8(poly8x8_t a) {
   return vreinterpret_s32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_p16
+// CHECK-LABEL: test_vreinterpret_s32_p16:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_p16(poly16x4_t a) {
   return vreinterpret_s32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s32_p64
+// CHECK-LABEL: test_vreinterpret_s32_p64:
 // CHECK-NEXT: ret
 int32x2_t test_vreinterpret_s32_p64(poly64x1_t a) {
   return vreinterpret_s32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_s8
+// CHECK-LABEL: test_vreinterpret_s64_s8:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_s8(int8x8_t a) {
   return vreinterpret_s64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_s16
+// CHECK-LABEL: test_vreinterpret_s64_s16:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_s16(int16x4_t a) {
   return vreinterpret_s64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_s32
+// CHECK-LABEL: test_vreinterpret_s64_s32:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_s32(int32x2_t a) {
   return vreinterpret_s64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_u8
+// CHECK-LABEL: test_vreinterpret_s64_u8:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_u8(uint8x8_t a) {
   return vreinterpret_s64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_u16
+// CHECK-LABEL: test_vreinterpret_s64_u16:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_u16(uint16x4_t a) {
   return vreinterpret_s64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_u32
+// CHECK-LABEL: test_vreinterpret_s64_u32:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_u32(uint32x2_t a) {
   return vreinterpret_s64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_u64
+// CHECK-LABEL: test_vreinterpret_s64_u64:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_u64(uint64x1_t a) {
   return vreinterpret_s64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_f16
+// CHECK-LABEL: test_vreinterpret_s64_f16:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_f16(float16x4_t a) {
   return vreinterpret_s64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_f32
+// CHECK-LABEL: test_vreinterpret_s64_f32:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_f32(float32x2_t a) {
   return vreinterpret_s64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_f64
+// CHECK-LABEL: test_vreinterpret_s64_f64:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_f64(float64x1_t a) {
   return vreinterpret_s64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_p8
+// CHECK-LABEL: test_vreinterpret_s64_p8:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_p8(poly8x8_t a) {
   return vreinterpret_s64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_p16
+// CHECK-LABEL: test_vreinterpret_s64_p16:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_p16(poly16x4_t a) {
   return vreinterpret_s64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_s64_p64
+// CHECK-LABEL: test_vreinterpret_s64_p64:
 // CHECK-NEXT: ret
 int64x1_t test_vreinterpret_s64_p64(poly64x1_t a) {
   return vreinterpret_s64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_s8
+// CHECK-LABEL: test_vreinterpret_u8_s8:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_s8(int8x8_t a) {
   return vreinterpret_u8_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_s16
+// CHECK-LABEL: test_vreinterpret_u8_s16:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_s16(int16x4_t a) {
   return vreinterpret_u8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_s32
+// CHECK-LABEL: test_vreinterpret_u8_s32:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_s32(int32x2_t a) {
   return vreinterpret_u8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_s64
+// CHECK-LABEL: test_vreinterpret_u8_s64:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_s64(int64x1_t a) {
   return vreinterpret_u8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_u16
+// CHECK-LABEL: test_vreinterpret_u8_u16:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_u16(uint16x4_t a) {
   return vreinterpret_u8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_u32
+// CHECK-LABEL: test_vreinterpret_u8_u32:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_u32(uint32x2_t a) {
   return vreinterpret_u8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_u64
+// CHECK-LABEL: test_vreinterpret_u8_u64:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_u64(uint64x1_t a) {
   return vreinterpret_u8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_f16
+// CHECK-LABEL: test_vreinterpret_u8_f16:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_f16(float16x4_t a) {
   return vreinterpret_u8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_f32
+// CHECK-LABEL: test_vreinterpret_u8_f32:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_f32(float32x2_t a) {
   return vreinterpret_u8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_f64
+// CHECK-LABEL: test_vreinterpret_u8_f64:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_f64(float64x1_t a) {
   return vreinterpret_u8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_p8
+// CHECK-LABEL: test_vreinterpret_u8_p8:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_p8(poly8x8_t a) {
   return vreinterpret_u8_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_p16
+// CHECK-LABEL: test_vreinterpret_u8_p16:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_p16(poly16x4_t a) {
   return vreinterpret_u8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u8_p64
+// CHECK-LABEL: test_vreinterpret_u8_p64:
 // CHECK-NEXT: ret
 uint8x8_t test_vreinterpret_u8_p64(poly64x1_t a) {
   return vreinterpret_u8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_s8
+// CHECK-LABEL: test_vreinterpret_u16_s8:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_s8(int8x8_t a) {
   return vreinterpret_u16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_s16
+// CHECK-LABEL: test_vreinterpret_u16_s16:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_s16(int16x4_t a) {
   return vreinterpret_u16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_s32
+// CHECK-LABEL: test_vreinterpret_u16_s32:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_s32(int32x2_t a) {
   return vreinterpret_u16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_s64
+// CHECK-LABEL: test_vreinterpret_u16_s64:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_s64(int64x1_t a) {
   return vreinterpret_u16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_u8
+// CHECK-LABEL: test_vreinterpret_u16_u8:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_u8(uint8x8_t a) {
   return vreinterpret_u16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_u32
+// CHECK-LABEL: test_vreinterpret_u16_u32:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_u32(uint32x2_t a) {
   return vreinterpret_u16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_u64
+// CHECK-LABEL: test_vreinterpret_u16_u64:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_u64(uint64x1_t a) {
   return vreinterpret_u16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_f16
+// CHECK-LABEL: test_vreinterpret_u16_f16:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_f16(float16x4_t a) {
   return vreinterpret_u16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_f32
+// CHECK-LABEL: test_vreinterpret_u16_f32:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_f32(float32x2_t a) {
   return vreinterpret_u16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_f64
+// CHECK-LABEL: test_vreinterpret_u16_f64:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_f64(float64x1_t a) {
   return vreinterpret_u16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_p8
+// CHECK-LABEL: test_vreinterpret_u16_p8:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_p8(poly8x8_t a) {
   return vreinterpret_u16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_p16
+// CHECK-LABEL: test_vreinterpret_u16_p16:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_p16(poly16x4_t a) {
   return vreinterpret_u16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u16_p64
+// CHECK-LABEL: test_vreinterpret_u16_p64:
 // CHECK-NEXT: ret
 uint16x4_t test_vreinterpret_u16_p64(poly64x1_t a) {
   return vreinterpret_u16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_s8
+// CHECK-LABEL: test_vreinterpret_u32_s8:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_s8(int8x8_t a) {
   return vreinterpret_u32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_s16
+// CHECK-LABEL: test_vreinterpret_u32_s16:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_s16(int16x4_t a) {
   return vreinterpret_u32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_s32
+// CHECK-LABEL: test_vreinterpret_u32_s32:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_s32(int32x2_t a) {
   return vreinterpret_u32_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_s64
+// CHECK-LABEL: test_vreinterpret_u32_s64:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_s64(int64x1_t a) {
   return vreinterpret_u32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_u8
+// CHECK-LABEL: test_vreinterpret_u32_u8:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_u8(uint8x8_t a) {
   return vreinterpret_u32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_u16
+// CHECK-LABEL: test_vreinterpret_u32_u16:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_u16(uint16x4_t a) {
   return vreinterpret_u32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_u64
+// CHECK-LABEL: test_vreinterpret_u32_u64:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_u64(uint64x1_t a) {
   return vreinterpret_u32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_f16
+// CHECK-LABEL: test_vreinterpret_u32_f16:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_f16(float16x4_t a) {
   return vreinterpret_u32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_f32
+// CHECK-LABEL: test_vreinterpret_u32_f32:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_f32(float32x2_t a) {
   return vreinterpret_u32_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_f64
+// CHECK-LABEL: test_vreinterpret_u32_f64:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_f64(float64x1_t a) {
   return vreinterpret_u32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_p8
+// CHECK-LABEL: test_vreinterpret_u32_p8:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_p8(poly8x8_t a) {
   return vreinterpret_u32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_p16
+// CHECK-LABEL: test_vreinterpret_u32_p16:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_p16(poly16x4_t a) {
   return vreinterpret_u32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u32_p64
+// CHECK-LABEL: test_vreinterpret_u32_p64:
 // CHECK-NEXT: ret
 uint32x2_t test_vreinterpret_u32_p64(poly64x1_t a) {
   return vreinterpret_u32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_s8
+// CHECK-LABEL: test_vreinterpret_u64_s8:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_s8(int8x8_t a) {
   return vreinterpret_u64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_s16
+// CHECK-LABEL: test_vreinterpret_u64_s16:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_s16(int16x4_t a) {
   return vreinterpret_u64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_s32
+// CHECK-LABEL: test_vreinterpret_u64_s32:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_s32(int32x2_t a) {
   return vreinterpret_u64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_s64
+// CHECK-LABEL: test_vreinterpret_u64_s64:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_s64(int64x1_t a) {
   return vreinterpret_u64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_u8
+// CHECK-LABEL: test_vreinterpret_u64_u8:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_u8(uint8x8_t a) {
   return vreinterpret_u64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_u16
+// CHECK-LABEL: test_vreinterpret_u64_u16:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_u16(uint16x4_t a) {
   return vreinterpret_u64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_u32
+// CHECK-LABEL: test_vreinterpret_u64_u32:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_u32(uint32x2_t a) {
   return vreinterpret_u64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_f16
+// CHECK-LABEL: test_vreinterpret_u64_f16:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_f16(float16x4_t a) {
   return vreinterpret_u64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_f32
+// CHECK-LABEL: test_vreinterpret_u64_f32:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_f32(float32x2_t a) {
   return vreinterpret_u64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_f64
+// CHECK-LABEL: test_vreinterpret_u64_f64:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_f64(float64x1_t a) {
   return vreinterpret_u64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_p8
+// CHECK-LABEL: test_vreinterpret_u64_p8:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_p8(poly8x8_t a) {
   return vreinterpret_u64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_p16
+// CHECK-LABEL: test_vreinterpret_u64_p16:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_p16(poly16x4_t a) {
   return vreinterpret_u64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_u64_p64
+// CHECK-LABEL: test_vreinterpret_u64_p64:
 // CHECK-NEXT: ret
 uint64x1_t test_vreinterpret_u64_p64(poly64x1_t a) {
   return vreinterpret_u64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_s8
+// CHECK-LABEL: test_vreinterpret_f16_s8:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_s8(int8x8_t a) {
   return vreinterpret_f16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_s16
+// CHECK-LABEL: test_vreinterpret_f16_s16:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_s16(int16x4_t a) {
   return vreinterpret_f16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_s32
+// CHECK-LABEL: test_vreinterpret_f16_s32:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_s32(int32x2_t a) {
   return vreinterpret_f16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_s64
+// CHECK-LABEL: test_vreinterpret_f16_s64:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_s64(int64x1_t a) {
   return vreinterpret_f16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_u8
+// CHECK-LABEL: test_vreinterpret_f16_u8:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_u8(uint8x8_t a) {
   return vreinterpret_f16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_u16
+// CHECK-LABEL: test_vreinterpret_f16_u16:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_u16(uint16x4_t a) {
   return vreinterpret_f16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_u32
+// CHECK-LABEL: test_vreinterpret_f16_u32:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_u32(uint32x2_t a) {
   return vreinterpret_f16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_u64
+// CHECK-LABEL: test_vreinterpret_f16_u64:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_u64(uint64x1_t a) {
   return vreinterpret_f16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_f32
+// CHECK-LABEL: test_vreinterpret_f16_f32:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_f32(float32x2_t a) {
   return vreinterpret_f16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_f64
+// CHECK-LABEL: test_vreinterpret_f16_f64:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_f64(float64x1_t a) {
   return vreinterpret_f16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_p8
+// CHECK-LABEL: test_vreinterpret_f16_p8:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_p8(poly8x8_t a) {
   return vreinterpret_f16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_p16
+// CHECK-LABEL: test_vreinterpret_f16_p16:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_p16(poly16x4_t a) {
   return vreinterpret_f16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f16_p64
+// CHECK-LABEL: test_vreinterpret_f16_p64:
 // CHECK-NEXT: ret
 float16x4_t test_vreinterpret_f16_p64(poly64x1_t a) {
   return vreinterpret_f16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_s8
+// CHECK-LABEL: test_vreinterpret_f32_s8:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_s8(int8x8_t a) {
   return vreinterpret_f32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_s16
+// CHECK-LABEL: test_vreinterpret_f32_s16:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_s16(int16x4_t a) {
   return vreinterpret_f32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_s32
+// CHECK-LABEL: test_vreinterpret_f32_s32:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_s32(int32x2_t a) {
   return vreinterpret_f32_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_s64
+// CHECK-LABEL: test_vreinterpret_f32_s64:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_s64(int64x1_t a) {
   return vreinterpret_f32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_u8
+// CHECK-LABEL: test_vreinterpret_f32_u8:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_u8(uint8x8_t a) {
   return vreinterpret_f32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_u16
+// CHECK-LABEL: test_vreinterpret_f32_u16:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_u16(uint16x4_t a) {
   return vreinterpret_f32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_u32
+// CHECK-LABEL: test_vreinterpret_f32_u32:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_u32(uint32x2_t a) {
   return vreinterpret_f32_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_u64
+// CHECK-LABEL: test_vreinterpret_f32_u64:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_u64(uint64x1_t a) {
   return vreinterpret_f32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_f16
+// CHECK-LABEL: test_vreinterpret_f32_f16:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_f16(float16x4_t a) {
   return vreinterpret_f32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_f64
+// CHECK-LABEL: test_vreinterpret_f32_f64:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_f64(float64x1_t a) {
   return vreinterpret_f32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_p8
+// CHECK-LABEL: test_vreinterpret_f32_p8:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_p8(poly8x8_t a) {
   return vreinterpret_f32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_p16
+// CHECK-LABEL: test_vreinterpret_f32_p16:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_p16(poly16x4_t a) {
   return vreinterpret_f32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f32_p64
+// CHECK-LABEL: test_vreinterpret_f32_p64:
 // CHECK-NEXT: ret
 float32x2_t test_vreinterpret_f32_p64(poly64x1_t a) {
   return vreinterpret_f32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_s8
+// CHECK-LABEL: test_vreinterpret_f64_s8:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_s8(int8x8_t a) {
   return vreinterpret_f64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_s16
+// CHECK-LABEL: test_vreinterpret_f64_s16:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_s16(int16x4_t a) {
   return vreinterpret_f64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_s32
+// CHECK-LABEL: test_vreinterpret_f64_s32:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_s32(int32x2_t a) {
   return vreinterpret_f64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_s64
+// CHECK-LABEL: test_vreinterpret_f64_s64:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_s64(int64x1_t a) {
   return vreinterpret_f64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_u8
+// CHECK-LABEL: test_vreinterpret_f64_u8:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_u8(uint8x8_t a) {
   return vreinterpret_f64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_u16
+// CHECK-LABEL: test_vreinterpret_f64_u16:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_u16(uint16x4_t a) {
   return vreinterpret_f64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_u32
+// CHECK-LABEL: test_vreinterpret_f64_u32:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_u32(uint32x2_t a) {
   return vreinterpret_f64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_u64
+// CHECK-LABEL: test_vreinterpret_f64_u64:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_u64(uint64x1_t a) {
   return vreinterpret_f64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_f16
+// CHECK-LABEL: test_vreinterpret_f64_f16:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_f16(float16x4_t a) {
   return vreinterpret_f64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_f32
+// CHECK-LABEL: test_vreinterpret_f64_f32:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_f32(float32x2_t a) {
   return vreinterpret_f64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_p8
+// CHECK-LABEL: test_vreinterpret_f64_p8:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_p8(poly8x8_t a) {
   return vreinterpret_f64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_p16
+// CHECK-LABEL: test_vreinterpret_f64_p16:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_p16(poly16x4_t a) {
   return vreinterpret_f64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_f64_p64
+// CHECK-LABEL: test_vreinterpret_f64_p64:
 // CHECK-NEXT: ret
 float64x1_t test_vreinterpret_f64_p64(poly64x1_t a) {
   return vreinterpret_f64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_s8
+// CHECK-LABEL: test_vreinterpret_p8_s8:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_s8(int8x8_t a) {
   return vreinterpret_p8_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_s16
+// CHECK-LABEL: test_vreinterpret_p8_s16:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_s16(int16x4_t a) {
   return vreinterpret_p8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_s32
+// CHECK-LABEL: test_vreinterpret_p8_s32:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_s32(int32x2_t a) {
   return vreinterpret_p8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_s64
+// CHECK-LABEL: test_vreinterpret_p8_s64:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_s64(int64x1_t a) {
   return vreinterpret_p8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_u8
+// CHECK-LABEL: test_vreinterpret_p8_u8:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_u8(uint8x8_t a) {
   return vreinterpret_p8_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_u16
+// CHECK-LABEL: test_vreinterpret_p8_u16:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_u16(uint16x4_t a) {
   return vreinterpret_p8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_u32
+// CHECK-LABEL: test_vreinterpret_p8_u32:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_u32(uint32x2_t a) {
   return vreinterpret_p8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_u64
+// CHECK-LABEL: test_vreinterpret_p8_u64:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_u64(uint64x1_t a) {
   return vreinterpret_p8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_f16
+// CHECK-LABEL: test_vreinterpret_p8_f16:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_f16(float16x4_t a) {
   return vreinterpret_p8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_f32
+// CHECK-LABEL: test_vreinterpret_p8_f32:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_f32(float32x2_t a) {
   return vreinterpret_p8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_f64
+// CHECK-LABEL: test_vreinterpret_p8_f64:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_f64(float64x1_t a) {
   return vreinterpret_p8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_p16
+// CHECK-LABEL: test_vreinterpret_p8_p16:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_p16(poly16x4_t a) {
   return vreinterpret_p8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p8_p64
+// CHECK-LABEL: test_vreinterpret_p8_p64:
 // CHECK-NEXT: ret
 poly8x8_t test_vreinterpret_p8_p64(poly64x1_t a) {
   return vreinterpret_p8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_s8
+// CHECK-LABEL: test_vreinterpret_p16_s8:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_s8(int8x8_t a) {
   return vreinterpret_p16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_s16
+// CHECK-LABEL: test_vreinterpret_p16_s16:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_s16(int16x4_t a) {
   return vreinterpret_p16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_s32
+// CHECK-LABEL: test_vreinterpret_p16_s32:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_s32(int32x2_t a) {
   return vreinterpret_p16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_s64
+// CHECK-LABEL: test_vreinterpret_p16_s64:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_s64(int64x1_t a) {
   return vreinterpret_p16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_u8
+// CHECK-LABEL: test_vreinterpret_p16_u8:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_u8(uint8x8_t a) {
   return vreinterpret_p16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_u16
+// CHECK-LABEL: test_vreinterpret_p16_u16:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_u16(uint16x4_t a) {
   return vreinterpret_p16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_u32
+// CHECK-LABEL: test_vreinterpret_p16_u32:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_u32(uint32x2_t a) {
   return vreinterpret_p16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_u64
+// CHECK-LABEL: test_vreinterpret_p16_u64:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_u64(uint64x1_t a) {
   return vreinterpret_p16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_f16
+// CHECK-LABEL: test_vreinterpret_p16_f16:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_f16(float16x4_t a) {
   return vreinterpret_p16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_f32
+// CHECK-LABEL: test_vreinterpret_p16_f32:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_f32(float32x2_t a) {
   return vreinterpret_p16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_f64
+// CHECK-LABEL: test_vreinterpret_p16_f64:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_f64(float64x1_t a) {
   return vreinterpret_p16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_p8
+// CHECK-LABEL: test_vreinterpret_p16_p8:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_p8(poly8x8_t a) {
   return vreinterpret_p16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p16_p64
+// CHECK-LABEL: test_vreinterpret_p16_p64:
 // CHECK-NEXT: ret
 poly16x4_t test_vreinterpret_p16_p64(poly64x1_t a) {
   return vreinterpret_p16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_s8
+// CHECK-LABEL: test_vreinterpret_p64_s8:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_s8(int8x8_t a) {
   return vreinterpret_p64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_s16
+// CHECK-LABEL: test_vreinterpret_p64_s16:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_s16(int16x4_t a) {
   return vreinterpret_p64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_s32
+// CHECK-LABEL: test_vreinterpret_p64_s32:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_s32(int32x2_t a) {
   return vreinterpret_p64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_s64
+// CHECK-LABEL: test_vreinterpret_p64_s64:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_s64(int64x1_t a) {
   return vreinterpret_p64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_u8
+// CHECK-LABEL: test_vreinterpret_p64_u8:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_u8(uint8x8_t a) {
   return vreinterpret_p64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_u16
+// CHECK-LABEL: test_vreinterpret_p64_u16:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_u16(uint16x4_t a) {
   return vreinterpret_p64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_u32
+// CHECK-LABEL: test_vreinterpret_p64_u32:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_u32(uint32x2_t a) {
   return vreinterpret_p64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_u64
+// CHECK-LABEL: test_vreinterpret_p64_u64:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_u64(uint64x1_t a) {
   return vreinterpret_p64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_f16
+// CHECK-LABEL: test_vreinterpret_p64_f16:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_f16(float16x4_t a) {
   return vreinterpret_p64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_f32
+// CHECK-LABEL: test_vreinterpret_p64_f32:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_f32(float32x2_t a) {
   return vreinterpret_p64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_f64
+// CHECK-LABEL: test_vreinterpret_p64_f64:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_f64(float64x1_t a) {
   return vreinterpret_p64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_p8
+// CHECK-LABEL: test_vreinterpret_p64_p8:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_p8(poly8x8_t a) {
   return vreinterpret_p64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpret_p64_p16
+// CHECK-LABEL: test_vreinterpret_p64_p16:
 // CHECK-NEXT: ret
 poly64x1_t test_vreinterpret_p64_p16(poly16x4_t a) {
   return vreinterpret_p64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_s16
+// CHECK-LABEL: test_vreinterpretq_s8_s16:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_s16(int16x8_t a) {
   return vreinterpretq_s8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_s32
+// CHECK-LABEL: test_vreinterpretq_s8_s32:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_s32(int32x4_t a) {
   return vreinterpretq_s8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_s64
+// CHECK-LABEL: test_vreinterpretq_s8_s64:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_s64(int64x2_t a) {
   return vreinterpretq_s8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_u8
+// CHECK-LABEL: test_vreinterpretq_s8_u8:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_u8(uint8x16_t a) {
   return vreinterpretq_s8_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_u16
+// CHECK-LABEL: test_vreinterpretq_s8_u16:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_u16(uint16x8_t a) {
   return vreinterpretq_s8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_u32
+// CHECK-LABEL: test_vreinterpretq_s8_u32:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_u32(uint32x4_t a) {
   return vreinterpretq_s8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_u64
+// CHECK-LABEL: test_vreinterpretq_s8_u64:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_u64(uint64x2_t a) {
   return vreinterpretq_s8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_f16
+// CHECK-LABEL: test_vreinterpretq_s8_f16:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_f16(float16x8_t a) {
   return vreinterpretq_s8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_f32
+// CHECK-LABEL: test_vreinterpretq_s8_f32:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_f32(float32x4_t a) {
   return vreinterpretq_s8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_f64
+// CHECK-LABEL: test_vreinterpretq_s8_f64:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_f64(float64x2_t a) {
   return vreinterpretq_s8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_p8
+// CHECK-LABEL: test_vreinterpretq_s8_p8:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_p8(poly8x16_t a) {
   return vreinterpretq_s8_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_p16
+// CHECK-LABEL: test_vreinterpretq_s8_p16:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_p16(poly16x8_t a) {
   return vreinterpretq_s8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s8_p64
+// CHECK-LABEL: test_vreinterpretq_s8_p64:
 // CHECK-NEXT: ret
 int8x16_t test_vreinterpretq_s8_p64(poly64x2_t a) {
   return vreinterpretq_s8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_s8
+// CHECK-LABEL: test_vreinterpretq_s16_s8:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_s8(int8x16_t a) {
   return vreinterpretq_s16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_s32
+// CHECK-LABEL: test_vreinterpretq_s16_s32:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_s32(int32x4_t a) {
   return vreinterpretq_s16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_s64
+// CHECK-LABEL: test_vreinterpretq_s16_s64:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_s64(int64x2_t a) {
   return vreinterpretq_s16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_u8
+// CHECK-LABEL: test_vreinterpretq_s16_u8:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_u8(uint8x16_t a) {
   return vreinterpretq_s16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_u16
+// CHECK-LABEL: test_vreinterpretq_s16_u16:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_u16(uint16x8_t a) {
   return vreinterpretq_s16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_u32
+// CHECK-LABEL: test_vreinterpretq_s16_u32:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_u32(uint32x4_t a) {
   return vreinterpretq_s16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_u64
+// CHECK-LABEL: test_vreinterpretq_s16_u64:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_u64(uint64x2_t a) {
   return vreinterpretq_s16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_f16
+// CHECK-LABEL: test_vreinterpretq_s16_f16:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_f16(float16x8_t a) {
   return vreinterpretq_s16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_f32
+// CHECK-LABEL: test_vreinterpretq_s16_f32:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_f32(float32x4_t a) {
   return vreinterpretq_s16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_f64
+// CHECK-LABEL: test_vreinterpretq_s16_f64:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_f64(float64x2_t a) {
   return vreinterpretq_s16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_p8
+// CHECK-LABEL: test_vreinterpretq_s16_p8:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_p8(poly8x16_t a) {
   return vreinterpretq_s16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_p16
+// CHECK-LABEL: test_vreinterpretq_s16_p16:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_p16(poly16x8_t a) {
   return vreinterpretq_s16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s16_p64
+// CHECK-LABEL: test_vreinterpretq_s16_p64:
 // CHECK-NEXT: ret
 int16x8_t test_vreinterpretq_s16_p64(poly64x2_t a) {
   return vreinterpretq_s16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_s8
+// CHECK-LABEL: test_vreinterpretq_s32_s8:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_s8(int8x16_t a) {
   return vreinterpretq_s32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_s16
+// CHECK-LABEL: test_vreinterpretq_s32_s16:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_s16(int16x8_t a) {
   return vreinterpretq_s32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_s64
+// CHECK-LABEL: test_vreinterpretq_s32_s64:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_s64(int64x2_t a) {
   return vreinterpretq_s32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_u8
+// CHECK-LABEL: test_vreinterpretq_s32_u8:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_u8(uint8x16_t a) {
   return vreinterpretq_s32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_u16
+// CHECK-LABEL: test_vreinterpretq_s32_u16:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_u16(uint16x8_t a) {
   return vreinterpretq_s32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_u32
+// CHECK-LABEL: test_vreinterpretq_s32_u32:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_u32(uint32x4_t a) {
   return vreinterpretq_s32_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_u64
+// CHECK-LABEL: test_vreinterpretq_s32_u64:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_u64(uint64x2_t a) {
   return vreinterpretq_s32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_f16
+// CHECK-LABEL: test_vreinterpretq_s32_f16:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_f16(float16x8_t a) {
   return vreinterpretq_s32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_f32
+// CHECK-LABEL: test_vreinterpretq_s32_f32:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_f32(float32x4_t a) {
   return vreinterpretq_s32_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_f64
+// CHECK-LABEL: test_vreinterpretq_s32_f64:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_f64(float64x2_t a) {
   return vreinterpretq_s32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_p8
+// CHECK-LABEL: test_vreinterpretq_s32_p8:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_p8(poly8x16_t a) {
   return vreinterpretq_s32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_p16
+// CHECK-LABEL: test_vreinterpretq_s32_p16:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_p16(poly16x8_t a) {
   return vreinterpretq_s32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s32_p64
+// CHECK-LABEL: test_vreinterpretq_s32_p64:
 // CHECK-NEXT: ret
 int32x4_t test_vreinterpretq_s32_p64(poly64x2_t a) {
   return vreinterpretq_s32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_s8
+// CHECK-LABEL: test_vreinterpretq_s64_s8:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_s8(int8x16_t a) {
   return vreinterpretq_s64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_s16
+// CHECK-LABEL: test_vreinterpretq_s64_s16:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_s16(int16x8_t a) {
   return vreinterpretq_s64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_s32
+// CHECK-LABEL: test_vreinterpretq_s64_s32:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_s32(int32x4_t a) {
   return vreinterpretq_s64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_u8
+// CHECK-LABEL: test_vreinterpretq_s64_u8:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_u8(uint8x16_t a) {
   return vreinterpretq_s64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_u16
+// CHECK-LABEL: test_vreinterpretq_s64_u16:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_u16(uint16x8_t a) {
   return vreinterpretq_s64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_u32
+// CHECK-LABEL: test_vreinterpretq_s64_u32:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_u32(uint32x4_t a) {
   return vreinterpretq_s64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_u64
+// CHECK-LABEL: test_vreinterpretq_s64_u64:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_u64(uint64x2_t a) {
   return vreinterpretq_s64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_f16
+// CHECK-LABEL: test_vreinterpretq_s64_f16:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_f16(float16x8_t a) {
   return vreinterpretq_s64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_f32
+// CHECK-LABEL: test_vreinterpretq_s64_f32:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_f32(float32x4_t a) {
   return vreinterpretq_s64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_f64
+// CHECK-LABEL: test_vreinterpretq_s64_f64:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_f64(float64x2_t a) {
   return vreinterpretq_s64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_p8
+// CHECK-LABEL: test_vreinterpretq_s64_p8:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_p8(poly8x16_t a) {
   return vreinterpretq_s64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_p16
+// CHECK-LABEL: test_vreinterpretq_s64_p16:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_p16(poly16x8_t a) {
   return vreinterpretq_s64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_s64_p64
+// CHECK-LABEL: test_vreinterpretq_s64_p64:
 // CHECK-NEXT: ret
 int64x2_t test_vreinterpretq_s64_p64(poly64x2_t a) {
   return vreinterpretq_s64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_s8
+// CHECK-LABEL: test_vreinterpretq_u8_s8:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_s8(int8x16_t a) {
   return vreinterpretq_u8_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_s16
+// CHECK-LABEL: test_vreinterpretq_u8_s16:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_s16(int16x8_t a) {
   return vreinterpretq_u8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_s32
+// CHECK-LABEL: test_vreinterpretq_u8_s32:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_s32(int32x4_t a) {
   return vreinterpretq_u8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_s64
+// CHECK-LABEL: test_vreinterpretq_u8_s64:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_s64(int64x2_t a) {
   return vreinterpretq_u8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_u16
+// CHECK-LABEL: test_vreinterpretq_u8_u16:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_u16(uint16x8_t a) {
   return vreinterpretq_u8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_u32
+// CHECK-LABEL: test_vreinterpretq_u8_u32:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_u32(uint32x4_t a) {
   return vreinterpretq_u8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_u64
+// CHECK-LABEL: test_vreinterpretq_u8_u64:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_u64(uint64x2_t a) {
   return vreinterpretq_u8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_f16
+// CHECK-LABEL: test_vreinterpretq_u8_f16:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_f16(float16x8_t a) {
   return vreinterpretq_u8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_f32
+// CHECK-LABEL: test_vreinterpretq_u8_f32:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_f32(float32x4_t a) {
   return vreinterpretq_u8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_f64
+// CHECK-LABEL: test_vreinterpretq_u8_f64:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_f64(float64x2_t a) {
   return vreinterpretq_u8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_p8
+// CHECK-LABEL: test_vreinterpretq_u8_p8:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_p8(poly8x16_t a) {
   return vreinterpretq_u8_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_p16
+// CHECK-LABEL: test_vreinterpretq_u8_p16:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_p16(poly16x8_t a) {
   return vreinterpretq_u8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u8_p64
+// CHECK-LABEL: test_vreinterpretq_u8_p64:
 // CHECK-NEXT: ret
 uint8x16_t test_vreinterpretq_u8_p64(poly64x2_t a) {
   return vreinterpretq_u8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_s8
+// CHECK-LABEL: test_vreinterpretq_u16_s8:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_s8(int8x16_t a) {
   return vreinterpretq_u16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_s16
+// CHECK-LABEL: test_vreinterpretq_u16_s16:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_s16(int16x8_t a) {
   return vreinterpretq_u16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_s32
+// CHECK-LABEL: test_vreinterpretq_u16_s32:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_s32(int32x4_t a) {
   return vreinterpretq_u16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_s64
+// CHECK-LABEL: test_vreinterpretq_u16_s64:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_s64(int64x2_t a) {
   return vreinterpretq_u16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_u8
+// CHECK-LABEL: test_vreinterpretq_u16_u8:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_u8(uint8x16_t a) {
   return vreinterpretq_u16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_u32
+// CHECK-LABEL: test_vreinterpretq_u16_u32:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_u32(uint32x4_t a) {
   return vreinterpretq_u16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_u64
+// CHECK-LABEL: test_vreinterpretq_u16_u64:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_u64(uint64x2_t a) {
   return vreinterpretq_u16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_f16
+// CHECK-LABEL: test_vreinterpretq_u16_f16:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_f16(float16x8_t a) {
   return vreinterpretq_u16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_f32
+// CHECK-LABEL: test_vreinterpretq_u16_f32:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_f32(float32x4_t a) {
   return vreinterpretq_u16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_f64
+// CHECK-LABEL: test_vreinterpretq_u16_f64:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_f64(float64x2_t a) {
   return vreinterpretq_u16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_p8
+// CHECK-LABEL: test_vreinterpretq_u16_p8:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_p8(poly8x16_t a) {
   return vreinterpretq_u16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_p16
+// CHECK-LABEL: test_vreinterpretq_u16_p16:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_p16(poly16x8_t a) {
   return vreinterpretq_u16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u16_p64
+// CHECK-LABEL: test_vreinterpretq_u16_p64:
 // CHECK-NEXT: ret
 uint16x8_t test_vreinterpretq_u16_p64(poly64x2_t a) {
   return vreinterpretq_u16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_s8
+// CHECK-LABEL: test_vreinterpretq_u32_s8:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_s8(int8x16_t a) {
   return vreinterpretq_u32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_s16
+// CHECK-LABEL: test_vreinterpretq_u32_s16:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_s16(int16x8_t a) {
   return vreinterpretq_u32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_s32
+// CHECK-LABEL: test_vreinterpretq_u32_s32:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_s32(int32x4_t a) {
   return vreinterpretq_u32_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_s64
+// CHECK-LABEL: test_vreinterpretq_u32_s64:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_s64(int64x2_t a) {
   return vreinterpretq_u32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_u8
+// CHECK-LABEL: test_vreinterpretq_u32_u8:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_u8(uint8x16_t a) {
   return vreinterpretq_u32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_u16
+// CHECK-LABEL: test_vreinterpretq_u32_u16:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_u16(uint16x8_t a) {
   return vreinterpretq_u32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_u64
+// CHECK-LABEL: test_vreinterpretq_u32_u64:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_u64(uint64x2_t a) {
   return vreinterpretq_u32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_f16
+// CHECK-LABEL: test_vreinterpretq_u32_f16:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_f16(float16x8_t a) {
   return vreinterpretq_u32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_f32
+// CHECK-LABEL: test_vreinterpretq_u32_f32:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_f32(float32x4_t a) {
   return vreinterpretq_u32_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_f64
+// CHECK-LABEL: test_vreinterpretq_u32_f64:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_f64(float64x2_t a) {
   return vreinterpretq_u32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_p8
+// CHECK-LABEL: test_vreinterpretq_u32_p8:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_p8(poly8x16_t a) {
   return vreinterpretq_u32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_p16
+// CHECK-LABEL: test_vreinterpretq_u32_p16:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_p16(poly16x8_t a) {
   return vreinterpretq_u32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u32_p64
+// CHECK-LABEL: test_vreinterpretq_u32_p64:
 // CHECK-NEXT: ret
 uint32x4_t test_vreinterpretq_u32_p64(poly64x2_t a) {
   return vreinterpretq_u32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_s8
+// CHECK-LABEL: test_vreinterpretq_u64_s8:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_s8(int8x16_t a) {
   return vreinterpretq_u64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_s16
+// CHECK-LABEL: test_vreinterpretq_u64_s16:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_s16(int16x8_t a) {
   return vreinterpretq_u64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_s32
+// CHECK-LABEL: test_vreinterpretq_u64_s32:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_s32(int32x4_t a) {
   return vreinterpretq_u64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_s64
+// CHECK-LABEL: test_vreinterpretq_u64_s64:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_s64(int64x2_t a) {
   return vreinterpretq_u64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_u8
+// CHECK-LABEL: test_vreinterpretq_u64_u8:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_u8(uint8x16_t a) {
   return vreinterpretq_u64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_u16
+// CHECK-LABEL: test_vreinterpretq_u64_u16:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_u16(uint16x8_t a) {
   return vreinterpretq_u64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_u32
+// CHECK-LABEL: test_vreinterpretq_u64_u32:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_u32(uint32x4_t a) {
   return vreinterpretq_u64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_f16
+// CHECK-LABEL: test_vreinterpretq_u64_f16:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_f16(float16x8_t a) {
   return vreinterpretq_u64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_f32
+// CHECK-LABEL: test_vreinterpretq_u64_f32:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_f32(float32x4_t a) {
   return vreinterpretq_u64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_f64
+// CHECK-LABEL: test_vreinterpretq_u64_f64:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_f64(float64x2_t a) {
   return vreinterpretq_u64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_p8
+// CHECK-LABEL: test_vreinterpretq_u64_p8:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_p8(poly8x16_t a) {
   return vreinterpretq_u64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_p16
+// CHECK-LABEL: test_vreinterpretq_u64_p16:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_p16(poly16x8_t a) {
   return vreinterpretq_u64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_u64_p64
+// CHECK-LABEL: test_vreinterpretq_u64_p64:
 // CHECK-NEXT: ret
 uint64x2_t test_vreinterpretq_u64_p64(poly64x2_t a) {
   return vreinterpretq_u64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_s8
+// CHECK-LABEL: test_vreinterpretq_f16_s8:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_s8(int8x16_t a) {
   return vreinterpretq_f16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_s16
+// CHECK-LABEL: test_vreinterpretq_f16_s16:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_s16(int16x8_t a) {
   return vreinterpretq_f16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_s32
+// CHECK-LABEL: test_vreinterpretq_f16_s32:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_s32(int32x4_t a) {
   return vreinterpretq_f16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_s64
+// CHECK-LABEL: test_vreinterpretq_f16_s64:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_s64(int64x2_t a) {
   return vreinterpretq_f16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_u8
+// CHECK-LABEL: test_vreinterpretq_f16_u8:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_u8(uint8x16_t a) {
   return vreinterpretq_f16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_u16
+// CHECK-LABEL: test_vreinterpretq_f16_u16:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_u16(uint16x8_t a) {
   return vreinterpretq_f16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_u32
+// CHECK-LABEL: test_vreinterpretq_f16_u32:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_u32(uint32x4_t a) {
   return vreinterpretq_f16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_u64
+// CHECK-LABEL: test_vreinterpretq_f16_u64:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_u64(uint64x2_t a) {
   return vreinterpretq_f16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_f32
+// CHECK-LABEL: test_vreinterpretq_f16_f32:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_f32(float32x4_t a) {
   return vreinterpretq_f16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_f64
+// CHECK-LABEL: test_vreinterpretq_f16_f64:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_f64(float64x2_t a) {
   return vreinterpretq_f16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_p8
+// CHECK-LABEL: test_vreinterpretq_f16_p8:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_p8(poly8x16_t a) {
   return vreinterpretq_f16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_p16
+// CHECK-LABEL: test_vreinterpretq_f16_p16:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_p16(poly16x8_t a) {
   return vreinterpretq_f16_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f16_p64
+// CHECK-LABEL: test_vreinterpretq_f16_p64:
 // CHECK-NEXT: ret
 float16x8_t test_vreinterpretq_f16_p64(poly64x2_t a) {
   return vreinterpretq_f16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_s8
+// CHECK-LABEL: test_vreinterpretq_f32_s8:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_s8(int8x16_t a) {
   return vreinterpretq_f32_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_s16
+// CHECK-LABEL: test_vreinterpretq_f32_s16:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_s16(int16x8_t a) {
   return vreinterpretq_f32_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_s32
+// CHECK-LABEL: test_vreinterpretq_f32_s32:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_s32(int32x4_t a) {
   return vreinterpretq_f32_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_s64
+// CHECK-LABEL: test_vreinterpretq_f32_s64:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_s64(int64x2_t a) {
   return vreinterpretq_f32_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_u8
+// CHECK-LABEL: test_vreinterpretq_f32_u8:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_u8(uint8x16_t a) {
   return vreinterpretq_f32_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_u16
+// CHECK-LABEL: test_vreinterpretq_f32_u16:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_u16(uint16x8_t a) {
   return vreinterpretq_f32_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_u32
+// CHECK-LABEL: test_vreinterpretq_f32_u32:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_u32(uint32x4_t a) {
   return vreinterpretq_f32_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_u64
+// CHECK-LABEL: test_vreinterpretq_f32_u64:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_u64(uint64x2_t a) {
   return vreinterpretq_f32_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_f16
+// CHECK-LABEL: test_vreinterpretq_f32_f16:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_f16(float16x8_t a) {
   return vreinterpretq_f32_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_f64
+// CHECK-LABEL: test_vreinterpretq_f32_f64:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_f64(float64x2_t a) {
   return vreinterpretq_f32_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_p8
+// CHECK-LABEL: test_vreinterpretq_f32_p8:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_p8(poly8x16_t a) {
   return vreinterpretq_f32_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_p16
+// CHECK-LABEL: test_vreinterpretq_f32_p16:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_p16(poly16x8_t a) {
   return vreinterpretq_f32_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f32_p64
+// CHECK-LABEL: test_vreinterpretq_f32_p64:
 // CHECK-NEXT: ret
 float32x4_t test_vreinterpretq_f32_p64(poly64x2_t a) {
   return vreinterpretq_f32_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_s8
+// CHECK-LABEL: test_vreinterpretq_f64_s8:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_s8(int8x16_t a) {
   return vreinterpretq_f64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_s16
+// CHECK-LABEL: test_vreinterpretq_f64_s16:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_s16(int16x8_t a) {
   return vreinterpretq_f64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_s32
+// CHECK-LABEL: test_vreinterpretq_f64_s32:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_s32(int32x4_t a) {
   return vreinterpretq_f64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_s64
+// CHECK-LABEL: test_vreinterpretq_f64_s64:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_s64(int64x2_t a) {
   return vreinterpretq_f64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_u8
+// CHECK-LABEL: test_vreinterpretq_f64_u8:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_u8(uint8x16_t a) {
   return vreinterpretq_f64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_u16
+// CHECK-LABEL: test_vreinterpretq_f64_u16:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_u16(uint16x8_t a) {
   return vreinterpretq_f64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_u32
+// CHECK-LABEL: test_vreinterpretq_f64_u32:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_u32(uint32x4_t a) {
   return vreinterpretq_f64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_u64
+// CHECK-LABEL: test_vreinterpretq_f64_u64:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_u64(uint64x2_t a) {
   return vreinterpretq_f64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_f16
+// CHECK-LABEL: test_vreinterpretq_f64_f16:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_f16(float16x8_t a) {
   return vreinterpretq_f64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_f32
+// CHECK-LABEL: test_vreinterpretq_f64_f32:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_f32(float32x4_t a) {
   return vreinterpretq_f64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_p8
+// CHECK-LABEL: test_vreinterpretq_f64_p8:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_p8(poly8x16_t a) {
   return vreinterpretq_f64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_p16
+// CHECK-LABEL: test_vreinterpretq_f64_p16:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_p16(poly16x8_t a) {
   return vreinterpretq_f64_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_f64_p64
+// CHECK-LABEL: test_vreinterpretq_f64_p64:
 // CHECK-NEXT: ret
 float64x2_t test_vreinterpretq_f64_p64(poly64x2_t a) {
   return vreinterpretq_f64_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_s8
+// CHECK-LABEL: test_vreinterpretq_p8_s8:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_s8(int8x16_t a) {
   return vreinterpretq_p8_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_s16
+// CHECK-LABEL: test_vreinterpretq_p8_s16:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_s16(int16x8_t a) {
   return vreinterpretq_p8_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_s32
+// CHECK-LABEL: test_vreinterpretq_p8_s32:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_s32(int32x4_t a) {
   return vreinterpretq_p8_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_s64
+// CHECK-LABEL: test_vreinterpretq_p8_s64:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_s64(int64x2_t a) {
   return vreinterpretq_p8_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_u8
+// CHECK-LABEL: test_vreinterpretq_p8_u8:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_u8(uint8x16_t a) {
   return vreinterpretq_p8_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_u16
+// CHECK-LABEL: test_vreinterpretq_p8_u16:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_u16(uint16x8_t a) {
   return vreinterpretq_p8_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_u32
+// CHECK-LABEL: test_vreinterpretq_p8_u32:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_u32(uint32x4_t a) {
   return vreinterpretq_p8_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_u64
+// CHECK-LABEL: test_vreinterpretq_p8_u64:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_u64(uint64x2_t a) {
   return vreinterpretq_p8_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_f16
+// CHECK-LABEL: test_vreinterpretq_p8_f16:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_f16(float16x8_t a) {
   return vreinterpretq_p8_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_f32
+// CHECK-LABEL: test_vreinterpretq_p8_f32:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_f32(float32x4_t a) {
   return vreinterpretq_p8_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_f64
+// CHECK-LABEL: test_vreinterpretq_p8_f64:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_f64(float64x2_t a) {
   return vreinterpretq_p8_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_p16
+// CHECK-LABEL: test_vreinterpretq_p8_p16:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_p16(poly16x8_t a) {
   return vreinterpretq_p8_p16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p8_p64
+// CHECK-LABEL: test_vreinterpretq_p8_p64:
 // CHECK-NEXT: ret
 poly8x16_t test_vreinterpretq_p8_p64(poly64x2_t a) {
   return vreinterpretq_p8_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_s8
+// CHECK-LABEL: test_vreinterpretq_p16_s8:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_s8(int8x16_t a) {
   return vreinterpretq_p16_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_s16
+// CHECK-LABEL: test_vreinterpretq_p16_s16:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_s16(int16x8_t a) {
   return vreinterpretq_p16_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_s32
+// CHECK-LABEL: test_vreinterpretq_p16_s32:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_s32(int32x4_t a) {
   return vreinterpretq_p16_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_s64
+// CHECK-LABEL: test_vreinterpretq_p16_s64:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_s64(int64x2_t a) {
   return vreinterpretq_p16_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_u8
+// CHECK-LABEL: test_vreinterpretq_p16_u8:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_u8(uint8x16_t a) {
   return vreinterpretq_p16_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_u16
+// CHECK-LABEL: test_vreinterpretq_p16_u16:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_u16(uint16x8_t a) {
   return vreinterpretq_p16_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_u32
+// CHECK-LABEL: test_vreinterpretq_p16_u32:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_u32(uint32x4_t a) {
   return vreinterpretq_p16_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_u64
+// CHECK-LABEL: test_vreinterpretq_p16_u64:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_u64(uint64x2_t a) {
   return vreinterpretq_p16_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_f16
+// CHECK-LABEL: test_vreinterpretq_p16_f16:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_f16(float16x8_t a) {
   return vreinterpretq_p16_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_f32
+// CHECK-LABEL: test_vreinterpretq_p16_f32:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_f32(float32x4_t a) {
   return vreinterpretq_p16_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_f64
+// CHECK-LABEL: test_vreinterpretq_p16_f64:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_f64(float64x2_t a) {
   return vreinterpretq_p16_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_p8
+// CHECK-LABEL: test_vreinterpretq_p16_p8:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_p8(poly8x16_t a) {
   return vreinterpretq_p16_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p16_p64
+// CHECK-LABEL: test_vreinterpretq_p16_p64:
 // CHECK-NEXT: ret
 poly16x8_t test_vreinterpretq_p16_p64(poly64x2_t a) {
   return vreinterpretq_p16_p64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_s8
+// CHECK-LABEL: test_vreinterpretq_p64_s8:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_s8(int8x16_t a) {
   return vreinterpretq_p64_s8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_s16
+// CHECK-LABEL: test_vreinterpretq_p64_s16:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_s16(int16x8_t a) {
   return vreinterpretq_p64_s16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_s32
+// CHECK-LABEL: test_vreinterpretq_p64_s32:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_s32(int32x4_t a) {
   return vreinterpretq_p64_s32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_s64
+// CHECK-LABEL: test_vreinterpretq_p64_s64:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_s64(int64x2_t a) {
   return vreinterpretq_p64_s64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_u8
+// CHECK-LABEL: test_vreinterpretq_p64_u8:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_u8(uint8x16_t a) {
   return vreinterpretq_p64_u8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_u16
+// CHECK-LABEL: test_vreinterpretq_p64_u16:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_u16(uint16x8_t a) {
   return vreinterpretq_p64_u16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_u32
+// CHECK-LABEL: test_vreinterpretq_p64_u32:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_u32(uint32x4_t a) {
   return vreinterpretq_p64_u32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_u64
+// CHECK-LABEL: test_vreinterpretq_p64_u64:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_u64(uint64x2_t a) {
   return vreinterpretq_p64_u64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_f16
+// CHECK-LABEL: test_vreinterpretq_p64_f16:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_f16(float16x8_t a) {
   return vreinterpretq_p64_f16(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_f32
+// CHECK-LABEL: test_vreinterpretq_p64_f32:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_f32(float32x4_t a) {
   return vreinterpretq_p64_f32(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_f64
+// CHECK-LABEL: test_vreinterpretq_p64_f64:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_f64(float64x2_t a) {
   return vreinterpretq_p64_f64(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_p8
+// CHECK-LABEL: test_vreinterpretq_p64_p8:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_p8(poly8x16_t a) {
   return vreinterpretq_p64_p8(a);
 }
 
-// CHECK-LABEL: test_vreinterpretq_p64_p16
+// CHECK-LABEL: test_vreinterpretq_p64_p16:
 // CHECK-NEXT: ret
 poly64x2_t test_vreinterpretq_p64_p16(poly16x8_t a) {
   return vreinterpretq_p64_p16(a);
@@ -11345,6 +11264,48 @@
   // CHECK: usqadd d{{[0-9]+}}, d{{[0-9]+}}
 }
 
+uint8x8_t test_vsqadd_u8(uint8x8_t a, int8x8_t b) {
+  // CHECK-LABEL: test_vsqadd_u8
+  return vsqadd_u8(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
+}
+
+uint8x16_t test_vsqaddq_u8(uint8x16_t a, int8x16_t b) {
+  // CHECK-LABEL: test_vsqaddq_u8
+  return vsqaddq_u8(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
+}
+
+uint16x4_t test_vsqadd_u16(uint16x4_t a, int16x4_t b) {
+  // CHECK-LABEL: test_vsqadd_u16
+  return vsqadd_u16(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
+}
+
+uint16x8_t test_vsqaddq_u16(uint16x8_t a, int16x8_t b) {
+  // CHECK-LABEL: test_vsqaddq_u16
+  return vsqaddq_u16(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
+}
+
+uint32x2_t test_vsqadd_u32(uint32x2_t a, int32x2_t b) {
+  // CHECK-LABEL: test_vsqadd_u32
+  return vsqadd_u32(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32x4_t test_vsqaddq_u32(uint32x4_t a, int32x4_t b) {
+  // CHECK-LABEL: test_vsqaddq_u32
+  return vsqaddq_u32(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
+}
+
+uint64x2_t test_vsqaddq_u64(uint64x2_t a, int64x2_t b) {
+  // CHECK-LABEL: test_vsqaddq_u64
+  return vsqaddq_u64(a, b);
+  // CHECK: usqadd {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+}
+
 int64x1_t test_vabs_s64(int64x1_t a) {
   // CHECK-LABEL: test_vabs_s64
   return vabs_s64(a);
@@ -11559,13 +11520,13 @@
 int64x1_t test_vcvt_s64_f64(float64x1_t a) {
   // CHECK-LABEL: test_vcvt_s64_f64
   return vcvt_s64_f64(a);
-  // CHECK: fcvtzs d{{[0-9]+}}, d{{[0-9]+}}
+  // CHECK: fcvtzs {{[xd][0-9]+}}, d{{[0-9]+}}
 }
 
 uint64x1_t test_vcvt_u64_f64(float64x1_t a) {
   // CHECK-LABEL: test_vcvt_u64_f64
   return vcvt_u64_f64(a);
-  // CHECK: fcvtzu d{{[0-9]+}}, d{{[0-9]+}}
+  // CHECK: fcvtzu {{[xd][0-9]+}}, d{{[0-9]+}}
 }
 
 int64x1_t test_vcvtn_s64_f64(float64x1_t a) {
@@ -11619,13 +11580,13 @@
 float64x1_t test_vcvt_f64_s64(int64x1_t a) {
   // CHECK-LABEL: test_vcvt_f64_s64
   return vcvt_f64_s64(a);
-  // CHECK: scvtf d{{[0-9]+}}, d{{[0-9]+}}
+  // CHECK: scvtf d{{[0-9]+}}, {{[xd][0-9]+}}
 }
 
 float64x1_t test_vcvt_f64_u64(uint64x1_t a) {
   // CHECK-LABEL: test_vcvt_f64_u64
   return vcvt_f64_u64(a);
-  // CHECK: ucvtf d{{[0-9]+}}, d{{[0-9]+}}
+  // CHECK: ucvtf d{{[0-9]+}}, {{[xd][0-9]+}}
 }
 
 int64x1_t test_vcvt_n_s64_f64(float64x1_t a) {
@@ -11723,3 +11684,51 @@
   return vrsqrts_f64(a, b);
   // CHECK: frsqrts d{{[0-9]+}}, d{{[0-9]+}}, d{{[0-9]+}}
 }
+
+int32_t test_vminv_s32(int32x2_t a) {
+  // CHECK-LABEL: test_vminv_s32
+  return vminv_s32(a);
+  // CHECK: sminp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32_t test_vminv_u32(uint32x2_t a) {
+  // CHECK-LABEL: test_vminv_u32
+  return vminv_u32(a);
+  // CHECK: uminp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32_t test_vmaxv_s32(int32x2_t a) {
+  // CHECK-LABEL: test_vmaxv_s32
+  return vmaxv_s32(a);
+  // CHECK: smaxp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32_t test_vmaxv_u32(uint32x2_t a) {
+  // CHECK-LABEL: test_vmaxv_u32
+  return vmaxv_u32(a);
+  // CHECK: umaxp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int32_t test_vaddv_s32(int32x2_t a) {
+  // CHECK-LABEL: test_vaddv_s32
+  return vaddv_s32(a);
+  // CHECK: addp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+uint32_t test_vaddv_u32(uint32x2_t a) {
+  // CHECK-LABEL: test_vaddv_u32
+  return vaddv_u32(a);
+  // CHECK: addp {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, {{v[0-9]+}}.2s
+}
+
+int64_t test_vaddlv_s32(int32x2_t a) {
+  // CHECK-LABEL: test_vaddlv_s32
+  return vaddlv_s32(a);
+  // CHECK: saddlp {{v[0-9]+}}.1d, {{v[0-9]+}}.2s
+}
+
+uint64_t test_vaddlv_u32(uint32x2_t a) {
+  // CHECK-LABEL: test_vaddlv_u32
+  return vaddlv_u32(a);
+  // CHECK: uaddlp {{v[0-9]+}}.1d, {{v[0-9]+}}.2s
+}
diff --git a/test/CodeGen/aarch64-neon-ld-sideeffect.c b/test/CodeGen/aarch64-neon-ld-sideeffect.c
new file mode 100644
index 0000000..1160133
--- /dev/null
+++ b/test/CodeGen/aarch64-neon-ld-sideeffect.c
@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -emit-llvm -O0 -o - %s | FileCheck %s
+
+#include <arm_neon.h>
+
+void *foo(void);
+
+float32x2_t bar(void) {
+  // CHECK-LABEL: @bar
+  return vld1_f32(foo());
+  // CHECK: call i8* @foo
+  // CHECK-NOT: call i8* @foo
+  // CHECK: call <2 x float> @llvm.arm.neon.vld1
+}
diff --git a/test/CodeGen/aarch64-neon-ldst-one.c b/test/CodeGen/aarch64-neon-ldst-one.c
index f629260..2d63628 100644
--- a/test/CodeGen/aarch64-neon-ldst-one.c
+++ b/test/CodeGen/aarch64-neon-ldst-one.c
@@ -1,2047 +1,2052 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 #include <arm_neon.h>
 
-uint8x16_t test_vld1q_dup_u8(uint8_t const *a) {
+uint8x16_t test_vld1q_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_u8
   return vld1q_dup_u8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x8_t test_vld1q_dup_u16(uint16_t const *a) {
+uint16x8_t test_vld1q_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_u16
   return vld1q_dup_u16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x4_t test_vld1q_dup_u32(uint32_t const *a) {
+uint32x4_t test_vld1q_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_u32
   return vld1q_dup_u32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x2_t test_vld1q_dup_u64(uint64_t const *a) {
+uint64x2_t test_vld1q_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_u64
   return vld1q_dup_u64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-int8x16_t test_vld1q_dup_s8(int8_t const *a) {
+int8x16_t test_vld1q_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_s8
   return vld1q_dup_s8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x8_t test_vld1q_dup_s16(int16_t const *a) {
+int16x8_t test_vld1q_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_s16
   return vld1q_dup_s16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x4_t test_vld1q_dup_s32(int32_t const *a) {
+int32x4_t test_vld1q_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_s32
   return vld1q_dup_s32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x2_t test_vld1q_dup_s64(int64_t const *a) {
+int64x2_t test_vld1q_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_s64
   return vld1q_dup_s64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-float16x8_t test_vld1q_dup_f16(float16_t const *a) {
+float16x8_t test_vld1q_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_f16
   return vld1q_dup_f16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x4_t test_vld1q_dup_f32(float32_t const *a) {
+float32x4_t test_vld1q_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_f32
   return vld1q_dup_f32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x2_t test_vld1q_dup_f64(float64_t const *a) {
+float64x2_t test_vld1q_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_f64
   return vld1q_dup_f64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x16_t test_vld1q_dup_p8(poly8_t const *a) {
+poly8x16_t test_vld1q_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_p8
   return vld1q_dup_p8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x8_t test_vld1q_dup_p16(poly16_t const *a) {
+poly16x8_t test_vld1q_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_p16
   return vld1q_dup_p16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x2_t test_vld1q_dup_p64(poly64_t const *a) {
+poly64x2_t test_vld1q_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld1q_dup_p64
   return vld1q_dup_p64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x8_t test_vld1_dup_u8(uint8_t const *a) {
+uint8x8_t test_vld1_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld1_dup_u8
   return vld1_dup_u8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x4_t test_vld1_dup_u16(uint16_t const *a) {
+uint16x4_t test_vld1_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld1_dup_u16
   return vld1_dup_u16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x2_t test_vld1_dup_u32(uint32_t const *a) {
+uint32x2_t test_vld1_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld1_dup_u32
   return vld1_dup_u32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x1_t test_vld1_dup_u64(uint64_t const *a) {
+uint64x1_t test_vld1_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld1_dup_u64
   return vld1_dup_u64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-int8x8_t test_vld1_dup_s8(int8_t const *a) {
+int8x8_t test_vld1_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld1_dup_s8
   return vld1_dup_s8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x4_t test_vld1_dup_s16(int16_t const *a) {
+int16x4_t test_vld1_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld1_dup_s16
   return vld1_dup_s16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x2_t test_vld1_dup_s32(int32_t const *a) {
+int32x2_t test_vld1_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld1_dup_s32
   return vld1_dup_s32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x1_t test_vld1_dup_s64(int64_t const *a) {
+int64x1_t test_vld1_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld1_dup_s64
   return vld1_dup_s64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-float16x4_t test_vld1_dup_f16(float16_t const *a) {
+float16x4_t test_vld1_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld1_dup_f16
   return vld1_dup_f16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x2_t test_vld1_dup_f32(float32_t const *a) {
+float32x2_t test_vld1_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld1_dup_f32
   return vld1_dup_f32(a);
-  // CHECK: ld1r {v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x1_t test_vld1_dup_f64(float64_t const *a) {
+float64x1_t test_vld1_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld1_dup_f64
   return vld1_dup_f64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x8_t test_vld1_dup_p8(poly8_t const *a) {
+poly8x8_t test_vld1_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld1_dup_p8
   return vld1_dup_p8(a);
-  // CHECK: ld1r {v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x4_t test_vld1_dup_p16(poly16_t const *a) {
+poly16x4_t test_vld1_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld1_dup_p16
   return vld1_dup_p16(a);
-  // CHECK: ld1r {v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1r {{{ *v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x1_t test_vld1_dup_p64(poly64_t const *a) {
+poly64x1_t test_vld1_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld1_dup_p64
   return vld1_dup_p64(a);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x16x2_t test_vld2q_dup_u8(uint8_t const *a) {
+uint8x16x2_t test_vld2q_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_u8
   return vld2q_dup_u8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x8x2_t test_vld2q_dup_u16(uint16_t const *a) {
+uint16x8x2_t test_vld2q_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_u16
   return vld2q_dup_u16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x4x2_t test_vld2q_dup_u32(uint32_t const *a) {
+uint32x4x2_t test_vld2q_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_u32
   return vld2q_dup_u32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x2x2_t test_vld2q_dup_u64(uint64_t const *a) {
+uint64x2x2_t test_vld2q_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_u64
   return vld2q_dup_u64(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-int8x16x2_t test_vld2q_dup_s8(int8_t const *a) {
+int8x16x2_t test_vld2q_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_s8
   return vld2q_dup_s8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x8x2_t test_vld2q_dup_s16(int16_t const *a) {
+int16x8x2_t test_vld2q_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_s16
   return vld2q_dup_s16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x4x2_t test_vld2q_dup_s32(int32_t const *a) {
+int32x4x2_t test_vld2q_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_s32
   return vld2q_dup_s32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x2x2_t test_vld2q_dup_s64(int64_t const *a) {
+int64x2x2_t test_vld2q_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_s64
   return vld2q_dup_s64(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-float16x8x2_t test_vld2q_dup_f16(float16_t const *a) {
+float16x8x2_t test_vld2q_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_f16
   return vld2q_dup_f16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x4x2_t test_vld2q_dup_f32(float32_t const *a) {
+float32x4x2_t test_vld2q_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_f32
   return vld2q_dup_f32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x2x2_t test_vld2q_dup_f64(float64_t const *a) {
+float64x2x2_t test_vld2q_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_f64
   return vld2q_dup_f64(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x16x2_t test_vld2q_dup_p8(poly8_t const *a) {
+poly8x16x2_t test_vld2q_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_p8
   return vld2q_dup_p8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x8x2_t test_vld2q_dup_p16(poly16_t const *a) {
+poly16x8x2_t test_vld2q_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_p16
   return vld2q_dup_p16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x2x2_t test_vld2q_dup_p64(poly64_t const *a) {
+poly64x2x2_t test_vld2q_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld2q_dup_p64
   return vld2q_dup_p64(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x8x2_t test_vld2_dup_u8(uint8_t const *a) {
+uint8x8x2_t test_vld2_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld2_dup_u8
   return vld2_dup_u8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x4x2_t test_vld2_dup_u16(uint16_t const *a) {
+uint16x4x2_t test_vld2_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld2_dup_u16
   return vld2_dup_u16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x2x2_t test_vld2_dup_u32(uint32_t const *a) {
+uint32x2x2_t test_vld2_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld2_dup_u32
   return vld2_dup_u32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x1x2_t test_vld2_dup_u64(uint64_t const *a) {
+uint64x1x2_t test_vld2_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld2_dup_u64
   return vld2_dup_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2r}} {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-int8x8x2_t test_vld2_dup_s8(int8_t const *a) {
+int8x8x2_t test_vld2_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld2_dup_s8
   return vld2_dup_s8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x4x2_t test_vld2_dup_s16(int16_t const *a) {
+int16x4x2_t test_vld2_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld2_dup_s16
   return vld2_dup_s16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x2x2_t test_vld2_dup_s32(int32_t const *a) {
+int32x2x2_t test_vld2_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld2_dup_s32
   return vld2_dup_s32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x1x2_t test_vld2_dup_s64(int64_t const *a) {
+int64x1x2_t test_vld2_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld2_dup_s64
   return vld2_dup_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2r}} {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-float16x4x2_t test_vld2_dup_f16(float16_t const *a) {
+float16x4x2_t test_vld2_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld2_dup_f16
   return vld2_dup_f16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x2x2_t test_vld2_dup_f32(float32_t const *a) {
+float32x2x2_t test_vld2_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld2_dup_f32
   return vld2_dup_f32(a);
-  // CHECK: ld2r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x1x2_t test_vld2_dup_f64(float64_t const *a) {
+float64x1x2_t test_vld2_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld2_dup_f64
   return vld2_dup_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2r}} {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x8x2_t test_vld2_dup_p8(poly8_t const *a) {
+poly8x8x2_t test_vld2_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld2_dup_p8
   return vld2_dup_p8(a);
-  // CHECK: ld2r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x4x2_t test_vld2_dup_p16(poly16_t const *a) {
+poly16x4x2_t test_vld2_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld2_dup_p16
   return vld2_dup_p16(a);
-  // CHECK: ld2r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2r {{{ *v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x1x2_t test_vld2_dup_p64(poly64_t const *a) {
+poly64x1x2_t test_vld2_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld2_dup_p64
   return vld2_dup_p64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld2r}} {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x16x3_t test_vld3q_dup_u8(uint8_t const *a) {
+uint8x16x3_t test_vld3q_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_u8
   return vld3q_dup_u8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
+  // CHECK: ld3r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint16x8x3_t test_vld3q_dup_u16(uint16_t const *a) {
+uint16x8x3_t test_vld3q_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_u16
   return vld3q_dup_u16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
+  // CHECK: ld3r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint32x4x3_t test_vld3q_dup_u32(uint32_t const *a) {
+uint32x4x3_t test_vld3q_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_u32
   return vld3q_dup_u32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
+  // CHECK: ld3r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint64x2x3_t test_vld3q_dup_u64(uint64_t const *a) {
+uint64x2x3_t test_vld3q_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_u64
   return vld3q_dup_u64(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
+  // CHECK: ld3r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int8x16x3_t test_vld3q_dup_s8(int8_t const *a) {
+int8x16x3_t test_vld3q_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_s8
   return vld3q_dup_s8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
+  // CHECK: ld3r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int16x8x3_t test_vld3q_dup_s16(int16_t const *a) {
+int16x8x3_t test_vld3q_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_s16
   return vld3q_dup_s16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
+  // CHECK: ld3r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int32x4x3_t test_vld3q_dup_s32(int32_t const *a) {
+int32x4x3_t test_vld3q_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_s32
   return vld3q_dup_s32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
+  // CHECK: ld3r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int64x2x3_t test_vld3q_dup_s64(int64_t const *a) {
+int64x2x3_t test_vld3q_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_s64
   return vld3q_dup_s64(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
+  // CHECK: ld3r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float16x8x3_t test_vld3q_dup_f16(float16_t const *a) {
+float16x8x3_t test_vld3q_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_f16
   return vld3q_dup_f16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
+  // CHECK: ld3r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float32x4x3_t test_vld3q_dup_f32(float32_t const *a) {
+float32x4x3_t test_vld3q_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_f32
   return vld3q_dup_f32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s},
+  // CHECK: ld3r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float64x2x3_t test_vld3q_dup_f64(float64_t const *a) {
+float64x2x3_t test_vld3q_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_f64
   return vld3q_dup_f64(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
+  // CHECK: ld3r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly8x16x3_t test_vld3q_dup_p8(poly8_t const *a) {
+poly8x16x3_t test_vld3q_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_p8
   return vld3q_dup_p8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b},
+  // CHECK: ld3r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly16x8x3_t test_vld3q_dup_p16(poly16_t const *a) {
+poly16x8x3_t test_vld3q_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_p16
   return vld3q_dup_p16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h},
+  // CHECK: ld3r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly64x2x3_t test_vld3q_dup_p64(poly64_t const *a) {
+poly64x2x3_t test_vld3q_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld3q_dup_p64
   return vld3q_dup_p64(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d},
+  // CHECK: ld3r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint8x8x3_t test_vld3_dup_u8(uint8_t const *a) {
+uint8x8x3_t test_vld3_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld3_dup_u8
   return vld3_dup_u8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
+  // CHECK: ld3r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint16x4x3_t test_vld3_dup_u16(uint16_t const *a) {
+uint16x4x3_t test_vld3_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld3_dup_u16
   return vld3_dup_u16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
+  // CHECK: ld3r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint32x2x3_t test_vld3_dup_u32(uint32_t const *a) {
+uint32x2x3_t test_vld3_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld3_dup_u32
   return vld3_dup_u32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
+  // CHECK: ld3r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint64x1x3_t test_vld3_dup_u64(uint64_t const *a) {
+uint64x1x3_t test_vld3_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld3_dup_u64
   return vld3_dup_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
+  // CHECK: {{ld1|ld3r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int8x8x3_t test_vld3_dup_s8(int8_t const *a) {
+int8x8x3_t test_vld3_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld3_dup_s8
   return vld3_dup_s8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
+  // CHECK: ld3r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int16x4x3_t test_vld3_dup_s16(int16_t const *a) {
+int16x4x3_t test_vld3_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld3_dup_s16
   return vld3_dup_s16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
+  // CHECK: ld3r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int32x2x3_t test_vld3_dup_s32(int32_t const *a) {
+int32x2x3_t test_vld3_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld3_dup_s32
   return vld3_dup_s32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
+  // CHECK: ld3r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-int64x1x3_t test_vld3_dup_s64(int64_t const *a) {
+int64x1x3_t test_vld3_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld3_dup_s64
   return vld3_dup_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
+  // CHECK: {{ld1|ld3r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float16x4x3_t test_vld3_dup_f16(float16_t const *a) {
+float16x4x3_t test_vld3_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld3_dup_f16
   return vld3_dup_f16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
+  // CHECK: ld3r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float32x2x3_t test_vld3_dup_f32(float32_t const *a) {
+float32x2x3_t test_vld3_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld3_dup_f32
   return vld3_dup_f32(a);
-  // CHECK: ld3r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s},
+  // CHECK: ld3r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-float64x1x3_t test_vld3_dup_f64(float64_t const *a) {
+float64x1x3_t test_vld3_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld3_dup_f64
   return vld3_dup_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
+  // CHECK: {{ld1|ld3r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly8x8x3_t test_vld3_dup_p8(poly8_t const *a) {
+poly8x8x3_t test_vld3_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld3_dup_p8
   return vld3_dup_p8(a);
-  // CHECK: ld3r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b},
+  // CHECK: ld3r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly16x4x3_t test_vld3_dup_p16(poly16_t const *a) {
+poly16x4x3_t test_vld3_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld3_dup_p16
   return vld3_dup_p16(a);
-  // CHECK: ld3r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h},
+  // CHECK: ld3r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-poly64x1x3_t test_vld3_dup_p64(poly64_t const *a) {
+poly64x1x3_t test_vld3_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld3_dup_p64
   return vld3_dup_p64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d},
+  // CHECK: {{ld1|ld3r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}},
   // [{{x[0-9]+|sp}}]
 }
 
-uint8x16x4_t test_vld4q_dup_u8(uint8_t const *a) {
+uint8x16x4_t test_vld4q_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_u8
   return vld4q_dup_u8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x8x4_t test_vld4q_dup_u16(uint16_t const *a) {
+uint16x8x4_t test_vld4q_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_u16
   return vld4q_dup_u16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x4x4_t test_vld4q_dup_u32(uint32_t const *a) {
+uint32x4x4_t test_vld4q_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_u32
   return vld4q_dup_u32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x2x4_t test_vld4q_dup_u64(uint64_t const *a) {
+uint64x2x4_t test_vld4q_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_u64
   return vld4q_dup_u64(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-int8x16x4_t test_vld4q_dup_s8(int8_t const *a) {
+int8x16x4_t test_vld4q_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_s8
   return vld4q_dup_s8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x8x4_t test_vld4q_dup_s16(int16_t const *a) {
+int16x8x4_t test_vld4q_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_s16
   return vld4q_dup_s16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x4x4_t test_vld4q_dup_s32(int32_t const *a) {
+int32x4x4_t test_vld4q_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_s32
   return vld4q_dup_s32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x2x4_t test_vld4q_dup_s64(int64_t const *a) {
+int64x2x4_t test_vld4q_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_s64
   return vld4q_dup_s64(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-float16x8x4_t test_vld4q_dup_f16(float16_t const *a) {
+float16x8x4_t test_vld4q_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_f16
   return vld4q_dup_f16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x4x4_t test_vld4q_dup_f32(float32_t const *a) {
+float32x4x4_t test_vld4q_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_f32
   return vld4q_dup_f32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4s, v{{[0-9]+}}.4s, v{{[0-9]+}}.4s,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s, v[0-9]+.4s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x2x4_t test_vld4q_dup_f64(float64_t const *a) {
+float64x2x4_t test_vld4q_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_f64
   return vld4q_dup_f64(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.4s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x16x4_t test_vld4q_dup_p8(poly8_t const *a) {
+poly8x16x4_t test_vld4q_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_p8
   return vld4q_dup_p8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.16b, v{{[0-9]+}}.16b, v{{[0-9]+}}.16b,
-  // v{{[0-9]+}}.16b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x8x4_t test_vld4q_dup_p16(poly16_t const *a) {
+poly16x8x4_t test_vld4q_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_p16
   return vld4q_dup_p16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8h, v{{[0-9]+}}.8h, v{{[0-9]+}}.8h,
-  // v{{[0-9]+}}.8h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h, v[0-9]+.8h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x2x4_t test_vld4q_dup_p64(poly64_t const *a) {
+poly64x2x4_t test_vld4q_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld4q_dup_p64
   return vld4q_dup_p64(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2d, v{{[0-9]+}}.2d, v{{[0-9]+}}.2d,
-  // v{{[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x8x4_t test_vld4_dup_u8(uint8_t const *a) {
+uint8x8x4_t test_vld4_dup_u8(uint8_t  *a) {
   // CHECK-LABEL: test_vld4_dup_u8
   return vld4_dup_u8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x4x4_t test_vld4_dup_u16(uint16_t const *a) {
+uint16x4x4_t test_vld4_dup_u16(uint16_t  *a) {
   // CHECK-LABEL: test_vld4_dup_u16
   return vld4_dup_u16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint32x2x4_t test_vld4_dup_u32(uint32_t const *a) {
+uint32x2x4_t test_vld4_dup_u32(uint32_t  *a) {
   // CHECK-LABEL: test_vld4_dup_u32
   return vld4_dup_u32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint64x1x4_t test_vld4_dup_u64(uint64_t const *a) {
+uint64x1x4_t test_vld4_dup_u64(uint64_t  *a) {
   // CHECK-LABEL: test_vld4_dup_u64
   return vld4_dup_u64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-int8x8x4_t test_vld4_dup_s8(int8_t const *a) {
+int8x8x4_t test_vld4_dup_s8(int8_t  *a) {
   // CHECK-LABEL: test_vld4_dup_s8
   return vld4_dup_s8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-int16x4x4_t test_vld4_dup_s16(int16_t const *a) {
+int16x4x4_t test_vld4_dup_s16(int16_t  *a) {
   // CHECK-LABEL: test_vld4_dup_s16
   return vld4_dup_s16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-int32x2x4_t test_vld4_dup_s32(int32_t const *a) {
+int32x2x4_t test_vld4_dup_s32(int32_t  *a) {
   // CHECK-LABEL: test_vld4_dup_s32
   return vld4_dup_s32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-int64x1x4_t test_vld4_dup_s64(int64_t const *a) {
+int64x1x4_t test_vld4_dup_s64(int64_t  *a) {
   // CHECK-LABEL: test_vld4_dup_s64
   return vld4_dup_s64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-float16x4x4_t test_vld4_dup_f16(float16_t const *a) {
+float16x4x4_t test_vld4_dup_f16(float16_t  *a) {
   // CHECK-LABEL: test_vld4_dup_f16
   return vld4_dup_f16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-float32x2x4_t test_vld4_dup_f32(float32_t const *a) {
+float32x2x4_t test_vld4_dup_f32(float32_t  *a) {
   // CHECK-LABEL: test_vld4_dup_f32
   return vld4_dup_f32(a);
-  // CHECK: ld4r {v{{[0-9]+}}.2s, v{{[0-9]+}}.2s, v{{[0-9]+}}.2s,
-  // v{{[0-9]+}}.2s}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s, v[0-9]+.2s *}}}, [{{x[0-9]+|sp}}]
 }
 
-float64x1x4_t test_vld4_dup_f64(float64_t const *a) {
+float64x1x4_t test_vld4_dup_f64(float64_t  *a) {
   // CHECK-LABEL: test_vld4_dup_f64
   return vld4_dup_f64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x8x4_t test_vld4_dup_p8(poly8_t const *a) {
+poly8x8x4_t test_vld4_dup_p8(poly8_t  *a) {
   // CHECK-LABEL: test_vld4_dup_p8
   return vld4_dup_p8(a);
-  // CHECK: ld4r {v{{[0-9]+}}.8b, v{{[0-9]+}}.8b, v{{[0-9]+}}.8b,
-  // v{{[0-9]+}}.8b}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b, v[0-9]+.8b *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly16x4x4_t test_vld4_dup_p16(poly16_t const *a) {
+poly16x4x4_t test_vld4_dup_p16(poly16_t  *a) {
   // CHECK-LABEL: test_vld4_dup_p16
   return vld4_dup_p16(a);
-  // CHECK: ld4r {v{{[0-9]+}}.4h, v{{[0-9]+}}.4h, v{{[0-9]+}}.4h,
-  // v{{[0-9]+}}.4h}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4r {{{ *v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h, v[0-9]+.4h *}}}, [{{x[0-9]+|sp}}]
 }
 
-poly64x1x4_t test_vld4_dup_p64(poly64_t const *a) {
+poly64x1x4_t test_vld4_dup_p64(poly64_t  *a) {
   // CHECK-LABEL: test_vld4_dup_p64
   return vld4_dup_p64(a);
-  // CHECK: ld1 {v{{[0-9]+}}.1d, v{{[0-9]+}}.1d, v{{[0-9]+}}.1d,
-  // v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1|ld4r}} {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
-uint8x16_t test_vld1q_lane_u8(uint8_t const *a, uint8x16_t b) {
+uint8x16_t test_vld1q_lane_u8(uint8_t  *a, uint8x16_t b) {
   // CHECK-LABEL: test_vld1q_lane_u8
   return vld1q_lane_u8(a, b, 15);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-uint16x8_t test_vld1q_lane_u16(uint16_t const *a, uint16x8_t b) {
+uint16x8_t test_vld1q_lane_u16(uint16_t  *a, uint16x8_t b) {
   // CHECK-LABEL: test_vld1q_lane_u16
   return vld1q_lane_u16(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint32x4_t test_vld1q_lane_u32(uint32_t const *a, uint32x4_t b) {
+uint32x4_t test_vld1q_lane_u32(uint32_t  *a, uint32x4_t b) {
   // CHECK-LABEL: test_vld1q_lane_u32
   return vld1q_lane_u32(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint64x2_t test_vld1q_lane_u64(uint64_t const *a, uint64x2_t b) {
+uint64x2_t test_vld1q_lane_u64(uint64_t  *a, uint64x2_t b) {
   // CHECK-LABEL: test_vld1q_lane_u64
   return vld1q_lane_u64(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int8x16_t test_vld1q_lane_s8(int8_t const *a, int8x16_t b) {
+int8x16_t test_vld1q_lane_s8(int8_t  *a, int8x16_t b) {
   // CHECK-LABEL: test_vld1q_lane_s8
   return vld1q_lane_s8(a, b, 15);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-int16x8_t test_vld1q_lane_s16(int16_t const *a, int16x8_t b) {
+int16x8_t test_vld1q_lane_s16(int16_t  *a, int16x8_t b) {
   // CHECK-LABEL: test_vld1q_lane_s16
   return vld1q_lane_s16(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int32x4_t test_vld1q_lane_s32(int32_t const *a, int32x4_t b) {
+int32x4_t test_vld1q_lane_s32(int32_t  *a, int32x4_t b) {
   // CHECK-LABEL: test_vld1q_lane_s32
   return vld1q_lane_s32(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int64x2_t test_vld1q_lane_s64(int64_t const *a, int64x2_t b) {
+int64x2_t test_vld1q_lane_s64(int64_t  *a, int64x2_t b) {
   // CHECK-LABEL: test_vld1q_lane_s64
   return vld1q_lane_s64(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float16x8_t test_vld1q_lane_f16(float16_t const *a, float16x8_t b) {
+float16x8_t test_vld1q_lane_f16(float16_t  *a, float16x8_t b) {
   // CHECK-LABEL: test_vld1q_lane_f16
   return vld1q_lane_f16(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-float32x4_t test_vld1q_lane_f32(float32_t const *a, float32x4_t b) {
+float32x4_t test_vld1q_lane_f32(float32_t  *a, float32x4_t b) {
   // CHECK-LABEL: test_vld1q_lane_f32
   return vld1q_lane_f32(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float64x2_t test_vld1q_lane_f64(float64_t const *a, float64x2_t b) {
+float64x2_t test_vld1q_lane_f64(float64_t  *a, float64x2_t b) {
   // CHECK-LABEL: test_vld1q_lane_f64
   return vld1q_lane_f64(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-poly8x16_t test_vld1q_lane_p8(poly8_t const *a, poly8x16_t b) {
+poly8x16_t test_vld1q_lane_p8(poly8_t  *a, poly8x16_t b) {
   // CHECK-LABEL: test_vld1q_lane_p8
   return vld1q_lane_p8(a, b, 15);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-poly16x8_t test_vld1q_lane_p16(poly16_t const *a, poly16x8_t b) {
+poly16x8_t test_vld1q_lane_p16(poly16_t  *a, poly16x8_t b) {
   // CHECK-LABEL: test_vld1q_lane_p16
   return vld1q_lane_p16(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly64x2_t test_vld1q_lane_p64(poly64_t const *a, poly64x2_t b) {
+poly64x2_t test_vld1q_lane_p64(poly64_t  *a, poly64x2_t b) {
   // CHECK-LABEL: test_vld1q_lane_p64
   return vld1q_lane_p64(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint8x8_t test_vld1_lane_u8(uint8_t const *a, uint8x8_t b) {
+uint8x8_t test_vld1_lane_u8(uint8_t  *a, uint8x8_t b) {
   // CHECK-LABEL: test_vld1_lane_u8
   return vld1_lane_u8(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint16x4_t test_vld1_lane_u16(uint16_t const *a, uint16x4_t b) {
+uint16x4_t test_vld1_lane_u16(uint16_t  *a, uint16x4_t b) {
   // CHECK-LABEL: test_vld1_lane_u16
   return vld1_lane_u16(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint32x2_t test_vld1_lane_u32(uint32_t const *a, uint32x2_t b) {
+uint32x2_t test_vld1_lane_u32(uint32_t  *a, uint32x2_t b) {
   // CHECK-LABEL: test_vld1_lane_u32
   return vld1_lane_u32(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint64x1_t test_vld1_lane_u64(uint64_t const *a, uint64x1_t b) {
+uint64x1_t test_vld1_lane_u64(uint64_t  *a, uint64x1_t b) {
   // CHECK-LABEL: test_vld1_lane_u64
   return vld1_lane_u64(a, b, 0);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-int8x8_t test_vld1_lane_s8(int8_t const *a, int8x8_t b) {
+int8x8_t test_vld1_lane_s8(int8_t  *a, int8x8_t b) {
   // CHECK-LABEL: test_vld1_lane_s8
   return vld1_lane_s8(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int16x4_t test_vld1_lane_s16(int16_t const *a, int16x4_t b) {
+int16x4_t test_vld1_lane_s16(int16_t  *a, int16x4_t b) {
   // CHECK-LABEL: test_vld1_lane_s16
   return vld1_lane_s16(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int32x2_t test_vld1_lane_s32(int32_t const *a, int32x2_t b) {
+int32x2_t test_vld1_lane_s32(int32_t  *a, int32x2_t b) {
   // CHECK-LABEL: test_vld1_lane_s32
   return vld1_lane_s32(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int64x1_t test_vld1_lane_s64(int64_t const *a, int64x1_t b) {
+int64x1_t test_vld1_lane_s64(int64_t  *a, int64x1_t b) {
   // CHECK-LABEL: test_vld1_lane_s64
   return vld1_lane_s64(a, b, 0);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-float16x4_t test_vld1_lane_f16(float16_t const *a, float16x4_t b) {
+float16x4_t test_vld1_lane_f16(float16_t  *a, float16x4_t b) {
   // CHECK-LABEL: test_vld1_lane_f16
   return vld1_lane_f16(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float32x2_t test_vld1_lane_f32(float32_t const *a, float32x2_t b) {
+float32x2_t test_vld1_lane_f32(float32_t  *a, float32x2_t b) {
   // CHECK-LABEL: test_vld1_lane_f32
   return vld1_lane_f32(a, b, 1);
-  // CHECK: ld1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float64x1_t test_vld1_lane_f64(float64_t const *a, float64x1_t b) {
+float64x1_t test_vld1_lane_f64(float64_t  *a, float64x1_t b) {
   // CHECK-LABEL: test_vld1_lane_f64
   return vld1_lane_f64(a, b, 0);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-poly8x8_t test_vld1_lane_p8(poly8_t const *a, poly8x8_t b) {
+poly8x8_t test_vld1_lane_p8(poly8_t  *a, poly8x8_t b) {
   // CHECK-LABEL: test_vld1_lane_p8
   return vld1_lane_p8(a, b, 7);
-  // CHECK: ld1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly16x4_t test_vld1_lane_p16(poly16_t const *a, poly16x4_t b) {
+poly16x4_t test_vld1_lane_p16(poly16_t  *a, poly16x4_t b) {
   // CHECK-LABEL: test_vld1_lane_p16
   return vld1_lane_p16(a, b, 3);
-  // CHECK: ld1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-poly64x1_t test_vld1_lane_p64(poly64_t const *a, poly64x1_t b) {
+poly64x1_t test_vld1_lane_p64(poly64_t  *a, poly64x1_t b) {
   // CHECK-LABEL: test_vld1_lane_p64
   return vld1_lane_p64(a, b, 0);
-  // CHECK: ld1r {v{{[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: {{ld1r {v[0-9]+.1d}|ldr d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-uint16x8x2_t test_vld2q_lane_u16(uint16_t const *a, uint16x8x2_t b) {
+int8x16x2_t test_vld2q_lane_s8(int8_t const * ptr, int8x16x2_t src) {
+  // CHECK-LABEL: test_vld2q_lane_s8
+  return vld2q_lane_s8(ptr, src, 15);
+  // CHECK: ld2 {{{ *v[0-9]+.b,  v[0-9]+.b *}}}[15], [x0]
+}
+
+uint8x16x2_t test_vld2q_lane_u8(uint8_t const * ptr, uint8x16x2_t src) {
+  // CHECK-LABEL: test_vld2q_lane_u8
+  return vld2q_lane_u8(ptr, src, 15);
+  // CHECK: ld2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[15], [x0]
+}
+
+poly8x16x2_t test_vld2q_lane_p8(poly8_t const * ptr, poly8x16x2_t src) {
+  // CHECK-LABEL: test_vld2q_lane_p8
+  return vld2q_lane_p8(ptr, src, 15);
+  // CHECK: ld2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[15], [x0]
+}
+
+int8x16x3_t test_vld3q_lane_s8(int8_t const * ptr, int8x16x3_t src) {
+  // CHECK-LABEL: test_vld3q_lane_s8
+  return vld3q_lane_s8(ptr, src, 15);
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [x0]
+}
+
+uint8x16x3_t test_vld3q_lane_u8(uint8_t const * ptr, uint8x16x3_t src) {
+  // CHECK-LABEL: test_vld3q_lane_u8
+  return vld3q_lane_u8(ptr, src, 15);
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [x0]
+}
+
+uint16x8x2_t test_vld2q_lane_u16(uint16_t  *a, uint16x8x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_u16
   return vld2q_lane_u16(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint32x4x2_t test_vld2q_lane_u32(uint32_t const *a, uint32x4x2_t b) {
+uint32x4x2_t test_vld2q_lane_u32(uint32_t  *a, uint32x4x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_u32
   return vld2q_lane_u32(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint64x2x2_t test_vld2q_lane_u64(uint64_t const *a, uint64x2x2_t b) {
+uint64x2x2_t test_vld2q_lane_u64(uint64_t  *a, uint64x2x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_u64
   return vld2q_lane_u64(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int16x8x2_t test_vld2q_lane_s16(int16_t const *a, int16x8x2_t b) {
+int16x8x2_t test_vld2q_lane_s16(int16_t  *a, int16x8x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_s16
   return vld2q_lane_s16(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int32x4x2_t test_vld2q_lane_s32(int32_t const *a, int32x4x2_t b) {
+int32x4x2_t test_vld2q_lane_s32(int32_t  *a, int32x4x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_s32
   return vld2q_lane_s32(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int64x2x2_t test_vld2q_lane_s64(int64_t const *a, int64x2x2_t b) {
+int64x2x2_t test_vld2q_lane_s64(int64_t  *a, int64x2x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_s64
   return vld2q_lane_s64(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float16x8x2_t test_vld2q_lane_f16(float16_t const *a, float16x8x2_t b) {
+float16x8x2_t test_vld2q_lane_f16(float16_t  *a, float16x8x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_f16
   return vld2q_lane_f16(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-float32x4x2_t test_vld2q_lane_f32(float32_t const *a, float32x4x2_t b) {
+float32x4x2_t test_vld2q_lane_f32(float32_t  *a, float32x4x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_f32
   return vld2q_lane_f32(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float64x2x2_t test_vld2q_lane_f64(float64_t const *a, float64x2x2_t b) {
+float64x2x2_t test_vld2q_lane_f64(float64_t  *a, float64x2x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_f64
   return vld2q_lane_f64(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-poly16x8x2_t test_vld2q_lane_p16(poly16_t const *a, poly16x8x2_t b) {
+poly16x8x2_t test_vld2q_lane_p16(poly16_t  *a, poly16x8x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_p16
   return vld2q_lane_p16(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly64x2x2_t test_vld2q_lane_p64(poly64_t const *a, poly64x2x2_t b) {
+poly64x2x2_t test_vld2q_lane_p64(poly64_t  *a, poly64x2x2_t b) {
   // CHECK-LABEL: test_vld2q_lane_p64
   return vld2q_lane_p64(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint8x8x2_t test_vld2_lane_u8(uint8_t const *a, uint8x8x2_t b) {
+uint8x8x2_t test_vld2_lane_u8(uint8_t  *a, uint8x8x2_t b) {
   // CHECK-LABEL: test_vld2_lane_u8
   return vld2_lane_u8(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint16x4x2_t test_vld2_lane_u16(uint16_t const *a, uint16x4x2_t b) {
+uint16x4x2_t test_vld2_lane_u16(uint16_t  *a, uint16x4x2_t b) {
   // CHECK-LABEL: test_vld2_lane_u16
   return vld2_lane_u16(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint32x2x2_t test_vld2_lane_u32(uint32_t const *a, uint32x2x2_t b) {
+uint32x2x2_t test_vld2_lane_u32(uint32_t  *a, uint32x2x2_t b) {
   // CHECK-LABEL: test_vld2_lane_u32
   return vld2_lane_u32(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint64x1x2_t test_vld2_lane_u64(uint64_t const *a, uint64x1x2_t b) {
+uint64x1x2_t test_vld2_lane_u64(uint64_t  *a, uint64x1x2_t b) {
   // CHECK-LABEL: test_vld2_lane_u64
   return vld2_lane_u64(a, b, 0);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-int8x8x2_t test_vld2_lane_s8(int8_t const *a, int8x8x2_t b) {
+int8x8x2_t test_vld2_lane_s8(int8_t  *a, int8x8x2_t b) {
   // CHECK-LABEL: test_vld2_lane_s8
   return vld2_lane_s8(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int16x4x2_t test_vld2_lane_s16(int16_t const *a, int16x4x2_t b) {
+int16x4x2_t test_vld2_lane_s16(int16_t  *a, int16x4x2_t b) {
   // CHECK-LABEL: test_vld2_lane_s16
   return vld2_lane_s16(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int32x2x2_t test_vld2_lane_s32(int32_t const *a, int32x2x2_t b) {
+int32x2x2_t test_vld2_lane_s32(int32_t  *a, int32x2x2_t b) {
   // CHECK-LABEL: test_vld2_lane_s32
   return vld2_lane_s32(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int64x1x2_t test_vld2_lane_s64(int64_t const *a, int64x1x2_t b) {
+int64x1x2_t test_vld2_lane_s64(int64_t  *a, int64x1x2_t b) {
   // CHECK-LABEL: test_vld2_lane_s64
   return vld2_lane_s64(a, b, 0);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-float16x4x2_t test_vld2_lane_f16(float16_t const *a, float16x4x2_t b) {
+float16x4x2_t test_vld2_lane_f16(float16_t  *a, float16x4x2_t b) {
   // CHECK-LABEL: test_vld2_lane_f16
   return vld2_lane_f16(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float32x2x2_t test_vld2_lane_f32(float32_t const *a, float32x2x2_t b) {
+float32x2x2_t test_vld2_lane_f32(float32_t  *a, float32x2x2_t b) {
   // CHECK-LABEL: test_vld2_lane_f32
   return vld2_lane_f32(a, b, 1);
-  // CHECK: ld2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float64x1x2_t test_vld2_lane_f64(float64_t const *a, float64x1x2_t b) {
+float64x1x2_t test_vld2_lane_f64(float64_t  *a, float64x1x2_t b) {
   // CHECK-LABEL: test_vld2_lane_f64
   return vld2_lane_f64(a, b, 0);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-poly8x8x2_t test_vld2_lane_p8(poly8_t const *a, poly8x8x2_t b) {
+poly8x8x2_t test_vld2_lane_p8(poly8_t  *a, poly8x8x2_t b) {
   // CHECK-LABEL: test_vld2_lane_p8
   return vld2_lane_p8(a, b, 7);
-  // CHECK: ld2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly16x4x2_t test_vld2_lane_p16(poly16_t const *a, poly16x4x2_t b) {
+poly16x4x2_t test_vld2_lane_p16(poly16_t  *a, poly16x4x2_t b) {
   // CHECK-LABEL: test_vld2_lane_p16
   return vld2_lane_p16(a, b, 3);
-  // CHECK: ld2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-poly64x1x2_t test_vld2_lane_p64(poly64_t const *a, poly64x1x2_t b) {
+poly64x1x2_t test_vld2_lane_p64(poly64_t  *a, poly64x1x2_t b) {
   // CHECK-LABEL: test_vld2_lane_p64
   return vld2_lane_p64(a, b, 0);
-  // CHECK: ld2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-uint16x8x3_t test_vld3q_lane_u16(uint16_t const *a, uint16x8x3_t b) {
+uint16x8x3_t test_vld3q_lane_u16(uint16_t  *a, uint16x8x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_u16
   return vld3q_lane_u16(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint32x4x3_t test_vld3q_lane_u32(uint32_t const *a, uint32x4x3_t b) {
+uint32x4x3_t test_vld3q_lane_u32(uint32_t  *a, uint32x4x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_u32
   return vld3q_lane_u32(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint64x2x3_t test_vld3q_lane_u64(uint64_t const *a, uint64x2x3_t b) {
+uint64x2x3_t test_vld3q_lane_u64(uint64_t  *a, uint64x2x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_u64
   return vld3q_lane_u64(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int16x8x3_t test_vld3q_lane_s16(int16_t const *a, int16x8x3_t b) {
+int16x8x3_t test_vld3q_lane_s16(int16_t  *a, int16x8x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_s16
   return vld3q_lane_s16(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int32x4x3_t test_vld3q_lane_s32(int32_t const *a, int32x4x3_t b) {
+int32x4x3_t test_vld3q_lane_s32(int32_t  *a, int32x4x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_s32
   return vld3q_lane_s32(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int64x2x3_t test_vld3q_lane_s64(int64_t const *a, int64x2x3_t b) {
+int64x2x3_t test_vld3q_lane_s64(int64_t  *a, int64x2x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_s64
   return vld3q_lane_s64(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float16x8x3_t test_vld3q_lane_f16(float16_t const *a, float16x8x3_t b) {
+float16x8x3_t test_vld3q_lane_f16(float16_t  *a, float16x8x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_f16
   return vld3q_lane_f16(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-float32x4x3_t test_vld3q_lane_f32(float32_t const *a, float32x4x3_t b) {
+float32x4x3_t test_vld3q_lane_f32(float32_t  *a, float32x4x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_f32
   return vld3q_lane_f32(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float64x2x3_t test_vld3q_lane_f64(float64_t const *a, float64x2x3_t b) {
+float64x2x3_t test_vld3q_lane_f64(float64_t  *a, float64x2x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_f64
   return vld3q_lane_f64(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-poly8x16x3_t test_vld3q_lane_p8(poly8_t const *a, poly8x16x3_t b) {
+poly8x16x3_t test_vld3q_lane_p8(poly8_t  *a, poly8x16x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_p8
   return vld3q_lane_p8(a, b, 15);
-  // CHECK: ld3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-poly16x8x3_t test_vld3q_lane_p16(poly16_t const *a, poly16x8x3_t b) {
+poly16x8x3_t test_vld3q_lane_p16(poly16_t  *a, poly16x8x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_p16
   return vld3q_lane_p16(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly64x2x3_t test_vld3q_lane_p64(poly64_t const *a, poly64x2x3_t b) {
+poly64x2x3_t test_vld3q_lane_p64(poly64_t  *a, poly64x2x3_t b) {
   // CHECK-LABEL: test_vld3q_lane_p64
   return vld3q_lane_p64(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint8x8x3_t test_vld3_lane_u8(uint8_t const *a, uint8x8x3_t b) {
+uint8x8x3_t test_vld3_lane_u8(uint8_t  *a, uint8x8x3_t b) {
   // CHECK-LABEL: test_vld3_lane_u8
   return vld3_lane_u8(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint16x4x3_t test_vld3_lane_u16(uint16_t const *a, uint16x4x3_t b) {
+uint16x4x3_t test_vld3_lane_u16(uint16_t  *a, uint16x4x3_t b) {
   // CHECK-LABEL: test_vld3_lane_u16
   return vld3_lane_u16(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint32x2x3_t test_vld3_lane_u32(uint32_t const *a, uint32x2x3_t b) {
+uint32x2x3_t test_vld3_lane_u32(uint32_t  *a, uint32x2x3_t b) {
   // CHECK-LABEL: test_vld3_lane_u32
   return vld3_lane_u32(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint64x1x3_t test_vld3_lane_u64(uint64_t const *a, uint64x1x3_t b) {
+uint64x1x3_t test_vld3_lane_u64(uint64_t  *a, uint64x1x3_t b) {
   // CHECK-LABEL: test_vld3_lane_u64
   return vld3_lane_u64(a, b, 0);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-int8x8x3_t test_vld3_lane_s8(int8_t const *a, int8x8x3_t b) {
+int8x8x3_t test_vld3_lane_s8(int8_t  *a, int8x8x3_t b) {
   // CHECK-LABEL: test_vld3_lane_s8
   return vld3_lane_s8(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int16x4x3_t test_vld3_lane_s16(int16_t const *a, int16x4x3_t b) {
+int16x4x3_t test_vld3_lane_s16(int16_t  *a, int16x4x3_t b) {
   // CHECK-LABEL: test_vld3_lane_s16
   return vld3_lane_s16(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int32x2x3_t test_vld3_lane_s32(int32_t const *a, int32x2x3_t b) {
+int32x2x3_t test_vld3_lane_s32(int32_t  *a, int32x2x3_t b) {
   // CHECK-LABEL: test_vld3_lane_s32
   return vld3_lane_s32(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int64x1x3_t test_vld3_lane_s64(int64_t const *a, int64x1x3_t b) {
+int64x1x3_t test_vld3_lane_s64(int64_t  *a, int64x1x3_t b) {
   // CHECK-LABEL: test_vld3_lane_s64
   return vld3_lane_s64(a, b, 0);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-float16x4x3_t test_vld3_lane_f16(float16_t const *a, float16x4x3_t b) {
+float16x4x3_t test_vld3_lane_f16(float16_t  *a, float16x4x3_t b) {
   // CHECK-LABEL: test_vld3_lane_f16
   return vld3_lane_f16(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float32x2x3_t test_vld3_lane_f32(float32_t const *a, float32x2x3_t b) {
+float32x2x3_t test_vld3_lane_f32(float32_t  *a, float32x2x3_t b) {
   // CHECK-LABEL: test_vld3_lane_f32
   return vld3_lane_f32(a, b, 1);
-  // CHECK: ld3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float64x1x3_t test_vld3_lane_f64(float64_t const *a, float64x1x3_t b) {
+float64x1x3_t test_vld3_lane_f64(float64_t  *a, float64x1x3_t b) {
   // CHECK-LABEL: test_vld3_lane_f64
   return vld3_lane_f64(a, b, 0);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-poly8x8x3_t test_vld3_lane_p8(poly8_t const *a, poly8x8x3_t b) {
+poly8x8x3_t test_vld3_lane_p8(poly8_t  *a, poly8x8x3_t b) {
   // CHECK-LABEL: test_vld3_lane_p8
   return vld3_lane_p8(a, b, 7);
-  // CHECK: ld3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly16x4x3_t test_vld3_lane_p16(poly16_t const *a, poly16x4x3_t b) {
+poly16x4x3_t test_vld3_lane_p16(poly16_t  *a, poly16x4x3_t b) {
   // CHECK-LABEL: test_vld3_lane_p16
   return vld3_lane_p16(a, b, 3);
-  // CHECK: ld3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-poly64x1x3_t test_vld3_lane_p64(poly64_t const *a, poly64x1x3_t b) {
+poly64x1x3_t test_vld3_lane_p64(poly64_t  *a, poly64x1x3_t b) {
   // CHECK-LABEL: test_vld3_lane_p64
   return vld3_lane_p64(a, b, 0);
-  // CHECK: ld3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-uint8x16x4_t test_vld4q_lane_u8(uint8_t const *a, uint8x16x4_t b) {
+uint8x16x4_t test_vld4q_lane_u8(uint8_t  *a, uint8x16x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_u8
   return vld4q_lane_u8(a, b, 15);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-uint16x8x4_t test_vld4q_lane_u16(uint16_t const *a, uint16x8x4_t b) {
+uint16x8x4_t test_vld4q_lane_u16(uint16_t  *a, uint16x8x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_u16
   return vld4q_lane_u16(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint32x4x4_t test_vld4q_lane_u32(uint32_t const *a, uint32x4x4_t b) {
+uint32x4x4_t test_vld4q_lane_u32(uint32_t  *a, uint32x4x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_u32
   return vld4q_lane_u32(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint64x2x4_t test_vld4q_lane_u64(uint64_t const *a, uint64x2x4_t b) {
+uint64x2x4_t test_vld4q_lane_u64(uint64_t  *a, uint64x2x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_u64
   return vld4q_lane_u64(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int8x16x4_t test_vld4q_lane_s8(int8_t const *a, int8x16x4_t b) {
+int8x16x4_t test_vld4q_lane_s8(int8_t  *a, int8x16x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_s8
   return vld4q_lane_s8(a, b, 15);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-int16x8x4_t test_vld4q_lane_s16(int16_t const *a, int16x8x4_t b) {
+int16x8x4_t test_vld4q_lane_s16(int16_t  *a, int16x8x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_s16
   return vld4q_lane_s16(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int32x4x4_t test_vld4q_lane_s32(int32_t const *a, int32x4x4_t b) {
+int32x4x4_t test_vld4q_lane_s32(int32_t  *a, int32x4x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_s32
   return vld4q_lane_s32(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int64x2x4_t test_vld4q_lane_s64(int64_t const *a, int64x2x4_t b) {
+int64x2x4_t test_vld4q_lane_s64(int64_t  *a, int64x2x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_s64
   return vld4q_lane_s64(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float16x8x4_t test_vld4q_lane_f16(float16_t const *a, float16x8x4_t b) {
+float16x8x4_t test_vld4q_lane_f16(float16_t  *a, float16x8x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_f16
   return vld4q_lane_f16(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-float32x4x4_t test_vld4q_lane_f32(float32_t const *a, float32x4x4_t b) {
+float32x4x4_t test_vld4q_lane_f32(float32_t  *a, float32x4x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_f32
   return vld4q_lane_f32(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float64x2x4_t test_vld4q_lane_f64(float64_t const *a, float64x2x4_t b) {
+float64x2x4_t test_vld4q_lane_f64(float64_t  *a, float64x2x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_f64
   return vld4q_lane_f64(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-poly8x16x4_t test_vld4q_lane_p8(poly8_t const *a, poly8x16x4_t b) {
+poly8x16x4_t test_vld4q_lane_p8(poly8_t  *a, poly8x16x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_p8
   return vld4q_lane_p8(a, b, 15);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-poly16x8x4_t test_vld4q_lane_p16(poly16_t const *a, poly16x8x4_t b) {
+poly16x8x4_t test_vld4q_lane_p16(poly16_t  *a, poly16x8x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_p16
   return vld4q_lane_p16(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly64x2x4_t test_vld4q_lane_p64(poly64_t const *a, poly64x2x4_t b) {
+poly64x2x4_t test_vld4q_lane_p64(poly64_t  *a, poly64x2x4_t b) {
   // CHECK-LABEL: test_vld4q_lane_p64
   return vld4q_lane_p64(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint8x8x4_t test_vld4_lane_u8(uint8_t const *a, uint8x8x4_t b) {
+uint8x8x4_t test_vld4_lane_u8(uint8_t  *a, uint8x8x4_t b) {
   // CHECK-LABEL: test_vld4_lane_u8
   return vld4_lane_u8(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-uint16x4x4_t test_vld4_lane_u16(uint16_t const *a, uint16x4x4_t b) {
+uint16x4x4_t test_vld4_lane_u16(uint16_t  *a, uint16x4x4_t b) {
   // CHECK-LABEL: test_vld4_lane_u16
   return vld4_lane_u16(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-uint32x2x4_t test_vld4_lane_u32(uint32_t const *a, uint32x2x4_t b) {
+uint32x2x4_t test_vld4_lane_u32(uint32_t  *a, uint32x2x4_t b) {
   // CHECK-LABEL: test_vld4_lane_u32
   return vld4_lane_u32(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-uint64x1x4_t test_vld4_lane_u64(uint64_t const *a, uint64x1x4_t b) {
+uint64x1x4_t test_vld4_lane_u64(uint64_t  *a, uint64x1x4_t b) {
   // CHECK-LABEL: test_vld4_lane_u64
   return vld4_lane_u64(a, b, 0);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-int8x8x4_t test_vld4_lane_s8(int8_t const *a, int8x8x4_t b) {
+int8x8x4_t test_vld4_lane_s8(int8_t  *a, int8x8x4_t b) {
   // CHECK-LABEL: test_vld4_lane_s8
   return vld4_lane_s8(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-int16x4x4_t test_vld4_lane_s16(int16_t const *a, int16x4x4_t b) {
+int16x4x4_t test_vld4_lane_s16(int16_t  *a, int16x4x4_t b) {
   // CHECK-LABEL: test_vld4_lane_s16
   return vld4_lane_s16(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-int32x2x4_t test_vld4_lane_s32(int32_t const *a, int32x2x4_t b) {
+int32x2x4_t test_vld4_lane_s32(int32_t  *a, int32x2x4_t b) {
   // CHECK-LABEL: test_vld4_lane_s32
   return vld4_lane_s32(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-int64x1x4_t test_vld4_lane_s64(int64_t const *a, int64x1x4_t b) {
+int64x1x4_t test_vld4_lane_s64(int64_t  *a, int64x1x4_t b) {
   // CHECK-LABEL: test_vld4_lane_s64
   return vld4_lane_s64(a, b, 0);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-float16x4x4_t test_vld4_lane_f16(float16_t const *a, float16x4x4_t b) {
+float16x4x4_t test_vld4_lane_f16(float16_t  *a, float16x4x4_t b) {
   // CHECK-LABEL: test_vld4_lane_f16
   return vld4_lane_f16(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-float32x2x4_t test_vld4_lane_f32(float32_t const *a, float32x2x4_t b) {
+float32x2x4_t test_vld4_lane_f32(float32_t  *a, float32x2x4_t b) {
   // CHECK-LABEL: test_vld4_lane_f32
   return vld4_lane_f32(a, b, 1);
-  // CHECK: ld4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-float64x1x4_t test_vld4_lane_f64(float64_t const *a, float64x1x4_t b) {
+float64x1x4_t test_vld4_lane_f64(float64_t  *a, float64x1x4_t b) {
   // CHECK-LABEL: test_vld4_lane_f64
   return vld4_lane_f64(a, b, 0);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-poly8x8x4_t test_vld4_lane_p8(poly8_t const *a, poly8x8x4_t b) {
+poly8x8x4_t test_vld4_lane_p8(poly8_t  *a, poly8x8x4_t b) {
   // CHECK-LABEL: test_vld4_lane_p8
   return vld4_lane_p8(a, b, 7);
-  // CHECK: ld4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-poly16x4x4_t test_vld4_lane_p16(poly16_t const *a, poly16x4x4_t b) {
+poly16x4x4_t test_vld4_lane_p16(poly16_t  *a, poly16x4x4_t b) {
   // CHECK-LABEL: test_vld4_lane_p16
   return vld4_lane_p16(a, b, 3);
-  // CHECK: ld4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-poly64x1x4_t test_vld4_lane_p64(poly64_t const *a, poly64x1x4_t b) {
+poly64x1x4_t test_vld4_lane_p64(poly64_t  *a, poly64x1x4_t b) {
   // CHECK-LABEL: test_vld4_lane_p64
   return vld4_lane_p64(a, b, 0);
-  // CHECK: ld4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_u8(uint8_t const *a, uint8x16_t b) {
+void test_vst1q_lane_u8(uint8_t  *a, uint8x16_t b) {
   // CHECK-LABEL: test_vst1q_lane_u8
   vst1q_lane_u8(a, b, 15);
-  // CHECK: st1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_u16(uint16_t const *a, uint16x8_t b) {
+void test_vst1q_lane_u16(uint16_t  *a, uint16x8_t b) {
   // CHECK-LABEL: test_vst1q_lane_u16
   vst1q_lane_u16(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_u32(uint32_t const *a, uint32x4_t b) {
+void test_vst1q_lane_u32(uint32_t  *a, uint32x4_t b) {
   // CHECK-LABEL: test_vst1q_lane_u32
   vst1q_lane_u32(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_u64(uint64_t const *a, uint64x2_t b) {
+void test_vst1q_lane_u64(uint64_t  *a, uint64x2_t b) {
   // CHECK-LABEL: test_vst1q_lane_u64
   vst1q_lane_u64(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_s8(int8_t const *a, int8x16_t b) {
+void test_vst1q_lane_s8(int8_t  *a, int8x16_t b) {
   // CHECK-LABEL: test_vst1q_lane_s8
   vst1q_lane_s8(a, b, 15);
-  // CHECK: st1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_s16(int16_t const *a, int16x8_t b) {
+void test_vst1q_lane_s16(int16_t  *a, int16x8_t b) {
   // CHECK-LABEL: test_vst1q_lane_s16
   vst1q_lane_s16(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_s32(int32_t const *a, int32x4_t b) {
+void test_vst1q_lane_s32(int32_t  *a, int32x4_t b) {
   // CHECK-LABEL: test_vst1q_lane_s32
   vst1q_lane_s32(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_s64(int64_t const *a, int64x2_t b) {
+void test_vst1q_lane_s64(int64_t  *a, int64x2_t b) {
   // CHECK-LABEL: test_vst1q_lane_s64
   vst1q_lane_s64(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_f16(float16_t const *a, float16x8_t b) {
+void test_vst1q_lane_f16(float16_t  *a, float16x8_t b) {
   // CHECK-LABEL: test_vst1q_lane_f16
   vst1q_lane_f16(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_f32(float32_t const *a, float32x4_t b) {
+void test_vst1q_lane_f32(float32_t  *a, float32x4_t b) {
   // CHECK-LABEL: test_vst1q_lane_f32
   vst1q_lane_f32(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_f64(float64_t const *a, float64x2_t b) {
+void test_vst1q_lane_f64(float64_t  *a, float64x2_t b) {
   // CHECK-LABEL: test_vst1q_lane_f64
   vst1q_lane_f64(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-poly8x16_t test_vst1q_lane_p8(poly8_t const *a, poly8x16_t b) {
+void test_vst1q_lane_p8(poly8_t  *a, poly8x16_t b) {
   // CHECK-LABEL: test_vst1q_lane_p8
   vst1q_lane_p8(a, b, 15);
-  // CHECK: st1 {v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_p16(poly16_t const *a, poly16x8_t b) {
+void test_vst1q_lane_p16(poly16_t  *a, poly16x8_t b) {
   // CHECK-LABEL: test_vst1q_lane_p16
   vst1q_lane_p16(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1q_lane_p64(poly64_t const *a, poly64x2_t b) {
+void test_vst1q_lane_p64(poly64_t  *a, poly64x2_t b) {
   // CHECK-LABEL: test_vst1q_lane_p64
   vst1q_lane_p64(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_u8(uint8_t const *a, uint8x8_t b) {
+void test_vst1_lane_u8(uint8_t  *a, uint8x8_t b) {
   // CHECK-LABEL: test_vst1_lane_u8
   vst1_lane_u8(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_u16(uint16_t const *a, uint16x4_t b) {
+void test_vst1_lane_u16(uint16_t  *a, uint16x4_t b) {
   // CHECK-LABEL: test_vst1_lane_u16
   vst1_lane_u16(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_u32(uint32_t const *a, uint32x2_t b) {
+void test_vst1_lane_u32(uint32_t  *a, uint32x2_t b) {
   // CHECK-LABEL: test_vst1_lane_u32
   vst1_lane_u32(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_u64(uint64_t const *a, uint64x1_t b) {
+void test_vst1_lane_u64(uint64_t  *a, uint64x1_t b) {
   // CHECK-LABEL: test_vst1_lane_u64
   vst1_lane_u64(a, b, 0);
-  // CHECK: st1 {v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_s8(int8_t const *a, int8x8_t b) {
+void test_vst1_lane_s8(int8_t  *a, int8x8_t b) {
   // CHECK-LABEL: test_vst1_lane_s8
   vst1_lane_s8(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_s16(int16_t const *a, int16x4_t b) {
+void test_vst1_lane_s16(int16_t  *a, int16x4_t b) {
   // CHECK-LABEL: test_vst1_lane_s16
   vst1_lane_s16(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_s32(int32_t const *a, int32x2_t b) {
+void test_vst1_lane_s32(int32_t  *a, int32x2_t b) {
   // CHECK-LABEL: test_vst1_lane_s32
   vst1_lane_s32(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_s64(int64_t const *a, int64x1_t b) {
+void test_vst1_lane_s64(int64_t  *a, int64x1_t b) {
   // CHECK-LABEL: test_vst1_lane_s64
   vst1_lane_s64(a, b, 0);
-  // CHECK: st1 {v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_f16(float16_t const *a, float16x4_t b) {
+void test_vst1_lane_f16(float16_t  *a, float16x4_t b) {
   // CHECK-LABEL: test_vst1_lane_f16
   vst1_lane_f16(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_f32(float32_t const *a, float32x2_t b) {
+void test_vst1_lane_f32(float32_t  *a, float32x2_t b) {
   // CHECK-LABEL: test_vst1_lane_f32
   vst1_lane_f32(a, b, 1);
-  // CHECK: st1 {v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_f64(float64_t const *a, float64x1_t b) {
+void test_vst1_lane_f64(float64_t  *a, float64x1_t b) {
   // CHECK-LABEL: test_vst1_lane_f64
   vst1_lane_f64(a, b, 0);
-  // CHECK: st1 {v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: {{st1 {v[0-9]+.d}\[0]|str d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_p8(poly8_t const *a, poly8x8_t b) {
+void test_vst1_lane_p8(poly8_t  *a, poly8x8_t b) {
   // CHECK-LABEL: test_vst1_lane_p8
   vst1_lane_p8(a, b, 7);
-  // CHECK: st1 {v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_p16(poly16_t const *a, poly16x4_t b) {
+void test_vst1_lane_p16(poly16_t  *a, poly16x4_t b) {
   // CHECK-LABEL: test_vst1_lane_p16
   vst1_lane_p16(a, b, 3);
-  // CHECK: st1 {v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst1_lane_p64(poly64_t const *a, poly64x1_t b) {
+void test_vst1_lane_p64(poly64_t  *a, poly64x1_t b) {
   // CHECK-LABEL: test_vst1_lane_p64
   vst1_lane_p64(a, b, 0);
-  // CHECK: st1 {v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st1 {{{ *v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_u8(uint8_t const *a, uint8x16x2_t b) {
+void test_vst2q_lane_u8(uint8_t  *a, uint8x16x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_u8
   vst2q_lane_u8(a, b, 15);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_u16(uint16_t const *a, uint16x8x2_t b) {
+void test_vst2q_lane_u16(uint16_t  *a, uint16x8x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_u16
   vst2q_lane_u16(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_u32(uint32_t const *a, uint32x4x2_t b) {
+void test_vst2q_lane_u32(uint32_t  *a, uint32x4x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_u32
   vst2q_lane_u32(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_u64(uint64_t const *a, uint64x2x2_t b) {
+void test_vst2q_lane_u64(uint64_t  *a, uint64x2x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_u64
   vst2q_lane_u64(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_s8(int8_t const *a, int8x16x2_t b) {
+void test_vst2q_lane_s8(int8_t  *a, int8x16x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_s8
   vst2q_lane_s8(a, b, 15);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_s16(int16_t const *a, int16x8x2_t b) {
+void test_vst2q_lane_s16(int16_t  *a, int16x8x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_s16
   vst2q_lane_s16(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_s32(int32_t const *a, int32x4x2_t b) {
+void test_vst2q_lane_s32(int32_t  *a, int32x4x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_s32
   vst2q_lane_s32(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_s64(int64_t const *a, int64x2x2_t b) {
+void test_vst2q_lane_s64(int64_t  *a, int64x2x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_s64
   vst2q_lane_s64(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_f16(float16_t const *a, float16x8x2_t b) {
+void test_vst2q_lane_f16(float16_t  *a, float16x8x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_f16
   vst2q_lane_f16(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_f32(float32_t const *a, float32x4x2_t b) {
+void test_vst2q_lane_f32(float32_t  *a, float32x4x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_f32
   vst2q_lane_f32(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_f64(float64_t const *a, float64x2x2_t b) {
+void test_vst2q_lane_f64(float64_t  *a, float64x2x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_f64
   vst2q_lane_f64(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_p8(poly8_t const *a, poly8x16x2_t b) {
+void test_vst2q_lane_p8(poly8_t  *a, poly8x16x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_p8
   vst2q_lane_p8(a, b, 15);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_p16(poly16_t const *a, poly16x8x2_t b) {
+void test_vst2q_lane_p16(poly16_t  *a, poly16x8x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_p16
   vst2q_lane_p16(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2q_lane_p64(poly64_t const *a, poly64x2x2_t b) {
+void test_vst2q_lane_p64(poly64_t  *a, poly64x2x2_t b) {
   // CHECK-LABEL: test_vst2q_lane_p64
   vst2q_lane_p64(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_u8(uint8_t const *a, uint8x8x2_t b) {
+void test_vst2_lane_u8(uint8_t  *a, uint8x8x2_t b) {
   // CHECK-LABEL: test_vst2_lane_u8
   vst2_lane_u8(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_u16(uint16_t const *a, uint16x4x2_t b) {
+void test_vst2_lane_u16(uint16_t  *a, uint16x4x2_t b) {
   // CHECK-LABEL: test_vst2_lane_u16
   vst2_lane_u16(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_u32(uint32_t const *a, uint32x2x2_t b) {
+void test_vst2_lane_u32(uint32_t  *a, uint32x2x2_t b) {
   // CHECK-LABEL: test_vst2_lane_u32
   vst2_lane_u32(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_u64(uint64_t const *a, uint64x1x2_t b) {
+void test_vst2_lane_u64(uint64_t  *a, uint64x1x2_t b) {
   // CHECK-LABEL: test_vst2_lane_u64
   vst2_lane_u64(a, b, 0);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_s8(int8_t const *a, int8x8x2_t b) {
+void test_vst2_lane_s8(int8_t  *a, int8x8x2_t b) {
   // CHECK-LABEL: test_vst2_lane_s8
   vst2_lane_s8(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_s16(int16_t const *a, int16x4x2_t b) {
+void test_vst2_lane_s16(int16_t  *a, int16x4x2_t b) {
   // CHECK-LABEL: test_vst2_lane_s16
   vst2_lane_s16(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_s32(int32_t const *a, int32x2x2_t b) {
+void test_vst2_lane_s32(int32_t  *a, int32x2x2_t b) {
   // CHECK-LABEL: test_vst2_lane_s32
   vst2_lane_s32(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_s64(int64_t const *a, int64x1x2_t b) {
+void test_vst2_lane_s64(int64_t  *a, int64x1x2_t b) {
   // CHECK-LABEL: test_vst2_lane_s64
   vst2_lane_s64(a, b, 0);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_f16(float16_t const *a, float16x4x2_t b) {
+void test_vst2_lane_f16(float16_t  *a, float16x4x2_t b) {
   // CHECK-LABEL: test_vst2_lane_f16
   vst2_lane_f16(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_f32(float32_t const *a, float32x2x2_t b) {
+void test_vst2_lane_f32(float32_t  *a, float32x2x2_t b) {
   // CHECK-LABEL: test_vst2_lane_f32
   vst2_lane_f32(a, b, 1);
-  // CHECK: st2 {v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_f64(float64_t const *a, float64x1x2_t b) {
+void test_vst2_lane_f64(float64_t  *a, float64x1x2_t b) {
   // CHECK-LABEL: test_vst2_lane_f64
   vst2_lane_f64(a, b, 0);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_p8(poly8_t const *a, poly8x8x2_t b) {
+void test_vst2_lane_p8(poly8_t  *a, poly8x8x2_t b) {
   // CHECK-LABEL: test_vst2_lane_p8
   vst2_lane_p8(a, b, 7);
-  // CHECK: st2 {v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_p16(poly16_t const *a, poly16x4x2_t b) {
+void test_vst2_lane_p16(poly16_t  *a, poly16x4x2_t b) {
   // CHECK-LABEL: test_vst2_lane_p16
   vst2_lane_p16(a, b, 3);
-  // CHECK: st2 {v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst2_lane_p64(poly64_t const *a, poly64x1x2_t b) {
+void test_vst2_lane_p64(poly64_t  *a, poly64x1x2_t b) {
   // CHECK-LABEL: test_vst2_lane_p64
   vst2_lane_p64(a, b, 0);
-  // CHECK: st2 {v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st2 {{{ *v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_u8(uint8_t const *a, uint8x16x3_t b) {
+void test_vst3q_lane_u8(uint8_t  *a, uint8x16x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_u8
   vst3q_lane_u8(a, b, 15);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_u16(uint16_t const *a, uint16x8x3_t b) {
+void test_vst3q_lane_u16(uint16_t  *a, uint16x8x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_u16
   vst3q_lane_u16(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_u32(uint32_t const *a, uint32x4x3_t b) {
+void test_vst3q_lane_u32(uint32_t  *a, uint32x4x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_u32
   vst3q_lane_u32(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_u64(uint64_t const *a, uint64x2x3_t b) {
+void test_vst3q_lane_u64(uint64_t  *a, uint64x2x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_u64
   vst3q_lane_u64(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_s8(int8_t const *a, int8x16x3_t b) {
+void test_vst3q_lane_s8(int8_t  *a, int8x16x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_s8
   vst3q_lane_s8(a, b, 15);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_s16(int16_t const *a, int16x8x3_t b) {
+void test_vst3q_lane_s16(int16_t  *a, int16x8x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_s16
   vst3q_lane_s16(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_s32(int32_t const *a, int32x4x3_t b) {
+void test_vst3q_lane_s32(int32_t  *a, int32x4x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_s32
   vst3q_lane_s32(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_s64(int64_t const *a, int64x2x3_t b) {
+void test_vst3q_lane_s64(int64_t  *a, int64x2x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_s64
   vst3q_lane_s64(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_f16(float16_t const *a, float16x8x3_t b) {
+void test_vst3q_lane_f16(float16_t  *a, float16x8x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_f16
   vst3q_lane_f16(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_f32(float32_t const *a, float32x4x3_t b) {
+void test_vst3q_lane_f32(float32_t  *a, float32x4x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_f32
   vst3q_lane_f32(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_f64(float64_t const *a, float64x2x3_t b) {
+void test_vst3q_lane_f64(float64_t  *a, float64x2x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_f64
   vst3q_lane_f64(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_p8(poly8_t const *a, poly8x16x3_t b) {
+void test_vst3q_lane_p8(poly8_t  *a, poly8x16x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_p8
   vst3q_lane_p8(a, b, 15);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_p16(poly16_t const *a, poly16x8x3_t b) {
+void test_vst3q_lane_p16(poly16_t  *a, poly16x8x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_p16
   vst3q_lane_p16(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3q_lane_p64(poly64_t const *a, poly64x2x3_t b) {
+void test_vst3q_lane_p64(poly64_t  *a, poly64x2x3_t b) {
   // CHECK-LABEL: test_vst3q_lane_p64
   vst3q_lane_p64(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_u8(uint8_t const *a, uint8x8x3_t b) {
+void test_vst3_lane_u8(uint8_t  *a, uint8x8x3_t b) {
   // CHECK-LABEL: test_vst3_lane_u8
   vst3_lane_u8(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_u16(uint16_t const *a, uint16x4x3_t b) {
+void test_vst3_lane_u16(uint16_t  *a, uint16x4x3_t b) {
   // CHECK-LABEL: test_vst3_lane_u16
   vst3_lane_u16(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_u32(uint32_t const *a, uint32x2x3_t b) {
+void test_vst3_lane_u32(uint32_t  *a, uint32x2x3_t b) {
   // CHECK-LABEL: test_vst3_lane_u32
   vst3_lane_u32(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_u64(uint64_t const *a, uint64x1x3_t b) {
+void test_vst3_lane_u64(uint64_t  *a, uint64x1x3_t b) {
   // CHECK-LABEL: test_vst3_lane_u64
   vst3_lane_u64(a, b, 0);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_s8(int8_t const *a, int8x8x3_t b) {
+void test_vst3_lane_s8(int8_t  *a, int8x8x3_t b) {
   // CHECK-LABEL: test_vst3_lane_s8
   vst3_lane_s8(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_s16(int16_t const *a, int16x4x3_t b) {
+void test_vst3_lane_s16(int16_t  *a, int16x4x3_t b) {
   // CHECK-LABEL: test_vst3_lane_s16
   vst3_lane_s16(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_s32(int32_t const *a, int32x2x3_t b) {
+void test_vst3_lane_s32(int32_t  *a, int32x2x3_t b) {
   // CHECK-LABEL: test_vst3_lane_s32
   vst3_lane_s32(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_s64(int64_t const *a, int64x1x3_t b) {
+void test_vst3_lane_s64(int64_t  *a, int64x1x3_t b) {
   // CHECK-LABEL: test_vst3_lane_s64
   vst3_lane_s64(a, b, 0);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_f16(float16_t const *a, float16x4x3_t b) {
+void test_vst3_lane_f16(float16_t  *a, float16x4x3_t b) {
   // CHECK-LABEL: test_vst3_lane_f16
   vst3_lane_f16(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_f32(float32_t const *a, float32x2x3_t b) {
+void test_vst3_lane_f32(float32_t  *a, float32x2x3_t b) {
   // CHECK-LABEL: test_vst3_lane_f32
   vst3_lane_f32(a, b, 1);
-  // CHECK: st3 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_f64(float64_t const *a, float64x1x3_t b) {
+void test_vst3_lane_f64(float64_t  *a, float64x1x3_t b) {
   // CHECK-LABEL: test_vst3_lane_f64
   vst3_lane_f64(a, b, 0);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_p8(poly8_t const *a, poly8x8x3_t b) {
+void test_vst3_lane_p8(poly8_t  *a, poly8x8x3_t b) {
   // CHECK-LABEL: test_vst3_lane_p8
   vst3_lane_p8(a, b, 7);
-  // CHECK: st3 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_p16(poly16_t const *a, poly16x4x3_t b) {
+void test_vst3_lane_p16(poly16_t  *a, poly16x4x3_t b) {
   // CHECK-LABEL: test_vst3_lane_p16
   vst3_lane_p16(a, b, 3);
-  // CHECK: st3 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst3_lane_p64(poly64_t const *a, poly64x1x3_t b) {
+void test_vst3_lane_p64(poly64_t  *a, poly64x1x3_t b) {
   // CHECK-LABEL: test_vst3_lane_p64
   vst3_lane_p64(a, b, 0);
-  // CHECK: st3 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st3 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_u8(uint16_t const *a, uint8x16x4_t b) {
+void test_vst4q_lane_u8(uint8_t  *a, uint8x16x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_u8
   vst4q_lane_u8(a, b, 15);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_u16(uint16_t const *a, uint16x8x4_t b) {
+void test_vst4q_lane_u16(uint16_t  *a, uint16x8x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_u16
   vst4q_lane_u16(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_u32(uint32_t const *a, uint32x4x4_t b) {
+void test_vst4q_lane_u32(uint32_t  *a, uint32x4x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_u32
   vst4q_lane_u32(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_u64(uint64_t const *a, uint64x2x4_t b) {
+void test_vst4q_lane_u64(uint64_t  *a, uint64x2x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_u64
   vst4q_lane_u64(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_s8(int16_t const *a, int8x16x4_t b) {
+void test_vst4q_lane_s8(int8_t  *a, int8x16x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_s8
   vst4q_lane_s8(a, b, 15);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_s16(int16_t const *a, int16x8x4_t b) {
+void test_vst4q_lane_s16(int16_t  *a, int16x8x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_s16
   vst4q_lane_s16(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_s32(int32_t const *a, int32x4x4_t b) {
+void test_vst4q_lane_s32(int32_t  *a, int32x4x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_s32
   vst4q_lane_s32(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_s64(int64_t const *a, int64x2x4_t b) {
+void test_vst4q_lane_s64(int64_t  *a, int64x2x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_s64
   vst4q_lane_s64(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_f16(float16_t const *a, float16x8x4_t b) {
+void test_vst4q_lane_f16(float16_t  *a, float16x8x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_f16
   vst4q_lane_f16(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_f32(float32_t const *a, float32x4x4_t b) {
+void test_vst4q_lane_f32(float32_t  *a, float32x4x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_f32
   vst4q_lane_f32(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_f64(float64_t const *a, float64x2x4_t b) {
+void test_vst4q_lane_f64(float64_t  *a, float64x2x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_f64
   vst4q_lane_f64(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_p8(poly16_t const *a, poly8x16x4_t b) {
+void test_vst4q_lane_p8(poly8_t  *a, poly8x16x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_p8
   vst4q_lane_p8(a, b, 15);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[15], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[15], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_p16(poly16_t const *a, poly16x8x4_t b) {
+void test_vst4q_lane_p16(poly16_t  *a, poly16x8x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_p16
   vst4q_lane_p16(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4q_lane_p64(poly64_t const *a, poly64x2x4_t b) {
+void test_vst4q_lane_p64(poly64_t  *a, poly64x2x4_t b) {
   // CHECK-LABEL: test_vst4q_lane_p64
   vst4q_lane_p64(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_u8(uint8_t const *a, uint8x8x4_t b) {
+void test_vst4_lane_u8(uint8_t  *a, uint8x8x4_t b) {
   // CHECK-LABEL: test_vst4_lane_u8
   vst4_lane_u8(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_u16(uint16_t const *a, uint16x4x4_t b) {
+void test_vst4_lane_u16(uint16_t  *a, uint16x4x4_t b) {
   // CHECK-LABEL: test_vst4_lane_u16
   vst4_lane_u16(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_u32(uint32_t const *a, uint32x2x4_t b) {
+void test_vst4_lane_u32(uint32_t  *a, uint32x2x4_t b) {
   // CHECK-LABEL: test_vst4_lane_u32
   vst4_lane_u32(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_u64(uint64_t const *a, uint64x1x4_t b) {
+void test_vst4_lane_u64(uint64_t  *a, uint64x1x4_t b) {
   // CHECK-LABEL: test_vst4_lane_u64
   vst4_lane_u64(a, b, 0);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_s8(int8_t const *a, int8x8x4_t b) {
+void test_vst4_lane_s8(int8_t  *a, int8x8x4_t b) {
   // CHECK-LABEL: test_vst4_lane_s8
   vst4_lane_s8(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_s16(int16_t const *a, int16x4x4_t b) {
+void test_vst4_lane_s16(int16_t  *a, int16x4x4_t b) {
   // CHECK-LABEL: test_vst4_lane_s16
   vst4_lane_s16(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_s32(int32_t const *a, int32x2x4_t b) {
+void test_vst4_lane_s32(int32_t  *a, int32x2x4_t b) {
   // CHECK-LABEL: test_vst4_lane_s32
   vst4_lane_s32(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_s64(int64_t const *a, int64x1x4_t b) {
+void test_vst4_lane_s64(int64_t  *a, int64x1x4_t b) {
   // CHECK-LABEL: test_vst4_lane_s64
   vst4_lane_s64(a, b, 0);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_f16(float16_t const *a, float16x4x4_t b) {
+void test_vst4_lane_f16(float16_t  *a, float16x4x4_t b) {
   // CHECK-LABEL: test_vst4_lane_f16
   vst4_lane_f16(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_f32(float32_t const *a, float32x2x4_t b) {
+void test_vst4_lane_f32(float32_t  *a, float32x2x4_t b) {
   // CHECK-LABEL: test_vst4_lane_f32
   vst4_lane_f32(a, b, 1);
-  // CHECK: st4 {v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s, v{{[0-9]+}}.s}[1], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.s, v[0-9]+.s, v[0-9]+.s, v[0-9]+.s *}}}[1], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_f64(float64_t const *a, float64x1x4_t b) {
+void test_vst4_lane_f64(float64_t  *a, float64x1x4_t b) {
   // CHECK-LABEL: test_vst4_lane_f64
   vst4_lane_f64(a, b, 0);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_p8(poly8_t const *a, poly8x8x4_t b) {
+void test_vst4_lane_p8(poly8_t  *a, poly8x8x4_t b) {
   // CHECK-LABEL: test_vst4_lane_p8
   vst4_lane_p8(a, b, 7);
-  // CHECK: st4 {v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b, v{{[0-9]+}}.b}[7], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.b, v[0-9]+.b, v[0-9]+.b, v[0-9]+.b *}}}[7], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_p16(poly16_t const *a, poly16x4x4_t b) {
+void test_vst4_lane_p16(poly16_t  *a, poly16x4x4_t b) {
   // CHECK-LABEL: test_vst4_lane_p16
   vst4_lane_p16(a, b, 3);
-  // CHECK: st4 {v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h, v{{[0-9]+}}.h}[3], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.h, v[0-9]+.h, v[0-9]+.h, v[0-9]+.h *}}}[3], [{{x[0-9]+|sp}}]
 }
 
-void test_vst4_lane_p64(poly64_t const *a, poly64x1x4_t b) {
+void test_vst4_lane_p64(poly64_t  *a, poly64x1x4_t b) {
   // CHECK-LABEL: test_vst4_lane_p64
   vst4_lane_p64(a, b, 0);
-  // CHECK: st4 {v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d, v{{[0-9]+}}.d}[0], [{{x[0-9]+|sp}}]
+  // CHECK: st4 {{{ *v[0-9]+.d, v[0-9]+.d, v[0-9]+.d, v[0-9]+.d *}}}[0], [{{x[0-9]+|sp}}]
 }
diff --git a/test/CodeGen/aarch64-neon-misc.c b/test/CodeGen/aarch64-neon-misc.c
index 08174d9..dce00f3 100644
--- a/test/CodeGen/aarch64-neon-misc.c
+++ b/test/CodeGen/aarch64-neon-misc.c
@@ -1,2005 +1,2044 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
-// CHECK: test_vceqz_s8
-// CHECK: cmeq  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x0
+// CHECK-LABEL: test_vceqz_s8
+// CHECK: cmeq  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
 uint8x8_t test_vceqz_s8(int8x8_t a) {
   return vceqz_s8(a);
 }
 
-// CHECK: test_vceqz_s16
-// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vceqz_s16
+// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vceqz_s16(int16x4_t a) {
   return vceqz_s16(a);
 }
 
-// CHECK: test_vceqz_s32
-// CHECK: cmeq  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0x0
+// CHECK-LABEL: test_vceqz_s32
+// CHECK: cmeq  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #{{0x0|0}}
 uint32x2_t test_vceqz_s32(int32x2_t a) {
   return vceqz_s32(a);
 }
 
-// CHECK: test_vceqz_s64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vceqz_s64
+// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vceqz_s64(int64x1_t a) {
   return vceqz_s64(a);
 }
 
-// CHECK: test_vceqz_u64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vceqz_u64
+// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vceqz_u64(uint64x1_t a) {
   return vceqz_u64(a);
 }
 
-// CHECK: test_vceqz_p64
-// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vceqz_p64
+// CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vceqz_p64(poly64x1_t a) {
   return vceqz_p64(a);
 }
 
-// CHECK: test_vceqzq_s8
-// CHECK: cmeq  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x0
+// CHECK-LABEL: test_vceqzq_s8
+// CHECK: cmeq  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
 uint8x16_t test_vceqzq_s8(int8x16_t a) {
   return vceqzq_s8(a);
 }
 
-// CHECK: test_vceqzq_s16
-// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vceqzq_s16
+// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vceqzq_s16(int16x8_t a) {
   return vceqzq_s16(a);
 }
 
-// CHECK: test_vceqzq_s32
-// CHECK: cmeq  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0x0
+// CHECK-LABEL: test_vceqzq_s32
+// CHECK: cmeq  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #{{0x0|0}}
 uint32x4_t test_vceqzq_s32(int32x4_t a) {
   return vceqzq_s32(a);
 }
 
-// CHECK: test_vceqzq_s64
-// CHECK: cmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0x0
+// CHECK-LABEL: test_vceqzq_s64
+// CHECK: cmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #{{0x0|0}}
 uint64x2_t test_vceqzq_s64(int64x2_t a) {
   return vceqzq_s64(a);
 }
 
-// CHECK: test_vceqz_u8
-// CHECK: cmeq  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x0
+// CHECK-LABEL: test_vceqz_u8
+// CHECK: cmeq  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
 uint8x8_t test_vceqz_u8(uint8x8_t a) {
   return vceqz_u8(a);
 }
 
-// CHECK: test_vceqz_u16
-// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vceqz_u16
+// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vceqz_u16(uint16x4_t a) {
   return vceqz_u16(a);
 }
 
-// CHECK: test_vceqz_u32
-// CHECK: cmeq  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0x0
+// CHECK-LABEL: test_vceqz_u32
+// CHECK: cmeq  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #{{0x0|0}}
 uint32x2_t test_vceqz_u32(uint32x2_t a) {
   return vceqz_u32(a);
 }
 
-// CHECK: test_vceqzq_u8
-// CHECK: cmeq  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x0
+// CHECK-LABEL: test_vceqzq_u8
+// CHECK: cmeq  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
 uint8x16_t test_vceqzq_u8(uint8x16_t a) {
   return vceqzq_u8(a);
 }
 
-// CHECK: test_vceqzq_u16
-// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vceqzq_u16
+// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vceqzq_u16(uint16x8_t a) {
   return vceqzq_u16(a);
 }
 
-// CHECK: test_vceqzq_u32
-// CHECK: cmeq  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0x0
+// CHECK-LABEL: test_vceqzq_u32
+// CHECK: cmeq  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #{{0x0|0}}
 uint32x4_t test_vceqzq_u32(uint32x4_t a) {
   return vceqzq_u32(a);
 }
 
-// CHECK: test_vceqzq_u64
-// CHECK: cmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0x0
+// CHECK-LABEL: test_vceqzq_u64
+// CHECK: cmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #{{0x0|0}}
 uint64x2_t test_vceqzq_u64(uint64x2_t a) {
   return vceqzq_u64(a);
 }
 
-// CHECK: test_vceqz_f32
+// CHECK-LABEL: test_vceqz_f32
 // CHECK: fcmeq  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
 uint32x2_t test_vceqz_f32(float32x2_t a) {
   return vceqz_f32(a);
 }
 
-// CHECK: test_vceqz_f64
+// CHECK-LABEL: test_vceqz_f64
 // CHECK: fcmeq  {{d[0-9]+}}, {{d[0-9]+}}, #0
 uint64x1_t test_vceqz_f64(float64x1_t a) {
   return vceqz_f64(a);
 }
 
-// CHECK: test_vceqzq_f32
+// CHECK-LABEL: test_vceqzq_f32
 // CHECK: fcmeq  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
 uint32x4_t test_vceqzq_f32(float32x4_t a) {
   return vceqzq_f32(a);
 }
 
-// CHECK: test_vceqz_p16
-// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vceqz_p8
+// CHECK: cmeq  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
+uint8x8_t test_vceqz_p8(poly8x8_t a) {
+  return vceqz_p8(a);
+}
+
+// CHECK-LABEL: test_vceqzq_p8
+// CHECK: cmeq  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
+uint8x16_t test_vceqzq_p8(poly8x16_t a) {
+  return vceqzq_p8(a);
+}
+
+// CHECK-LABEL: test_vceqz_p16
+// CHECK: cmeq  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vceqz_p16(poly16x4_t a) {
   return vceqz_p16(a);
 }
 
-// CHECK: test_vceqzq_p16
-// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vceqzq_p16
+// CHECK: cmeq  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vceqzq_p16(poly16x8_t a) {
   return vceqzq_p16(a);
 }
 
-// CHECK: test_vceqzq_f64
+// CHECK-LABEL: test_vceqzq_f64
 // CHECK: fcmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vceqzq_f64(float64x2_t a) {
   return vceqzq_f64(a);
 }
 
-// CHECK: test_vceqzq_p64
+// CHECK-LABEL: test_vceqzq_p64
 // CHECK: cmeq  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vceqzq_p64(poly64x2_t a) {
   return vceqzq_p64(a);
 }
 
-// CHECK: test_vcgez_s8
-// CHECK: cmge  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x0
+// CHECK-LABEL: test_vcgez_s8
+// CHECK: cmge  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
 uint8x8_t test_vcgez_s8(int8x8_t a) {
   return vcgez_s8(a);
 }
 
-// CHECK: test_vcgez_s16
-// CHECK: cmge  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vcgez_s16
+// CHECK: cmge  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vcgez_s16(int16x4_t a) {
   return vcgez_s16(a);
 }
 
-// CHECK: test_vcgez_s32
-// CHECK: cmge  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0x0
+// CHECK-LABEL: test_vcgez_s32
+// CHECK: cmge  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #{{0x0|0}}
 uint32x2_t test_vcgez_s32(int32x2_t a) {
   return vcgez_s32(a);
 }
 
-// CHECK: test_vcgez_s64
-// CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vcgez_s64
+// CHECK: cmge {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vcgez_s64(int64x1_t a) {
   return vcgez_s64(a);
 }
 
-// CHECK: test_vcgezq_s8
-// CHECK: cmge  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x0
+// CHECK-LABEL: test_vcgezq_s8
+// CHECK: cmge  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
 uint8x16_t test_vcgezq_s8(int8x16_t a) {
   return vcgezq_s8(a);
 }
 
-// CHECK: test_vcgezq_s16
-// CHECK: cmge  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vcgezq_s16
+// CHECK: cmge  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vcgezq_s16(int16x8_t a) {
   return vcgezq_s16(a);
 }
 
-// CHECK: test_vcgezq_s32
-// CHECK: cmge  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0x0
+// CHECK-LABEL: test_vcgezq_s32
+// CHECK: cmge  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #{{0x0|0}}
 uint32x4_t test_vcgezq_s32(int32x4_t a) {
   return vcgezq_s32(a);
 }
 
-// CHECK: test_vcgezq_s64
-// CHECK: cmge  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0x0
+// CHECK-LABEL: test_vcgezq_s64
+// CHECK: cmge  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #{{0x0|0}}
 uint64x2_t test_vcgezq_s64(int64x2_t a) {
   return vcgezq_s64(a);
 }
 
-// CHECK: test_vcgez_f32
+// CHECK-LABEL: test_vcgez_f32
 // CHECK: fcmge  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
 uint32x2_t test_vcgez_f32(float32x2_t a) {
   return vcgez_f32(a);
 }
 
-// CHECK: test_vcgezq_f32
+// CHECK-LABEL: test_vcgez_f64
+// CHECK: fcmge  {{d[0-9]+}}, {{d[0-9]+}}, #0
+uint64x1_t test_vcgez_f64(float64x1_t a) {
+  return vcgez_f64(a);
+}
+
+// CHECK-LABEL: test_vcgezq_f32
 // CHECK: fcmge  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
 uint32x4_t test_vcgezq_f32(float32x4_t a) {
   return vcgezq_f32(a);
 }
 
-// CHECK: test_vcgezq_f64
+// CHECK-LABEL: test_vcgezq_f64
 // CHECK: fcmge  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vcgezq_f64(float64x2_t a) {
   return vcgezq_f64(a);
 }
 
-// CHECK: test_vclez_s8
-// CHECK: cmle  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x0
+// CHECK-LABEL: test_vclez_s8
+// CHECK: cmle  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
 uint8x8_t test_vclez_s8(int8x8_t a) {
   return vclez_s8(a);
 }
 
-// CHECK: test_vclez_s16
-// CHECK: cmle  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vclez_s16
+// CHECK: cmle  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vclez_s16(int16x4_t a) {
   return vclez_s16(a);
 }
 
-// CHECK: test_vclez_s32
-// CHECK: cmle  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0x0
+// CHECK-LABEL: test_vclez_s32
+// CHECK: cmle  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #{{0x0|0}}
 uint32x2_t test_vclez_s32(int32x2_t a) {
   return vclez_s32(a);
 }
 
-// CHECK: test_vclez_s64
-// CHECK: cmle {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vclez_s64
+// CHECK: cmle {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vclez_s64(int64x1_t a) {
   return vclez_s64(a);
 }
 
-// CHECK: test_vclezq_s8
-// CHECK: cmle  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x0
+// CHECK-LABEL: test_vclezq_s8
+// CHECK: cmle  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
 uint8x16_t test_vclezq_s8(int8x16_t a) {
   return vclezq_s8(a);
 }
 
-// CHECK: test_vclezq_s16
-// CHECK: cmle  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vclezq_s16
+// CHECK: cmle  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vclezq_s16(int16x8_t a) {
   return vclezq_s16(a);
 }
 
-// CHECK: test_vclezq_s32
-// CHECK: cmle  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0x0
+// CHECK-LABEL: test_vclezq_s32
+// CHECK: cmle  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #{{0x0|0}}
 uint32x4_t test_vclezq_s32(int32x4_t a) {
   return vclezq_s32(a);
 }
 
-// CHECK: test_vclezq_s64
-// CHECK: cmle  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0x0
+// CHECK-LABEL: test_vclezq_s64
+// CHECK: cmle  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #{{0x0|0}}
 uint64x2_t test_vclezq_s64(int64x2_t a) {
   return vclezq_s64(a);
 }
 
-// CHECK: test_vclez_f32
+// CHECK-LABEL: test_vclez_f32
 // CHECK: fcmle  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
 uint32x2_t test_vclez_f32(float32x2_t a) {
   return vclez_f32(a);
 }
 
-// CHECK: test_vclezq_f32
+// CHECK-LABEL: test_vclez_f64
+// CHECK: fcmle  {{d[0-9]+}}, {{d[0-9]+}}, #0
+uint64x1_t test_vclez_f64(float64x1_t a) {
+  return vclez_f64(a);
+}
+
+// CHECK-LABEL: test_vclezq_f32
 // CHECK: fcmle  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
 uint32x4_t test_vclezq_f32(float32x4_t a) {
   return vclezq_f32(a);
 }
 
-// CHECK: test_vclezq_f64
+// CHECK-LABEL: test_vclezq_f64
 // CHECK: fcmle  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vclezq_f64(float64x2_t a) {
   return vclezq_f64(a);
 }
 
-// CHECK: test_vcgtz_s8
-// CHECK: cmgt  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0x0
+// CHECK-LABEL: test_vcgtz_s8
+// CHECK: cmgt  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #{{0x0|0}}
 uint8x8_t test_vcgtz_s8(int8x8_t a) {
   return vcgtz_s8(a);
 }
 
-// CHECK: test_vcgtz_s16
-// CHECK: cmgt  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0x0
+// CHECK-LABEL: test_vcgtz_s16
+// CHECK: cmgt  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #{{0x0|0}}
 uint16x4_t test_vcgtz_s16(int16x4_t a) {
   return vcgtz_s16(a);
 }
 
-// CHECK: test_vcgtz_s32
-// CHECK: cmgt  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0x0
+// CHECK-LABEL: test_vcgtz_s32
+// CHECK: cmgt  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #{{0x0|0}}
 uint32x2_t test_vcgtz_s32(int32x2_t a) {
   return vcgtz_s32(a);
 }
 
-// CHECK: test_vcgtz_s64
-// CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, #0x0
+// CHECK-LABEL: test_vcgtz_s64
+// CHECK: cmgt {{d[0-9]+}}, {{d[0-9]+}}, #{{0x0|0}}
 uint64x1_t test_vcgtz_s64(int64x1_t a) {
   return vcgtz_s64(a);
 }
 
-// CHECK: test_vcgtzq_s8
-// CHECK: cmgt  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x0
+// CHECK-LABEL: test_vcgtzq_s8
+// CHECK: cmgt  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #{{0x0|0}}
 uint8x16_t test_vcgtzq_s8(int8x16_t a) {
   return vcgtzq_s8(a);
 }
 
-// CHECK: test_vcgtzq_s16
-// CHECK: cmgt  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0x0
+// CHECK-LABEL: test_vcgtzq_s16
+// CHECK: cmgt  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #{{0x0|0}}
 uint16x8_t test_vcgtzq_s16(int16x8_t a) {
   return vcgtzq_s16(a);
 }
 
-// CHECK: test_vcgtzq_s32
-// CHECK: cmgt  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0x0
+// CHECK-LABEL: test_vcgtzq_s32
+// CHECK: cmgt  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #{{0x0|0}}
 uint32x4_t test_vcgtzq_s32(int32x4_t a) {
   return vcgtzq_s32(a);
 }
 
-// CHECK: test_vcgtzq_s64
-// CHECK: cmgt  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0x0
+// CHECK-LABEL: test_vcgtzq_s64
+// CHECK: cmgt  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #{{0x0|0}}
 uint64x2_t test_vcgtzq_s64(int64x2_t a) {
   return vcgtzq_s64(a);
 }
 
-// CHECK: test_vcgtz_f32
+// CHECK-LABEL: test_vcgtz_f32
 // CHECK: fcmgt  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
 uint32x2_t test_vcgtz_f32(float32x2_t a) {
   return vcgtz_f32(a);
 }
 
-// CHECK: test_vcgtzq_f32
+// CHECK-LABEL: test_vcgtz_f64
+// CHECK: fcmgt  {{d[0-9]+}}, {{d[0-9]+}}, #0
+uint64x1_t test_vcgtz_f64(float64x1_t a) {
+  return vcgtz_f64(a);
+}
+
+// CHECK-LABEL: test_vcgtzq_f32
 // CHECK: fcmgt  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
 uint32x4_t test_vcgtzq_f32(float32x4_t a) {
   return vcgtzq_f32(a);
 }
 
-// CHECK: test_vcgtzq_f64
+// CHECK-LABEL: test_vcgtzq_f64
 // CHECK: fcmgt  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vcgtzq_f64(float64x2_t a) {
   return vcgtzq_f64(a);
 }
 
-// CHECK: test_vcltz_s8
-// CHECK: cmlt  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #0
+// CHECK-LABEL: test_vcltz_s8
+// CHECK: sshr  {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, #7
 uint8x8_t test_vcltz_s8(int8x8_t a) {
   return vcltz_s8(a);
 }
 
-// CHECK: test_vcltz_s16
-// CHECK: cmlt  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #0
+// CHECK-LABEL: test_vcltz_s16
+// CHECK: sshr  {{v[0-9]+}}.4h, {{v[0-9]+}}.4h, #15
 uint16x4_t test_vcltz_s16(int16x4_t a) {
   return vcltz_s16(a);
 }
 
-// CHECK: test_vcltz_s32
-// CHECK: cmlt  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
+// CHECK-LABEL: test_vcltz_s32
+// CHECK: sshr  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #31
 uint32x2_t test_vcltz_s32(int32x2_t a) {
   return vcltz_s32(a);
 }
 
-// CHECK: test_vcltz_s64
-// CHECK: cmlt {{d[0-9]+}}, {{d[0-9]+}}, #0
+// CHECK-LABEL: test_vcltz_s64
+// CHECK: sshr {{d[0-9]+}}, {{d[0-9]+}}, #63
 uint64x1_t test_vcltz_s64(int64x1_t a) {
   return vcltz_s64(a);
 }
 
-// CHECK: test_vcltzq_s8
-// CHECK: cmlt  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0
+// CHECK-LABEL: test_vcltzq_s8
+// CHECK: sshr  {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #7
 uint8x16_t test_vcltzq_s8(int8x16_t a) {
   return vcltzq_s8(a);
 }
 
-// CHECK: test_vcltzq_s16
-// CHECK: cmlt  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #0
+// CHECK-LABEL: test_vcltzq_s16
+// CHECK: sshr  {{v[0-9]+}}.8h, {{v[0-9]+}}.8h, #15
 uint16x8_t test_vcltzq_s16(int16x8_t a) {
   return vcltzq_s16(a);
 }
 
-// CHECK: test_vcltzq_s32
-// CHECK: cmlt  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
+// CHECK-LABEL: test_vcltzq_s32
+// CHECK: sshr  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #31
 uint32x4_t test_vcltzq_s32(int32x4_t a) {
   return vcltzq_s32(a);
 }
 
-// CHECK: test_vcltzq_s64
-// CHECK: cmlt  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
+// CHECK-LABEL: test_vcltzq_s64
+// CHECK: sshr  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #63
 uint64x2_t test_vcltzq_s64(int64x2_t a) {
   return vcltzq_s64(a);
 }
 
-// CHECK: test_vcltz_f32
+// CHECK-LABEL: test_vcltz_f32
 // CHECK: fcmlt  {{v[0-9]+}}.2s, {{v[0-9]+}}.2s, #0
 uint32x2_t test_vcltz_f32(float32x2_t a) {
   return vcltz_f32(a);
 }
+ 
+// CHECK-LABEL: test_vcltz_f64
+// CHECK: fcmlt  {{d[0-9]+}}, {{d[0-9]+}}, #0
+uint64x1_t test_vcltz_f64(float64x1_t a) {
+  return vcltz_f64(a);
+}
 
-// CHECK: test_vcltzq_f32
+// CHECK-LABEL: test_vcltzq_f32
 // CHECK: fcmlt  {{v[0-9]+}}.4s, {{v[0-9]+}}.4s, #0
 uint32x4_t test_vcltzq_f32(float32x4_t a) {
   return vcltzq_f32(a);
 }
 
-// CHECK: test_vcltzq_f64
+// CHECK-LABEL: test_vcltzq_f64
 // CHECK: fcmlt  {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #0
 uint64x2_t test_vcltzq_f64(float64x2_t a) {
   return vcltzq_f64(a);
 }
 
-// CHECK: test_vrev16_s8
+// CHECK-LABEL: test_vrev16_s8
 // CHECK: rev16 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 int8x8_t test_vrev16_s8(int8x8_t a) {
   return vrev16_s8(a);
 }
 
-// CHECK: test_vrev16_u8
+// CHECK-LABEL: test_vrev16_u8
 // CHECK: rev16 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 uint8x8_t test_vrev16_u8(uint8x8_t a) {
   return vrev16_u8(a);
 }
 
-// CHECK: test_vrev16_p8
+// CHECK-LABEL: test_vrev16_p8
 // CHECK: rev16 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 poly8x8_t test_vrev16_p8(poly8x8_t a) {
   return vrev16_p8(a);
 }
 
-// CHECK: test_vrev16q_s8
+// CHECK-LABEL: test_vrev16q_s8
 // CHECK: rev16 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 int8x16_t test_vrev16q_s8(int8x16_t a) {
   return vrev16q_s8(a);
 }
 
-// CHECK: test_vrev16q_u8
+// CHECK-LABEL: test_vrev16q_u8
 // CHECK: rev16 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 uint8x16_t test_vrev16q_u8(uint8x16_t a) {
   return vrev16q_u8(a);
 }
 
-// CHECK: test_vrev16q_p8
+// CHECK-LABEL: test_vrev16q_p8
 // CHECK: rev16 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 poly8x16_t test_vrev16q_p8(poly8x16_t a) {
   return vrev16q_p8(a);
 }
 
-// CHECK: test_vrev32_s8
+// CHECK-LABEL: test_vrev32_s8
 // CHECK: rev32 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 int8x8_t test_vrev32_s8(int8x8_t a) {
   return vrev32_s8(a);
 }
 
-// CHECK: test_vrev32_s16
+// CHECK-LABEL: test_vrev32_s16
 // CHECK: rev32 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 int16x4_t test_vrev32_s16(int16x4_t a) {
   return vrev32_s16(a);
 }
 
-// CHECK: test_vrev32_u8
+// CHECK-LABEL: test_vrev32_u8
 // CHECK: rev32 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 uint8x8_t test_vrev32_u8(uint8x8_t a) {
   return vrev32_u8(a);
 }
 
-// CHECK: test_vrev32_u16
+// CHECK-LABEL: test_vrev32_u16
 // CHECK: rev32 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 uint16x4_t test_vrev32_u16(uint16x4_t a) {
   return vrev32_u16(a);
 }
 
-// CHECK: test_vrev32_p8
+// CHECK-LABEL: test_vrev32_p8
 // CHECK: rev32 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 poly8x8_t test_vrev32_p8(poly8x8_t a) {
   return vrev32_p8(a);
 }
 
-// CHECK: test_vrev32_p16
+// CHECK-LABEL: test_vrev32_p16
 // CHECK: rev32 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 poly16x4_t test_vrev32_p16(poly16x4_t a) {
   return vrev32_p16(a);
 }
 
-// CHECK: test_vrev32q_s8
+// CHECK-LABEL: test_vrev32q_s8
 // CHECK: rev32 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 int8x16_t test_vrev32q_s8(int8x16_t a) {
   return vrev32q_s8(a);
 }
 
-// CHECK: test_vrev32q_s16
+// CHECK-LABEL: test_vrev32q_s16
 // CHECK: rev32 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 int16x8_t test_vrev32q_s16(int16x8_t a) {
   return vrev32q_s16(a);
 }
 
-// CHECK: test_vrev32q_u8
+// CHECK-LABEL: test_vrev32q_u8
 // CHECK: rev32 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 uint8x16_t test_vrev32q_u8(uint8x16_t a) {
   return vrev32q_u8(a);
 }
 
-// CHECK: test_vrev32q_u16
+// CHECK-LABEL: test_vrev32q_u16
 // CHECK: rev32 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 uint16x8_t test_vrev32q_u16(uint16x8_t a) {
   return vrev32q_u16(a);
 }
 
-// CHECK: test_vrev32q_p8
+// CHECK-LABEL: test_vrev32q_p8
 // CHECK: rev32 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 poly8x16_t test_vrev32q_p8(poly8x16_t a) {
   return vrev32q_p8(a);
 }
 
-// CHECK: test_vrev32q_p16
+// CHECK-LABEL: test_vrev32q_p16
 // CHECK: rev32 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 poly16x8_t test_vrev32q_p16(poly16x8_t a) {
   return vrev32q_p16(a);
 }
 
-// CHECK: test_vrev64_s8
+// CHECK-LABEL: test_vrev64_s8
 // CHECK: rev64 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 int8x8_t test_vrev64_s8(int8x8_t a) {
   return vrev64_s8(a);
 }
 
-// CHECK: test_vrev64_s16
+// CHECK-LABEL: test_vrev64_s16
 // CHECK: rev64 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 int16x4_t test_vrev64_s16(int16x4_t a) {
   return vrev64_s16(a);
 }
 
-// CHECK: test_vrev64_s32
+// CHECK-LABEL: test_vrev64_s32
 // CHECK: rev64 v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 int32x2_t test_vrev64_s32(int32x2_t a) {
   return vrev64_s32(a);
 }
 
-// CHECK: test_vrev64_u8
+// CHECK-LABEL: test_vrev64_u8
 // CHECK: rev64 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 uint8x8_t test_vrev64_u8(uint8x8_t a) {
   return vrev64_u8(a);
 }
 
-// CHECK: test_vrev64_u16
+// CHECK-LABEL: test_vrev64_u16
 // CHECK: rev64 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 uint16x4_t test_vrev64_u16(uint16x4_t a) {
   return vrev64_u16(a);
 }
 
-// CHECK: test_vrev64_u32
+// CHECK-LABEL: test_vrev64_u32
 // CHECK: rev64 v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 uint32x2_t test_vrev64_u32(uint32x2_t a) {
   return vrev64_u32(a);
 }
 
-// CHECK: test_vrev64_p8
+// CHECK-LABEL: test_vrev64_p8
 // CHECK: rev64 v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 poly8x8_t test_vrev64_p8(poly8x8_t a) {
   return vrev64_p8(a);
 }
 
-// CHECK: test_vrev64_p16
+// CHECK-LABEL: test_vrev64_p16
 // CHECK: rev64 v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 poly16x4_t test_vrev64_p16(poly16x4_t a) {
   return vrev64_p16(a);
 }
 
-// CHECK: test_vrev64_f32
+// CHECK-LABEL: test_vrev64_f32
 // CHECK: rev64 v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 float32x2_t test_vrev64_f32(float32x2_t a) {
   return vrev64_f32(a);
 }
 
-// CHECK: test_vrev64q_s8
+// CHECK-LABEL: test_vrev64q_s8
 // CHECK: rev64 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 int8x16_t test_vrev64q_s8(int8x16_t a) {
   return vrev64q_s8(a);
 }
 
-// CHECK: test_vrev64q_s16
+// CHECK-LABEL: test_vrev64q_s16
 // CHECK: rev64 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 int16x8_t test_vrev64q_s16(int16x8_t a) {
   return vrev64q_s16(a);
 }
 
-// CHECK: test_vrev64q_s32
+// CHECK-LABEL: test_vrev64q_s32
 // CHECK: rev64 v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 int32x4_t test_vrev64q_s32(int32x4_t a) {
   return vrev64q_s32(a);
 }
 
-// CHECK: test_vrev64q_u8
+// CHECK-LABEL: test_vrev64q_u8
 // CHECK: rev64 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 uint8x16_t test_vrev64q_u8(uint8x16_t a) {
   return vrev64q_u8(a);
 }
 
-// CHECK: test_vrev64q_u16
+// CHECK-LABEL: test_vrev64q_u16
 // CHECK: rev64 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 uint16x8_t test_vrev64q_u16(uint16x8_t a) {
   return vrev64q_u16(a);
 }
 
-// CHECK: test_vrev64q_u32
+// CHECK-LABEL: test_vrev64q_u32
 // CHECK: rev64 v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 uint32x4_t test_vrev64q_u32(uint32x4_t a) {
   return vrev64q_u32(a);
 }
 
-// CHECK: test_vrev64q_p8
+// CHECK-LABEL: test_vrev64q_p8
 // CHECK: rev64 v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 poly8x16_t test_vrev64q_p8(poly8x16_t a) {
   return vrev64q_p8(a);
 }
 
-// CHECK: test_vrev64q_p16
+// CHECK-LABEL: test_vrev64q_p16
 // CHECK: rev64 v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 poly16x8_t test_vrev64q_p16(poly16x8_t a) {
   return vrev64q_p16(a);
 }
 
-// CHECK: test_vrev64q_f32
+// CHECK-LABEL: test_vrev64q_f32
 // CHECK: rev64 v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 float32x4_t test_vrev64q_f32(float32x4_t a) {
   return vrev64q_f32(a);
 }
 
 int16x4_t test_vpaddl_s8(int8x8_t a) {
-  // CHECK: test_vpaddl_s8
+  // CHECK-LABEL: test_vpaddl_s8
   return vpaddl_s8(a);
   // CHECK: saddlp v{{[0-9]+}}.4h, v{{[0-9]+}}.8b
 }
 
 int32x2_t test_vpaddl_s16(int16x4_t a) {
-  // CHECK: test_vpaddl_s16
+  // CHECK-LABEL: test_vpaddl_s16
   return vpaddl_s16(a);
   // CHECK: saddlp v{{[0-9]+}}.2s, v{{[0-9]+}}.4h
 }
 
 int64x1_t test_vpaddl_s32(int32x2_t a) {
-  // CHECK: test_vpaddl_s32
+  // CHECK-LABEL: test_vpaddl_s32
   return vpaddl_s32(a);
   // CHECK: saddlp v{{[0-9]+}}.1d, v{{[0-9]+}}.2s
 }
 
 uint16x4_t test_vpaddl_u8(uint8x8_t a) {
-  // CHECK: test_vpaddl_u8
+  // CHECK-LABEL: test_vpaddl_u8
   return vpaddl_u8(a);
   // CHECK: uaddlp v{{[0-9]+}}.4h, v{{[0-9]+}}.8b
 }
 
 uint32x2_t test_vpaddl_u16(uint16x4_t a) {
-  // CHECK: test_vpaddl_u16
+  // CHECK-LABEL: test_vpaddl_u16
   return vpaddl_u16(a);
   // CHECK: uaddlp v{{[0-9]+}}.2s, v{{[0-9]+}}.4h
 }
 
 uint64x1_t test_vpaddl_u32(uint32x2_t a) {
-  // CHECK: test_vpaddl_u32
+  // CHECK-LABEL: test_vpaddl_u32
   return vpaddl_u32(a);
   // CHECK: uaddlp v{{[0-9]+}}.1d, v{{[0-9]+}}.2s
 }
 
 int16x8_t test_vpaddlq_s8(int8x16_t a) {
-  // CHECK: test_vpaddlq_s8
+  // CHECK-LABEL: test_vpaddlq_s8
   return vpaddlq_s8(a);
   // CHECK: saddlp v{{[0-9]+}}.8h, v{{[0-9]+}}.16b
 }
 
 int32x4_t test_vpaddlq_s16(int16x8_t a) {
-  // CHECK: test_vpaddlq_s16
+  // CHECK-LABEL: test_vpaddlq_s16
   return vpaddlq_s16(a);
   // CHECK: saddlp v{{[0-9]+}}.4s, v{{[0-9]+}}.8h
 }
 
 int64x2_t test_vpaddlq_s32(int32x4_t a) {
-  // CHECK: test_vpaddlq_s32
+  // CHECK-LABEL: test_vpaddlq_s32
   return vpaddlq_s32(a);
   // CHECK: saddlp v{{[0-9]+}}.2d, v{{[0-9]+}}.4s
 }
 
 uint16x8_t test_vpaddlq_u8(uint8x16_t a) {
-  // CHECK: test_vpaddlq_u8
+  // CHECK-LABEL: test_vpaddlq_u8
   return vpaddlq_u8(a);
   // CHECK: uaddlp v{{[0-9]+}}.8h, v{{[0-9]+}}.16b
 }
 
 uint32x4_t test_vpaddlq_u16(uint16x8_t a) {
-  // CHECK: test_vpaddlq_u16
+  // CHECK-LABEL: test_vpaddlq_u16
   return vpaddlq_u16(a);
   // CHECK: uaddlp v{{[0-9]+}}.4s, v{{[0-9]+}}.8h
 }
 
 uint64x2_t test_vpaddlq_u32(uint32x4_t a) {
-  // CHECK: test_vpaddlq_u32
+  // CHECK-LABEL: test_vpaddlq_u32
   return vpaddlq_u32(a);
   // CHECK: uaddlp v{{[0-9]+}}.2d, v{{[0-9]+}}.4s
 }
 
 int16x4_t test_vpadal_s8(int16x4_t a, int8x8_t b) {
-  // CHECK: test_vpadal_s8
+  // CHECK-LABEL: test_vpadal_s8
   return vpadal_s8(a, b);
   // CHECK: sadalp v{{[0-9]+}}.4h, v{{[0-9]+}}.8b
 }
 
 int32x2_t test_vpadal_s16(int32x2_t a, int16x4_t b) {
-  // CHECK: test_vpadal_s16
+  // CHECK-LABEL: test_vpadal_s16
   return vpadal_s16(a, b);
   // CHECK: sadalp v{{[0-9]+}}.2s, v{{[0-9]+}}.4h
 }
 
 int64x1_t test_vpadal_s32(int64x1_t a, int32x2_t b) {
-  // CHECK: test_vpadal_s32
+  // CHECK-LABEL: test_vpadal_s32
   return vpadal_s32(a, b);
   // CHECK: sadalp v{{[0-9]+}}.1d, v{{[0-9]+}}.2s
 }
 
 uint16x4_t test_vpadal_u8(uint16x4_t a, uint8x8_t b) {
-  // CHECK: test_vpadal_u8
+  // CHECK-LABEL: test_vpadal_u8
   return vpadal_u8(a, b);
   // CHECK: uadalp v{{[0-9]+}}.4h, v{{[0-9]+}}.8b
 }
 
 uint32x2_t test_vpadal_u16(uint32x2_t a, uint16x4_t b) {
-  // CHECK: test_vpadal_u16
+  // CHECK-LABEL: test_vpadal_u16
   return vpadal_u16(a, b);
   // CHECK: uadalp v{{[0-9]+}}.2s, v{{[0-9]+}}.4h
 }
 
 uint64x1_t test_vpadal_u32(uint64x1_t a, uint32x2_t b) {
-  // CHECK: test_vpadal_u32
+  // CHECK-LABEL: test_vpadal_u32
   return vpadal_u32(a, b);
   // CHECK: uadalp v{{[0-9]+}}.1d, v{{[0-9]+}}.2s
 }
 
 int16x8_t test_vpadalq_s8(int16x8_t a, int8x16_t b) {
-  // CHECK: test_vpadalq_s8
+  // CHECK-LABEL: test_vpadalq_s8
   return vpadalq_s8(a, b);
   // CHECK: sadalp v{{[0-9]+}}.8h, v{{[0-9]+}}.16b
 }
 
 int32x4_t test_vpadalq_s16(int32x4_t a, int16x8_t b) {
-  // CHECK: test_vpadalq_s16
+  // CHECK-LABEL: test_vpadalq_s16
   return vpadalq_s16(a, b);
   // CHECK: sadalp v{{[0-9]+}}.4s, v{{[0-9]+}}.8h
 }
 
 int64x2_t test_vpadalq_s32(int64x2_t a, int32x4_t b) {
-  // CHECK: test_vpadalq_s32
+  // CHECK-LABEL: test_vpadalq_s32
   return vpadalq_s32(a, b);
   // CHECK: sadalp v{{[0-9]+}}.2d, v{{[0-9]+}}.4s
 }
 
 uint16x8_t test_vpadalq_u8(uint16x8_t a, uint8x16_t b) {
-  // CHECK: test_vpadalq_u8
+  // CHECK-LABEL: test_vpadalq_u8
   return vpadalq_u8(a, b);
   // CHECK: uadalp v{{[0-9]+}}.8h, v{{[0-9]+}}.16b
 }
 
 uint32x4_t test_vpadalq_u16(uint32x4_t a, uint16x8_t b) {
-  // CHECK: test_vpadalq_u16
+  // CHECK-LABEL: test_vpadalq_u16
   return vpadalq_u16(a, b);
   // CHECK: uadalp v{{[0-9]+}}.4s, v{{[0-9]+}}.8h
 }
 
 uint64x2_t test_vpadalq_u32(uint64x2_t a, uint32x4_t b) {
-  // CHECK: test_vpadalq_u32
+  // CHECK-LABEL: test_vpadalq_u32
   return vpadalq_u32(a, b);
   // CHECK: uadalp v{{[0-9]+}}.2d, v{{[0-9]+}}.4s
 }
 
 int8x8_t test_vqabs_s8(int8x8_t a) {
-  // CHECK: test_vqabs_s8
+  // CHECK-LABEL: test_vqabs_s8
   return vqabs_s8(a);
   // CHECK: sqabs v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vqabsq_s8(int8x16_t a) {
-  // CHECK: test_vqabsq_s8
+  // CHECK-LABEL: test_vqabsq_s8
   return vqabsq_s8(a);
   // CHECK: sqabs v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vqabs_s16(int16x4_t a) {
-  // CHECK: test_vqabs_s16
+  // CHECK-LABEL: test_vqabs_s16
   return vqabs_s16(a);
   // CHECK: sqabs v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vqabsq_s16(int16x8_t a) {
-  // CHECK: test_vqabsq_s16
+  // CHECK-LABEL: test_vqabsq_s16
   return vqabsq_s16(a);
   // CHECK: sqabs v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vqabs_s32(int32x2_t a) {
-  // CHECK: test_vqabs_s32
+  // CHECK-LABEL: test_vqabs_s32
   return vqabs_s32(a);
   // CHECK: sqabs v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vqabsq_s32(int32x4_t a) {
-  // CHECK: test_vqabsq_s32
+  // CHECK-LABEL: test_vqabsq_s32
   return vqabsq_s32(a);
   // CHECK: sqabs v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vqabsq_s64(int64x2_t a) {
-  // CHECK: test_vqabsq_s64
+  // CHECK-LABEL: test_vqabsq_s64
   return vqabsq_s64(a);
   // CHECK: sqabs v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vqneg_s8(int8x8_t a) {
-  // CHECK: test_vqneg_s8
+  // CHECK-LABEL: test_vqneg_s8
   return vqneg_s8(a);
   // CHECK: sqneg v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vqnegq_s8(int8x16_t a) {
-  // CHECK: test_vqnegq_s8
+  // CHECK-LABEL: test_vqnegq_s8
   return vqnegq_s8(a);
   // CHECK: sqneg v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vqneg_s16(int16x4_t a) {
-  // CHECK: test_vqneg_s16
+  // CHECK-LABEL: test_vqneg_s16
   return vqneg_s16(a);
   // CHECK: sqneg v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vqnegq_s16(int16x8_t a) {
-  // CHECK: test_vqnegq_s16
+  // CHECK-LABEL: test_vqnegq_s16
   return vqnegq_s16(a);
   // CHECK: sqneg v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vqneg_s32(int32x2_t a) {
-  // CHECK: test_vqneg_s32
+  // CHECK-LABEL: test_vqneg_s32
   return vqneg_s32(a);
   // CHECK: sqneg v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vqnegq_s32(int32x4_t a) {
-  // CHECK: test_vqnegq_s32
+  // CHECK-LABEL: test_vqnegq_s32
   return vqnegq_s32(a);
   // CHECK: sqneg v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vqnegq_s64(int64x2_t a) {
-  // CHECK: test_vqnegq_s64
+  // CHECK-LABEL: test_vqnegq_s64
   return vqnegq_s64(a);
   // CHECK: sqneg v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vneg_s8(int8x8_t a) {
-  // CHECK: test_vneg_s8
+  // CHECK-LABEL: test_vneg_s8
   return vneg_s8(a);
   // CHECK: neg v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vnegq_s8(int8x16_t a) {
-  // CHECK: test_vnegq_s8
+  // CHECK-LABEL: test_vnegq_s8
   return vnegq_s8(a);
   // CHECK: neg v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vneg_s16(int16x4_t a) {
-  // CHECK: test_vneg_s16
+  // CHECK-LABEL: test_vneg_s16
   return vneg_s16(a);
   // CHECK: neg v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vnegq_s16(int16x8_t a) {
-  // CHECK: test_vnegq_s16
+  // CHECK-LABEL: test_vnegq_s16
   return vnegq_s16(a);
   // CHECK: neg v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vneg_s32(int32x2_t a) {
-  // CHECK: test_vneg_s32
+  // CHECK-LABEL: test_vneg_s32
   return vneg_s32(a);
   // CHECK: neg v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vnegq_s32(int32x4_t a) {
-  // CHECK: test_vnegq_s32
+  // CHECK-LABEL: test_vnegq_s32
   return vnegq_s32(a);
   // CHECK: neg v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vnegq_s64(int64x2_t a) {
-  // CHECK: test_vnegq_s64
+  // CHECK-LABEL: test_vnegq_s64
   return vnegq_s64(a);
   // CHECK: neg v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vneg_f32(float32x2_t a) {
-  // CHECK: test_vneg_f32
+  // CHECK-LABEL: test_vneg_f32
   return vneg_f32(a);
   // CHECK: fneg v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vnegq_f32(float32x4_t a) {
-  // CHECK: test_vnegq_f32
+  // CHECK-LABEL: test_vnegq_f32
   return vnegq_f32(a);
   // CHECK: fneg v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vnegq_f64(float64x2_t a) {
-  // CHECK: test_vnegq_f64
+  // CHECK-LABEL: test_vnegq_f64
   return vnegq_f64(a);
   // CHECK: fneg v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vabs_s8(int8x8_t a) {
-  // CHECK: test_vabs_s8
+  // CHECK-LABEL: test_vabs_s8
   return vabs_s8(a);
   // CHECK: abs v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vabsq_s8(int8x16_t a) {
-  // CHECK: test_vabsq_s8
+  // CHECK-LABEL: test_vabsq_s8
   return vabsq_s8(a);
   // CHECK: abs v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vabs_s16(int16x4_t a) {
-  // CHECK: test_vabs_s16
+  // CHECK-LABEL: test_vabs_s16
   return vabs_s16(a);
   // CHECK: abs v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vabsq_s16(int16x8_t a) {
-  // CHECK: test_vabsq_s16
+  // CHECK-LABEL: test_vabsq_s16
   return vabsq_s16(a);
   // CHECK: abs v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vabs_s32(int32x2_t a) {
-  // CHECK: test_vabs_s32
+  // CHECK-LABEL: test_vabs_s32
   return vabs_s32(a);
   // CHECK: abs v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vabsq_s32(int32x4_t a) {
-  // CHECK: test_vabsq_s32
+  // CHECK-LABEL: test_vabsq_s32
   return vabsq_s32(a);
   // CHECK: abs v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vabsq_s64(int64x2_t a) {
-  // CHECK: test_vabsq_s64
+  // CHECK-LABEL: test_vabsq_s64
   return vabsq_s64(a);
   // CHECK: abs v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vabs_f32(float32x2_t a) {
-  // CHECK: test_vabs_f32
+  // CHECK-LABEL: test_vabs_f32
   return vabs_f32(a);
   // CHECK: fabs v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vabsq_f32(float32x4_t a) {
-  // CHECK: test_vabsq_f32
+  // CHECK-LABEL: test_vabsq_f32
   return vabsq_f32(a);
   // CHECK: fabs v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vabsq_f64(float64x2_t a) {
-  // CHECK: test_vabsq_f64
+  // CHECK-LABEL: test_vabsq_f64
   return vabsq_f64(a);
   // CHECK: fabs v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vuqadd_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vuqadd_s8
+  // CHECK-LABEL: test_vuqadd_s8
   return vuqadd_s8(a, b);
   // CHECK: suqadd v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vuqaddq_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vuqaddq_s8
+  // CHECK-LABEL: test_vuqaddq_s8
   return vuqaddq_s8(a, b);
   // CHECK: suqadd v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vuqadd_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vuqadd_s16
+  // CHECK-LABEL: test_vuqadd_s16
   return vuqadd_s16(a, b);
   // CHECK: suqadd v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vuqaddq_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vuqaddq_s16
+  // CHECK-LABEL: test_vuqaddq_s16
   return vuqaddq_s16(a, b);
   // CHECK: suqadd v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vuqadd_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vuqadd_s32
+  // CHECK-LABEL: test_vuqadd_s32
   return vuqadd_s32(a, b);
   // CHECK: suqadd v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vuqaddq_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vuqaddq_s32
+  // CHECK-LABEL: test_vuqaddq_s32
   return vuqaddq_s32(a, b);
   // CHECK: suqadd v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vuqaddq_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vuqaddq_s64
+  // CHECK-LABEL: test_vuqaddq_s64
   return vuqaddq_s64(a, b);
   // CHECK: suqadd v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vcls_s8(int8x8_t a) {
-  // CHECK: test_vcls_s8
+  // CHECK-LABEL: test_vcls_s8
   return vcls_s8(a);
   // CHECK: cls v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vclsq_s8(int8x16_t a) {
-  // CHECK: test_vclsq_s8
+  // CHECK-LABEL: test_vclsq_s8
   return vclsq_s8(a);
   // CHECK: cls v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vcls_s16(int16x4_t a) {
-  // CHECK: test_vcls_s16
+  // CHECK-LABEL: test_vcls_s16
   return vcls_s16(a);
   // CHECK: cls v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vclsq_s16(int16x8_t a) {
-  // CHECK: test_vclsq_s16
+  // CHECK-LABEL: test_vclsq_s16
   return vclsq_s16(a);
   // CHECK: cls v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vcls_s32(int32x2_t a) {
-  // CHECK: test_vcls_s32
+  // CHECK-LABEL: test_vcls_s32
   return vcls_s32(a);
   // CHECK: cls v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vclsq_s32(int32x4_t a) {
-  // CHECK: test_vclsq_s32
+  // CHECK-LABEL: test_vclsq_s32
   return vclsq_s32(a);
   // CHECK: cls v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int8x8_t test_vclz_s8(int8x8_t a) {
-  // CHECK: test_vclz_s8
+  // CHECK-LABEL: test_vclz_s8
   return vclz_s8(a);
   // CHECK: clz v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vclzq_s8(int8x16_t a) {
-  // CHECK: test_vclzq_s8
+  // CHECK-LABEL: test_vclzq_s8
   return vclzq_s8(a);
   // CHECK: clz v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vclz_s16(int16x4_t a) {
-  // CHECK: test_vclz_s16
+  // CHECK-LABEL: test_vclz_s16
   return vclz_s16(a);
   // CHECK: clz v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 int16x8_t test_vclzq_s16(int16x8_t a) {
-  // CHECK: test_vclzq_s16
+  // CHECK-LABEL: test_vclzq_s16
   return vclzq_s16(a);
   // CHECK: clz v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 int32x2_t test_vclz_s32(int32x2_t a) {
-  // CHECK: test_vclz_s32
+  // CHECK-LABEL: test_vclz_s32
   return vclz_s32(a);
   // CHECK: clz v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vclzq_s32(int32x4_t a) {
-  // CHECK: test_vclzq_s32
+  // CHECK-LABEL: test_vclzq_s32
   return vclzq_s32(a);
   // CHECK: clz v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint8x8_t test_vclz_u8(uint8x8_t a) {
-  // CHECK: test_vclz_u8
+  // CHECK-LABEL: test_vclz_u8
   return vclz_u8(a);
   // CHECK: clz v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint8x16_t test_vclzq_u8(uint8x16_t a) {
-  // CHECK: test_vclzq_u8
+  // CHECK-LABEL: test_vclzq_u8
   return vclzq_u8(a);
   // CHECK: clz v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint16x4_t test_vclz_u16(uint16x4_t a) {
-  // CHECK: test_vclz_u16
+  // CHECK-LABEL: test_vclz_u16
   return vclz_u16(a);
   // CHECK: clz v{{[0-9]+}}.4h, v{{[0-9]+}}.4h
 }
 
 uint16x8_t test_vclzq_u16(uint16x8_t a) {
-  // CHECK: test_vclzq_u16
+  // CHECK-LABEL: test_vclzq_u16
   return vclzq_u16(a);
   // CHECK: clz v{{[0-9]+}}.8h, v{{[0-9]+}}.8h
 }
 
 uint32x2_t test_vclz_u32(uint32x2_t a) {
-  // CHECK: test_vclz_u32
+  // CHECK-LABEL: test_vclz_u32
   return vclz_u32(a);
   // CHECK: clz v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vclzq_u32(uint32x4_t a) {
-  // CHECK: test_vclzq_u32
+  // CHECK-LABEL: test_vclzq_u32
   return vclzq_u32(a);
   // CHECK: clz v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int8x8_t test_vcnt_s8(int8x8_t a) {
-  // CHECK: test_vcnt_s8
+  // CHECK-LABEL: test_vcnt_s8
   return vcnt_s8(a);
   // CHECK: cnt v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vcntq_s8(int8x16_t a) {
-  // CHECK: test_vcntq_s8
+  // CHECK-LABEL: test_vcntq_s8
   return vcntq_s8(a);
   // CHECK: cnt v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint8x8_t test_vcnt_u8(uint8x8_t a) {
-  // CHECK: test_vcnt_u8
+  // CHECK-LABEL: test_vcnt_u8
   return vcnt_u8(a);
   // CHECK: cnt v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint8x16_t test_vcntq_u8(uint8x16_t a) {
-  // CHECK: test_vcntq_u8
+  // CHECK-LABEL: test_vcntq_u8
   return vcntq_u8(a);
   // CHECK: cnt v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 poly8x8_t test_vcnt_p8(poly8x8_t a) {
-  // CHECK: test_vcnt_p8
+  // CHECK-LABEL: test_vcnt_p8
   return vcnt_p8(a);
   // CHECK: cnt v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 poly8x16_t test_vcntq_p8(poly8x16_t a) {
-  // CHECK: test_vcntq_p8
+  // CHECK-LABEL: test_vcntq_p8
   return vcntq_p8(a);
   // CHECK: cnt v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int8x8_t test_vmvn_s8(int8x8_t a) {
-  // CHECK: test_vmvn_s8
+  // CHECK-LABEL: test_vmvn_s8
   return vmvn_s8(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vmvnq_s8(int8x16_t a) {
-  // CHECK: test_vmvnq_s8
+  // CHECK-LABEL: test_vmvnq_s8
   return vmvnq_s8(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int16x4_t test_vmvn_s16(int16x4_t a) {
-  // CHECK: test_vmvn_s16
+  // CHECK-LABEL: test_vmvn_s16
   return vmvn_s16(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int16x8_t test_vmvnq_s16(int16x8_t a) {
-  // CHECK: test_vmvnq_s16
+  // CHECK-LABEL: test_vmvnq_s16
   return vmvnq_s16(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int32x2_t test_vmvn_s32(int32x2_t a) {
-  // CHECK: test_vmvn_s32
+  // CHECK-LABEL: test_vmvn_s32
   return vmvn_s32(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int32x4_t test_vmvnq_s32(int32x4_t a) {
-  // CHECK: test_vmvnq_s32
+  // CHECK-LABEL: test_vmvnq_s32
   return vmvnq_s32(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint8x8_t test_vmvn_u8(uint8x8_t a) {
-  // CHECK: test_vmvn_u8
+  // CHECK-LABEL: test_vmvn_u8
   return vmvn_u8(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint8x16_t test_vmvnq_u8(uint8x16_t a) {
-  // CHECK: test_vmvnq_u8
+  // CHECK-LABEL: test_vmvnq_u8
   return vmvnq_u8(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint16x4_t test_vmvn_u16(uint16x4_t a) {
-  // CHECK: test_vmvn_u16
+  // CHECK-LABEL: test_vmvn_u16
   return vmvn_u16(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint16x8_t test_vmvnq_u16(uint16x8_t a) {
-  // CHECK: test_vmvnq_u16
+  // CHECK-LABEL: test_vmvnq_u16
   return vmvnq_u16(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint32x2_t test_vmvn_u32(uint32x2_t a) {
-  // CHECK: test_vmvn_u32
+  // CHECK-LABEL: test_vmvn_u32
   return vmvn_u32(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint32x4_t test_vmvnq_u32(uint32x4_t a) {
-  // CHECK: test_vmvnq_u32
+  // CHECK-LABEL: test_vmvnq_u32
   return vmvnq_u32(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 poly8x8_t test_vmvn_p8(poly8x8_t a) {
-  // CHECK: test_vmvn_p8
+  // CHECK-LABEL: test_vmvn_p8
   return vmvn_p8(a);
   // CHECK: not v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 poly8x16_t test_vmvnq_p8(poly8x16_t a) {
-  // CHECK: test_vmvnq_p8
+  // CHECK-LABEL: test_vmvnq_p8
   return vmvnq_p8(a);
   // CHECK: not v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int8x8_t test_vrbit_s8(int8x8_t a) {
-  // CHECK: test_vrbit_s8
+  // CHECK-LABEL: test_vrbit_s8
   return vrbit_s8(a);
   // CHECK: rbit v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 int8x16_t test_vrbitq_s8(int8x16_t a) {
-  // CHECK: test_vrbitq_s8
+  // CHECK-LABEL: test_vrbitq_s8
   return vrbitq_s8(a);
   // CHECK: rbit v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 uint8x8_t test_vrbit_u8(uint8x8_t a) {
-  // CHECK: test_vrbit_u8
+  // CHECK-LABEL: test_vrbit_u8
   return vrbit_u8(a);
   // CHECK: rbit v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 uint8x16_t test_vrbitq_u8(uint8x16_t a) {
-  // CHECK: test_vrbitq_u8
+  // CHECK-LABEL: test_vrbitq_u8
   return vrbitq_u8(a);
   // CHECK: rbit v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 poly8x8_t test_vrbit_p8(poly8x8_t a) {
-  // CHECK: test_vrbit_p8
+  // CHECK-LABEL: test_vrbit_p8
   return vrbit_p8(a);
   // CHECK: rbit v{{[0-9]+}}.8b, v{{[0-9]+}}.8b
 }
 
 poly8x16_t test_vrbitq_p8(poly8x16_t a) {
-  // CHECK: test_vrbitq_p8
+  // CHECK-LABEL: test_vrbitq_p8
   return vrbitq_p8(a);
   // CHECK: rbit v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
 }
 
 int8x8_t test_vmovn_s16(int16x8_t a) {
-  // CHECK: test_vmovn_s16
+  // CHECK-LABEL: test_vmovn_s16
   return vmovn_s16(a);
   // CHECK: xtn v{{[0-9]+}}.8b, v{{[0-9]+}}.8h
 }
 
 int16x4_t test_vmovn_s32(int32x4_t a) {
-  // CHECK: test_vmovn_s32
+  // CHECK-LABEL: test_vmovn_s32
   return vmovn_s32(a);
   // CHECK: xtn v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 int32x2_t test_vmovn_s64(int64x2_t a) {
-  // CHECK: test_vmovn_s64
+  // CHECK-LABEL: test_vmovn_s64
   return vmovn_s64(a);
   // CHECK: xtn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 uint8x8_t test_vmovn_u16(uint16x8_t a) {
-  // CHECK: test_vmovn_u16
+  // CHECK-LABEL: test_vmovn_u16
   return vmovn_u16(a);
   // CHECK: xtn v{{[0-9]+}}.8b, v{{[0-9]+}}.8h
 }
 
 uint16x4_t test_vmovn_u32(uint32x4_t a) {
-  // CHECK: test_vmovn_u32
+  // CHECK-LABEL: test_vmovn_u32
   return vmovn_u32(a);
   // CHECK: xtn v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 uint32x2_t test_vmovn_u64(uint64x2_t a) {
-  // CHECK: test_vmovn_u64
+  // CHECK-LABEL: test_vmovn_u64
   return vmovn_u64(a);
   // CHECK: xtn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 int8x16_t test_vmovn_high_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vmovn_high_s16
+  // CHECK-LABEL: test_vmovn_high_s16
   return vmovn_high_s16(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.16b, v{{[0-9]+}}.8h
 }
 
 int16x8_t test_vmovn_high_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vmovn_high_s32
+  // CHECK-LABEL: test_vmovn_high_s32
   return vmovn_high_s32(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 int32x4_t test_vmovn_high_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vmovn_high_s64
+  // CHECK-LABEL: test_vmovn_high_s64
   return vmovn_high_s64(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 int8x16_t test_vmovn_high_u16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vmovn_high_u16
+  // CHECK-LABEL: test_vmovn_high_u16
   return vmovn_high_u16(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.16b, v{{[0-9]+}}.8h
 }
 
 int16x8_t test_vmovn_high_u32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vmovn_high_u32
+  // CHECK-LABEL: test_vmovn_high_u32
   return vmovn_high_u32(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 int32x4_t test_vmovn_high_u64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vmovn_high_u64
+  // CHECK-LABEL: test_vmovn_high_u64
   return vmovn_high_u64(a, b);
   // CHECK: xtn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vqmovun_s16(int16x8_t a) {
-  // CHECK: test_vqmovun_s16
+  // CHECK-LABEL: test_vqmovun_s16
   return vqmovun_s16(a);
   // CHECK: sqxtun v{{[0-9]+}}.8b, v{{[0-9]+}}.8h
 }
 
 int16x4_t test_vqmovun_s32(int32x4_t a) {
-  // CHECK: test_vqmovun_s32
+  // CHECK-LABEL: test_vqmovun_s32
   return vqmovun_s32(a);
   // CHECK: sqxtun v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 int32x2_t test_vqmovun_s64(int64x2_t a) {
-  // CHECK: test_vqmovun_s64
+  // CHECK-LABEL: test_vqmovun_s64
   return vqmovun_s64(a);
   // CHECK: sqxtun v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 int8x16_t test_vqmovun_high_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqmovun_high_s16
+  // CHECK-LABEL: test_vqmovun_high_s16
   return vqmovun_high_s16(a, b);
   // CHECK: sqxtun2 v{{[0-9]+}}.16b, v{{[0-9]+}}.8h
 }
 
 int16x8_t test_vqmovun_high_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqmovun_high_s32
+  // CHECK-LABEL: test_vqmovun_high_s32
   return vqmovun_high_s32(a, b);
   // CHECK: sqxtun2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 int32x4_t test_vqmovun_high_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqmovun_high_s64
+  // CHECK-LABEL: test_vqmovun_high_s64
   return vqmovun_high_s64(a, b);
   // CHECK: sqxtun2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 int8x8_t test_vqmovn_s16(int16x8_t a) {
-  // CHECK: test_vqmovn_s16
+  // CHECK-LABEL: test_vqmovn_s16
   return vqmovn_s16(a);
   // CHECK: sqxtn v{{[0-9]+}}.8b, v{{[0-9]+}}.8h
 }
 
 int16x4_t test_vqmovn_s32(int32x4_t a) {
-  // CHECK: test_vqmovn_s32
+  // CHECK-LABEL: test_vqmovn_s32
   return vqmovn_s32(a);
   // CHECK: sqxtn v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 int32x2_t test_vqmovn_s64(int64x2_t a) {
-  // CHECK: test_vqmovn_s64
+  // CHECK-LABEL: test_vqmovn_s64
   return vqmovn_s64(a);
   // CHECK: sqxtn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 int8x16_t test_vqmovn_high_s16(int8x8_t a, int16x8_t b) {
-  // CHECK: test_vqmovn_high_s16
+  // CHECK-LABEL: test_vqmovn_high_s16
   return vqmovn_high_s16(a, b);
   // CHECK: sqxtn2 v{{[0-9]+}}.16b, v{{[0-9]+}}.8h
 }
 
 int16x8_t test_vqmovn_high_s32(int16x4_t a, int32x4_t b) {
-  // CHECK: test_vqmovn_high_s32
+  // CHECK-LABEL: test_vqmovn_high_s32
   return vqmovn_high_s32(a, b);
   // CHECK: sqxtn2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 int32x4_t test_vqmovn_high_s64(int32x2_t a, int64x2_t b) {
-  // CHECK: test_vqmovn_high_s64
+  // CHECK-LABEL: test_vqmovn_high_s64
   return vqmovn_high_s64(a, b);
   // CHECK: sqxtn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 uint8x8_t test_vqmovn_u16(uint16x8_t a) {
-  // CHECK: test_vqmovn_u16
+  // CHECK-LABEL: test_vqmovn_u16
   return vqmovn_u16(a);
   // CHECK: uqxtn v{{[0-9]+}}.8b, v{{[0-9]+}}.8h
 }
 
 uint16x4_t test_vqmovn_u32(uint32x4_t a) {
-  // CHECK: test_vqmovn_u32
+  // CHECK-LABEL: test_vqmovn_u32
   return vqmovn_u32(a);
   // CHECK: uqxtn v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 uint32x2_t test_vqmovn_u64(uint64x2_t a) {
-  // CHECK: test_vqmovn_u64
+  // CHECK-LABEL: test_vqmovn_u64
   return vqmovn_u64(a);
   // CHECK: uqxtn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 uint8x16_t test_vqmovn_high_u16(uint8x8_t a, uint16x8_t b) {
-  // CHECK: test_vqmovn_high_u16
+  // CHECK-LABEL: test_vqmovn_high_u16
   return vqmovn_high_u16(a, b);
   // CHECK: uqxtn2 v{{[0-9]+}}.16b, v{{[0-9]+}}.8h
 }
 
 uint16x8_t test_vqmovn_high_u32(uint16x4_t a, uint32x4_t b) {
-  // CHECK: test_vqmovn_high_u32
+  // CHECK-LABEL: test_vqmovn_high_u32
   return vqmovn_high_u32(a, b);
   // CHECK: uqxtn2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 uint32x4_t test_vqmovn_high_u64(uint32x2_t a, uint64x2_t b) {
-  // CHECK: test_vqmovn_high_u64
+  // CHECK-LABEL: test_vqmovn_high_u64
   return vqmovn_high_u64(a, b);
   // CHECK: uqxtn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 int16x8_t test_vshll_n_s8(int8x8_t a) {
-  // CHECK: test_vshll_n_s8
+  // CHECK-LABEL: test_vshll_n_s8
   return vshll_n_s8(a, 8);
   // CHECK: shll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #8
 }
 
 int32x4_t test_vshll_n_s16(int16x4_t a) {
-  // CHECK: test_vshll_n_s16
+  // CHECK-LABEL: test_vshll_n_s16
   return vshll_n_s16(a, 16);
   // CHECK: shll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #16
 }
 
 int64x2_t test_vshll_n_s32(int32x2_t a) {
-  // CHECK: test_vshll_n_s32
+  // CHECK-LABEL: test_vshll_n_s32
   return vshll_n_s32(a, 32);
   // CHECK: shll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #32
 }
 
 uint16x8_t test_vshll_n_u8(uint8x8_t a) {
-  // CHECK: test_vshll_n_u8
+  // CHECK-LABEL: test_vshll_n_u8
   return vshll_n_u8(a, 8);
   // CHECK: shll {{v[0-9]+}}.8h, {{v[0-9]+}}.8b, #8
 }
 
 uint32x4_t test_vshll_n_u16(uint16x4_t a) {
-  // CHECK: test_vshll_n_u16
+  // CHECK-LABEL: test_vshll_n_u16
   return vshll_n_u16(a, 16);
   // CHECK: shll {{v[0-9]+}}.4s, {{v[0-9]+}}.4h, #16
 }
 
 uint64x2_t test_vshll_n_u32(uint32x2_t a) {
-  // CHECK: test_vshll_n_u32
+  // CHECK-LABEL: test_vshll_n_u32
   return vshll_n_u32(a, 32);
   // CHECK: shll {{v[0-9]+}}.2d, {{v[0-9]+}}.2s, #32
 }
 
 int16x8_t test_vshll_high_n_s8(int8x16_t a) {
-  // CHECK: test_vshll_high_n_s8
+  // CHECK-LABEL: test_vshll_high_n_s8
   return vshll_high_n_s8(a, 8);
   // CHECK: shll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #8
 }
 
 int32x4_t test_vshll_high_n_s16(int16x8_t a) {
-  // CHECK: test_vshll_high_n_s16
+  // CHECK-LABEL: test_vshll_high_n_s16
   return vshll_high_n_s16(a, 16);
   // CHECK: shll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #16
 }
 
 int64x2_t test_vshll_high_n_s32(int32x4_t a) {
-  // CHECK: test_vshll_high_n_s32
+  // CHECK-LABEL: test_vshll_high_n_s32
   return vshll_high_n_s32(a, 32);
   // CHECK: shll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #32
 }
 
 uint16x8_t test_vshll_high_n_u8(uint8x16_t a) {
-  // CHECK: test_vshll_high_n_u8
+  // CHECK-LABEL: test_vshll_high_n_u8
   return vshll_high_n_u8(a, 8);
   // CHECK: shll2 {{v[0-9]+}}.8h, {{v[0-9]+}}.16b, #8
 }
 
 uint32x4_t test_vshll_high_n_u16(uint16x8_t a) {
-  // CHECK: test_vshll_high_n_u16
+  // CHECK-LABEL: test_vshll_high_n_u16
   return vshll_high_n_u16(a, 16);
   // CHECK: shll2 {{v[0-9]+}}.4s, {{v[0-9]+}}.8h, #16
 }
 
 uint64x2_t test_vshll_high_n_u32(uint32x4_t a) {
-  // CHECK: test_vshll_high_n_u32
+  // CHECK-LABEL: test_vshll_high_n_u32
   return vshll_high_n_u32(a, 32);
   // CHECK: shll2 {{v[0-9]+}}.2d, {{v[0-9]+}}.4s, #32
 }
 
 float16x4_t test_vcvt_f16_f32(float32x4_t a) {
-  //CHECK: test_vcvt_f16_f32
+  //CHECK-LABEL: test_vcvt_f16_f32
   return vcvt_f16_f32(a);
   // CHECK: fcvtn v{{[0-9]+}}.4h, v{{[0-9]+}}.4s
 }
 
 float16x8_t test_vcvt_high_f16_f32(float16x4_t a, float32x4_t b) {
-  //CHECK: test_vcvt_high_f16_f32
+  //CHECK-LABEL: test_vcvt_high_f16_f32
   return vcvt_high_f16_f32(a, b);
   // CHECK: fcvtn2 v{{[0-9]+}}.8h, v{{[0-9]+}}.4s
 }
 
 float32x2_t test_vcvt_f32_f64(float64x2_t a) {
-  //CHECK: test_vcvt_f32_f64
+  //CHECK-LABEL: test_vcvt_f32_f64
   return vcvt_f32_f64(a);
   // CHECK: fcvtn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 float32x4_t test_vcvt_high_f32_f64(float32x2_t a, float64x2_t b) {
-  //CHECK: test_vcvt_high_f32_f64
+  //CHECK-LABEL: test_vcvt_high_f32_f64
   return vcvt_high_f32_f64(a, b);
   // CHECK: fcvtn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vcvtx_f32_f64(float64x2_t a) {
-  //CHECK: test_vcvtx_f32_f64
+  //CHECK-LABEL: test_vcvtx_f32_f64
   return vcvtx_f32_f64(a);
   // CHECK: fcvtxn v{{[0-9]+}}.2s, v{{[0-9]+}}.2d
 }
 
 float32x4_t test_vcvtx_high_f32_f64(float32x2_t a, float64x2_t b) {
-  //CHECK: test_vcvtx_high_f32_f64
+  //CHECK-LABEL: test_vcvtx_high_f32_f64
   return vcvtx_high_f32_f64(a, b);
   // CHECK: fcvtxn2 v{{[0-9]+}}.4s, v{{[0-9]+}}.2d
 }
 
 float32x4_t test_vcvt_f32_f16(float16x4_t a) {
-  //CHECK: test_vcvt_f32_f16
+  //CHECK-LABEL: test_vcvt_f32_f16
   return vcvt_f32_f16(a);
   // CHECK: fcvtl v{{[0-9]+}}.4s, v{{[0-9]+}}.4h
 }
 
 float32x4_t test_vcvt_high_f32_f16(float16x8_t a) {
-  //CHECK: test_vcvt_high_f32_f16
+  //CHECK-LABEL: test_vcvt_high_f32_f16
   return vcvt_high_f32_f16(a);
   // CHECK: fcvtl2 v{{[0-9]+}}.4s, v{{[0-9]+}}.8h
 }
 
 float64x2_t test_vcvt_f64_f32(float32x2_t a) {
-  //CHECK: test_vcvt_f64_f32
+  //CHECK-LABEL: test_vcvt_f64_f32
   return vcvt_f64_f32(a);
   // CHECK: fcvtl v{{[0-9]+}}.2d, v{{[0-9]+}}.2s
 }
 
 float64x2_t test_vcvt_high_f64_f32(float32x4_t a) {
-  //CHECK: test_vcvt_high_f64_f32
+  //CHECK-LABEL: test_vcvt_high_f64_f32
   return vcvt_high_f64_f32(a);
   // CHECK: fcvtl2 v{{[0-9]+}}.2d, v{{[0-9]+}}.4s
 }
 
 float32x2_t test_vrndn_f32(float32x2_t a) {
-  //CHECK: test_vrndn_f32
+  //CHECK-LABEL: test_vrndn_f32
   return vrndn_f32(a);
   // CHECK: frintn v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndnq_f32(float32x4_t a) {
-  //CHECK: test_vrndnq_f32
+  //CHECK-LABEL: test_vrndnq_f32
   return vrndnq_f32(a);
   // CHECK: frintn v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndnq_f64(float64x2_t a) {
-  //CHECK: test_vrndnq_f64
+  //CHECK-LABEL: test_vrndnq_f64
   return vrndnq_f64(a);
   // CHECK: frintn v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrnda_f32(float32x2_t a) {
-  //CHECK: test_vrnda_f32
+  //CHECK-LABEL: test_vrnda_f32
   return vrnda_f32(a);
   // CHECK: frinta v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndaq_f32(float32x4_t a) {
-  //CHECK: test_vrndaq_f32
+  //CHECK-LABEL: test_vrndaq_f32
   return vrndaq_f32(a);
   // CHECK: frinta v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndaq_f64(float64x2_t a) {
-  //CHECK: test_vrndaq_f64
+  //CHECK-LABEL: test_vrndaq_f64
   return vrndaq_f64(a);
   // CHECK: frinta v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrndp_f32(float32x2_t a) {
-  //CHECK: test_vrndp_f32
+  //CHECK-LABEL: test_vrndp_f32
   return vrndp_f32(a);
   // CHECK: frintp v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndpq_f32(float32x4_t a) {
-  //CHECK: test_vrndpq_f32
+  //CHECK-LABEL: test_vrndpq_f32
   return vrndpq_f32(a);
   // CHECK: frintp v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndpq_f64(float64x2_t a) {
-  //CHECK: test_vrndpq_f64
+  //CHECK-LABEL: test_vrndpq_f64
   return vrndpq_f64(a);
   // CHECK: frintp v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrndm_f32(float32x2_t a) {
-  //CHECK: test_vrndm_f32
+  //CHECK-LABEL: test_vrndm_f32
   return vrndm_f32(a);
   // CHECK: frintm v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndmq_f32(float32x4_t a) {
-  //CHECK: test_vrndmq_f32
+  //CHECK-LABEL: test_vrndmq_f32
   return vrndmq_f32(a);
   // CHECK: frintm v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndmq_f64(float64x2_t a) {
-  //CHECK: test_vrndmq_f64
+  //CHECK-LABEL: test_vrndmq_f64
   return vrndmq_f64(a);
   // CHECK: frintm v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrndx_f32(float32x2_t a) {
-  //CHECK: test_vrndx_f32
+  //CHECK-LABEL: test_vrndx_f32
   return vrndx_f32(a);
   // CHECK: frintx v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndxq_f32(float32x4_t a) {
-  //CHECK: test_vrndxq_f32
+  //CHECK-LABEL: test_vrndxq_f32
   return vrndxq_f32(a);
   // CHECK: frintx v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndxq_f64(float64x2_t a) {
-  //CHECK: test_vrndxq_f64
+  //CHECK-LABEL: test_vrndxq_f64
   return vrndxq_f64(a);
   // CHECK: frintx v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrnd_f32(float32x2_t a) {
-  //CHECK: test_vrnd_f32
+  //CHECK-LABEL: test_vrnd_f32
   return vrnd_f32(a);
   // CHECK: frintz v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndq_f32(float32x4_t a) {
-  //CHECK: test_vrndq_f32
+  //CHECK-LABEL: test_vrndq_f32
   return vrndq_f32(a);
   // CHECK: frintz v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndq_f64(float64x2_t a) {
-  //CHECK: test_vrndq_f64
+  //CHECK-LABEL: test_vrndq_f64
   return vrndq_f64(a);
   // CHECK: frintz v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrndi_f32(float32x2_t a) {
-  //CHECK: test_vrndi_f32
+  //CHECK-LABEL: test_vrndi_f32
   return vrndi_f32(a);
   // CHECK: frinti v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrndiq_f32(float32x4_t a) {
-  //CHECK: test_vrndiq_f32
+  //CHECK-LABEL: test_vrndiq_f32
   return vrndiq_f32(a);
   // CHECK: frinti v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrndiq_f64(float64x2_t a) {
-  //CHECK: test_vrndiq_f64
+  //CHECK-LABEL: test_vrndiq_f64
   return vrndiq_f64(a);
   // CHECK: frinti v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int32x2_t test_vcvt_s32_f32(float32x2_t a) {
-  //CHECK: test_vcvt_s32_f32
+  //CHECK-LABEL: test_vcvt_s32_f32
   return vcvt_s32_f32(a);
   // CHECK: fcvtzs v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vcvtq_s32_f32(float32x4_t a) {
-  //CHECK: test_vcvtq_s32_f32
+  //CHECK-LABEL: test_vcvtq_s32_f32
   return vcvtq_s32_f32(a);
   // CHECK: fcvtzs v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vcvtq_s64_f64(float64x2_t a) {
-  //CHECK: test_vcvtq_s64_f64
+  //CHECK-LABEL: test_vcvtq_s64_f64
   return vcvtq_s64_f64(a);
   // CHECK: fcvtzs v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vcvt_u32_f32(float32x2_t a) {
-  //CHECK: test_vcvt_u32_f32
+  //CHECK-LABEL: test_vcvt_u32_f32
   return vcvt_u32_f32(a);
   // CHECK: fcvtzu v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vcvtq_u32_f32(float32x4_t a) {
-  //CHECK: test_vcvtq_u32_f32
+  //CHECK-LABEL: test_vcvtq_u32_f32
   return vcvtq_u32_f32(a);
   // CHECK: fcvtzu v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint64x2_t test_vcvtq_u64_f64(float64x2_t a) {
-  //CHECK: test_vcvtq_u64_f64
+  //CHECK-LABEL: test_vcvtq_u64_f64
   return vcvtq_u64_f64(a);
   // CHECK: fcvtzu v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int32x2_t test_vcvtn_s32_f32(float32x2_t a) {
-  //CHECK: test_vcvtn_s32_f32
+  //CHECK-LABEL: test_vcvtn_s32_f32
   return vcvtn_s32_f32(a);
   // CHECK: fcvtns v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vcvtnq_s32_f32(float32x4_t a) {
-  //CHECK: test_vcvtnq_s32_f32
+  //CHECK-LABEL: test_vcvtnq_s32_f32
   return vcvtnq_s32_f32(a);
   // CHECK: fcvtns v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vcvtnq_s64_f64(float64x2_t a) {
-  //CHECK: test_vcvtnq_s64_f64
+  //CHECK-LABEL: test_vcvtnq_s64_f64
   return vcvtnq_s64_f64(a);
   // CHECK: fcvtns v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vcvtn_u32_f32(float32x2_t a) {
-  //CHECK: test_vcvtn_u32_f32
+  //CHECK-LABEL: test_vcvtn_u32_f32
   return vcvtn_u32_f32(a);
   // CHECK: fcvtnu v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vcvtnq_u32_f32(float32x4_t a) {
-  //CHECK: test_vcvtnq_u32_f32
+  //CHECK-LABEL: test_vcvtnq_u32_f32
   return vcvtnq_u32_f32(a);
   // CHECK: fcvtnu v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint64x2_t test_vcvtnq_u64_f64(float64x2_t a) {
-  //CHECK: test_vcvtnq_u64_f64
+  //CHECK-LABEL: test_vcvtnq_u64_f64
   return vcvtnq_u64_f64(a);
   // CHECK: fcvtnu v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int32x2_t test_vcvtp_s32_f32(float32x2_t a) {
-  //CHECK: test_vcvtp_s32_f32
+  //CHECK-LABEL: test_vcvtp_s32_f32
   return vcvtp_s32_f32(a);
   // CHECK: fcvtps v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vcvtpq_s32_f32(float32x4_t a) {
-  //CHECK: test_vcvtpq_s32_f32
+  //CHECK-LABEL: test_vcvtpq_s32_f32
   return vcvtpq_s32_f32(a);
   // CHECK: fcvtps v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vcvtpq_s64_f64(float64x2_t a) {
-  //CHECK: test_vcvtpq_s64_f64
+  //CHECK-LABEL: test_vcvtpq_s64_f64
   return vcvtpq_s64_f64(a);
   // CHECK: fcvtps v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vcvtp_u32_f32(float32x2_t a) {
-  //CHECK: test_vcvtp_u32_f32
+  //CHECK-LABEL: test_vcvtp_u32_f32
   return vcvtp_u32_f32(a);
   // CHECK: fcvtpu v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vcvtpq_u32_f32(float32x4_t a) {
-  //CHECK: test_vcvtpq_u32_f32
+  //CHECK-LABEL: test_vcvtpq_u32_f32
   return vcvtpq_u32_f32(a);
   // CHECK: fcvtpu v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint64x2_t test_vcvtpq_u64_f64(float64x2_t a) {
-  //CHECK: test_vcvtpq_u64_f64
+  //CHECK-LABEL: test_vcvtpq_u64_f64
   return vcvtpq_u64_f64(a);
   // CHECK: fcvtpu v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int32x2_t test_vcvtm_s32_f32(float32x2_t a) {
-  //CHECK: test_vcvtm_s32_f32
+  //CHECK-LABEL: test_vcvtm_s32_f32
   return vcvtm_s32_f32(a);
   // CHECK: fcvtms v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vcvtmq_s32_f32(float32x4_t a) {
-  //CHECK: test_vcvtmq_s32_f32
+  //CHECK-LABEL: test_vcvtmq_s32_f32
   return vcvtmq_s32_f32(a);
   // CHECK: fcvtms v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vcvtmq_s64_f64(float64x2_t a) {
-  //CHECK: test_vcvtmq_s64_f64
+  //CHECK-LABEL: test_vcvtmq_s64_f64
   return vcvtmq_s64_f64(a);
   // CHECK: fcvtms v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vcvtm_u32_f32(float32x2_t a) {
-  //CHECK: test_vcvtm_u32_f32
+  //CHECK-LABEL: test_vcvtm_u32_f32
   return vcvtm_u32_f32(a);
   // CHECK: fcvtmu v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vcvtmq_u32_f32(float32x4_t a) {
-  //CHECK: test_vcvtmq_u32_f32
+  //CHECK-LABEL: test_vcvtmq_u32_f32
   return vcvtmq_u32_f32(a);
   // CHECK: fcvtmu v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint64x2_t test_vcvtmq_u64_f64(float64x2_t a) {
-  //CHECK: test_vcvtmq_u64_f64
+  //CHECK-LABEL: test_vcvtmq_u64_f64
   return vcvtmq_u64_f64(a);
   // CHECK: fcvtmu v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 int32x2_t test_vcvta_s32_f32(float32x2_t a) {
-  //CHECK: test_vcvta_s32_f32
+  //CHECK-LABEL: test_vcvta_s32_f32
   return vcvta_s32_f32(a);
   // CHECK: fcvtas v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 int32x4_t test_vcvtaq_s32_f32(float32x4_t a) {
-  //CHECK: test_vcvtaq_s32_f32
+  //CHECK-LABEL: test_vcvtaq_s32_f32
   return vcvtaq_s32_f32(a);
   // CHECK: fcvtas v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 int64x2_t test_vcvtaq_s64_f64(float64x2_t a) {
-  //CHECK: test_vcvtaq_s64_f64
+  //CHECK-LABEL: test_vcvtaq_s64_f64
   return vcvtaq_s64_f64(a);
   // CHECK: fcvtas v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vcvta_u32_f32(float32x2_t a) {
-  //CHECK: test_vcvta_u32_f32
+  //CHECK-LABEL: test_vcvta_u32_f32
   return vcvta_u32_f32(a);
   // CHECK: fcvtau v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vcvtaq_u32_f32(float32x4_t a) {
-  //CHECK: test_vcvtaq_u32_f32
+  //CHECK-LABEL: test_vcvtaq_u32_f32
   return vcvtaq_u32_f32(a);
   // CHECK: fcvtau v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 uint64x2_t test_vcvtaq_u64_f64(float64x2_t a) {
-  //CHECK: test_vcvtaq_u64_f64
+  //CHECK-LABEL: test_vcvtaq_u64_f64
   return vcvtaq_u64_f64(a);
   // CHECK: fcvtau v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrsqrte_f32(float32x2_t a) {
-  //CHECK: test_vrsqrte_f32
+  //CHECK-LABEL: test_vrsqrte_f32
   return vrsqrte_f32(a);
   // CHECK: frsqrte v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrsqrteq_f32(float32x4_t a) {
-  //CHECK: test_vrsqrteq_f32
+  //CHECK-LABEL: test_vrsqrteq_f32
   return vrsqrteq_f32(a);
   // CHECK: frsqrte v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrsqrteq_f64(float64x2_t a) {
-  //CHECK: test_vrsqrteq_f64
+  //CHECK-LABEL: test_vrsqrteq_f64
   return vrsqrteq_f64(a);
   // CHECK: frsqrte v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vrecpe_f32(float32x2_t a) {
-  //CHECK: test_vrecpe_f32
+  //CHECK-LABEL: test_vrecpe_f32
   return vrecpe_f32(a);
   // CHECK: frecpe v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vrecpeq_f32(float32x4_t a) {
-  //CHECK: test_vrecpeq_f32
+  //CHECK-LABEL: test_vrecpeq_f32
   return vrecpeq_f32(a);
   // CHECK: frecpe v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vrecpeq_f64(float64x2_t a) {
-  //CHECK: test_vrecpeq_f64
+  //CHECK-LABEL: test_vrecpeq_f64
   return vrecpeq_f64(a);
   // CHECK: frecpe v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 uint32x2_t test_vrecpe_u32(uint32x2_t a) {
-  //CHECK: test_vrecpe_u32
+  //CHECK-LABEL: test_vrecpe_u32
   return vrecpe_u32(a);
   // CHECK: urecpe v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 uint32x4_t test_vrecpeq_u32(uint32x4_t a) {
-  //CHECK: test_vrecpeq_u32
+  //CHECK-LABEL: test_vrecpeq_u32
   return vrecpeq_u32(a);
   // CHECK: urecpe v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float32x2_t test_vsqrt_f32(float32x2_t a) {
-  //CHECK: test_vsqrt_f32
+  //CHECK-LABEL: test_vsqrt_f32
   return vsqrt_f32(a);
   // CHECK: fsqrt v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vsqrtq_f32(float32x4_t a) {
-  //CHECK: test_vsqrtq_f32
+  //CHECK-LABEL: test_vsqrtq_f32
   return vsqrtq_f32(a);
   // CHECK: fsqrt v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vsqrtq_f64(float64x2_t a) {
-  //CHECK: test_vsqrtq_f64
+  //CHECK-LABEL: test_vsqrtq_f64
   return vsqrtq_f64(a);
   // CHECK: fsqrt v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float32x2_t test_vcvt_f32_s32(int32x2_t a) {
-  //CHECK: test_vcvt_f32_s32
+  //CHECK-LABEL: test_vcvt_f32_s32
   return vcvt_f32_s32(a);
   //CHECK: scvtf v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x2_t test_vcvt_f32_u32(uint32x2_t a) {
-  //CHECK: test_vcvt_f32_u32
+  //CHECK-LABEL: test_vcvt_f32_u32
   return vcvt_f32_u32(a);
   //CHECK: ucvtf v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
 }
 
 float32x4_t test_vcvtq_f32_s32(int32x4_t a) {
-  //CHECK: test_vcvtq_f32_s32
+  //CHECK-LABEL: test_vcvtq_f32_s32
   return vcvtq_f32_s32(a);
   //CHECK: scvtf v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float32x4_t test_vcvtq_f32_u32(uint32x4_t a) {
-  //CHECK: test_vcvtq_f32_u32
+  //CHECK-LABEL: test_vcvtq_f32_u32
   return vcvtq_f32_u32(a);
   //CHECK: ucvtf v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
 }
 
 float64x2_t test_vcvtq_f64_s64(int64x2_t a) {
-  //CHECK: test_vcvtq_f64_s64
+  //CHECK-LABEL: test_vcvtq_f64_s64
   return vcvtq_f64_s64(a);
   //CHECK: scvtf v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
 
 float64x2_t test_vcvtq_f64_u64(uint64x2_t a) {
-  //CHECK: test_vcvtq_f64_u64
+  //CHECK-LABEL: test_vcvtq_f64_u64
   return vcvtq_f64_u64(a);
   //CHECK: ucvtf v{{[0-9]+}}.2d, v{{[0-9]+}}.2d
 }
diff --git a/test/CodeGen/aarch64-neon-perm.c b/test/CodeGen/aarch64-neon-perm.c
index 903570b..c9a3d60 100644
--- a/test/CodeGen/aarch64-neon-perm.c
+++ b/test/CodeGen/aarch64-neon-perm.c
@@ -1,1092 +1,1094 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
-
 #include <arm_neon.h>
 
 int8x8_t test_vuzp1_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vuzp1_s8
+  // CHECK-LABEL: test_vuzp1_s8
   return vuzp1_s8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vuzp1q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vuzp1q_s8
+  // CHECK-LABEL: test_vuzp1q_s8
   return vuzp1q_s8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vuzp1_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vuzp1_s16
+  // CHECK-LABEL: test_vuzp1_s16
   return vuzp1_s16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vuzp1q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vuzp1q_s16
+  // CHECK-LABEL: test_vuzp1q_s16
   return vuzp1q_s16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vuzp1_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vuzp1_s32
+  // CHECK-LABEL: test_vuzp1_s32
   return vuzp1_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vuzp1q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vuzp1q_s32
+  // CHECK-LABEL: test_vuzp1q_s32
   return vuzp1q_s32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vuzp1q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vuzp1q_s64
+  // CHECK-LABEL: test_vuzp1q_s64
   return vuzp1q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vuzp1_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vuzp1_u8
+  // CHECK-LABEL: test_vuzp1_u8
   return vuzp1_u8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vuzp1q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vuzp1q_u8
+  // CHECK-LABEL: test_vuzp1q_u8
   return vuzp1q_u8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vuzp1_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vuzp1_u16
+  // CHECK-LABEL: test_vuzp1_u16
   return vuzp1_u16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vuzp1q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vuzp1q_u16
+  // CHECK-LABEL: test_vuzp1q_u16
   return vuzp1q_u16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vuzp1_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vuzp1_u32
+  // CHECK-LABEL: test_vuzp1_u32
   return vuzp1_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vuzp1q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vuzp1q_u32
+  // CHECK-LABEL: test_vuzp1q_u32
   return vuzp1q_u32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vuzp1q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vuzp1q_u64
+  // CHECK-LABEL: test_vuzp1q_u64
   return vuzp1q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vuzp1_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vuzp1_f32
+  // CHECK-LABEL: test_vuzp1_f32
   return vuzp1_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vuzp1q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vuzp1q_f32
+  // CHECK-LABEL: test_vuzp1q_f32
   return vuzp1q_f32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vuzp1q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vuzp1q_f64
+  // CHECK-LABEL: test_vuzp1q_f64
   return vuzp1q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vuzp1_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vuzp1_p8
+  // CHECK-LABEL: test_vuzp1_p8
   return vuzp1_p8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vuzp1q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vuzp1q_p8
+  // CHECK-LABEL: test_vuzp1q_p8
   return vuzp1q_p8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vuzp1_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vuzp1_p16
+  // CHECK-LABEL: test_vuzp1_p16
   return vuzp1_p16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vuzp1q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vuzp1q_p16
+  // CHECK-LABEL: test_vuzp1q_p16
   return vuzp1q_p16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8_t test_vuzp2_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vuzp2_s8
+  // CHECK-LABEL: test_vuzp2_s8
   return vuzp2_s8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vuzp2q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vuzp2q_s8
+  // CHECK-LABEL: test_vuzp2q_s8
   return vuzp2q_s8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vuzp2_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vuzp2_s16
+  // CHECK-LABEL: test_vuzp2_s16
   return vuzp2_s16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vuzp2q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vuzp2q_s16
+  // CHECK-LABEL: test_vuzp2q_s16
   return vuzp2q_s16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vuzp2_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vuzp2_s32
+  // CHECK-LABEL: test_vuzp2_s32
   return vuzp2_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vuzp2q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vuzp2q_s32
+  // CHECK-LABEL: test_vuzp2q_s32
   return vuzp2q_s32(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vuzp2q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vuzp2q_s64
+  // CHECK-LABEL: test_vuzp2q_s64
   return vuzp2q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vuzp2_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vuzp2_u8
+  // CHECK-LABEL: test_vuzp2_u8
   return vuzp2_u8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vuzp2q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vuzp2q_u8
+  // CHECK-LABEL: test_vuzp2q_u8
   return vuzp2q_u8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vuzp2_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vuzp2_u16
+  // CHECK-LABEL: test_vuzp2_u16
   return vuzp2_u16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vuzp2q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vuzp2q_u16
+  // CHECK-LABEL: test_vuzp2q_u16
   return vuzp2q_u16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vuzp2_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vuzp2_u32
+  // CHECK-LABEL: test_vuzp2_u32
   return vuzp2_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vuzp2q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vuzp2q_u32
+  // CHECK-LABEL: test_vuzp2q_u32
   return vuzp2q_u32(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vuzp2q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vuzp2q_u64
+  // CHECK-LABEL: test_vuzp2q_u64
   return vuzp2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vuzp2_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vuzp2_f32
+  // CHECK-LABEL: test_vuzp2_f32
   return vuzp2_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vuzp2q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vuzp2q_f32
+  // CHECK-LABEL: test_vuzp2q_f32
   return vuzp2q_f32(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vuzp2q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vuzp2q_f64
+  // CHECK-LABEL: test_vuzp2q_f64
   return vuzp2q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vuzp2_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vuzp2_p8
+  // CHECK-LABEL: test_vuzp2_p8
   return vuzp2_p8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vuzp2q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vuzp2q_p8
+  // CHECK-LABEL: test_vuzp2q_p8
   return vuzp2q_p8(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vuzp2_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vuzp2_p16
+  // CHECK-LABEL: test_vuzp2_p16
   return vuzp2_p16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vuzp2q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vuzp2q_p16
+  // CHECK-LABEL: test_vuzp2q_p16
   return vuzp2q_p16(a, b);
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8_t test_vzip1_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vzip1_s8
+  // CHECK-LABEL: test_vzip1_s8
   return vzip1_s8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vzip1q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vzip1q_s8
+  // CHECK-LABEL: test_vzip1q_s8
   return vzip1q_s8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vzip1_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vzip1_s16
+  // CHECK-LABEL: test_vzip1_s16
   return vzip1_s16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vzip1q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vzip1q_s16
+  // CHECK-LABEL: test_vzip1q_s16
   return vzip1q_s16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vzip1_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vzip1_s32
+  // CHECK-LABEL: test_vzip1_s32
   return vzip1_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vzip1q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vzip1q_s32
+  // CHECK-LABEL: test_vzip1q_s32
   return vzip1q_s32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vzip1q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vzip1q_s64
+  // CHECK-LABEL: test_vzip1q_s64
   return vzip1q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vzip1_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vzip1_u8
+  // CHECK-LABEL: test_vzip1_u8
   return vzip1_u8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vzip1q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vzip1q_u8
+  // CHECK-LABEL: test_vzip1q_u8
   return vzip1q_u8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vzip1_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vzip1_u16
+  // CHECK-LABEL: test_vzip1_u16
   return vzip1_u16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vzip1q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vzip1q_u16
+  // CHECK-LABEL: test_vzip1q_u16
   return vzip1q_u16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vzip1_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vzip1_u32
+  // CHECK-LABEL: test_vzip1_u32
   return vzip1_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vzip1q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vzip1q_u32
+  // CHECK-LABEL: test_vzip1q_u32
   return vzip1q_u32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vzip1q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vzip1q_u64
+  // CHECK-LABEL: test_vzip1q_u64
   return vzip1q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vzip1_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vzip1_f32
+  // CHECK-LABEL: test_vzip1_f32
   return vzip1_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vzip1q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vzip1q_f32
+  // CHECK-LABEL: test_vzip1q_f32
   return vzip1q_f32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vzip1q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vzip1q_f64
+  // CHECK-LABEL: test_vzip1q_f64
   return vzip1q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vzip1_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vzip1_p8
+  // CHECK-LABEL: test_vzip1_p8
   return vzip1_p8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vzip1q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vzip1q_p8
+  // CHECK-LABEL: test_vzip1q_p8
   return vzip1q_p8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vzip1_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vzip1_p16
+  // CHECK-LABEL: test_vzip1_p16
   return vzip1_p16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vzip1q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vzip1q_p16
+  // CHECK-LABEL: test_vzip1q_p16
   return vzip1q_p16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8_t test_vzip2_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vzip2_s8
+  // CHECK-LABEL: test_vzip2_s8
   return vzip2_s8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vzip2q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vzip2q_s8
+  // CHECK-LABEL: test_vzip2q_s8
   return vzip2q_s8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vzip2_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vzip2_s16
+  // CHECK-LABEL: test_vzip2_s16
   return vzip2_s16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vzip2q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vzip2q_s16
+  // CHECK-LABEL: test_vzip2q_s16
   return vzip2q_s16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vzip2_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vzip2_s32
+  // CHECK-LABEL: test_vzip2_s32
   return vzip2_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vzip2q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vzip2q_s32
+  // CHECK-LABEL: test_vzip2q_s32
   return vzip2q_s32(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vzip2q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vzip2q_s64
+  // CHECK-LABEL: test_vzip2q_s64
   return vzip2q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vzip2_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vzip2_u8
+  // CHECK-LABEL: test_vzip2_u8
   return vzip2_u8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vzip2q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vzip2q_u8
+  // CHECK-LABEL: test_vzip2q_u8
   return vzip2q_u8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vzip2_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vzip2_u16
+  // CHECK-LABEL: test_vzip2_u16
   return vzip2_u16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vzip2q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vzip2q_u16
+  // CHECK-LABEL: test_vzip2q_u16
   return vzip2q_u16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vzip2_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vzip2_u32
+  // CHECK-LABEL: test_vzip2_u32
   return vzip2_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vzip2q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vzip2q_u32
+  // CHECK-LABEL: test_vzip2q_u32
   return vzip2q_u32(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vzip2q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vzip2q_u64
+  // CHECK-LABEL: test_vzip2q_u64
   return vzip2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vzip2_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vzip2_f32
+  // CHECK-LABEL: test_vzip2_f32
   return vzip2_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vzip2q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vzip2q_f32
+  // CHECK-LABEL: test_vzip2q_f32
   return vzip2q_f32(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vzip2q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vzip2q_f64
+  // CHECK-LABEL: test_vzip2q_f64
   return vzip2q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vzip2_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vzip2_p8
+  // CHECK-LABEL: test_vzip2_p8
   return vzip2_p8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vzip2q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vzip2q_p8
+  // CHECK-LABEL: test_vzip2q_p8
   return vzip2q_p8(a, b);
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vzip2_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vzip2_p16
+  // CHECK-LABEL: test_vzip2_p16
   return vzip2_p16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vzip2q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vzip2q_p16
+  // CHECK-LABEL: test_vzip2q_p16
   return vzip2q_p16(a, b);
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8_t test_vtrn1_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vtrn1_s8
+  // CHECK-LABEL: test_vtrn1_s8
   return vtrn1_s8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vtrn1q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vtrn1q_s8
+  // CHECK-LABEL: test_vtrn1q_s8
   return vtrn1q_s8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vtrn1_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vtrn1_s16
+  // CHECK-LABEL: test_vtrn1_s16
   return vtrn1_s16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vtrn1q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vtrn1q_s16
+  // CHECK-LABEL: test_vtrn1q_s16
   return vtrn1q_s16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vtrn1_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vtrn1_s32
+  // CHECK-LABEL: test_vtrn1_s32
   return vtrn1_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vtrn1q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vtrn1q_s32
+  // CHECK-LABEL: test_vtrn1q_s32
   return vtrn1q_s32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vtrn1q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vtrn1q_s64
+  // CHECK-LABEL: test_vtrn1q_s64
   return vtrn1q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vtrn1_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vtrn1_u8
+  // CHECK-LABEL: test_vtrn1_u8
   return vtrn1_u8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vtrn1q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vtrn1q_u8
+  // CHECK-LABEL: test_vtrn1q_u8
   return vtrn1q_u8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vtrn1_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vtrn1_u16
+  // CHECK-LABEL: test_vtrn1_u16
   return vtrn1_u16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vtrn1q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vtrn1q_u16
+  // CHECK-LABEL: test_vtrn1q_u16
   return vtrn1q_u16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vtrn1_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vtrn1_u32
+  // CHECK-LABEL: test_vtrn1_u32
   return vtrn1_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vtrn1q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vtrn1q_u32
+  // CHECK-LABEL: test_vtrn1q_u32
   return vtrn1q_u32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vtrn1q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vtrn1q_u64
+  // CHECK-LABEL: test_vtrn1q_u64
   return vtrn1q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vtrn1_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vtrn1_f32
+  // CHECK-LABEL: test_vtrn1_f32
   return vtrn1_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vtrn1q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vtrn1q_f32
+  // CHECK-LABEL: test_vtrn1q_f32
   return vtrn1q_f32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vtrn1q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vtrn1q_f64
+  // CHECK-LABEL: test_vtrn1q_f64
   return vtrn1q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK: {{ins v[0-9]+.d\[1\], v[0-9]+.d\[0\]|zip1 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vtrn1_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vtrn1_p8
+  // CHECK-LABEL: test_vtrn1_p8
   return vtrn1_p8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vtrn1q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vtrn1q_p8
+  // CHECK-LABEL: test_vtrn1q_p8
   return vtrn1q_p8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vtrn1_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vtrn1_p16
+  // CHECK-LABEL: test_vtrn1_p16
   return vtrn1_p16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vtrn1q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vtrn1q_p16
+  // CHECK-LABEL: test_vtrn1q_p16
   return vtrn1q_p16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8_t test_vtrn2_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vtrn2_s8
+  // CHECK-LABEL: test_vtrn2_s8
   return vtrn2_s8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vtrn2q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vtrn2q_s8
+  // CHECK-LABEL: test_vtrn2q_s8
   return vtrn2q_s8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 int16x4_t test_vtrn2_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vtrn2_s16
+  // CHECK-LABEL: test_vtrn2_s16
   return vtrn2_s16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 int16x8_t test_vtrn2q_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vtrn2q_s16
+  // CHECK-LABEL: test_vtrn2q_s16
   return vtrn2q_s16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int32x2_t test_vtrn2_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vtrn2_s32
+  // CHECK-LABEL: test_vtrn2_s32
   return vtrn2_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 int32x4_t test_vtrn2q_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vtrn2q_s32
+  // CHECK-LABEL: test_vtrn2q_s32
   return vtrn2q_s32(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 int64x2_t test_vtrn2q_s64(int64x2_t a, int64x2_t b) {
-  // CHECK: test_vtrn2q_s64
+  // CHECK-LABEL: test_vtrn2q_s64
   return vtrn2q_s64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 uint8x8_t test_vtrn2_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vtrn2_u8
+  // CHECK-LABEL: test_vtrn2_u8
   return vtrn2_u8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vtrn2q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vtrn2q_u8
+  // CHECK-LABEL: test_vtrn2q_u8
   return vtrn2q_u8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 uint16x4_t test_vtrn2_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vtrn2_u16
+  // CHECK-LABEL: test_vtrn2_u16
   return vtrn2_u16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 uint16x8_t test_vtrn2q_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vtrn2q_u16
+  // CHECK-LABEL: test_vtrn2q_u16
   return vtrn2q_u16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 uint32x2_t test_vtrn2_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vtrn2_u32
+  // CHECK-LABEL: test_vtrn2_u32
   return vtrn2_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 uint32x4_t test_vtrn2q_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vtrn2q_u32
+  // CHECK-LABEL: test_vtrn2q_u32
   return vtrn2q_u32(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 uint64x2_t test_vtrn2q_u64(uint64x2_t a, uint64x2_t b) {
-  // CHECK: test_vtrn2q_u64
+  // CHECK-LABEL: test_vtrn2q_u64
   return vtrn2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 float32x2_t test_vtrn2_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vtrn2_f32
+  // CHECK-LABEL: test_vtrn2_f32
   return vtrn2_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v0.2s, v0.2s, v1.2s}}
 }
 
 float32x4_t test_vtrn2q_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vtrn2q_f32
+  // CHECK-LABEL: test_vtrn2q_f32
   return vtrn2q_f32(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 
 float64x2_t test_vtrn2q_f64(float64x2_t a, float64x2_t b) {
-  // CHECK: test_vtrn2q_f64
+  // CHECK-LABEL: test_vtrn2q_f64
   return vtrn2q_f64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK: {{ins v[0-9]+.d\[0\], v[0-9]+.d\[1\]|zip2 v0.2d, v0.2d, v1.2d}}
 }
 
 poly8x8_t test_vtrn2_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vtrn2_p8
+  // CHECK-LABEL: test_vtrn2_p8
   return vtrn2_p8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vtrn2q_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vtrn2q_p8
+  // CHECK-LABEL: test_vtrn2q_p8
   return vtrn2q_p8(a, b);
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly16x4_t test_vtrn2_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vtrn2_p16
+  // CHECK-LABEL: test_vtrn2_p16
   return vtrn2_p16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 
 poly16x8_t test_vtrn2q_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vtrn2q_p16
+  // CHECK-LABEL: test_vtrn2q_p16
   return vtrn2q_p16(a, b);
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8x2_t test_vuzp_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vuzp_s8
+  // CHECK-LABEL: test_vuzp_s8
   return vuzp_s8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4x2_t test_vuzp_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vuzp_s16
+  // CHECK-LABEL: test_vuzp_s16
   return vuzp_s16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int32x2x2_t test_vuzp_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vuzp_s32
+  // CHECK-LABEL: test_vuzp_s32
   return vuzp_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vuzp_u8
+  // CHECK-LABEL: test_vuzp_u8
   return vuzp_u8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vuzp_u16
+  // CHECK-LABEL: test_vuzp_u16
   return vuzp_u16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint32x2x2_t test_vuzp_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vuzp_u32
+  // CHECK-LABEL: test_vuzp_u32
   return vuzp_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 float32x2x2_t test_vuzp_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vuzp_f32
+  // CHECK-LABEL: test_vuzp_f32
   return vuzp_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 poly8x8x2_t test_vuzp_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vuzp_p8
+  // CHECK-LABEL: test_vuzp_p8
   return vuzp_p8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: uzp2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 poly16x4x2_t test_vuzp_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vuzp_p16
+  // CHECK-LABEL: test_vuzp_p16
   return vuzp_p16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: uzp2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int8x16x2_t test_vuzpq_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vuzpq_s8
+  // CHECK-LABEL: test_vuzpq_s8
   return vuzpq_s8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int16x8x2_t test_vuzpq_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vuzpq_s16
+  // CHECK-LABEL: test_vuzpq_s16
   return vuzpq_s16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int32x4x2_t test_vuzpq_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vuzpq_s32
+  // CHECK-LABEL: test_vuzpq_s32
   return vuzpq_s32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint8x16x2_t test_vuzpq_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vuzpq_u8
+  // CHECK-LABEL: test_vuzpq_u8
   return vuzpq_u8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint16x8x2_t test_vuzpq_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vuzpq_u16
+  // CHECK-LABEL: test_vuzpq_u16
   return vuzpq_u16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vuzpq_u32
+  // CHECK-LABEL: test_vuzpq_u32
   return vuzpq_u32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 float32x4x2_t test_vuzpq_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vuzpq_f32
+  // CHECK-LABEL: test_vuzpq_f32
   return vuzpq_f32(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: uzp2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 poly8x16x2_t test_vuzpq_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vuzpq_p8
+  // CHECK-LABEL: test_vuzpq_p8
   return vuzpq_p8(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: uzp2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 poly16x8x2_t test_vuzpq_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vuzpq_p16
+  // CHECK-LABEL: test_vuzpq_p16
   return vuzpq_p16(a, b);
   // CHECK: uzp1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: uzp2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8x2_t test_vzip_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vzip_s8
+  // CHECK-LABEL: test_vzip_s8
   return vzip_s8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4x2_t test_vzip_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vzip_s16
+  // CHECK-LABEL: test_vzip_s16
   return vzip_s16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int32x2x2_t test_vzip_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vzip_s32
+  // CHECK-LABEL: test_vzip_s32
   return vzip_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vzip_u8
+  // CHECK-LABEL: test_vzip_u8
   return vzip_u8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vzip_u16
+  // CHECK-LABEL: test_vzip_u16
   return vzip_u16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint32x2x2_t test_vzip_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vzip_u32
+  // CHECK-LABEL: test_vzip_u32
   return vzip_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 float32x2x2_t test_vzip_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vzip_f32
+  // CHECK-LABEL: test_vzip_f32
   return vzip_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 poly8x8x2_t test_vzip_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vzip_p8
+  // CHECK-LABEL: test_vzip_p8
   return vzip_p8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: zip2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 poly16x4x2_t test_vzip_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vzip_p16
+  // CHECK-LABEL: test_vzip_p16
   return vzip_p16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: zip2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int8x16x2_t test_vzipq_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vzipq_s8
+  // CHECK-LABEL: test_vzipq_s8
   return vzipq_s8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int16x8x2_t test_vzipq_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vzipq_s16
+  // CHECK-LABEL: test_vzipq_s16
   return vzipq_s16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int32x4x2_t test_vzipq_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vzipq_s32
+  // CHECK-LABEL: test_vzipq_s32
   return vzipq_s32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint8x16x2_t test_vzipq_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vzipq_u8
+  // CHECK-LABEL: test_vzipq_u8
   return vzipq_u8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint16x8x2_t test_vzipq_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vzipq_u16
+  // CHECK-LABEL: test_vzipq_u16
   return vzipq_u16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vzipq_u32
+  // CHECK-LABEL: test_vzipq_u32
   return vzipq_u32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 float32x4x2_t test_vzipq_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vzipq_f32
+  // CHECK-LABEL: test_vzipq_f32
   return vzipq_f32(a, b);
   // CHECK: zip1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: zip2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 poly8x16x2_t test_vzipq_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vzipq_p8
+  // CHECK-LABEL: test_vzipq_p8
   return vzipq_p8(a, b);
   // CHECK: zip1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: zip2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 poly16x8x2_t test_vzipq_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vzipq_p16
+  // CHECK-LABEL: test_vzipq_p16
   return vzipq_p16(a, b);
   // CHECK: zip1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: zip2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 
 int8x8x2_t test_vtrn_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vtrn_s8
+  // CHECK-LABEL: test_vtrn_s8
   return vtrn_s8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int16x4x2_t test_vtrn_s16(int16x4_t a, int16x4_t b) {
-  // CHECK: test_vtrn_s16
+  // CHECK-LABEL: test_vtrn_s16
   return vtrn_s16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int32x2x2_t test_vtrn_s32(int32x2_t a, int32x2_t b) {
-  // CHECK: test_vtrn_s32
+  // CHECK-LABEL: test_vtrn_s32
   return vtrn_s32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vtrn_u8
+  // CHECK-LABEL: test_vtrn_u8
   return vtrn_u8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) {
-  // CHECK: test_vtrn_u16
+  // CHECK-LABEL: test_vtrn_u16
   return vtrn_u16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 uint32x2x2_t test_vtrn_u32(uint32x2_t a, uint32x2_t b) {
-  // CHECK: test_vtrn_u32
+  // CHECK-LABEL: test_vtrn_u32
   return vtrn_u32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 float32x2x2_t test_vtrn_f32(float32x2_t a, float32x2_t b) {
-  // CHECK: test_vtrn_f32
+  // CHECK-LABEL: test_vtrn_f32
   return vtrn_f32(a, b);
-  // CHECK: ins {{v[0-9]+}}.s[1], {{v[0-9]+}}.s[0]
-  // CHECK: ins {{v[0-9]+}}.s[0], {{v[0-9]+}}.s[1]
+  // CHECK: {{ins v[0-9]+.s\[1\], v[0-9]+.s\[0\]|zip1 v2.2s, v0.2s, v1.2s}}
+  // CHECK: {{ins v[0-9]+.s\[0\], v[0-9]+.s\[1\]|zip2 v1.2s, v0.2s, v1.2s}}
 }
 poly8x8x2_t test_vtrn_p8(poly8x8_t a, poly8x8_t b) {
-  // CHECK: test_vtrn_p8
+  // CHECK-LABEL: test_vtrn_p8
   return vtrn_p8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: trn2 {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 poly16x4x2_t test_vtrn_p16(poly16x4_t a, poly16x4_t b) {
-  // CHECK: test_vtrn_p16
+  // CHECK-LABEL: test_vtrn_p16
   return vtrn_p16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
   // CHECK: trn2 {{v[0-9]+}}.4h, {{v[0-9]+}}.4h
 }
 int8x16x2_t test_vtrnq_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vtrnq_s8
+  // CHECK-LABEL: test_vtrnq_s8
   return vtrnq_s8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 int16x8x2_t test_vtrnq_s16(int16x8_t a, int16x8_t b) {
-  // CHECK: test_vtrnq_s16
+  // CHECK-LABEL: test_vtrnq_s16
   return vtrnq_s16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 int32x4x2_t test_vtrnq_s32(int32x4_t a, int32x4_t b) {
-  // CHECK: test_vtrnq_s32
+  // CHECK-LABEL: test_vtrnq_s32
   return vtrnq_s32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 uint8x16x2_t test_vtrnq_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vtrnq_u8
+  // CHECK-LABEL: test_vtrnq_u8
   return vtrnq_u8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 uint16x8x2_t test_vtrnq_u16(uint16x8_t a, uint16x8_t b) {
-  // CHECK: test_vtrnq_u16
+  // CHECK-LABEL: test_vtrnq_u16
   return vtrnq_u16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
 }
 uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) {
-  // CHECK: test_vtrnq_u32
+  // CHECK-LABEL: test_vtrnq_u32
   return vtrnq_u32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 float32x4x2_t test_vtrnq_f32(float32x4_t a, float32x4_t b) {
-  // CHECK: test_vtrnq_f32
+  // CHECK-LABEL: test_vtrnq_f32
   return vtrnq_f32(a, b);
   // CHECK: trn1 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
   // CHECK: trn2 {{v[0-9]+}}.4s, {{v[0-9]+}}.4s
 }
 poly8x16x2_t test_vtrnq_p8(poly8x16_t a, poly8x16_t b) {
-  // CHECK: test_vtrnq_p8
+  // CHECK-LABEL: test_vtrnq_p8
   return vtrnq_p8(a, b);
   // CHECK: trn1 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
   // CHECK: trn2 {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 poly16x8x2_t test_vtrnq_p16(poly16x8_t a, poly16x8_t b) {
-  // CHECK: test_vtrnq_p16
+  // CHECK-LABEL: test_vtrnq_p16
   return vtrnq_p16(a, b);
   // CHECK: trn1 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
   // CHECK: trn2 {{v[0-9]+}}.8h, {{v[0-9]+}}.8h
diff --git a/test/CodeGen/aarch64-neon-scalar-copy.c b/test/CodeGen/aarch64-neon-scalar-copy.c
index 33e97c7..227c6e0 100644
--- a/test/CodeGen/aarch64-neon-scalar-copy.c
+++ b/test/CodeGen/aarch64-neon-scalar-copy.c
@@ -1,11 +1,14 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 
 #include <arm_neon.h>
 
-// CHECK: test_vdups_lane_f32
+// CHECK-LABEL: test_vdups_lane_f32
 float32_t test_vdups_lane_f32(float32x2_t a) {
   return vdups_lane_f32(a, 1);
 // CHECK: ret
@@ -13,7 +16,7 @@
 }
 
 
-// CHECK: test_vdupd_lane_f64
+// CHECK-LABEL: test_vdupd_lane_f64
 float64_t test_vdupd_lane_f64(float64x1_t a) {
   return vdupd_lane_f64(a, 0);
 // CHECK: ret
@@ -21,7 +24,7 @@
 }
 
 
-// CHECK: test_vdups_laneq_f32
+// CHECK-LABEL: test_vdups_laneq_f32
 float32_t test_vdups_laneq_f32(float32x4_t a) {
   return vdups_laneq_f32(a, 3);
 // CHECK: ret
@@ -29,7 +32,7 @@
 }
 
 
-// CHECK: test_vdupd_laneq_f64
+// CHECK-LABEL: test_vdupd_laneq_f64
 float64_t test_vdupd_laneq_f64(float64x2_t a) {
   return vdupd_laneq_f64(a, 1);
 // CHECK: ret
@@ -37,135 +40,135 @@
 }
 
 
-// CHECK: test_vdupb_lane_s8
+// CHECK-LABEL: test_vdupb_lane_s8
 int8_t test_vdupb_lane_s8(int8x8_t a) {
   return vdupb_lane_s8(a, 7);
-// CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
+// CHECK: {{umov|smov}} {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
 
-// CHECK: test_vduph_lane_s16
+// CHECK-LABEL: test_vduph_lane_s16
 int16_t test_vduph_lane_s16(int16x4_t a) {
   return vduph_lane_s16(a, 3);
-// CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
+// CHECK: {{umov|smov}} {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
 
-// CHECK: test_vdups_lane_s32
+// CHECK-LABEL: test_vdups_lane_s32
 int32_t test_vdups_lane_s32(int32x2_t a) {
   return vdups_lane_s32(a, 1);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 
-// CHECK: test_vdupd_lane_s64
+// CHECK-LABEL: test_vdupd_lane_s64
 int64_t test_vdupd_lane_s64(int64x1_t a) {
   return vdupd_lane_s64(a, 0);
 // CHECK: fmov {{x[0-9]+}}, {{d[0-9]+}}
 }
 
 
-// CHECK: test_vdupb_lane_u8
+// CHECK-LABEL: test_vdupb_lane_u8
 uint8_t test_vdupb_lane_u8(uint8x8_t a) {
   return vdupb_lane_u8(a, 7);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
 
-// CHECK: test_vduph_lane_u16
+// CHECK-LABEL: test_vduph_lane_u16
 uint16_t test_vduph_lane_u16(uint16x4_t a) {
   return vduph_lane_u16(a, 3);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
 
-// CHECK: test_vdups_lane_u32
+// CHECK-LABEL: test_vdups_lane_u32
 uint32_t test_vdups_lane_u32(uint32x2_t a) {
   return vdups_lane_u32(a, 1);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 
-// CHECK: test_vdupd_lane_u64
+// CHECK-LABEL: test_vdupd_lane_u64
 uint64_t test_vdupd_lane_u64(uint64x1_t a) {
   return vdupd_lane_u64(a, 0);
 // CHECK: fmov {{x[0-9]+}}, {{d[0-9]+}}
 }
 
-// CHECK: test_vdupb_laneq_s8
+// CHECK-LABEL: test_vdupb_laneq_s8
 int8_t test_vdupb_laneq_s8(int8x16_t a) {
   return vdupb_laneq_s8(a, 15);
-// CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[15]
+// CHECK: {{umov|smov}} {{w[0-9]+}}, {{v[0-9]+}}.b[15]
 }
 
 
-// CHECK: test_vduph_laneq_s16
+// CHECK-LABEL: test_vduph_laneq_s16
 int16_t test_vduph_laneq_s16(int16x8_t a) {
   return vduph_laneq_s16(a, 7);
-// CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[7]
+// CHECK: {{umov|smov}} {{w[0-9]+}}, {{v[0-9]+}}.h[7]
 }
 
 
-// CHECK: test_vdups_laneq_s32
+// CHECK-LABEL: test_vdups_laneq_s32
 int32_t test_vdups_laneq_s32(int32x4_t a) {
   return vdups_laneq_s32(a, 3);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
 
-// CHECK: test_vdupd_laneq_s64
+// CHECK-LABEL: test_vdupd_laneq_s64
 int64_t test_vdupd_laneq_s64(int64x2_t a) {
   return vdupd_laneq_s64(a, 1);
 // CHECK: umov {{x[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 
-// CHECK: test_vdupb_laneq_u8
+// CHECK-LABEL: test_vdupb_laneq_u8
 uint8_t test_vdupb_laneq_u8(uint8x16_t a) {
   return vdupb_laneq_u8(a, 15);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[15]
 }
 
 
-// CHECK: test_vduph_laneq_u16
+// CHECK-LABEL: test_vduph_laneq_u16
 uint16_t test_vduph_laneq_u16(uint16x8_t a) {
   return vduph_laneq_u16(a, 7);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[7]
 }
 
 
-// CHECK: test_vdups_laneq_u32
+// CHECK-LABEL: test_vdups_laneq_u32
 uint32_t test_vdups_laneq_u32(uint32x4_t a) {
   return vdups_laneq_u32(a, 3);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
 
-// CHECK: test_vdupd_laneq_u64
+// CHECK-LABEL: test_vdupd_laneq_u64
 uint64_t test_vdupd_laneq_u64(uint64x2_t a) {
   return vdupd_laneq_u64(a, 1);
 // CHECK: umov {{x[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vdupb_lane_p8
+// CHECK-LABEL: test_vdupb_lane_p8
 poly8_t test_vdupb_lane_p8(poly8x8_t a) {
   return vdupb_lane_p8(a, 7);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[7]
 }
 
-// CHECK: test_vduph_lane_p16
+// CHECK-LABEL: test_vduph_lane_p16
 poly16_t test_vduph_lane_p16(poly16x4_t a) {
   return vduph_lane_p16(a, 3);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vdupb_laneq_p8
+// CHECK-LABEL: test_vdupb_laneq_p8
 poly8_t test_vdupb_laneq_p8(poly8x16_t a) {
   return vdupb_laneq_p8(a, 15);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.b[15]
 }
 
-// CHECK: test_vduph_laneq_p16
+// CHECK-LABEL: test_vduph_laneq_p16
 poly16_t test_vduph_laneq_p16(poly16x8_t a) {
   return vduph_laneq_p16(a, 7);
 // CHECK: umov {{w[0-9]+}}, {{v[0-9]+}}.h[7]
diff --git a/test/CodeGen/aarch64-neon-scalar-x-indexed-elem.c b/test/CodeGen/aarch64-neon-scalar-x-indexed-elem.c
index 4f0771a..0293e98 100644
--- a/test/CodeGen/aarch64-neon-scalar-x-indexed-elem.c
+++ b/test/CodeGen/aarch64-neon-scalar-x-indexed-elem.c
@@ -1,6 +1,9 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
@@ -8,248 +11,283 @@
 
 
 float32_t test_vmuls_lane_f32(float32_t a, float32x2_t b) {
-  // CHECK: test_vmuls_lane_f32
+  // CHECK-LABEL: test_vmuls_lane_f32
   return vmuls_lane_f32(a, b, 1);
   // CHECK: fmul {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 float64_t test_vmuld_lane_f64(float64_t a, float64x1_t b) {
-  // CHECK: test_vmuld_lane_f64
+  // CHECK-LABEL: test_vmuld_lane_f64
   return vmuld_lane_f64(a, b, 0);
-  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
 float32_t test_vmuls_laneq_f32(float32_t a, float32x4_t b) {
-  // CHECK: test_vmuls_laneq_f32
+  // CHECK-LABEL: test_vmuls_laneq_f32
   return vmuls_laneq_f32(a, b, 3);
   // CHECK: fmul {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
 float64_t test_vmuld_laneq_f64(float64_t a, float64x2_t b) {
-  // CHECK: test_vmuld_laneq_f64
+  // CHECK-LABEL: test_vmuld_laneq_f64
   return vmuld_laneq_f64(a, b, 1);
   // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 float64x1_t test_vmul_n_f64(float64x1_t a, float64_t b) {
-  // CHECK: test_vmul_n_f64
+  // CHECK-LABEL: test_vmul_n_f64
   return vmul_n_f64(a, b);
-  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: fmul {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
 float32_t test_vmulxs_lane_f32(float32_t a, float32x2_t b) {
-// CHECK: test_vmulxs_lane_f32
+// CHECK-LABEL: test_vmulxs_lane_f32
   return vmulxs_lane_f32(a, b, 1);
 // CHECK: fmulx {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 float32_t test_vmulxs_laneq_f32(float32_t a, float32x4_t b) {
-// CHECK: test_vmulxs_laneq_f32
+// CHECK-LABEL: test_vmulxs_laneq_f32
   return vmulxs_laneq_f32(a, b, 3);
 // CHECK: fmulx {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
 float64_t test_vmulxd_lane_f64(float64_t a, float64x1_t b) {
-// CHECK: test_vmulxd_lane_f64
+// CHECK-LABEL: test_vmulxd_lane_f64
   return vmulxd_lane_f64(a, b, 0);
-// CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+// CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
 float64_t test_vmulxd_laneq_f64(float64_t a, float64x2_t b) {
-// CHECK: test_vmulxd_laneq_f64
+// CHECK-LABEL: test_vmulxd_laneq_f64
   return vmulxd_laneq_f64(a, b, 1);
 // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vmulx_lane_f64
+// CHECK-LABEL: test_vmulx_lane_f64
 float64x1_t test_vmulx_lane_f64(float64x1_t a, float64x1_t b) {
   return vmulx_lane_f64(a, b, 0);
-  // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
 
-// CHECK: test_vmulx_laneq_f64_0
+// CHECK-LABEL: test_vmulx_laneq_f64_0
 float64x1_t test_vmulx_laneq_f64_0(float64x1_t a, float64x2_t b) {
   return vmulx_laneq_f64(a, b, 0);
   // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vmulx_laneq_f64_1
+// CHECK-LABEL: test_vmulx_laneq_f64_1
 float64x1_t test_vmulx_laneq_f64_1(float64x1_t a, float64x2_t b) {
   return vmulx_laneq_f64(a, b, 1);
   // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 
-// CHECK: test_vfmas_lane_f32
+// CHECK-LABEL: test_vfmas_lane_f32
 float32_t test_vfmas_lane_f32(float32_t a, float32_t b, float32x2_t c) {
   return vfmas_lane_f32(a, b, c, 1);
   // CHECK: fmla {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
-// CHECK: test_vfmad_lane_f64
+// CHECK-LABEL: test_vfmad_lane_f64
 float64_t test_vfmad_lane_f64(float64_t a, float64_t b, float64x1_t c) {
   return vfmad_lane_f64(a, b, c, 0);
-  // CHECK: fmla {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: {{fmla|fmadd}} {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
-// CHECK: test_vfmad_laneq_f64
+// CHECK-LABEL: test_vfmad_laneq_f64
 float64_t test_vfmad_laneq_f64(float64_t a, float64_t b, float64x2_t c) {
   return vfmad_laneq_f64(a, b, c, 1);
   // CHECK: fmla {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
-// CHECK: test_vfmss_lane_f32
+// CHECK-LABEL: test_vfmss_lane_f32
 float32_t test_vfmss_lane_f32(float32_t a, float32_t b, float32x2_t c) {
   return vfmss_lane_f32(a, b, c, 1);
   // CHECK: fmls {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
-// CHECK: test_vfma_lane_f64
+// CHECK-LABEL: test_vfma_lane_f64
 float64x1_t test_vfma_lane_f64(float64x1_t a, float64x1_t b, float64x1_t v) {
   return vfma_lane_f64(a, b, v, 0);
-  // CHECK: fmla {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: {{fmla|fmadd}} {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
-// CHECK: test_vfms_lane_f64
+// CHECK-LABEL: test_vfms_lane_f64
 float64x1_t test_vfms_lane_f64(float64x1_t a, float64x1_t b, float64x1_t v) {
   return vfms_lane_f64(a, b, v, 0);
-  // CHECK: fmls {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
+  // CHECK: {{fmls|fmsub}} {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+.d\[0\]|d[0-9]+}}
 }
 
-// CHECK: test_vfma_laneq_f64
+// CHECK-LABEL: test_vfma_laneq_f64
 float64x1_t test_vfma_laneq_f64(float64x1_t a, float64x1_t b, float64x2_t v) {
   return vfma_laneq_f64(a, b, v, 0);
   // CHECK: fmla {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vfms_laneq_f64
+// CHECK-LABEL: test_vfms_laneq_f64
 float64x1_t test_vfms_laneq_f64(float64x1_t a, float64x1_t b, float64x2_t v) {
   return vfms_laneq_f64(a, b, v, 0);
   // CHECK: fmls {{d[0-9]+}}, {{d[0-9]+}}, {{v[0-9]+}}.d[0]
 }
 
-// CHECK: test_vqdmullh_lane_s16
+// CHECK-LABEL: test_vqdmullh_lane_s16
 int32_t test_vqdmullh_lane_s16(int16_t a, int16x4_t b) {
   return vqdmullh_lane_s16(a, b, 3);
-  // CHECK: sqdmull {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[3]
+  // CHECK: sqdmull {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9].4h}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vqdmulls_lane_s32
+// CHECK-LABEL: test_vqdmulls_lane_s32
 int64_t test_vqdmulls_lane_s32(int32_t a, int32x2_t b) {
   return vqdmulls_lane_s32(a, b, 1);
   // CHECK: sqdmull {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
-// CHECK: test_vqdmullh_laneq_s16
+// CHECK-LABEL: test_vqdmullh_laneq_s16
 int32_t test_vqdmullh_laneq_s16(int16_t a, int16x8_t b) {
   return vqdmullh_laneq_s16(a, b, 7);
-  // CHECK: sqdmull {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[7]
+  // CHECK: sqdmull {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[7]
 }
 
-// CHECK: test_vqdmulls_laneq_s32
+// CHECK-LABEL: test_vqdmulls_laneq_s32
 int64_t test_vqdmulls_laneq_s32(int32_t a, int32x4_t b) {
   return vqdmulls_laneq_s32(a, b, 3);
   // CHECK: sqdmull {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
-// CHECK: test_vqdmulhh_lane_s16
+// CHECK-LABEL: test_vqdmulhh_lane_s16
 int16_t test_vqdmulhh_lane_s16(int16_t a, int16x4_t b) {
   return vqdmulhh_lane_s16(a, b, 3);
-// CHECK: sqdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[3]
+// CHECK: sqdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vqdmulhs_lane_s32
+// CHECK-LABEL: test_vqdmulhs_lane_s32
 int32_t test_vqdmulhs_lane_s32(int32_t a, int32x2_t b) {
   return vqdmulhs_lane_s32(a, b, 1);
 // CHECK: sqdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 
-// CHECK: test_vqdmulhh_laneq_s16
+// CHECK-LABEL: test_vqdmulhh_laneq_s16
 int16_t test_vqdmulhh_laneq_s16(int16_t a, int16x8_t b) {
   return vqdmulhh_laneq_s16(a, b, 7);
-// CHECK: sqdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[7]
+// CHECK: sqdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[7]
 }
 
 
-// CHECK: test_vqdmulhs_laneq_s32
+// CHECK-LABEL: test_vqdmulhs_laneq_s32
 int32_t test_vqdmulhs_laneq_s32(int32_t a, int32x4_t b) {
   return vqdmulhs_laneq_s32(a, b, 3);
 // CHECK: sqdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
-// CHECK: test_vqrdmulhh_lane_s16
+// CHECK-LABEL: test_vqrdmulhh_lane_s16
 int16_t test_vqrdmulhh_lane_s16(int16_t a, int16x4_t b) {
   return vqrdmulhh_lane_s16(a, b, 3);
-// CHECK: sqrdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[3]
+// CHECK: sqrdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vqrdmulhs_lane_s32
+// CHECK-LABEL: test_vqrdmulhs_lane_s32
 int32_t test_vqrdmulhs_lane_s32(int32_t a, int32x2_t b) {
   return vqrdmulhs_lane_s32(a, b, 1);
 // CHECK: sqrdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
 
-// CHECK: test_vqrdmulhh_laneq_s16
+// CHECK-LABEL: test_vqrdmulhh_laneq_s16
 int16_t test_vqrdmulhh_laneq_s16(int16_t a, int16x8_t b) {
   return vqrdmulhh_laneq_s16(a, b, 7);
-// CHECK: sqrdmulh {{h[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[7]
+// CHECK: sqrdmulh {{h[0-9]+|v[0-9]+.4h}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[7]
 }
 
 
-// CHECK: test_vqrdmulhs_laneq_s32
+// CHECK-LABEL: test_vqrdmulhs_laneq_s32
 int32_t test_vqrdmulhs_laneq_s32(int32_t a, int32x4_t b) {
   return vqrdmulhs_laneq_s32(a, b, 3);
 // CHECK: sqrdmulh {{s[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
-// CHECK: test_vqdmlalh_lane_s16
+// CHECK-LABEL: test_vqdmlalh_lane_s16
 int32_t test_vqdmlalh_lane_s16(int32_t a, int16_t b, int16x4_t c) {
   return vqdmlalh_lane_s16(a, b, c, 3);
-// CHECK: sqdmlal {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[3]
+// CHECK: sqdmlal {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vqdmlals_lane_s32
+// CHECK-LABEL: test_vqdmlals_lane_s32
 int64_t test_vqdmlals_lane_s32(int64_t a, int32_t b, int32x2_t c) {
   return vqdmlals_lane_s32(a, b, c, 1);
 // CHECK: sqdmlal {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
-// CHECK: test_vqdmlalh_laneq_s16
+// CHECK-LABEL: test_vqdmlalh_laneq_s16
 int32_t test_vqdmlalh_laneq_s16(int32_t a, int16_t b, int16x8_t c) {
   return vqdmlalh_laneq_s16(a, b, c, 7);
-// CHECK: sqdmlal {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[7]
+// CHECK: sqdmlal {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[7]
 }
 
-// CHECK: test_vqdmlals_laneq_s32
+// CHECK-LABEL: test_vqdmlals_laneq_s32
 int64_t test_vqdmlals_laneq_s32(int64_t a, int32_t b, int32x4_t c) {
   return vqdmlals_laneq_s32(a, b, c, 3);
 // CHECK: sqdmlal {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
-// CHECK: test_vqdmlslh_lane_s16
+// CHECK-LABEL: test_vqdmlslh_lane_s16
 int32_t test_vqdmlslh_lane_s16(int32_t a, int16_t b, int16x4_t c) {
   return vqdmlslh_lane_s16(a, b, c, 3);
-// CHECK: sqdmlsl {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[3]
+// CHECK: sqdmlsl {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[3]
 }
 
-// CHECK: test_vqdmlsls_lane_s32
+// CHECK-LABEL: test_vqdmlsls_lane_s32
 int64_t test_vqdmlsls_lane_s32(int64_t a, int32_t b, int32x2_t c) {
   return vqdmlsls_lane_s32(a, b, c, 1);
 // CHECK: sqdmlsl {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[1]
 }
 
-// CHECK: test_vqdmlslh_laneq_s16
+// CHECK-LABEL: test_vqdmlslh_laneq_s16
 int32_t test_vqdmlslh_laneq_s16(int32_t a, int16_t b, int16x8_t c) {
   return vqdmlslh_laneq_s16(a, b, c, 7);
-// CHECK: sqdmlsl {{s[0-9]+}}, {{h[0-9]+}}, {{v[0-9]+}}.h[7]
+// CHECK: sqdmlsl {{s[0-9]+|v[0-9]+.4s}}, {{h[0-9]+|v[0-9]+.4h}}, {{v[0-9]+}}.h[7]
 }
 
-// CHECK: test_vqdmlsls_laneq_s32
+// CHECK-LABEL: test_vqdmlsls_laneq_s32
 int64_t test_vqdmlsls_laneq_s32(int64_t a, int32_t b, int32x4_t c) {
   return vqdmlsls_laneq_s32(a, b, c, 3);
 // CHECK: sqdmlsl {{d[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}.s[3]
 }
 
+// CHECK-LABEL: test_vmulx_lane_f64_0:
+float64x1_t test_vmulx_lane_f64_0() {
+      float64x1_t arg1;
+      float64x1_t arg2;
+      float64x1_t result;
+      float64_t sarg1, sarg2, sres;
+      arg1 = vcreate_f64(UINT64_C(0x3fd6304bc43ab5c2));
+      arg2 = vcreate_f64(UINT64_C(0x3fee211e215aeef3));
+      result = vmulx_lane_f64(arg1, arg2, 0);
+// CHECK: adrp x[[ADDRLO:[0-9]+]]
+// CHECK: ldr d0, [x[[ADDRLO]],
+// CHECK: adrp x[[ADDRLO:[0-9]+]]
+// CHECK: ldr d1, [x[[ADDRLO]],
+// CHECK: fmulx d0, d1, d0
+      return result;
+}
+
+// CHECK-LABEL: test_vmulx_laneq_f64_2:
+float64x1_t test_vmulx_laneq_f64_2() {
+      float64x1_t arg1;
+      float64x1_t arg2;
+      float64x2_t arg3;
+      float64x1_t result;
+      float64_t sarg1, sarg2, sres;
+      arg1 = vcreate_f64(UINT64_C(0x3fd6304bc43ab5c2));
+      arg2 = vcreate_f64(UINT64_C(0x3fee211e215aeef3));
+      arg3 = vcombine_f64(arg1, arg2);
+      result = vmulx_laneq_f64(arg1, arg3, 1);
+// CHECK: adrp x[[ADDRLO:[0-9]+]]
+// CHECK: ldr d0, [x[[ADDRLO]],
+// CHECK: adrp x[[ADDRLO:[0-9]+]]
+// CHECK: ldr d1, [x[[ADDRLO]],
+// CHECK: fmulx d0, d1, d0
+      return result;
+}
diff --git a/test/CodeGen/aarch64-neon-shifts.c b/test/CodeGen/aarch64-neon-shifts.c
index 4777f18..99adff1 100644
--- a/test/CodeGen/aarch64-neon-shifts.c
+++ b/test/CodeGen/aarch64-neon-shifts.c
@@ -1,6 +1,9 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -emit-llvm -O1 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -emit-llvm -O1 -o - %s | FileCheck %s
 
 #include <arm_neon.h>
 
diff --git a/test/CodeGen/aarch64-neon-tbl.c b/test/CodeGen/aarch64-neon-tbl.c
index db78a7a..93cba1d 100644
--- a/test/CodeGen/aarch64-neon-tbl.c
+++ b/test/CodeGen/aarch64-neon-tbl.c
@@ -1,463 +1,466 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int8x8_t test_vtbl1_s8(int8x8_t a, int8x8_t b) {
-  // CHECK: test_vtbl1_s8
+  // CHECK-LABEL: test_vtbl1_s8
   return vtbl1_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbl1_s8(int8x16_t a, int8x8_t b) {
-  // CHECK: test_vqtbl1_s8
+  // CHECK-LABEL: test_vqtbl1_s8
   return vqtbl1_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbl2_s8(int8x8x2_t a, int8x8_t b) {
-  // CHECK: test_vtbl2_s8
+  // CHECK-LABEL: test_vtbl2_s8
   return vtbl2_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbl2_s8(int8x16x2_t a, int8x8_t b) {
-  // CHECK: test_vqtbl2_s8
+  // CHECK-LABEL: test_vqtbl2_s8
   return vqtbl2_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbl3_s8(int8x8x3_t a, int8x8_t b) {
-  // CHECK: test_vtbl3_s8
+  // CHECK-LABEL: test_vtbl3_s8
   return vtbl3_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbl3_s8(int8x16x3_t a, int8x8_t b) {
-  // CHECK: test_vqtbl3_s8
+  // CHECK-LABEL: test_vqtbl3_s8
   return vqtbl3_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbl4_s8(int8x8x4_t a, int8x8_t b) {
-  // CHECK: test_vtbl4_s8
+  // CHECK-LABEL: test_vtbl4_s8
   return vtbl4_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbl4_s8(int8x16x4_t a, int8x8_t b) {
-  // CHECK: test_vqtbl4_s8
+  // CHECK-LABEL: test_vqtbl4_s8
   return vqtbl4_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vqtbl1q_s8(int8x16_t a, int8x16_t b) {
-  // CHECK: test_vqtbl1q_s8
+  // CHECK-LABEL: test_vqtbl1q_s8
   return vqtbl1q_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbl2q_s8(int8x16x2_t a, int8x16_t b) {
-  // CHECK: test_vqtbl2q_s8
+  // CHECK-LABEL: test_vqtbl2q_s8
   return vqtbl2q_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbl3q_s8(int8x16x3_t a, int8x16_t b) {
-  // CHECK: test_vqtbl3q_s8
+  // CHECK-LABEL: test_vqtbl3q_s8
   return vqtbl3q_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbl4q_s8(int8x16x4_t a, int8x16_t b) {
-  // CHECK: test_vqtbl4q_s8
+  // CHECK-LABEL: test_vqtbl4q_s8
   return vqtbl4q_s8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x8_t test_vtbx1_s8(int8x8_t a, int8x8_t b, int8x8_t c) {
-  // CHECK: test_vtbx1_s8
+  // CHECK-LABEL: test_vtbx1_s8
   return vtbx1_s8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbx2_s8(int8x8_t a, int8x8x2_t b, int8x8_t c) {
-  // CHECK: test_vtbx2_s8
+  // CHECK-LABEL: test_vtbx2_s8
   return vtbx2_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbx3_s8(int8x8_t a, int8x8x3_t b, int8x8_t c) {
-  // CHECK: test_vtbx3_s8
+  // CHECK-LABEL: test_vtbx3_s8
   return vtbx3_s8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vtbx4_s8(int8x8_t a, int8x8x4_t b, int8x8_t c) {
-  // CHECK: test_vtbx4_s8
+  // CHECK-LABEL: test_vtbx4_s8
   return vtbx4_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbx1_s8(int8x8_t a, int8x16_t b, int8x8_t c) {
-  // CHECK: test_vqtbx1_s8
+  // CHECK-LABEL: test_vqtbx1_s8
   return vqtbx1_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbx2_s8(int8x8_t a, int8x16x2_t b, int8x8_t c) {
-  // CHECK: test_vqtbx2_s8
+  // CHECK-LABEL: test_vqtbx2_s8
   return vqtbx2_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbx3_s8(int8x8_t a, int8x16x3_t b, int8x8_t c) {
-  // CHECK: test_vqtbx3_s8
+  // CHECK-LABEL: test_vqtbx3_s8
   return vqtbx3_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x8_t test_vqtbx4_s8(int8x8_t a, int8x16x4_t b, int8x8_t c) {
-  // CHECK: test_vqtbx4_s8
+  // CHECK-LABEL: test_vqtbx4_s8
   return vqtbx4_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 int8x16_t test_vqtbx1q_s8(int8x16_t a, int8x16_t b, int8x16_t c) {
-  // CHECK: test_vqtbx1q_s8
+  // CHECK-LABEL: test_vqtbx1q_s8
   return vqtbx1q_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbx2q_s8(int8x16_t a, int8x16x2_t b, int8x16_t c) {
-  // CHECK: test_vqtbx2q_s8
+  // CHECK-LABEL: test_vqtbx2q_s8
   return vqtbx2q_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbx3q_s8(int8x16_t a, int8x16x3_t b, int8x16_t c) {
-  // CHECK: test_vqtbx3q_s8
+  // CHECK-LABEL: test_vqtbx3q_s8
   return vqtbx3q_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 int8x16_t test_vqtbx4q_s8(int8x16_t a, int8x16x4_t b, int8x16_t c) {
-  // CHECK: test_vqtbx4q_s8
+  // CHECK-LABEL: test_vqtbx4q_s8
   return vqtbx4q_s8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x8_t test_vtbl1_u8(uint8x8_t a, uint8x8_t b) {
-  // CHECK: test_vtbl1_u8
+  // CHECK-LABEL: test_vtbl1_u8
   return vtbl1_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbl1_u8(uint8x16_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl1_u8
+  // CHECK-LABEL: test_vqtbl1_u8
   return vqtbl1_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbl2_u8(uint8x8x2_t a, uint8x8_t b) {
-  // CHECK: test_vtbl2_u8
+  // CHECK-LABEL: test_vtbl2_u8
   return vtbl2_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbl2_u8(uint8x16x2_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl2_u8
+  // CHECK-LABEL: test_vqtbl2_u8
   return vqtbl2_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbl3_u8(uint8x8x3_t a, uint8x8_t b) {
-  // CHECK: test_vtbl3_u8
+  // CHECK-LABEL: test_vtbl3_u8
   return vtbl3_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbl3_u8(uint8x16x3_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl3_u8
+  // CHECK-LABEL: test_vqtbl3_u8
   return vqtbl3_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbl4_u8(uint8x8x4_t a, uint8x8_t b) {
-  // CHECK: test_vtbl4_u8
+  // CHECK-LABEL: test_vtbl4_u8
   return vtbl4_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbl4_u8(uint8x16x4_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl4_u8
+  // CHECK-LABEL: test_vqtbl4_u8
   return vqtbl4_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vqtbl1q_u8(uint8x16_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl1q_u8
+  // CHECK-LABEL: test_vqtbl1q_u8
   return vqtbl1q_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbl2q_u8(uint8x16x2_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl2q_u8
+  // CHECK-LABEL: test_vqtbl2q_u8
   return vqtbl2q_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbl3q_u8(uint8x16x3_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl3q_u8
+  // CHECK-LABEL: test_vqtbl3q_u8
   return vqtbl3q_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbl4q_u8(uint8x16x4_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl4q_u8
+  // CHECK-LABEL: test_vqtbl4q_u8
   return vqtbl4q_u8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x8_t test_vtbx1_u8(uint8x8_t a, uint8x8_t b, uint8x8_t c) {
-  // CHECK: test_vtbx1_u8
+  // CHECK-LABEL: test_vtbx1_u8
   return vtbx1_u8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbx2_u8(uint8x8_t a, uint8x8x2_t b, uint8x8_t c) {
-  // CHECK: test_vtbx2_u8
+  // CHECK-LABEL: test_vtbx2_u8
   return vtbx2_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbx3_u8(uint8x8_t a, uint8x8x3_t b, uint8x8_t c) {
-  // CHECK: test_vtbx3_u8
+  // CHECK-LABEL: test_vtbx3_u8
   return vtbx3_u8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vtbx4_u8(uint8x8_t a, uint8x8x4_t b, uint8x8_t c) {
-  // CHECK: test_vtbx4_u8
+  // CHECK-LABEL: test_vtbx4_u8
   return vtbx4_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbx1_u8(uint8x8_t a, uint8x16_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx1_u8
+  // CHECK-LABEL: test_vqtbx1_u8
   return vqtbx1_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbx2_u8(uint8x8_t a, uint8x16x2_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx2_u8
+  // CHECK-LABEL: test_vqtbx2_u8
   return vqtbx2_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbx3_u8(uint8x8_t a, uint8x16x3_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx3_u8
+  // CHECK-LABEL: test_vqtbx3_u8
   return vqtbx3_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x8_t test_vqtbx4_u8(uint8x8_t a, uint8x16x4_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx4_u8
+  // CHECK-LABEL: test_vqtbx4_u8
   return vqtbx4_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 uint8x16_t test_vqtbx1q_u8(uint8x16_t a, uint8x16_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx1q_u8
+  // CHECK-LABEL: test_vqtbx1q_u8
   return vqtbx1q_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbx2q_u8(uint8x16_t a, uint8x16x2_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx2q_u8
+  // CHECK-LABEL: test_vqtbx2q_u8
   return vqtbx2q_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbx3q_u8(uint8x16_t a, uint8x16x3_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx3q_u8
+  // CHECK-LABEL: test_vqtbx3q_u8
   return vqtbx3q_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 uint8x16_t test_vqtbx4q_u8(uint8x16_t a, uint8x16x4_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx4q_u8
+  // CHECK-LABEL: test_vqtbx4q_u8
   return vqtbx4q_u8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x8_t test_vtbl1_p8(poly8x8_t a, uint8x8_t b) {
-  // CHECK: test_vtbl1_p8
+  // CHECK-LABEL: test_vtbl1_p8
   return vtbl1_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbl1_p8(poly8x16_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl1_p8
+  // CHECK-LABEL: test_vqtbl1_p8
   return vqtbl1_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbl2_p8(poly8x8x2_t a, uint8x8_t b) {
-  // CHECK: test_vtbl2_p8
+  // CHECK-LABEL: test_vtbl2_p8
   return vtbl2_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbl2_p8(poly8x16x2_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl2_p8
+  // CHECK-LABEL: test_vqtbl2_p8
   return vqtbl2_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbl3_p8(poly8x8x3_t a, uint8x8_t b) {
-  // CHECK: test_vtbl3_p8
+  // CHECK-LABEL: test_vtbl3_p8
   return vtbl3_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbl3_p8(poly8x16x3_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl3_p8
+  // CHECK-LABEL: test_vqtbl3_p8
   return vqtbl3_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbl4_p8(poly8x8x4_t a, uint8x8_t b) {
-  // CHECK: test_vtbl4_p8
+  // CHECK-LABEL: test_vtbl4_p8
   return vtbl4_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbl4_p8(poly8x16x4_t a, uint8x8_t b) {
-  // CHECK: test_vqtbl4_p8
+  // CHECK-LABEL: test_vqtbl4_p8
   return vqtbl4_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vqtbl1q_p8(poly8x16_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl1q_p8
+  // CHECK-LABEL: test_vqtbl1q_p8
   return vqtbl1q_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbl2q_p8(poly8x16x2_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl2q_p8
+  // CHECK-LABEL: test_vqtbl2q_p8
   return vqtbl2q_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbl3q_p8(poly8x16x3_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl3q_p8
+  // CHECK-LABEL: test_vqtbl3q_p8
   return vqtbl3q_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbl4q_p8(poly8x16x4_t a, uint8x16_t b) {
-  // CHECK: test_vqtbl4q_p8
+  // CHECK-LABEL: test_vqtbl4q_p8
   return vqtbl4q_p8(a, b);
-  // CHECK: tbl {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbl {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x8_t test_vtbx1_p8(poly8x8_t a, poly8x8_t b, uint8x8_t c) {
-  // CHECK: test_vtbx1_p8
+  // CHECK-LABEL: test_vtbx1_p8
   return vtbx1_p8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbx2_p8(poly8x8_t a, poly8x8x2_t b, uint8x8_t c) {
-  // CHECK: test_vtbx2_p8
+  // CHECK-LABEL: test_vtbx2_p8
   return vtbx2_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbx3_p8(poly8x8_t a, poly8x8x3_t b, uint8x8_t c) {
-  // CHECK: test_vtbx3_p8
+  // CHECK-LABEL: test_vtbx3_p8
   return vtbx3_p8(a, b, c);
-  // CHECK: movi {{v[0-9]+}}.8b, #0
+  // CHECK: movi {{v[0-9]+.8b|d[0-9]+}}, #0
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
-  // CHECK: tbl {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbl {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
   // CHECK: cmhs {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vtbx4_p8(poly8x8_t a, poly8x8x4_t b, uint8x8_t c) {
-  // CHECK: test_vtbx4_p8
+  // CHECK-LABEL: test_vtbx4_p8
   return vtbx4_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbx1_p8(poly8x8_t a, uint8x16_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx1_p8
+  // CHECK-LABEL: test_vqtbx1_p8
   return vqtbx1_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbx2_p8(poly8x8_t a, poly8x16x2_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx2_p8
+  // CHECK-LABEL: test_vqtbx2_p8
   return vqtbx2_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbx3_p8(poly8x8_t a, poly8x16x3_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx3_p8
+  // CHECK-LABEL: test_vqtbx3_p8
   return vqtbx3_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x8_t test_vqtbx4_p8(poly8x8_t a, poly8x16x4_t b, uint8x8_t c) {
-  // CHECK: test_vqtbx4_p8
+  // CHECK-LABEL: test_vqtbx4_p8
   return vqtbx4_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.8b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.8b
+  // CHECK: tbx {{v[0-9]+}}.8b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.8b
 }
 
 poly8x16_t test_vqtbx1q_p8(poly8x16_t a, uint8x16_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx1q_p8
+  // CHECK-LABEL: test_vqtbx1q_p8
   return vqtbx1q_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbx2q_p8(poly8x16_t a, poly8x16x2_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx2q_p8
+  // CHECK-LABEL: test_vqtbx2q_p8
   return vqtbx2q_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbx3q_p8(poly8x16_t a, poly8x16x3_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx3q_p8
+  // CHECK-LABEL: test_vqtbx3q_p8
   return vqtbx3q_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
 
 poly8x16_t test_vqtbx4q_p8(poly8x16_t a, poly8x16x4_t b, uint8x16_t c) {
-  // CHECK: test_vqtbx4q_p8
+  // CHECK-LABEL: test_vqtbx4q_p8
   return vqtbx4q_p8(a, b, c);
-  // CHECK: tbx {{v[0-9]+}}.16b, {{{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b}, {{v[0-9]+}}.16b
+  // CHECK: tbx {{v[0-9]+}}.16b, {{{ ?v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b, v[0-9]+.16b ?}}}, {{v[0-9]+}}.16b
 }
diff --git a/test/CodeGen/aarch64-neon-vcombine.c b/test/CodeGen/aarch64-neon-vcombine.c
index 3e170c8..381b052 100644
--- a/test/CodeGen/aarch64-neon-vcombine.c
+++ b/test/CodeGen/aarch64-neon-vcombine.c
@@ -1,6 +1,8 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
 // RUN:   -S -O3 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu -S -O3 -o - %s | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
diff --git a/test/CodeGen/aarch64-neon-vget-hilo.c b/test/CodeGen/aarch64-neon-vget-hilo.c
index 012b0bb..4d80d41 100644
--- a/test/CodeGen/aarch64-neon-vget-hilo.c
+++ b/test/CodeGen/aarch64-neon-vget-hilo.c
@@ -1,176 +1,193 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
-// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix CHECK-COMMON --check-prefix CHECK-AARCH64
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix CHECK-COMMON --check-prefix CHECK-ARM64
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 int8x8_t test_vget_high_s8(int8x16_t a) {
-  // CHECK-LABEL: test_vget_high_s8:
+  // CHECK-COMMON-LABEL: test_vget_high_s8:
   return vget_high_s8(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 int16x4_t test_vget_high_s16(int16x8_t a) {
-  // CHECK-LABEL: test_vget_high_s16:
+  // CHECK-COMMON-LABEL: test_vget_high_s16:
   return vget_high_s16(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 int32x2_t test_vget_high_s32(int32x4_t a) {
-  // CHECK-LABEL: test_vget_high_s32:
+  // CHECK-COMMON-LABEL: test_vget_high_s32:
   return vget_high_s32(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 int64x1_t test_vget_high_s64(int64x2_t a) {
-  // CHECK-LABEL: test_vget_high_s64:
+  // CHECK-COMMON-LABEL: test_vget_high_s64:
   return vget_high_s64(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 uint8x8_t test_vget_high_u8(uint8x16_t a) {
-  // CHECK-LABEL: test_vget_high_u8:
+  // CHECK-COMMON-LABEL: test_vget_high_u8:
   return vget_high_u8(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 uint16x4_t test_vget_high_u16(uint16x8_t a) {
-  // CHECK-LABEL: test_vget_high_u16:
+  // CHECK-COMMON-LABEL: test_vget_high_u16:
   return vget_high_u16(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 uint32x2_t test_vget_high_u32(uint32x4_t a) {
-  // CHECK-LABEL: test_vget_high_u32:
+  // CHECK-COMMON-LABEL: test_vget_high_u32:
   return vget_high_u32(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 uint64x1_t test_vget_high_u64(uint64x2_t a) {
-  // CHECK-LABEL: test_vget_high_u64:
+  // CHECK-COMMON-LABEL: test_vget_high_u64:
   return vget_high_u64(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 poly64x1_t test_vget_high_p64(poly64x2_t a) {
-  // CHECK-LABEL: test_vget_high_p64:
+  // CHECK-COMMON-LABEL: test_vget_high_p64:
   return vget_high_p64(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 float16x4_t test_vget_high_f16(float16x8_t a) {
-  // CHECK-LABEL: test_vget_high_f16:
+  // CHECK-COMMON-LABEL: test_vget_high_f16:
   return vget_high_f16(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 float32x2_t test_vget_high_f32(float32x4_t a) {
-  // CHECK-LABEL: test_vget_high_f32:
+  // CHECK-COMMON-LABEL: test_vget_high_f32:
   return vget_high_f32(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 poly8x8_t test_vget_high_p8(poly8x16_t a) {
-  // CHECK-LABEL: test_vget_high_p8:
+  // CHECK-COMMON-LABEL: test_vget_high_p8:
   return vget_high_p8(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 poly16x4_t test_vget_high_p16(poly16x8_t a) {
-  // CHECK-LABEL: test_vget_high_p16
+  // CHECK-COMMON-LABEL: test_vget_high_p16
   return vget_high_p16(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 float64x1_t test_vget_high_f64(float64x2_t a) {
-  // CHECK-LABEL: test_vget_high_f64
+  // CHECK-COMMON-LABEL: test_vget_high_f64
   return vget_high_f64(a);
-  // CHECK: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: dup d0, {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: ext v0.16b, v0.16b, v0.16b, #8
 }
 
 int8x8_t test_vget_low_s8(int8x16_t a) {
-  // CHECK-LABEL: test_vget_low_s8:
+  // CHECK-COMMON-LABEL: test_vget_low_s8:
   return vget_low_s8(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 int16x4_t test_vget_low_s16(int16x8_t a) {
-  // CHECK-LABEL: test_vget_low_s16:
+  // CHECK-COMMON-LABEL: test_vget_low_s16:
   return vget_low_s16(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 int32x2_t test_vget_low_s32(int32x4_t a) {
-  // CHECK-LABEL: test_vget_low_s32:
+  // CHECK-COMMON-LABEL: test_vget_low_s32:
   return vget_low_s32(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 int64x1_t test_vget_low_s64(int64x2_t a) {
-  // CHECK-LABEL: test_vget_low_s64:
+  // CHECK-COMMON-LABEL: test_vget_low_s64:
   return vget_low_s64(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 uint8x8_t test_vget_low_u8(uint8x16_t a) {
-  // CHECK-LABEL: test_vget_low_u8:
+  // CHECK-COMMON-LABEL: test_vget_low_u8:
   return vget_low_u8(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 uint16x4_t test_vget_low_u16(uint16x8_t a) {
-  // CHECK-LABEL: test_vget_low_u16:
+  // CHECK-COMMON-LABEL: test_vget_low_u16:
   return vget_low_u16(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 uint32x2_t test_vget_low_u32(uint32x4_t a) {
-  // CHECK-LABEL: test_vget_low_u32:
+  // CHECK-COMMON-LABEL: test_vget_low_u32:
   return vget_low_u32(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 uint64x1_t test_vget_low_u64(uint64x2_t a) {
-  // CHECK-LABEL: test_vget_low_u64:
+  // CHECK-COMMON-LABEL: test_vget_low_u64:
   return vget_low_u64(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 poly64x1_t test_vget_low_p64(poly64x2_t a) {
-  // CHECK-LABEL: test_vget_low_p64:
+  // CHECK-COMMON-LABEL: test_vget_low_p64:
   return vget_low_p64(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 float16x4_t test_vget_low_f16(float16x8_t a) {
-  // CHECK-LABEL: test_vget_low_f16:
+  // CHECK-COMMON-LABEL: test_vget_low_f16:
   return vget_low_f16(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 float32x2_t test_vget_low_f32(float32x4_t a) {
-  // CHECK-LABEL: test_vget_low_f32:
+  // CHECK-COMMON-LABEL: test_vget_low_f32:
   return vget_low_f32(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 poly8x8_t test_vget_low_p8(poly8x16_t a) {
-  // CHECK-LABEL: test_vget_low_p8:
+  // CHECK-COMMON-LABEL: test_vget_low_p8:
   return vget_low_p8(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 poly16x4_t test_vget_low_p16(poly16x8_t a) {
-  // CHECK-LABEL: test_vget_low_p16:
+  // CHECK-COMMON-LABEL: test_vget_low_p16:
   return vget_low_p16(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
 float64x1_t test_vget_low_f64(float64x2_t a) {
-  // CHECK-LABEL: test_vget_low_f64:
+  // CHECK-COMMON-LABEL: test_vget_low_f64:
   return vget_low_f64(a);
-  // CHECK-NEXT: ret
+  // CHECK-COMMON-NEXT: ret
 }
 
diff --git a/test/CodeGen/aarch64-poly128.c b/test/CodeGen/aarch64-poly128.c
new file mode 100644
index 0000000..609e596
--- /dev/null
+++ b/test/CodeGen/aarch64-poly128.c
@@ -0,0 +1,215 @@
+// REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:  -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK \
+// RUN:  --check-prefix=CHECK-AARCH64
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:  -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK \
+// RUN:  --check-prefix=CHECK-ARM64
+
+// Test new aarch64 intrinsics with poly128
+// FIXME: Currently, poly128_t equals to uint128, which will be spilt into
+// two 64-bit GPR(eg X0, X1). Now moving data from X0, X1 to FPR128 will
+// introduce 2 store and 1 load instructions(store X0, X1 to memory and
+// then load back to Q0). If target has NEON, this is better replaced by
+// FMOV or INS.
+
+#include <arm_neon.h>
+
+void test_vstrq_p128(poly128_t * ptr, poly128_t val) {
+  // CHECK-LABEL: test_vstrq_p128
+  vstrq_p128(ptr, val);
+// CHECK-AARCH64: str {{x[0-9]+}}, [{{x[0-9]+}}, #8]
+// CHECK-AARCH64-NEXT: str  {{x[0-9]+}}, [{{x[0-9]+}}]
+
+  // CHECK-ARM64: stp {{x[0-9]+}}, {{x[0-9]+}}, [x0]
+}
+
+poly128_t test_vldrq_p128(poly128_t * ptr) {
+  // CHECK-LABEL: test_vldrq_p128
+  return vldrq_p128(ptr);
+  // CHECK-AARCH64: ldr {{x[0-9]+}}, [{{x[0-9]+}}]
+  // CHECK-AARCH64-NEXT: ldr {{x[0-9]+}}, [{{x[0-9]+}}, #8]
+
+  // CHECK-ARM64: ldp {{x[0-9]+}}, {{x[0-9]+}}, [x0]
+}
+
+void test_ld_st_p128(poly128_t * ptr) {
+  // CHECK-LABEL: test_ld_st_p128
+   vstrq_p128(ptr+1, vldrq_p128(ptr));
+ // CHECK-AARCH64: ldr {{q[0-9]+}}, [{{x[0-9]+}}]
+ // CHECK-AARCH64-NEXT: str {{q[0-9]+}}, [{{x[0-9]+}}, #16]
+
+ // CHECK-ARM64: ldp [[PLO:x[0-9]+]], [[PHI:x[0-9]+]], [{{x[0-9]+}}]
+ // CHECK-ARM64-NEXT: stp [[PLO]], [[PHI]], [{{x[0-9]+}}, #16]
+}
+
+poly128_t test_vmull_p64(poly64_t a, poly64_t b) {
+  // CHECK-LABEL: test_vmull_p64
+  return vmull_p64(a, b);
+  // CHECK: pmull {{v[0-9]+}}.1q, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d
+}
+
+poly128_t test_vmull_high_p64(poly64x2_t a, poly64x2_t b) {
+  // CHECK-LABEL: test_vmull_high_p64
+  return vmull_high_p64(a, b);
+  // CHECK: pmull2 {{v[0-9]+}}.1q, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_s8
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_s8(int8x16_t a) {
+  return vreinterpretq_p128_s8(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_s16
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_s16(int16x8_t a) {
+  return vreinterpretq_p128_s16(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_s32
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_s32(int32x4_t a) {
+  return vreinterpretq_p128_s32(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_s64
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_s64(int64x2_t a) {
+  return vreinterpretq_p128_s64(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_u8
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_u8(uint8x16_t a) {
+  return vreinterpretq_p128_u8(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_u16
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_u16(uint16x8_t a) {
+  return vreinterpretq_p128_u16(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_u32
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_u32(uint32x4_t a) {
+  return vreinterpretq_p128_u32(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_u64
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_u64(uint64x2_t a) {
+  return vreinterpretq_p128_u64(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_f32
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_f32(float32x4_t a) {
+  return vreinterpretq_p128_f32(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_f64
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_f64(float64x2_t a) {
+  return vreinterpretq_p128_f64(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_p8
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_p8(poly8x16_t a) {
+  return vreinterpretq_p128_p8(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_p16
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_p16(poly16x8_t a) {
+  return vreinterpretq_p128_p16(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p128_p64
+// CHECK: ret
+poly128_t test_vreinterpretq_p128_p64(poly64x2_t a) {
+  return vreinterpretq_p128_p64(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_s8_p128
+// CHECK: ret
+int8x16_t test_vreinterpretq_s8_p128(poly128_t a) {
+  return vreinterpretq_s8_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_s16_p128
+// CHECK: ret
+int16x8_t test_vreinterpretq_s16_p128(poly128_t  a) {
+  return vreinterpretq_s16_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_s32_p128
+// CHECK: ret
+int32x4_t test_vreinterpretq_s32_p128(poly128_t a) {
+  return vreinterpretq_s32_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_s64_p128
+// CHECK: ret
+int64x2_t test_vreinterpretq_s64_p128(poly128_t  a) {
+  return vreinterpretq_s64_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_u8_p128
+// CHECK: ret
+uint8x16_t test_vreinterpretq_u8_p128(poly128_t  a) {
+  return vreinterpretq_u8_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_u16_p128
+// CHECK: ret
+uint16x8_t test_vreinterpretq_u16_p128(poly128_t  a) {
+  return vreinterpretq_u16_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_u32_p128
+// CHECK: ret
+uint32x4_t test_vreinterpretq_u32_p128(poly128_t  a) {
+  return vreinterpretq_u32_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_u64_p128
+// CHECK: ret
+uint64x2_t test_vreinterpretq_u64_p128(poly128_t  a) {
+  return vreinterpretq_u64_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_f32_p128
+// CHECK: ret
+float32x4_t test_vreinterpretq_f32_p128(poly128_t  a) {
+  return vreinterpretq_f32_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_f64_p128
+// CHECK: ret
+float64x2_t test_vreinterpretq_f64_p128(poly128_t  a) {
+  return vreinterpretq_f64_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p8_p128
+// CHECK: ret
+poly8x16_t test_vreinterpretq_p8_p128(poly128_t  a) {
+  return vreinterpretq_p8_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p16_p128
+// CHECK: ret
+poly16x8_t test_vreinterpretq_p16_p128(poly128_t  a) {
+  return vreinterpretq_p16_p128(a);
+}
+
+// CHECK-LABEL: test_vreinterpretq_p64_p128
+// CHECK: ret
+poly64x2_t test_vreinterpretq_p64_p128(poly128_t  a) {
+  return vreinterpretq_p64_p128(a);
+}
+
+
diff --git a/test/CodeGen/aarch64-poly64.c b/test/CodeGen/aarch64-poly64.c
index 3e19501..c071147 100644
--- a/test/CodeGen/aarch64-poly64.c
+++ b/test/CodeGen/aarch64-poly64.c
@@ -1,282 +1,299 @@
 // REQUIRES: aarch64-registered-target
+// REQUIRES: arm64-registered-target
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
-// RUN:   -ffp-contract=fast -S -O3 -o - %s | FileCheck %s
+// RUN:  -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK \
+// RUN:  --check-prefix=CHECK-AARCH64
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:  -ffp-contract=fast -S -O3 -o - %s | FileCheck %s --check-prefix=CHECK \
+// RUN:  --check-prefix=CHECK-ARM64
 
 // Test new aarch64 intrinsics with poly64
 
 #include <arm_neon.h>
 
 uint64x1_t test_vceq_p64(poly64x1_t a, poly64x1_t b) {
-  // CHECK: test_vceq_p64
+  // CHECK-LABEL: test_vceq_p64
   return vceq_p64(a, b);
   // CHECK: cmeq {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x2_t test_vceqq_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vceqq_p64
+  // CHECK-LABEL: test_vceqq_p64
   return vceqq_p64(a, b);
   // CHECK: cmeq {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 uint64x1_t test_vtst_p64(poly64x1_t a, poly64x1_t b) {
-  // CHECK: test_vtst_p64
+  // CHECK-LABEL: test_vtst_p64
   return vtst_p64(a, b);
   // CHECK: cmtst {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
 }
 
 uint64x2_t test_vtstq_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vtstq_p64
+  // CHECK-LABEL: test_vtstq_p64
   return vtstq_p64(a, b);
   // CHECK: cmtst {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x1_t test_vbsl_p64(poly64x1_t a, poly64x1_t b, poly64x1_t c) {
-  // CHECK: test_vbsl_p64
+  // CHECK-LABEL: test_vbsl_p64
   return vbsl_p64(a, b, c);
   // CHECK: bsl {{v[0-9]+}}.8b, {{v[0-9]+}}.8b, {{v[0-9]+}}.8b
 }
 
 poly64x2_t test_vbslq_p64(poly64x2_t a, poly64x2_t b, poly64x2_t c) {
-  // CHECK: test_vbslq_p64
+  // CHECK-LABEL: test_vbslq_p64
   return vbslq_p64(a, b, c);
   // CHECK: bsl {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b
 }
 
 poly64_t test_vget_lane_p64(poly64x1_t v) {
-  // CHECK: test_vget_lane_p64
+  // CHECK-LABEL: test_vget_lane_p64
   return vget_lane_p64(v, 0);
   // CHECK: fmov  {{x[0-9]+}}, {{d[0-9]+}}
 }
 
 poly64_t test_vgetq_lane_p64(poly64x2_t v) {
-  // CHECK: test_vgetq_lane_p64
+  // CHECK-LABEL: test_vgetq_lane_p64
   return vgetq_lane_p64(v, 1);
   // CHECK: umov  {{x[0-9]+}}, {{v[0-9]+}}.d[1]
 }
 
 poly64x1_t test_vset_lane_p64(poly64_t a, poly64x1_t v) {
-  // CHECK: test_vset_lane_p64
+  // CHECK-LABEL: test_vset_lane_p64
   return vset_lane_p64(a, v, 0);
   // CHECK: fmov  {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 poly64x2_t test_vsetq_lane_p64(poly64_t a, poly64x2_t v) {
-  // CHECK: test_vsetq_lane_p64
+  // CHECK-LABEL: test_vsetq_lane_p64
   return vsetq_lane_p64(a, v, 1);
   // CHECK: ins  {{v[0-9]+}}.d[1], {{x[0-9]+}}
 }
 
 poly64x1_t test_vcopy_lane_p64(poly64x1_t a, poly64x1_t b) {
-  // CHECK: test_vcopy_lane_p64
+  // CHECK-LABEL: test_vcopy_lane_p64
   return vcopy_lane_p64(a, 0, b, 0);
-  // CHECK: fmov  {{d[0-9]+}}, {{d[0-9]+}}
+  // CHECK-AARCH64: fmov  {{d[0-9]+}}, {{d[0-9]+}}
+
+  // CHECK-ARM64: orr v0.16b, v1.16b, v1.16b
 }
 
 poly64x2_t test_vcopyq_lane_p64(poly64x2_t a, poly64x1_t b) {
-  // CHECK: test_vcopyq_lane_p64
+  // CHECK-LABEL: test_vcopyq_lane_p64
   return vcopyq_lane_p64(a, 1, b, 0);
   // CHECK: ins  {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
 }
 
 poly64x2_t test_vcopyq_laneq_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vcopyq_laneq_p64
+  // CHECK-LABEL: test_vcopyq_laneq_p64
   return vcopyq_laneq_p64(a, 1, b, 1);
-  // CHECK: ins  {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: ins  {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[1]
 }
 
 poly64x1_t test_vcreate_p64(uint64_t a) {
-  // CHECK: test_vcreate_p64
+  // CHECK-LABEL: test_vcreate_p64
   return vcreate_p64(a);
   // CHECK: fmov  {{d[0-9]+}}, {{x[0-9]+}}
 }
 
 poly64x1_t test_vdup_n_p64(poly64_t a) {
-  // CHECK: test_vdup_n_p64
+  // CHECK-LABEL: test_vdup_n_p64
   return vdup_n_p64(a);
   // CHECK: fmov {{d[0-9]+}}, {{x[0-9]+}}
 }
 poly64x2_t test_vdupq_n_p64(poly64_t a) {
-  // CHECK: test_vdup_n_p64
+  // CHECK-LABEL: test_vdupq_n_p64
   return vdupq_n_p64(a);
   // CHECK: dup {{v[0-9]+}}.2d, {{x[0-9]+}}
 }
 
 poly64x1_t test_vdup_lane_p64(poly64x1_t vec) {
-  // CHECK: test_vdup_lane_p64
+  // CHECK-LABEL: test_vdup_lane_p64
   return vdup_lane_p64(vec, 0);
   // CHECK: ret
 }
 
 poly64x2_t test_vdupq_lane_p64(poly64x1_t vec) {
-  // CHECK: test_vdupq_lane_p64
+  // CHECK-LABEL: test_vdupq_lane_p64
   return vdupq_lane_p64(vec, 0);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[0]
 }
 
 poly64x2_t test_vdupq_laneq_p64(poly64x2_t vec) {
-  // CHECK: test_vdupq_laneq_p64
+  // CHECK-LABEL: test_vdupq_laneq_p64
   return vdupq_laneq_p64(vec, 1);
   // CHECK: dup {{v[0-9]+}}.2d, {{v[0-9]+}}.d[1]
 }
 
 poly64x2_t test_vcombine_p64(poly64x1_t low, poly64x1_t high) {
-  // CHECK: test_vcombine_p64
+  // CHECK-LABEL: test_vcombine_p64
   return vcombine_p64(low, high);
   // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
 }
 
 poly64x1_t test_vld1_p64(poly64_t const * ptr) {
-  // CHECK: test_vld1_p64
+  // CHECK-LABEL: test_vld1_p64
   return vld1_p64(ptr);
-  // CHECK:  ld1 {{{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK-AARCH64:  ld1 {{{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK-ARM64: ldr {{d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2_t test_vld1q_p64(poly64_t const * ptr) {
-  // CHECK: test_vld1q_p64
+  // CHECK-LABEL: test_vld1q_p64
   return vld1q_p64(ptr);
-  // CHECK:  ld1 {{{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK-AARCH64:  ld1 {{{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK-ARM64: ldr {{q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1_p64(poly64_t * ptr, poly64x1_t val) {
-  // CHECK: test_vst1_p64
+  // CHECK-LABEL: test_vst1_p64
   return vst1_p64(ptr, val);
-  // CHECK:  st1 {{{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK-AARCH64:  st1 {{{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK-ARM64: str {{d[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst1q_p64(poly64_t * ptr, poly64x2_t val) {
-  // CHECK: test_vst1q_p64
+  // CHECK-LABEL: test_vst1q_p64
   return vst1q_p64(ptr, val);
-  // CHECK:  st1 {{{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK-AARCH64:  st1 {{{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK-ARM64: str {{q[0-9]+}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x2_t test_vld2_p64(poly64_t const * ptr) {
-  // CHECK: test_vld2_p64
+  // CHECK-LABEL: test_vld2_p64
   return vld2_p64(ptr);
-  // CHECK: ld1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x2_t test_vld2q_p64(poly64_t const * ptr) {
-  // CHECK: test_vld2q_p64
+  // CHECK-LABEL: test_vld2q_p64
   return vld2q_p64(ptr);
-  // CHECK: ld2 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld2 {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x3_t test_vld3_p64(poly64_t const * ptr) {
-  // CHECK: test_vld3_p64
+  // CHECK-LABEL: test_vld3_p64
   return vld3_p64(ptr);
-  // CHECK: ld1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x3_t test_vld3q_p64(poly64_t const * ptr) {
-  // CHECK: test_vld3q_p64
+  // CHECK-LABEL: test_vld3q_p64
   return vld3q_p64(ptr);
-  // CHECK: ld3 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld3 {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1x4_t test_vld4_p64(poly64_t const * ptr) {
-  // CHECK: test_vld4_p64
+  // CHECK-LABEL: test_vld4_p64
   return vld4_p64(ptr);
-  // CHECK: ld1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld1 {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x2x4_t test_vld4q_p64(poly64_t const * ptr) {
-  // CHECK: test_vld4q_p64
+  // CHECK-LABEL: test_vld4q_p64
   return vld4q_p64(ptr);
-  // CHECK: ld4 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK: ld4 {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2_p64(poly64_t * ptr, poly64x1x2_t val) {
-  // CHECK: test_vst2_p64
+  // CHECK-LABEL: test_vst2_p64
   return vst2_p64(ptr, val);
-  // CHECK:  st1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st1 {{{ *v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst2q_p64(poly64_t * ptr, poly64x2x2_t val) {
-  // CHECK: test_vst2q_p64
+  // CHECK-LABEL: test_vst2q_p64
   return vst2q_p64(ptr, val);
-  // CHECK:  st2 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st2 {{{ *v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3_p64(poly64_t * ptr, poly64x1x3_t val) {
-  // CHECK: test_vst3_p64
+  // CHECK-LABEL: test_vst3_p64
   return vst3_p64(ptr, val);
-  // CHECK:  st1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st1 {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst3q_p64(poly64_t * ptr, poly64x2x3_t val) {
-  // CHECK: test_vst3q_p64
+  // CHECK-LABEL: test_vst3q_p64
   return vst3q_p64(ptr, val);
-  // CHECK:  st3 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st3 {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4_p64(poly64_t * ptr, poly64x1x4_t val) {
-  // CHECK: test_vst4_p64
+  // CHECK-LABEL: test_vst4_p64
   return vst4_p64(ptr, val);
-  // CHECK:  st1 {{{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d, {{v[0-9]+}}.1d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st1 {{{ *v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d, v[0-9]+.1d *}}}, [{{x[0-9]+|sp}}]
 }
 
 void test_vst4q_p64(poly64_t * ptr, poly64x2x4_t val) {
-  // CHECK: test_vst4q_p64
+  // CHECK-LABEL: test_vst4q_p64
   return vst4q_p64(ptr, val);
-  // CHECK:  st4 {{{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d}, [{{x[0-9]+|sp}}]
+  // CHECK:  st4 {{{ *v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d, v[0-9]+.2d *}}}, [{{x[0-9]+|sp}}]
 }
 
 poly64x1_t test_vext_p64(poly64x1_t a, poly64x1_t b) {
-  // CHECK: test_vext_p64
+  // CHECK-LABEL: test_vext_p64
   return vext_u64(a, b, 0);
 
 }
 
 poly64x2_t test_vextq_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vextq_p64
+  // CHECK-LABEL: test_vextq_p64
   return vextq_p64(a, b, 1);
-  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, #0x8
+  // CHECK: ext {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{v[0-9]+}}.16b, {{#0x8|#8}}
 }
 
 poly64x2_t test_vzip1q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vzip1q_p64
+  // CHECK-LABEL: test_vzip1q_p64
   return vzip1q_p64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-ARM64: zip1 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x2_t test_vzip2q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vzip2q_p64
+  // CHECK-LABEL: test_vzip2q_p64
   return vzip2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: zip2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x2_t test_vuzp1q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vuzp1q_p64
+  // CHECK-LABEL: test_vuzp1q_p64
   return vuzp1q_p64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-ARM64: zip1 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x2_t test_vuzp2q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vuzp2q_p64
+  // CHECK-LABEL: test_vuzp2q_p64
   return vuzp2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: zip2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x2_t test_vtrn1q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vtrn1q_p64
+  // CHECK-LABEL: test_vtrn1q_p64
   return vtrn1q_p64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[1], {{v[0-9]+}}.d[0]
+  // CHECK-ARM64: zip1 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x2_t test_vtrn2q_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vtrn2q_p64
+  // CHECK-LABEL: test_vtrn2q_p64
   return vtrn2q_u64(a, b);
-  // CHECK: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-AARCH64: ins {{v[0-9]+}}.d[0], {{v[0-9]+}}.d[1]
+  // CHECK-ARM64: zip2 {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, {{v[0-9]+}}.2d
 }
 
 poly64x1_t test_vsri_n_p64(poly64x1_t a, poly64x1_t b) {
-  // CHECK: test_vsri_n_p64
+  // CHECK-LABEL: test_vsri_n_p64
   return vsri_n_p64(a, b, 33);
   // CHECK: sri {{d[0-9]+}}, {{d[0-9]+}}, #33
 }
 
 poly64x2_t test_vsriq_n_p64(poly64x2_t a, poly64x2_t b) {
-  // CHECK: test_vsriq_n_p64
+  // CHECK-LABEL: test_vsriq_n_p64
   return vsriq_n_p64(a, b, 64);
   // CHECK: sri {{v[0-9]+}}.2d, {{v[0-9]+}}.2d, #64
 }
diff --git a/test/CodeGen/aarch64-type-sizes.c b/test/CodeGen/aarch64-type-sizes.c
index 3b9c9fc..f9e55f7 100644
--- a/test/CodeGen/aarch64-type-sizes.c
+++ b/test/CodeGen/aarch64-type-sizes.c
@@ -1,7 +1,11 @@
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -emit-llvm -w -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -emit-llvm -w -o - %s | FileCheck --check-prefix=CHECK --check-prefix=CHECK-LE %s
 
+// RUN: %clang_cc1 -triple aarch64_be-none-linux-gnu -emit-llvm -w -o - %s | FileCheck --check-prefix=CHECK --check-prefix=CHECK-BE %s
 // char by definition has size 1
 
+// CHECK-LE: target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
+// CHECK-BE: target datalayout = "E-m:e-i64:64-i128:128-n32:64-S128"
+
 int check_short() {
   return sizeof(short);
 // CHECK: ret i32 2
diff --git a/test/CodeGen/aarch64-varargs.c b/test/CodeGen/aarch64-varargs.c
index 3d9cd86..053982b 100644
--- a/test/CodeGen/aarch64-varargs.c
+++ b/test/CodeGen/aarch64-varargs.c
@@ -1,4 +1,7 @@
-// RUN: %clang_cc1 -triple aarch64 -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64 -emit-llvm -o - %s | FileCheck -check-prefix=CHECK --check-prefix=CHECK-LE %s
+// RUN: %clang_cc1 -triple aarch64_be -emit-llvm -o - %s | FileCheck --check-prefix=CHECK --check-prefix=CHECK-BE %s 
+// RUN: %clang_cc1 -triple arm64-linux-gnu -emit-llvm -o - %s | FileCheck --check-prefix=CHECK --check-prefix=CHECK-LE %s
+
 #include <stdarg.h>
 
 // Obviously there's more than one way to implement va_arg. This test should at
@@ -22,6 +25,9 @@
 // CHECK: [[VAARG_IN_REG]]
 // CHECK: [[REG_TOP:%[a-z_0-9]+]] = load i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 1)
 // CHECK: [[REG_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[REG_TOP]], i32 [[GR_OFFS]]
+// CHECK-BE: [[REG_ADDR_VAL:%[0-9]+]] = ptrtoint i8* [[REG_ADDR]] to i64
+// CHECK-BE: [[REG_ADDR_VAL_ALIGNED:%[a-z_0-9]*]] = add i64 [[REG_ADDR_VAL]], 4
+// CHECK-BE: [[REG_ADDR:%[0-9]+]] = inttoptr i64 [[REG_ADDR_VAL_ALIGNED]] to i8*
 // CHECK: [[FROMREG_ADDR:%[a-z_0-9]+]] = bitcast i8* [[REG_ADDR]] to i32*
 // CHECK: br label %[[VAARG_END:[a-z._0-9]+]]
 
@@ -29,6 +35,9 @@
 // CHECK: [[STACK:%[a-z_0-9]+]] = load i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 0)
 // CHECK: [[NEW_STACK:%[a-z_0-9]+]] = getelementptr i8* [[STACK]], i32 8
 // CHECK: store i8* [[NEW_STACK]], i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 0)
+// CHECK-BE: [[STACK_VAL:%[0-9]+]] = ptrtoint i8* [[STACK]] to i64
+// CHECK-BE: [[STACK_VAL_ALIGNED:%[a-z_0-9]*]] = add i64 [[STACK_VAL]], 4
+// CHECK-BE: [[STACK:%[0-9]+]] = inttoptr i64 [[STACK_VAL_ALIGNED]] to i8*
 // CHECK: [[FROMSTACK_ADDR:%[a-z_0-9]+]] = bitcast i8* [[STACK]] to i32*
 // CHECK: br label %[[VAARG_END]]
 
@@ -154,7 +163,7 @@
   return va_arg(the_list, double);
 // CHECK: [[VR_OFFS:%[a-z_0-9]+]] = load i32* getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 4)
 // CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[VR_OFFS]], 0
-// CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK]], label %[[VAARG_MAYBE_REG]]
+// CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK:[a-z_.0-9]+]], label %[[VAARG_MAYBE_REG]]
 
 // CHECK: [[VAARG_MAYBE_REG]]
 // CHECK: [[NEW_REG_OFFS:%[a-z_0-9]+]] = add i32 [[VR_OFFS]], 16
@@ -165,8 +174,11 @@
 // CHECK: [[VAARG_IN_REG]]
 // CHECK: [[REG_TOP:%[a-z_0-9]+]] = load i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 2)
 // CHECK: [[REG_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[REG_TOP]], i32 [[VR_OFFS]]
+// CHECK-BE: [[REG_ADDR_VAL:%[0-9]+]] = ptrtoint i8* [[REG_ADDR]] to i64
+// CHECK-BE: [[REG_ADDR_VAL_ALIGNED:%[a-z_0-9]*]] = add i64 [[REG_ADDR_VAL]], 8
+// CHECK-BE: [[REG_ADDR:%[0-9]+]] = inttoptr i64 [[REG_ADDR_VAL_ALIGNED]] to i8*
 // CHECK: [[FROMREG_ADDR:%[a-z_0-9]+]] = bitcast i8* [[REG_ADDR]] to double*
-// CHECK: br label %[[VAARG_END]]
+// CHECK: br label %[[VAARG_END:[a-z._0-9]+]]
 
 // CHECK: [[VAARG_ON_STACK]]
 // CHECK: [[STACK:%[a-z_0-9]+]] = load i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 0)
@@ -201,12 +213,14 @@
 // CHECK: [[VAARG_IN_REG]]
 // CHECK: [[REG_TOP:%[a-z_0-9]+]] = load i8** getelementptr inbounds (%struct.__va_list* @the_list, i32 0, i32 2)
 // CHECK: [[FIRST_REG:%[a-z_0-9]+]] = getelementptr i8* [[REG_TOP]], i32 [[VR_OFFS]]
-// CHECK: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 0
+// CHECK-LE: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 0
+// CHECK-BE: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 12
 // CHECK: [[EL_TYPED:%[a-z_0-9]+]] = bitcast i8* [[EL_ADDR]] to float*
 // CHECK: [[EL_TMPADDR:%[a-z_0-9]+]] = getelementptr inbounds [2 x float]* %[[TMP_HFA:[a-z_.0-9]+]], i32 0, i32 0
 // CHECK: [[EL:%[a-z_0-9]+]] = load float* [[EL_TYPED]]
 // CHECK: store float [[EL]], float* [[EL_TMPADDR]]
-// CHECK: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 16
+// CHECK-LE: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 16
+// CHECK-BE: [[EL_ADDR:%[a-z_0-9]+]] = getelementptr i8* [[FIRST_REG]], i32 28
 // CHECK: [[EL_TYPED:%[a-z_0-9]+]] = bitcast i8* [[EL_ADDR]] to float*
 // CHECK: [[EL_TMPADDR:%[a-z_0-9]+]] = getelementptr inbounds [2 x float]* %[[TMP_HFA]], i32 0, i32 1
 // CHECK: [[EL:%[a-z_0-9]+]] = load float* [[EL_TYPED]]
diff --git a/test/CodeGen/alias.c b/test/CodeGen/alias.c
index efa94b3..4a89b13 100644
--- a/test/CodeGen/alias.c
+++ b/test/CodeGen/alias.c
@@ -14,6 +14,8 @@
 extern void f1(void);
 extern void f1(void) __attribute((alias("f0")));
 // CHECKBASIC-DAG: @f1 = alias void ()* @f0
+// CHECKBASIC-DAG: @test8_foo = alias weak bitcast (void ()* @test8_bar to void (...)*)
+// CHECKBASIC-DAG: @test8_zed = alias bitcast (void ()* @test8_bar to void (...)*)
 // CHECKBASIC: define void @f0() [[NUW:#[0-9]+]] {
 
 // Make sure that aliases cause referenced values to be emitted.
@@ -48,3 +50,7 @@
 // CHECKBASIC: attributes [[NUW]] = { nounwind{{.*}} }
 
 // CHECKCC: attributes [[NUW]] = { nounwind{{.*}} }
+
+void test8_bar() {}
+void test8_foo() __attribute__((weak, alias("test8_bar")));
+void test8_zed() __attribute__((alias("test8_foo")));
diff --git a/test/CodeGen/altivec.c b/test/CodeGen/altivec.c
index 6c924a7..c25d9f4 100644
--- a/test/CodeGen/altivec.c
+++ b/test/CodeGen/altivec.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -faltivec -triple powerpc-unknown-unknown -emit-llvm %s -o - | FileCheck %s
 
 // Check initialization
diff --git a/test/CodeGen/arm-aapcs-vfp.c b/test/CodeGen/arm-aapcs-vfp.c
index 0e102f3..cd9aaa4 100644
--- a/test/CodeGen/arm-aapcs-vfp.c
+++ b/test/CodeGen/arm-aapcs-vfp.c
@@ -12,7 +12,15 @@
 // RUN:  -ffreestanding \
 // RUN:  -emit-llvm -w -o - %s | FileCheck %s
 
+// RUN: %clang_cc1 -triple arm64-apple-darwin9 \
+// RUN:   -ffreestanding \
+// RUN:   -emit-llvm -w -o - %s | FileCheck -check-prefix=CHECK64 %s
+
+#ifdef __arm64__
 #include <arm_neon.h>
+#else
+#include <arm_neon.h>
+#endif
 
 struct homogeneous_struct {
   float f[2];
@@ -20,15 +28,22 @@
   float f4;
 };
 // CHECK: define arm_aapcs_vfpcc %struct.homogeneous_struct @test_struct(float %{{.*}}, float %{{.*}}, float %{{.*}}, float %{{.*}})
+// CHECK64: define %struct.homogeneous_struct @test_struct(float %{{.*}}, float %{{.*}}, float %{{.*}}, float %{{.*}})
 extern struct homogeneous_struct struct_callee(struct homogeneous_struct);
 struct homogeneous_struct test_struct(struct homogeneous_struct arg) {
   return struct_callee(arg);
 }
 
+// CHECK: define arm_aapcs_vfpcc void @test_struct_variadic(%struct.homogeneous_struct* {{.*}}, [4 x i32] %{{.*}}, ...)
+struct homogeneous_struct test_struct_variadic(struct homogeneous_struct arg, ...) {
+  return struct_callee(arg);
+}
+
 struct nested_array {
   double d[4];
 };
 // CHECK: define arm_aapcs_vfpcc void @test_array(double %{{.*}}, double %{{.*}}, double %{{.*}}, double %{{.*}})
+// CHECK64: define void @test_array(double %{{.*}}, double %{{.*}}, double %{{.*}}, double %{{.*}})
 extern void array_callee(struct nested_array);
 void test_array(struct nested_array arg) {
   array_callee(arg);
@@ -36,6 +51,7 @@
 
 extern void complex_callee(__complex__ double);
 // CHECK: define arm_aapcs_vfpcc void @test_complex(double %{{.*}}, double %{{.*}})
+// CHECK64: define void @test_complex(double %{{.*}}, double %{{.*}})
 void test_complex(__complex__ double cd) {
   complex_callee(cd);
 }
@@ -57,6 +73,9 @@
   float f4;
 };
 // CHECK: define arm_aapcs_vfpcc void @test_big([5 x i32] %{{.*}})
+// CHECK64: define void @test_big(%struct.big_struct* %{{.*}})
+// CHECK64: call void @llvm.memcpy
+// CHECK64: call void @big_callee(%struct.big_struct*
 extern void big_callee(struct big_struct);
 void test_big(struct big_struct arg) {
   big_callee(arg);
@@ -70,6 +89,7 @@
   int i2;
 };
 // CHECK: define arm_aapcs_vfpcc void @test_hetero([2 x i32] %{{.*}})
+// CHECK64: define void @test_hetero(i64 %{{.*}})
 extern void hetero_callee(struct heterogeneous_struct);
 void test_hetero(struct heterogeneous_struct arg) {
   hetero_callee(arg);
@@ -77,6 +97,7 @@
 
 // Neon multi-vector types are homogeneous aggregates.
 // CHECK: define arm_aapcs_vfpcc <16 x i8> @f0(<16 x i8> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}})
+// CHECK64: define <16 x i8> @f0(<16 x i8> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}})
 int8x16_t f0(int8x16x4_t v4) {
   return vaddq_s8(v4.val[0], v4.val[3]);
 }
@@ -90,11 +111,22 @@
   int16x4_t v4;
 };
 // CHECK: define arm_aapcs_vfpcc void @test_neon(<8 x i8> %{{.*}}, <8 x i8> %{{.*}}, <2 x i32> %{{.*}}, <4 x i16> %{{.*}})
+// CHECK64: define void @test_neon(<8 x i8> %{{.*}}, <8 x i8> %{{.*}}, <2 x i32> %{{.*}}, <4 x i16> %{{.*}})
 extern void neon_callee(struct neon_struct);
 void test_neon(struct neon_struct arg) {
   neon_callee(arg);
 }
 
-// CHECK-LABEL: define arm_aapcs_vfpcc void @f33(%struct.s33* byval %s)
+// CHECK-LABEL: define arm_aapcs_vfpcc void @f33(%struct.s33* byval align 1 %s)
 struct s33 { char buf[32*32]; };
 void f33(struct s33 s) { }
+
+typedef struct { long long x; int y; } struct_long_long_int;
+// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_1(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, i64 %k, i32 %l)
+void test_vfp_stack_gpr_split_1(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, long long k, int l) {}
+
+// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_2(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, [3 x i32], i64 %k.0, i32 %k.1)
+void test_vfp_stack_gpr_split_2(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, struct_long_long_int k) {}
+
+// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_3(%struct.struct_long_long_int* noalias sret %agg.result, double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, [3 x i32], i64 %k.0, i32 %k.1)
+struct_long_long_int test_vfp_stack_gpr_split_3(double a, double b, double c, double d, double e, double f, double g, double h, double i, struct_long_long_int k) {}
diff --git a/test/CodeGen/arm-arguments.c b/test/CodeGen/arm-arguments.c
index b6bac9a..2c5df91 100644
--- a/test/CodeGen/arm-arguments.c
+++ b/test/CodeGen/arm-arguments.c
@@ -176,8 +176,8 @@
 // PR13350
 struct s33 { char buf[32*32]; };
 void f33(struct s33 s) { }
-// APCS-GNU-LABEL: define void @f33(%struct.s33* byval %s)
-// AAPCS-LABEL: define arm_aapcscc void @f33(%struct.s33* byval %s)
+// APCS-GNU-LABEL: define void @f33(%struct.s33* byval align 1 %s)
+// AAPCS-LABEL: define arm_aapcscc void @f33(%struct.s33* byval align 1 %s)
 
 // PR14048
 struct s34 { char c; };
@@ -209,14 +209,14 @@
                             *(float32x4_t *)&s2);
   return v;
 }
-// APCS-GNU-LABEL: define <4 x float> @f35(i32 %i, %struct.s35* byval, %struct.s35* byval)
+// APCS-GNU-LABEL: define <4 x float> @f35(i32 %i, %struct.s35* byval align 16, %struct.s35* byval align 16)
 // APCS-GNU: %[[a:.*]] = alloca %struct.s35, align 16
 // APCS-GNU: %[[b:.*]] = bitcast %struct.s35* %[[a]] to i8*
 // APCS-GNU: %[[c:.*]] = bitcast %struct.s35* %0 to i8*
 // APCS-GNU: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %[[b]], i8* %[[c]]
 // APCS-GNU: %[[d:.*]] = bitcast %struct.s35* %[[a]] to <4 x float>*
 // APCS-GNU: load <4 x float>* %[[d]], align 16
-// AAPCS-LABEL: define arm_aapcscc <4 x float> @f35(i32 %i, %struct.s35* byval, %struct.s35* byval)
+// AAPCS-LABEL: define arm_aapcscc <4 x float> @f35(i32 %i, %struct.s35* byval align 16, %struct.s35* byval align 16)
 // AAPCS: %[[a:.*]] = alloca %struct.s35, align 16
 // AAPCS: %[[b:.*]] = bitcast %struct.s35* %[[a]] to i8*
 // AAPCS: %[[c:.*]] = bitcast %struct.s35* %0 to i8*
diff --git a/test/CodeGen/arm-byval-align.c b/test/CodeGen/arm-byval-align.c
new file mode 100644
index 0000000..aa22503
--- /dev/null
+++ b/test/CodeGen/arm-byval-align.c
@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 -triple=armv7-none-eabi < %s -S -emit-llvm | FileCheck %s
+
+struct foo {
+  long long a;
+  char b;
+  int c:16;
+  int d[16];
+};
+
+// CHECK: %struct.foo* byval align 8 %z
+long long bar(int a, int b, int c, int d, int e,
+              struct foo z) {
+  return z.a;
+}
diff --git a/test/CodeGen/arm-cortex-cpus.c b/test/CodeGen/arm-cortex-cpus.c
new file mode 100644
index 0000000..d8d9830
--- /dev/null
+++ b/test/CodeGen/arm-cortex-cpus.c
@@ -0,0 +1,11 @@
+// REQUIRES: arm-registered-target
+
+// Check that Cortex-M cores don't enable hwdiv-arm (and don't emit Tag_DIV_use)
+//
+// This target feature doesn't affect C predefines, nor the generated IR;
+// only the build attributes in generated assembly and object files are affected.
+
+// RUN: %clang -target arm-linux-gnueabi -mcpu=cortex-m3 -S %s -o - | FileCheck %s
+// RUN: %clang -target arm-linux-gnueabi -mcpu=cortex-m4 -S %s -o - | FileCheck %s
+// CHECK-NOT: .eabi_attribute	44
+
diff --git a/test/CodeGen/arm-homogenous.c b/test/CodeGen/arm-homogenous.c
index 5d21088..ad21444 100644
--- a/test/CodeGen/arm-homogenous.c
+++ b/test/CodeGen/arm-homogenous.c
@@ -1,6 +1,11 @@
 // REQUIRES: arm-registered-target
 // RUN: %clang_cc1 -triple armv7---eabi -target-abi aapcs -mfloat-abi hard -emit-llvm %s -o - | FileCheck %s
 
+// RUN: %clang_cc1 -triple arm64-apple-darwin9 -target-abi darwinpcs \
+// RUN:  -ffreestanding -emit-llvm -w -o - %s | FileCheck -check-prefix=CHECK64 %s
+
+// RUN: %clang_cc1 -triple arm64-linux-gnu -ffreestanding -emit-llvm -w -o - %s \
+// RUN:   | FileCheck --check-prefix=CHECK64-AAPCS %s
 typedef long long int64_t;
 typedef unsigned int uint32_t;
 
@@ -170,9 +175,21 @@
 void test_struct_of_four_doubles(void) {
 // CHECK: test_struct_of_four_doubles
 // CHECK: call arm_aapcs_vfpcc void @takes_struct_of_four_doubles(double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, [6 x float] undef, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}})
+// CHECK64: test_struct_of_four_doubles
+// CHECK64: call void @takes_struct_of_four_doubles(double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, [3 x float] undef, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}})
+// CHECK64-AAPCS: test_struct_of_four_doubles
+// CHECK64-AAPCS: call void @takes_struct_of_four_doubles(double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, [3 x float] undef, [4 x double] {{.*}}, double {{.*}})
   takes_struct_of_four_doubles(3.0, g_s4d, g_s4d, 4.0);
 }
 
+extern void takes_struct_of_four_doubles_variadic(double a, struct_of_four_doubles b, struct_of_four_doubles c, double d, ...);
+
+void test_struct_of_four_doubles_variadic(void) {
+// CHECK: test_struct_of_four_doubles_variadic
+// CHECK: call arm_aapcs_vfpcc void (double, [4 x i64], [4 x i64], double, ...)* @takes_struct_of_four_doubles_variadic(double {{.*}}, [4 x i64] {{.*}}, [4 x i64] {{.*}}, double {{.*}})
+  takes_struct_of_four_doubles_variadic(3.0, g_s4d, g_s4d, 4.0);
+}
+
 extern void takes_struct_with_backfill(float f1, double a, float f2, struct_of_four_doubles b, struct_of_four_doubles c, double d);
 void test_struct_with_backfill(void) {
 // CHECK: test_struct_with_backfill
@@ -194,9 +211,24 @@
 void test_struct_of_vecs(void) {
 // CHECK: test_struct_of_vecs
 // CHECK: call arm_aapcs_vfpcc void @takes_struct_of_vecs(double {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, [6 x float] undef, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, double {{.*}})
+// CHECK64: test_struct_of_vecs
+// CHECK64: call void @takes_struct_of_vecs(double {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, [3 x float] undef, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, double {{.*}})
+// CHECK64-AAPCS: test_struct_of_vecs
+// CHECK64-AAPCS: call void @takes_struct_of_vecs(double {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, [3 x float] undef, [4 x double] {{.*}})
   takes_struct_of_vecs(3.0, g_vec, g_vec, 4.0);
 }
 
+typedef struct {
+  double a;
+  long double b;
+} struct_of_double_and_long_double;
+struct_of_double_and_long_double g_dld;
+
+struct_of_double_and_long_double test_struct_of_double_and_long_double(void) {
+  return g_dld;
+}
+// CHECK: define arm_aapcs_vfpcc %struct.struct_of_double_and_long_double @test_struct_of_double_and_long_double()
+
 // FIXME: Tests necessary:
 //         - Vectors
 //         - C++ stuff
diff --git a/test/CodeGen/arm-neon-vcvtX.c b/test/CodeGen/arm-neon-vcvtX.c
new file mode 100644
index 0000000..ff8ce7e
--- /dev/null
+++ b/test/CodeGen/arm-neon-vcvtX.c
@@ -0,0 +1,99 @@
+// RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 -ffreestanding -O1 -emit-llvm %s -o - | FileCheck %s
+
+#include <arm_neon.h>
+
+int32x2_t test_vcvta_s32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvta_s32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtas.v2i32.v2f32(<2 x float> %a)
+  return vcvta_s32_f32(a);
+}
+
+uint32x2_t test_vcvta_u32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvta_u32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtau.v2i32.v2f32(<2 x float> %a)
+  return vcvta_u32_f32(a);
+}
+
+int32x4_t test_vcvtaq_s32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtaq_s32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtas.v4i32.v4f32(<4 x float> %a)
+  return vcvtaq_s32_f32(a);
+}
+
+uint32x4_t test_vcvtaq_u32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtaq_u32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtau.v4i32.v4f32(<4 x float> %a)
+  return vcvtaq_u32_f32(a);
+}
+
+int32x2_t test_vcvtn_s32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtn_s32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtns.v2i32.v2f32(<2 x float> %a)
+  return vcvtn_s32_f32(a);
+}
+
+uint32x2_t test_vcvtn_u32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtn_u32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtnu.v2i32.v2f32(<2 x float> %a)
+  return vcvtn_u32_f32(a);
+}
+
+int32x4_t test_vcvtnq_s32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtnq_s32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtns.v4i32.v4f32(<4 x float> %a)
+  return vcvtnq_s32_f32(a);
+}
+
+uint32x4_t test_vcvtnq_u32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtnq_u32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtnu.v4i32.v4f32(<4 x float> %a)
+  return vcvtnq_u32_f32(a);
+}
+
+int32x2_t test_vcvtp_s32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtp_s32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtps.v2i32.v2f32(<2 x float> %a)
+  return vcvtp_s32_f32(a);
+}
+
+uint32x2_t test_vcvtp_u32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtp_u32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtpu.v2i32.v2f32(<2 x float> %a)
+  return vcvtp_u32_f32(a);
+}
+
+int32x4_t test_vcvtpq_s32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtpq_s32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtps.v4i32.v4f32(<4 x float> %a)
+  return vcvtpq_s32_f32(a);
+}
+
+uint32x4_t test_vcvtpq_u32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtpq_u32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtpu.v4i32.v4f32(<4 x float> %a)
+  return vcvtpq_u32_f32(a);
+}
+
+int32x2_t test_vcvtm_s32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtm_s32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtms.v2i32.v2f32(<2 x float> %a)
+  return vcvtm_s32_f32(a);
+}
+
+uint32x2_t test_vcvtm_u32_f32(float32x2_t a) {
+  // CHECK-LABEL: test_vcvtm_u32_f32
+  // CHECK-LABEL: call <2 x i32> @llvm.arm.neon.vcvtmu.v2i32.v2f32(<2 x float> %a)
+  return vcvtm_u32_f32(a);
+}
+
+int32x4_t test_vcvtmq_s32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtmq_s32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtms.v4i32.v4f32(<4 x float> %a)
+  return vcvtmq_s32_f32(a);
+}
+
+uint32x4_t test_vcvtmq_u32_f32(float32x4_t a) {
+  // CHECK-LABEL: test_vcvtmq_u32_f32
+  // CHECK-LABEL: call <4 x i32> @llvm.arm.neon.vcvtmu.v4i32.v4f32(<4 x float> %a)
+  return vcvtmq_u32_f32(a);
+}
diff --git a/test/CodeGen/arm64-abi-vector.c b/test/CodeGen/arm64-abi-vector.c
new file mode 100644
index 0000000..502fb08
--- /dev/null
+++ b/test/CodeGen/arm64-abi-vector.c
@@ -0,0 +1,430 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7 -target-abi darwinpcs -emit-llvm -o - %s | FileCheck %s
+
+#include <stdarg.h>
+
+typedef __attribute__(( ext_vector_type(3) ))  char __char3;
+typedef __attribute__(( ext_vector_type(4) ))  char __char4;
+typedef __attribute__(( ext_vector_type(5) ))  char __char5;
+typedef __attribute__(( ext_vector_type(9) ))  char __char9;
+typedef __attribute__(( ext_vector_type(19) )) char __char19;
+typedef __attribute__(( ext_vector_type(3) ))  short __short3;
+typedef __attribute__(( ext_vector_type(5) ))  short __short5;
+typedef __attribute__(( ext_vector_type(3) ))  int __int3;
+typedef __attribute__(( ext_vector_type(5) ))  int __int5;
+typedef __attribute__(( ext_vector_type(3) ))  double __double3;
+
+double varargs_vec_3c(int fixed, ...) {
+// CHECK: varargs_vec_3c
+// CHECK: alloca <3 x i8>, align 4
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <3 x i8>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char3 c3 = va_arg(ap, __char3);
+  sum = sum + c3.x + c3.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_3c(__char3 *in) {
+// CHECK: test_3c
+// CHECK: call double (i32, ...)* @varargs_vec_3c(i32 3, i32 {{%.*}})
+  return varargs_vec_3c(3, *in);
+}
+
+double varargs_vec_4c(int fixed, ...) {
+// CHECK: varargs_vec_4c
+// CHECK: alloca <4 x i8>, align 4
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <4 x i8>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char4 c4 = va_arg(ap, __char4);
+  sum = sum + c4.x + c4.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_4c(__char4 *in) {
+// CHECK: test_4c
+// CHECK: call double (i32, ...)* @varargs_vec_4c(i32 4, i32 {{%.*}})
+  return varargs_vec_4c(4, *in);
+}
+
+double varargs_vec_5c(int fixed, ...) {
+// CHECK: varargs_vec_5c
+// CHECK: alloca <5 x i8>, align 8
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <5 x i8>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char5 c5 = va_arg(ap, __char5);
+  sum = sum + c5.x + c5.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_5c(__char5 *in) {
+// CHECK: test_5c
+// CHECK: call double (i32, ...)* @varargs_vec_5c(i32 5, <2 x i32> {{%.*}})
+  return varargs_vec_5c(5, *in);
+}
+
+double varargs_vec_9c(int fixed, ...) {
+// CHECK: varargs_vec_9c
+// CHECK: alloca <9 x i8>, align 16
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <9 x i8>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char9 c9 = va_arg(ap, __char9);
+  sum = sum + c9.x + c9.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_9c(__char9 *in) {
+// CHECK: test_9c
+// CHECK: call double (i32, ...)* @varargs_vec_9c(i32 9, <4 x i32> {{%.*}})
+  return varargs_vec_9c(9, *in);
+}
+
+double varargs_vec_19c(int fixed, ...) {
+// CHECK: varargs_vec_19c
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <19 x i8>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char19 c19 = va_arg(ap, __char19);
+  sum = sum + c19.x + c19.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_19c(__char19 *in) {
+// CHECK: test_19c
+// CHECK: call double (i32, ...)* @varargs_vec_19c(i32 19, <19 x i8>* {{%.*}})
+  return varargs_vec_19c(19, *in);
+}
+
+double varargs_vec_3s(int fixed, ...) {
+// CHECK: varargs_vec_3s
+// CHECK: alloca <3 x i16>, align 8
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <3 x i16>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __short3 c3 = va_arg(ap, __short3);
+  sum = sum + c3.x + c3.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_3s(__short3 *in) {
+// CHECK: test_3s
+// CHECK: call double (i32, ...)* @varargs_vec_3s(i32 3, <2 x i32> {{%.*}})
+  return varargs_vec_3s(3, *in);
+}
+
+double varargs_vec_5s(int fixed, ...) {
+// CHECK: varargs_vec_5s
+// CHECK: alloca <5 x i16>, align 16
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <5 x i16>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __short5 c5 = va_arg(ap, __short5);
+  sum = sum + c5.x + c5.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_5s(__short5 *in) {
+// CHECK: test_5s
+// CHECK: call double (i32, ...)* @varargs_vec_5s(i32 5, <4 x i32> {{%.*}})
+  return varargs_vec_5s(5, *in);
+}
+
+double varargs_vec_3i(int fixed, ...) {
+// CHECK: varargs_vec_3i
+// CHECK: alloca <3 x i32>, align 16
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <3 x i32>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __int3 c3 = va_arg(ap, __int3);
+  sum = sum + c3.x + c3.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_3i(__int3 *in) {
+// CHECK: test_3i
+// CHECK: call double (i32, ...)* @varargs_vec_3i(i32 3, <4 x i32> {{%.*}})
+  return varargs_vec_3i(3, *in);
+}
+
+double varargs_vec_5i(int fixed, ...) {
+// CHECK: varargs_vec_5i
+// CHECK: alloca <5 x i32>, align 16
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <5 x i32>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __int5 c5 = va_arg(ap, __int5);
+  sum = sum + c5.x + c5.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_5i(__int5 *in) {
+// CHECK: test_5i
+// CHECK: call double (i32, ...)* @varargs_vec_5i(i32 5, <5 x i32>* {{%.*}})
+  return varargs_vec_5i(5, *in);
+}
+
+double varargs_vec_3d(int fixed, ...) {
+// CHECK: varargs_vec_3d
+// CHECK: alloca <3 x double>, align 16
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <3 x double>*
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __double3 c3 = va_arg(ap, __double3);
+  sum = sum + c3.x + c3.y;
+  va_end(ap);
+  return sum;
+}
+
+double test_3d(__double3 *in) {
+// CHECK: test_3d
+// CHECK: call double (i32, ...)* @varargs_vec_3d(i32 3, <3 x double>* {{%.*}})
+  return varargs_vec_3d(3, *in);
+}
+
+double varargs_vec(int fixed, ...) {
+// CHECK: varargs_vec
+  va_list ap;
+  double sum = fixed;
+  va_start(ap, fixed);
+  __char3 c3 = va_arg(ap, __char3);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <3 x i8>*
+  sum = sum + c3.x + c3.y;
+  __char5 c5 = va_arg(ap, __char5);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <5 x i8>*
+  sum = sum + c5.x + c5.y;
+  __char9 c9 = va_arg(ap, __char9);
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <9 x i8>*
+  sum = sum + c9.x + c9.y;
+  __char19 c19 = va_arg(ap, __char19);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <19 x i8>*
+  sum = sum + c19.x + c19.y;
+  __short3 s3 = va_arg(ap, __short3);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: bitcast i8* [[AP_CUR]] to <3 x i16>*
+  sum = sum + s3.x + s3.y;
+  __short5 s5 = va_arg(ap, __short5);
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <5 x i16>*
+  sum = sum + s5.x + s5.y;
+  __int3 i3 = va_arg(ap, __int3);
+// CHECK: [[ALIGN:%.*]] = and i64 {{%.*}}, -16
+// CHECK: [[AP_ALIGN:%.*]] = inttoptr i64 [[ALIGN]] to i8*
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_ALIGN]], i32 16
+// CHECK: bitcast i8* [[AP_ALIGN]] to <3 x i32>*
+  sum = sum + i3.x + i3.y;
+  __int5 i5 = va_arg(ap, __int5);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <5 x i32>*
+  sum = sum + i5.x + i5.y;
+  __double3 d3 = va_arg(ap, __double3);
+// CHECK: [[AP_NEXT:%.*]] = getelementptr i8* [[AP_CUR:%.*]], i32 8
+// CHECK: [[VAR:%.*]] = bitcast i8* [[AP_CUR]] to i8**
+// CHECK: [[VAR2:%.*]] = load i8** [[VAR]]
+// CHECK: bitcast i8* [[VAR2]] to <3 x double>*
+  sum = sum + d3.x + d3.y;
+  va_end(ap);
+  return sum;
+}
+
+double test(__char3 *c3, __char5 *c5, __char9 *c9, __char19 *c19,
+            __short3 *s3, __short5 *s5, __int3 *i3, __int5 *i5,
+            __double3 *d3) {
+  double ret = varargs_vec(3, *c3, *c5, *c9, *c19, *s3, *s5, *i3, *i5, *d3);
+// CHECK: call double (i32, ...)* @varargs_vec(i32 3, i32 {{%.*}}, <2 x i32> {{%.*}}, <4 x i32> {{%.*}}, <19 x i8>* {{%.*}}, <2 x i32> {{%.*}}, <4 x i32> {{%.*}}, <4 x i32> {{%.*}}, <5 x i32>* {{%.*}}, <3 x double>* {{%.*}})
+  return ret;
+}
+
+__attribute__((noinline)) double args_vec_3c(int fixed, __char3 c3) {
+// CHECK: args_vec_3c
+// CHECK: [[C3:%.*]] = alloca <3 x i8>, align 4
+// CHECK: [[TMP:%.*]] = bitcast <3 x i8>* [[C3]] to i32*
+// CHECK: store i32 {{%.*}}, i32* [[TMP]]
+  double sum = fixed;
+  sum = sum + c3.x + c3.y;
+  return sum;
+}
+
+double fixed_3c(__char3 *in) {
+// CHECK: fixed_3c
+// CHECK: call double @args_vec_3c(i32 3, i32 {{%.*}})
+  return args_vec_3c(3, *in);
+}
+
+__attribute__((noinline)) double args_vec_5c(int fixed, __char5 c5) {
+// CHECK: args_vec_5c
+// CHECK: [[C5:%.*]] = alloca <5 x i8>, align 8
+// CHECK: [[TMP:%.*]] = bitcast <5 x i8>* [[C5]] to <2 x i32>*
+// CHECK: store <2 x i32> {{%.*}}, <2 x i32>* [[TMP]], align 1
+  double sum = fixed;
+  sum = sum + c5.x + c5.y;
+  return sum;
+}
+
+double fixed_5c(__char5 *in) {
+// CHECK: fixed_5c
+// CHECK: call double @args_vec_5c(i32 5, <2 x i32> {{%.*}})
+  return args_vec_5c(5, *in);
+}
+
+__attribute__((noinline)) double args_vec_9c(int fixed, __char9 c9) {
+// CHECK: args_vec_9c
+// CHECK: [[C9:%.*]] = alloca <9 x i8>, align 16
+// CHECK: [[TMP:%.*]] = bitcast <9 x i8>* [[C9]] to <4 x i32>*
+// CHECK: store <4 x i32> {{%.*}}, <4 x i32>* [[TMP]], align 1
+  double sum = fixed;
+  sum = sum + c9.x + c9.y;
+  return sum;
+}
+
+double fixed_9c(__char9 *in) {
+// CHECK: fixed_9c
+// CHECK: call double @args_vec_9c(i32 9, <4 x i32> {{%.*}})
+  return args_vec_9c(9, *in);
+}
+
+__attribute__((noinline)) double args_vec_19c(int fixed, __char19 c19) {
+// CHECK: args_vec_19c
+// CHECK: [[C19:%.*]] = load <19 x i8>* {{.*}}, align 16
+  double sum = fixed;
+  sum = sum + c19.x + c19.y;
+  return sum;
+}
+
+double fixed_19c(__char19 *in) {
+// CHECK: fixed_19c
+// CHECK: call double @args_vec_19c(i32 19, <19 x i8>* {{%.*}})
+  return args_vec_19c(19, *in);
+}
+
+__attribute__((noinline)) double args_vec_3s(int fixed, __short3 c3) {
+// CHECK: args_vec_3s
+// CHECK: [[C3:%.*]] = alloca <3 x i16>, align 8
+// CHECK: [[TMP:%.*]] = bitcast <3 x i16>* [[C3]] to <2 x i32>*
+// CHECK: store <2 x i32> {{%.*}}, <2 x i32>* [[TMP]], align 1
+  double sum = fixed;
+  sum = sum + c3.x + c3.y;
+  return sum;
+}
+
+double fixed_3s(__short3 *in) {
+// CHECK: fixed_3s
+// CHECK: call double @args_vec_3s(i32 3, <2 x i32> {{%.*}})
+  return args_vec_3s(3, *in);
+}
+
+__attribute__((noinline)) double args_vec_5s(int fixed, __short5 c5) {
+// CHECK: args_vec_5s
+// CHECK: [[C5:%.*]] = alloca <5 x i16>, align 16
+// CHECK: [[TMP:%.*]] = bitcast <5 x i16>* [[C5]] to <4 x i32>*
+// CHECK: store <4 x i32> {{%.*}}, <4 x i32>* [[TMP]], align 1
+  double sum = fixed;
+  sum = sum + c5.x + c5.y;
+  return sum;
+}
+
+double fixed_5s(__short5 *in) {
+// CHECK: fixed_5s
+// CHECK: call double @args_vec_5s(i32 5, <4 x i32> {{%.*}})
+  return args_vec_5s(5, *in);
+}
+
+__attribute__((noinline)) double args_vec_3i(int fixed, __int3 c3) {
+// CHECK: args_vec_3i
+// CHECK: [[C3:%.*]] = alloca <3 x i32>, align 16
+// CHECK: [[TMP:%.*]] = bitcast <3 x i32>* [[C3]] to <4 x i32>*
+// CHECK: store <4 x i32> {{%.*}}, <4 x i32>* [[TMP]], align 1
+  double sum = fixed;
+  sum = sum + c3.x + c3.y;
+  return sum;
+}
+
+double fixed_3i(__int3 *in) {
+// CHECK: fixed_3i
+// CHECK: call double @args_vec_3i(i32 3, <4 x i32> {{%.*}})
+  return args_vec_3i(3, *in);
+}
+
+__attribute__((noinline)) double args_vec_5i(int fixed, __int5 c5) {
+// CHECK: args_vec_5i
+// CHECK: [[C5:%.*]] = load <5 x i32>* {{%.*}}, align 16
+  double sum = fixed;
+  sum = sum + c5.x + c5.y;
+  return sum;
+}
+
+double fixed_5i(__int5 *in) {
+// CHECK: fixed_5i
+// CHECK: call double @args_vec_5i(i32 5, <5 x i32>* {{%.*}})
+  return args_vec_5i(5, *in);
+}
+
+__attribute__((noinline)) double args_vec_3d(int fixed, __double3 c3) {
+// CHECK: args_vec_3d
+// CHECK: [[CAST:%.*]] = bitcast <3 x double>* {{%.*}} to <4 x double>*
+// CHECK: [[LOAD:%.*]] = load <4 x double>* [[CAST]]
+// CHECK: shufflevector <4 x double> [[LOAD]], <4 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
+  double sum = fixed;
+  sum = sum + c3.x + c3.y;
+  return sum;
+}
+
+double fixed_3d(__double3 *in) {
+// CHECK: fixed_3d
+// CHECK: call double @args_vec_3d(i32 3, <3 x double>* {{%.*}})
+  return args_vec_3d(3, *in);
+}
diff --git a/test/CodeGen/arm64-arguments.c b/test/CodeGen/arm64-arguments.c
new file mode 100644
index 0000000..edfb628
--- /dev/null
+++ b/test/CodeGen/arm64-arguments.c
@@ -0,0 +1,713 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7 -target-abi darwinpcs -ffreestanding -emit-llvm -w -o - %s | FileCheck %s
+
+// CHECK: define signext i8 @f0()
+char f0(void) {
+  return 0;
+}
+
+// Struct as return type. Aggregates <= 16 bytes are passed directly and round
+// up to multiple of 8 bytes.
+// CHECK: define i64 @f1()
+struct s1 { char f0; };
+struct s1 f1(void) {}
+
+// CHECK: define i64 @f2()
+struct s2 { short f0; };
+struct s2 f2(void) {}
+
+// CHECK: define i64 @f3()
+struct s3 { int f0; };
+struct s3 f3(void) {}
+
+// CHECK: define i64 @f4()
+struct s4 { struct s4_0 { int f0; } f0; };
+struct s4 f4(void) {}
+
+// CHECK: define i64 @f5()
+struct s5 { struct { } f0; int f1; };
+struct s5 f5(void) {}
+
+// CHECK: define i64 @f6()
+struct s6 { int f0[1]; };
+struct s6 f6(void) {}
+
+// CHECK: define void @f7()
+struct s7 { struct { int : 0; } f0; };
+struct s7 f7(void) {}
+
+// CHECK: define void @f8()
+struct s8 { struct { int : 0; } f0[1]; };
+struct s8 f8(void) {}
+
+// CHECK: define i64 @f9()
+struct s9 { int f0; int : 0; };
+struct s9 f9(void) {}
+
+// CHECK: define i64 @f10()
+struct s10 { int f0; int : 0; int : 0; };
+struct s10 f10(void) {}
+
+// CHECK: define i64 @f11()
+struct s11 { int : 0; int f0; };
+struct s11 f11(void) {}
+
+// CHECK: define i64 @f12()
+union u12 { char f0; short f1; int f2; };
+union u12 f12(void) {}
+
+// Homogeneous Aggregate as return type will be passed directly.
+// CHECK: define %struct.s13 @f13()
+struct s13 { float f0; };
+struct s13 f13(void) {}
+// CHECK: define %union.u14 @f14()
+union u14 { float f0; };
+union u14 f14(void) {}
+
+// CHECK: define void @f15()
+void f15(struct s7 a0) {}
+
+// CHECK: define void @f16()
+void f16(struct s8 a0) {}
+
+// CHECK: define i64 @f17()
+struct s17 { short f0 : 13; char f1 : 4; };
+struct s17 f17(void) {}
+
+// CHECK: define i64 @f18()
+struct s18 { short f0; char f1 : 4; };
+struct s18 f18(void) {}
+
+// CHECK: define i64 @f19()
+struct s19 { int f0; struct s8 f1; };
+struct s19 f19(void) {}
+
+// CHECK: define i64 @f20()
+struct s20 { struct s8 f1; int f0; };
+struct s20 f20(void) {}
+
+// CHECK: define i64 @f21()
+struct s21 { struct {} f1; int f0 : 4; };
+struct s21 f21(void) {}
+
+// CHECK: define i64 @f22()
+// CHECK: define i64 @f23()
+// CHECK: define i64 @f24()
+// CHECK: define i128 @f25()
+// CHECK: define { float, float } @f26()
+// CHECK: define { double, double } @f27()
+_Complex char       f22(void) {}
+_Complex short      f23(void) {}
+_Complex int        f24(void) {}
+_Complex long long  f25(void) {}
+_Complex float      f26(void) {}
+_Complex double     f27(void) {}
+
+// CHECK: define i64 @f28()
+struct s28 { _Complex char f0; };
+struct s28 f28() {}
+
+// CHECK: define i64 @f29()
+struct s29 { _Complex short f0; };
+struct s29 f29() {}
+
+// CHECK: define i64 @f30()
+struct s30 { _Complex int f0; };
+struct s30 f30() {}
+
+struct s31 { char x; };
+void f31(struct s31 s) { }
+// CHECK: define void @f31(i64 %s.coerce)
+// CHECK: %s = alloca %struct.s31, align 8
+// CHECK: trunc i64 %s.coerce to i8
+// CHECK: store i8 %{{.*}},
+
+struct s32 { double x; };
+void f32(struct s32 s) { }
+// Expand Homogeneous Aggregate.
+// CHECK: @f32(double %{{.*}})
+
+// A composite type larger than 16 bytes should be passed indirectly.
+struct s33 { char buf[32*32]; };
+void f33(struct s33 s) { }
+// CHECK: define void @f33(%struct.s33* %s)
+
+struct s34 { char c; };
+void f34(struct s34 s);
+void g34(struct s34 *s) { f34(*s); }
+// CHECK: @g34(%struct.s34* %s)
+// CHECK: %[[a:.*]] = load i8* %{{.*}}
+// CHECK: zext i8 %[[a]] to i64
+// CHECK: call void @f34(i64 %{{.*}})
+
+/*
+ * Check that va_arg accesses stack according to ABI alignment
+ */
+long long t1(int i, ...) {
+    // CHECK: t1
+    __builtin_va_list ap;
+    __builtin_va_start(ap, i);
+    // CHECK-NOT: add i32 %{{.*}} 7
+    // CHECK-NOT: and i32 %{{.*}} -8
+    long long ll = __builtin_va_arg(ap, long long);
+    __builtin_va_end(ap);
+    return ll;
+}
+double t2(int i, ...) {
+    // CHECK: t2
+    __builtin_va_list ap;
+    __builtin_va_start(ap, i);
+    // CHECK-NOT: add i32 %{{.*}} 7
+    // CHECK-NOT: and i32 %{{.*}} -8
+    double ll = __builtin_va_arg(ap, double);
+    __builtin_va_end(ap);
+    return ll;
+}
+
+#include <arm_neon.h>
+
+// Homogeneous Vector Aggregate as return type and argument type.
+// CHECK: define %struct.int8x16x2_t @f0_0(<16 x i8> %{{.*}}, <16 x i8> %{{.*}})
+int8x16x2_t f0_0(int8x16_t a0, int8x16_t a1) {
+  return vzipq_s8(a0, a1);
+}
+
+// Test direct vector passing.
+typedef float T_float32x2 __attribute__ ((__vector_size__ (8)));
+typedef float T_float32x4 __attribute__ ((__vector_size__ (16)));
+typedef float T_float32x8 __attribute__ ((__vector_size__ (32)));
+typedef float T_float32x16 __attribute__ ((__vector_size__ (64)));
+
+// CHECK: define <2 x float> @f1_0(<2 x float> %{{.*}})
+T_float32x2 f1_0(T_float32x2 a0) { return a0; }
+// CHECK: define <4 x float> @f1_1(<4 x float> %{{.*}})
+T_float32x4 f1_1(T_float32x4 a0) { return a0; }
+// Vector with length bigger than 16-byte is illegal and is passed indirectly.
+// CHECK: define void @f1_2(<8 x float>* noalias sret  %{{.*}}, <8 x float>*)
+T_float32x8 f1_2(T_float32x8 a0) { return a0; }
+// CHECK: define void @f1_3(<16 x float>* noalias sret %{{.*}}, <16 x float>*)
+T_float32x16 f1_3(T_float32x16 a0) { return a0; }
+
+// Testing alignment with aggregates: HFA, aggregates with size <= 16 bytes and
+// aggregates with size > 16 bytes.
+struct s35
+{
+   float v[4]; //Testing HFA.
+} __attribute__((aligned(16)));
+typedef struct s35 s35_with_align;
+
+typedef __attribute__((neon_vector_type(4))) float float32x4_t;
+float32x4_t f35(int i, s35_with_align s1, s35_with_align s2) {
+// CHECK: define <4 x float> @f35(i32 %i, float %s1.0, float %s1.1, float %s1.2, float %s1.3, float %s2.0, float %s2.1, float %s2.2, float %s2.3)
+// CHECK: %s1 = alloca %struct.s35, align 16
+// CHECK: %s2 = alloca %struct.s35, align 16
+// CHECK: %[[a:.*]] = bitcast %struct.s35* %s1 to <4 x float>*
+// CHECK: load <4 x float>* %[[a]], align 16
+// CHECK: %[[b:.*]] = bitcast %struct.s35* %s2 to <4 x float>*
+// CHECK: load <4 x float>* %[[b]], align 16
+  float32x4_t v = vaddq_f32(*(float32x4_t *)&s1,
+                            *(float32x4_t *)&s2);
+  return v;
+}
+
+struct s36
+{
+   int v[4]; //Testing 16-byte aggregate.
+} __attribute__((aligned(16)));
+typedef struct s36 s36_with_align;
+
+typedef __attribute__((neon_vector_type(4))) int int32x4_t;
+int32x4_t f36(int i, s36_with_align s1, s36_with_align s2) {
+// CHECK: define <4 x i32> @f36(i32 %i, i128 %s1.coerce, i128 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s36, align 16
+// CHECK: %s2 = alloca %struct.s36, align 16
+// CHECK: store i128 %s1.coerce, i128* %{{.*}}, align 1
+// CHECK: store i128 %s2.coerce, i128* %{{.*}}, align 1
+// CHECK: %[[a:.*]] = bitcast %struct.s36* %s1 to <4 x i32>*
+// CHECK: load <4 x i32>* %[[a]], align 16
+// CHECK: %[[b:.*]] = bitcast %struct.s36* %s2 to <4 x i32>*
+// CHECK: load <4 x i32>* %[[b]], align 16
+  int32x4_t v = vaddq_s32(*(int32x4_t *)&s1,
+                          *(int32x4_t *)&s2);
+  return v;
+}
+
+struct s37
+{
+   int v[18]; //Testing large aggregate.
+} __attribute__((aligned(16)));
+typedef struct s37 s37_with_align;
+
+int32x4_t f37(int i, s37_with_align s1, s37_with_align s2) {
+// CHECK: define <4 x i32> @f37(i32 %i, %struct.s37* %s1, %struct.s37* %s2)
+// CHECK: %[[a:.*]] = bitcast %struct.s37* %s1 to <4 x i32>*
+// CHECK: load <4 x i32>* %[[a]], align 16
+// CHECK: %[[b:.*]] = bitcast %struct.s37* %s2 to <4 x i32>*
+// CHECK: load <4 x i32>* %[[b]], align 16
+  int32x4_t v = vaddq_s32(*(int32x4_t *)&s1,
+                          *(int32x4_t *)&s2);
+  return v;
+}
+s37_with_align g37;
+int32x4_t caller37() {
+// CHECK: caller37
+// CHECK: %[[a:.*]] = alloca %struct.s37, align 16
+// CHECK: %[[b:.*]] = alloca %struct.s37, align 16
+// CHECK: call void @llvm.memcpy
+// CHECK: call void @llvm.memcpy
+// CHECK: call <4 x i32> @f37(i32 3, %struct.s37* %[[a]], %struct.s37* %[[b]])
+  return f37(3, g37, g37);
+}
+
+// rdar://problem/12648441
+// Test passing structs with size < 8, < 16 and > 16
+// with alignment of 16 and without
+
+// structs with size <= 8 bytes, without alignment attribute
+// passed as i64 regardless of the align attribute
+struct s38
+{
+  int i;
+  short s;
+};
+typedef struct s38 s38_no_align;
+// passing structs in registers
+__attribute__ ((noinline))
+int f38(int i, s38_no_align s1, s38_no_align s2) {
+// CHECK: define i32 @f38(i32 %i, i64 %s1.coerce, i64 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s38, align 8
+// CHECK: %s2 = alloca %struct.s38, align 8
+// CHECK: store i64 %s1.coerce, i64* %{{.*}}, align 1
+// CHECK: store i64 %s2.coerce, i64* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s38* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s38* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s38* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s38* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s38_no_align g38;
+s38_no_align g38_2;
+int caller38() {
+// CHECK: define i32 @caller38()
+// CHECK: %[[a:.*]] = load i64* bitcast (%struct.s38* @g38 to i64*), align 1
+// CHECK: %[[b:.*]] = load i64* bitcast (%struct.s38* @g38_2 to i64*), align 1
+// CHECK: call i32 @f38(i32 3, i64 %[[a]], i64 %[[b]])
+  return f38(3, g38, g38_2);
+}
+// passing structs on stack
+__attribute__ ((noinline))
+int f38_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s38_no_align s1, s38_no_align s2) {
+// CHECK: define i32 @f38_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, i64 %s1.coerce, i64 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s38, align 8
+// CHECK: %s2 = alloca %struct.s38, align 8
+// CHECK: store i64 %s1.coerce, i64* %{{.*}}, align 1
+// CHECK: store i64 %s2.coerce, i64* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s38* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s38* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s38* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s38* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller38_stack() {
+// CHECK: define i32 @caller38_stack()
+// CHECK: %[[a:.*]] = load i64* bitcast (%struct.s38* @g38 to i64*), align 1
+// CHECK: %[[b:.*]] = load i64* bitcast (%struct.s38* @g38_2 to i64*), align 1
+// CHECK: call i32 @f38_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i64 %[[a]], i64 %[[b]])
+  return f38_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g38, g38_2);
+}
+
+// structs with size <= 8 bytes, with alignment attribute
+struct s39
+{
+  int i;
+  short s;
+} __attribute__((aligned(16)));
+typedef struct s39 s39_with_align;
+// passing aligned structs in registers
+__attribute__ ((noinline))
+int f39(int i, s39_with_align s1, s39_with_align s2) {
+// CHECK: define i32 @f39(i32 %i, i128 %s1.coerce, i128 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s39, align 16
+// CHECK: %s2 = alloca %struct.s39, align 16
+// CHECK: store i128 %s1.coerce, i128* %{{.*}}, align 1
+// CHECK: store i128 %s2.coerce, i128* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s39* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s39* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s39* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s39* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s39_with_align g39;
+s39_with_align g39_2;
+int caller39() {
+// CHECK: define i32 @caller39()
+// CHECK: %[[a:.*]] = load i128* bitcast (%struct.s39* @g39 to i128*), align 1
+// CHECK: %[[b:.*]] = load i128* bitcast (%struct.s39* @g39_2 to i128*), align 1
+// CHECK: call i32 @f39(i32 3, i128 %[[a]], i128 %[[b]])
+  return f39(3, g39, g39_2);
+}
+// passing aligned structs on stack
+__attribute__ ((noinline))
+int f39_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s39_with_align s1, s39_with_align s2) {
+// CHECK: define i32 @f39_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, i128 %s1.coerce, i128 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s39, align 16
+// CHECK: %s2 = alloca %struct.s39, align 16
+// CHECK: store i128 %s1.coerce, i128* %{{.*}}, align 1
+// CHECK: store i128 %s2.coerce, i128* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s39* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s39* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s39* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s39* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller39_stack() {
+// CHECK: define i32 @caller39_stack()
+// CHECK: %[[a:.*]] = load i128* bitcast (%struct.s39* @g39 to i128*), align 1
+// CHECK: %[[b:.*]] = load i128* bitcast (%struct.s39* @g39_2 to i128*), align 1
+// CHECK: call i32 @f39_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i128 %[[a]], i128 %[[b]])
+  return f39_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g39, g39_2);
+}
+
+// structs with size <= 16 bytes, without alignment attribute
+struct s40
+{
+  int i;
+  short s;
+  int i2;
+  short s2;
+};
+typedef struct s40 s40_no_align;
+// passing structs in registers
+__attribute__ ((noinline))
+int f40(int i, s40_no_align s1, s40_no_align s2) {
+// CHECK: define i32 @f40(i32 %i, [2 x i64] %s1.coerce, [2 x i64] %s2.coerce)
+// CHECK: %s1 = alloca %struct.s40, align 8
+// CHECK: %s2 = alloca %struct.s40, align 8
+// CHECK: store [2 x i64] %s1.coerce, [2 x i64]* %{{.*}}, align 1
+// CHECK: store [2 x i64] %s2.coerce, [2 x i64]* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s40* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s40* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s40* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s40* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s40_no_align g40;
+s40_no_align g40_2;
+int caller40() {
+// CHECK: define i32 @caller40()
+// CHECK: %[[a:.*]] = load [2 x i64]* bitcast (%struct.s40* @g40 to [2 x i64]*), align 1
+// CHECK: %[[b:.*]] = load [2 x i64]* bitcast (%struct.s40* @g40_2 to [2 x i64]*), align 1
+// CHECK: call i32 @f40(i32 3, [2 x i64] %[[a]], [2 x i64] %[[b]])
+  return f40(3, g40, g40_2);
+}
+// passing structs on stack
+__attribute__ ((noinline))
+int f40_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s40_no_align s1, s40_no_align s2) {
+// CHECK: define i32 @f40_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, [2 x i64] %s1.coerce, [2 x i64] %s2.coerce)
+// CHECK: %s1 = alloca %struct.s40, align 8
+// CHECK: %s2 = alloca %struct.s40, align 8
+// CHECK: store [2 x i64] %s1.coerce, [2 x i64]* %{{.*}}, align 1
+// CHECK: store [2 x i64] %s2.coerce, [2 x i64]* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s40* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s40* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s40* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s40* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller40_stack() {
+// CHECK: define i32 @caller40_stack()
+// CHECK: %[[a:.*]] = load [2 x i64]* bitcast (%struct.s40* @g40 to [2 x i64]*), align 1
+// CHECK: %[[b:.*]] = load [2 x i64]* bitcast (%struct.s40* @g40_2 to [2 x i64]*), align 1
+// CHECK: call i32 @f40_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, [2 x i64] %[[a]], [2 x i64] %[[b]])
+  return f40_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g40, g40_2);
+}
+
+// structs with size <= 16 bytes, with alignment attribute
+struct s41
+{
+  int i;
+  short s;
+  int i2;
+  short s2;
+} __attribute__((aligned(16)));
+typedef struct s41 s41_with_align;
+// passing aligned structs in registers
+__attribute__ ((noinline))
+int f41(int i, s41_with_align s1, s41_with_align s2) {
+// CHECK: define i32 @f41(i32 %i, i128 %s1.coerce, i128 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s41, align 16
+// CHECK: %s2 = alloca %struct.s41, align 16
+// CHECK: store i128 %s1.coerce, i128* %{{.*}}, align 1
+// CHECK: store i128 %s2.coerce, i128* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s41* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s41* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s41* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s41* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s41_with_align g41;
+s41_with_align g41_2;
+int caller41() {
+// CHECK: define i32 @caller41()
+// CHECK: %[[a:.*]] = load i128* bitcast (%struct.s41* @g41 to i128*), align 1
+// CHECK: %[[b:.*]] = load i128* bitcast (%struct.s41* @g41_2 to i128*), align 1
+// CHECK: call i32 @f41(i32 3, i128 %[[a]], i128 %[[b]])
+  return f41(3, g41, g41_2);
+}
+// passing aligned structs on stack
+__attribute__ ((noinline))
+int f41_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s41_with_align s1, s41_with_align s2) {
+// CHECK: define i32 @f41_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, i128 %s1.coerce, i128 %s2.coerce)
+// CHECK: %s1 = alloca %struct.s41, align 16
+// CHECK: %s2 = alloca %struct.s41, align 16
+// CHECK: store i128 %s1.coerce, i128* %{{.*}}, align 1
+// CHECK: store i128 %s2.coerce, i128* %{{.*}}, align 1
+// CHECK: getelementptr inbounds %struct.s41* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s41* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s41* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s41* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller41_stack() {
+// CHECK: define i32 @caller41_stack()
+// CHECK: %[[a:.*]] = load i128* bitcast (%struct.s41* @g41 to i128*), align 1
+// CHECK: %[[b:.*]] = load i128* bitcast (%struct.s41* @g41_2 to i128*), align 1
+// CHECK: call i32 @f41_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i128 %[[a]], i128 %[[b]])
+  return f41_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g41, g41_2);
+}
+
+// structs with size > 16 bytes, without alignment attribute
+struct s42
+{
+  int i;
+  short s;
+  int i2;
+  short s2;
+  int i3;
+  short s3;
+};
+typedef struct s42 s42_no_align;
+// passing structs in registers
+__attribute__ ((noinline))
+int f42(int i, s42_no_align s1, s42_no_align s2) {
+// CHECK: define i32 @f42(i32 %i, %struct.s42* %s1, %struct.s42* %s2)
+// CHECK: getelementptr inbounds %struct.s42* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s42* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s42* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s42* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s42_no_align g42;
+s42_no_align g42_2;
+int caller42() {
+// CHECK: define i32 @caller42()
+// CHECK: %[[a:.*]] = alloca %struct.s42, align 4
+// CHECK: %[[b:.*]] = alloca %struct.s42, align 4
+// CHECK: %[[c:.*]] = bitcast %struct.s42* %[[a]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: %[[d:.*]] = bitcast %struct.s42* %[[b]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: call i32 @f42(i32 3, %struct.s42* %[[a]], %struct.s42* %[[b]])
+  return f42(3, g42, g42_2);
+}
+// passing structs on stack
+__attribute__ ((noinline))
+int f42_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s42_no_align s1, s42_no_align s2) {
+// CHECK: define i32 @f42_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, %struct.s42* %s1, %struct.s42* %s2)
+// CHECK: getelementptr inbounds %struct.s42* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s42* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s42* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s42* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller42_stack() {
+// CHECK: define i32 @caller42_stack()
+// CHECK: %[[a:.*]] = alloca %struct.s42, align 4
+// CHECK: %[[b:.*]] = alloca %struct.s42, align 4
+// CHECK: %[[c:.*]] = bitcast %struct.s42* %[[a]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: %[[d:.*]] = bitcast %struct.s42* %[[b]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: call i32 @f42_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, %struct.s42* %[[a]], %struct.s42* %[[b]])
+  return f42_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g42, g42_2);
+}
+
+// structs with size > 16 bytes, with alignment attribute
+struct s43
+{
+  int i;
+  short s;
+  int i2;
+  short s2;
+  int i3;
+  short s3;
+} __attribute__((aligned(16)));
+typedef struct s43 s43_with_align;
+// passing aligned structs in registers
+__attribute__ ((noinline))
+int f43(int i, s43_with_align s1, s43_with_align s2) {
+// CHECK: define i32 @f43(i32 %i, %struct.s43* %s1, %struct.s43* %s2)
+// CHECK: getelementptr inbounds %struct.s43* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s43* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s43* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s43* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + s1.s + s2.s;
+}
+s43_with_align g43;
+s43_with_align g43_2;
+int caller43() {
+// CHECK: define i32 @caller43()
+// CHECK: %[[a:.*]] = alloca %struct.s43, align 16
+// CHECK: %[[b:.*]] = alloca %struct.s43, align 16
+// CHECK: %[[c:.*]] = bitcast %struct.s43* %[[a]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: %[[d:.*]] = bitcast %struct.s43* %[[b]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: call i32 @f43(i32 3, %struct.s43* %[[a]], %struct.s43* %[[b]])
+  return f43(3, g43, g43_2);
+}
+// passing aligned structs on stack
+__attribute__ ((noinline))
+int f43_stack(int i, int i2, int i3, int i4, int i5, int i6, int i7, int i8,
+              int i9, s43_with_align s1, s43_with_align s2) {
+// CHECK: define i32 @f43_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, i32 %i8, i32 %i9, %struct.s43* %s1, %struct.s43* %s2)
+// CHECK: getelementptr inbounds %struct.s43* %s1, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s43* %s2, i32 0, i32 0
+// CHECK: getelementptr inbounds %struct.s43* %s1, i32 0, i32 1
+// CHECK: getelementptr inbounds %struct.s43* %s2, i32 0, i32 1
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + s1.s + s2.s;
+}
+int caller43_stack() {
+// CHECK: define i32 @caller43_stack()
+// CHECK: %[[a:.*]] = alloca %struct.s43, align 16
+// CHECK: %[[b:.*]] = alloca %struct.s43, align 16
+// CHECK: %[[c:.*]] = bitcast %struct.s43* %[[a]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: %[[d:.*]] = bitcast %struct.s43* %[[b]] to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
+// CHECK: call i32 @f43_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, %struct.s43* %[[a]], %struct.s43* %[[b]])
+  return f43_stack(1, 2, 3, 4, 5, 6, 7, 8, 9, g43, g43_2);
+}
+
+// rdar://13668927
+// We should not split argument s1 between registers and stack.
+__attribute__ ((noinline))
+int f40_split(int i, int i2, int i3, int i4, int i5, int i6, int i7,
+              s40_no_align s1, s40_no_align s2) {
+// CHECK: define i32 @f40_split(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, [1 x i32], [2 x i64] %s1.coerce, [2 x i64] %s2.coerce)
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + s1.s + s2.s;
+}
+int caller40_split() {
+// CHECK: define i32 @caller40_split()
+// CHECK: call i32 @f40_split(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, [1 x i32] undef, [2 x i64] %{{.*}} [2 x i64] %{{.*}})
+  return f40_split(1, 2, 3, 4, 5, 6, 7, g40, g40_2);
+}
+
+__attribute__ ((noinline))
+int f41_split(int i, int i2, int i3, int i4, int i5, int i6, int i7,
+              s41_with_align s1, s41_with_align s2) {
+// CHECK: define i32 @f41_split(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6, i32 %i7, [1 x i32], i128 %s1.coerce, i128 %s2.coerce)
+  return s1.i + s2.i + i + i2 + i3 + i4 + i5 + i6 + i7 + s1.s + s2.s;
+}
+int caller41_split() {
+// CHECK: define i32 @caller41_split()
+// CHECK: call i32 @f41_split(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, [1 x i32] undef, i128 %{{.*}}, i128 %{{.*}})
+  return f41_split(1, 2, 3, 4, 5, 6, 7, g41, g41_2);
+}
+
+// Handle homogeneous aggregates properly in variadic functions.
+struct HFA {
+  float a, b, c, d;
+};
+
+float test_hfa(int n, ...) {
+// CHECK-LABEL: define float @test_hfa(i32 %n, ...)
+// CHECK: [[THELIST:%.*]] = alloca i8*
+// CHECK: [[CURLIST:%.*]] = load i8** [[THELIST]]
+
+  // HFA is not indirect, so occupies its full 16 bytes on the stack.
+// CHECK: [[NEXTLIST:%.*]] = getelementptr i8* [[CURLIST]], i32 16
+// CHECK: store i8* [[NEXTLIST]], i8** [[THELIST]]
+
+// CHECK: bitcast i8* [[CURLIST]] to %struct.HFA*
+  __builtin_va_list thelist;
+  __builtin_va_start(thelist, n);
+  struct HFA h = __builtin_va_arg(thelist, struct HFA);
+  return h.d;
+}
+
+struct TooBigHFA {
+  float a, b, c, d, e;
+};
+
+float test_toobig_hfa(int n, ...) {
+// CHECK-LABEL: define float @test_toobig_hfa(i32 %n, ...)
+// CHECK: [[THELIST:%.*]] = alloca i8*
+// CHECK: [[CURLIST:%.*]] = load i8** [[THELIST]]
+
+  // TooBigHFA is not actually an HFA, so gets passed indirectly. Only 8 bytes
+  // of stack consumed.
+// CHECK: [[NEXTLIST:%.*]] = getelementptr i8* [[CURLIST]], i32 8
+// CHECK: store i8* [[NEXTLIST]], i8** [[THELIST]]
+
+// CHECK: [[HFAPTRPTR:%.*]] = bitcast i8* [[CURLIST]] to i8**
+// CHECK: [[HFAPTR:%.*]] = load i8** [[HFAPTRPTR]]
+// CHECK: bitcast i8* [[HFAPTR]] to %struct.TooBigHFA*
+  __builtin_va_list thelist;
+  __builtin_va_start(thelist, n);
+  struct TooBigHFA h = __builtin_va_arg(thelist, struct TooBigHFA);
+  return h.d;
+}
+
+struct HVA {
+  int32x4_t a, b;
+};
+
+int32x4_t test_hva(int n, ...) {
+// CHECK-LABEL: define <4 x i32> @test_hva(i32 %n, ...)
+// CHECK: [[THELIST:%.*]] = alloca i8*
+// CHECK: [[CURLIST:%.*]] = load i8** [[THELIST]]
+
+  // HVA is not indirect, so occupies its full 16 bytes on the stack. but it
+  // must be properly aligned.
+// CHECK: [[ALIGN0:%.*]] = getelementptr i8* [[CURLIST]], i32 15
+// CHECK: [[ALIGN1:%.*]] = ptrtoint i8* [[ALIGN0]] to i64
+// CHECK: [[ALIGN2:%.*]] = and i64 [[ALIGN1]], -16
+// CHECK: [[ALIGNED_LIST:%.*]] = inttoptr i64 [[ALIGN2]] to i8*
+
+// CHECK: [[NEXTLIST:%.*]] = getelementptr i8* [[ALIGNED_LIST]], i32 32
+// CHECK: store i8* [[NEXTLIST]], i8** [[THELIST]]
+
+// CHECK: bitcast i8* [[ALIGNED_LIST]] to %struct.HVA*
+  __builtin_va_list thelist;
+  __builtin_va_start(thelist, n);
+  struct HVA h = __builtin_va_arg(thelist, struct HVA);
+  return h.b;
+}
+
+struct TooBigHVA {
+  int32x4_t a, b, c, d, e;
+};
+
+int32x4_t test_toobig_hva(int n, ...) {
+// CHECK-LABEL: define <4 x i32> @test_toobig_hva(i32 %n, ...)
+// CHECK: [[THELIST:%.*]] = alloca i8*
+// CHECK: [[CURLIST:%.*]] = load i8** [[THELIST]]
+
+  // TooBigHVA is not actually an HVA, so gets passed indirectly. Only 8 bytes
+  // of stack consumed.
+// CHECK: [[NEXTLIST:%.*]] = getelementptr i8* [[CURLIST]], i32 8
+// CHECK: store i8* [[NEXTLIST]], i8** [[THELIST]]
+
+// CHECK: [[HVAPTRPTR:%.*]] = bitcast i8* [[CURLIST]] to i8**
+// CHECK: [[HVAPTR:%.*]] = load i8** [[HVAPTRPTR]]
+// CHECK: bitcast i8* [[HVAPTR]] to %struct.TooBigHVA*
+  __builtin_va_list thelist;
+  __builtin_va_start(thelist, n);
+  struct TooBigHVA h = __builtin_va_arg(thelist, struct TooBigHVA);
+  return h.d;
+}
diff --git a/test/CodeGen/arm64-crc32.c b/test/CodeGen/arm64-crc32.c
new file mode 100644
index 0000000..cb31a78
--- /dev/null
+++ b/test/CodeGen/arm64-crc32.c
@@ -0,0 +1,55 @@
+// REQUIRES: arm64-registered-target
+// RUN: %clang_cc1 -triple arm64-none-linux-gnu \
+// RUN:   -O3 -S -emit-llvm -o - %s | FileCheck %s
+
+int crc32b(int a, char b)
+{
+        return __builtin_arm_crc32b(a,b);
+// CHECK: [[T0:%[0-9]+]] = zext i8 %b to i32
+// CHECK: call i32 @llvm.arm64.crc32b(i32 %a, i32 [[T0]])
+}
+
+int crc32cb(int a, char b)
+{
+        return __builtin_arm_crc32cb(a,b);
+// CHECK: [[T0:%[0-9]+]] = zext i8 %b to i32
+// CHECK: call i32 @llvm.arm64.crc32cb(i32 %a, i32 [[T0]])
+}
+
+int crc32h(int a, short b)
+{
+        return __builtin_arm_crc32h(a,b);
+// CHECK: [[T0:%[0-9]+]] = zext i16 %b to i32
+// CHECK: call i32 @llvm.arm64.crc32h(i32 %a, i32 [[T0]])
+}
+
+int crc32ch(int a, short b)
+{
+        return __builtin_arm_crc32ch(a,b);
+// CHECK: [[T0:%[0-9]+]] = zext i16 %b to i32
+// CHECK: call i32 @llvm.arm64.crc32ch(i32 %a, i32 [[T0]])
+}
+
+int crc32w(int a, int b)
+{
+        return __builtin_arm_crc32w(a,b);
+// CHECK: call i32 @llvm.arm64.crc32w(i32 %a, i32 %b)
+}
+
+int crc32cw(int a, int b)
+{
+        return __builtin_arm_crc32cw(a,b);
+// CHECK: call i32 @llvm.arm64.crc32cw(i32 %a, i32 %b)
+}
+
+int crc32d(int a, long b)
+{
+        return __builtin_arm_crc32d(a,b);
+// CHECK: call i32 @llvm.arm64.crc32x(i32 %a, i64 %b)
+}
+
+int crc32cd(int a, long b)
+{
+        return __builtin_arm_crc32cd(a,b);
+// CHECK: call i32 @llvm.arm64.crc32cx(i32 %a, i64 %b)
+}
diff --git a/test/CodeGen/arm64-lanes.c b/test/CodeGen/arm64-lanes.c
new file mode 100644
index 0000000..63d7d0c
--- /dev/null
+++ b/test/CodeGen/arm64-lanes.c
@@ -0,0 +1,63 @@
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -ffreestanding -emit-llvm -o - %s | FileCheck %s
+
+#include <arm_neon.h>
+
+// CHECK-LABEL: @test_vdupb_lane_s8
+int8_t test_vdupb_lane_s8(int8x8_t src) {
+  return vdupb_lane_s8(src, 2);
+  // CHECK: extractelement <8 x i8> %src, i32 2
+}
+
+// CHECK-LABEL: @test_vdupb_lane_u8
+uint8_t test_vdupb_lane_u8(uint8x8_t src) {
+  return vdupb_lane_u8(src, 2);
+  // CHECK: extractelement <8 x i8> %src, i32 2
+}
+
+// CHECK-LABEL: @test_vduph_lane_s16
+int16_t test_vduph_lane_s16(int16x4_t src) {
+  return vduph_lane_s16(src, 2);
+  // CHECK: extractelement <4 x i16> %src, i32 2
+}
+
+// CHECK-LABEL: @test_vduph_lane_u16
+uint16_t test_vduph_lane_u16(uint16x4_t src) {
+  return vduph_lane_u16(src, 2);
+  // CHECK: extractelement <4 x i16> %src, i32 2
+}
+
+// CHECK-LABEL: @test_vdups_lane_s32
+int32_t test_vdups_lane_s32(int32x2_t src) {
+  return vdups_lane_s32(src, 0);
+  // CHECK: extractelement <2 x i32> %src, i32 0
+}
+
+// CHECK-LABEL: @test_vdups_lane_u32
+uint32_t test_vdups_lane_u32(uint32x2_t src) {
+  return vdups_lane_u32(src, 0);
+  // CHECK: extractelement <2 x i32> %src, i32 0
+}
+
+// CHECK-LABEL: @test_vdups_lane_f32
+float32_t test_vdups_lane_f32(float32x2_t src) {
+  return vdups_lane_f32(src, 0);
+  // CHECK: extractelement <2 x float> %src, i32 0
+}
+
+// CHECK-LABEL: @test_vdupd_lane_s64
+int64_t test_vdupd_lane_s64(int64x1_t src) {
+  return vdupd_lane_s64(src, 0);
+  // CHECK: extractelement <1 x i64> %src, i32 0
+}
+
+// CHECK-LABEL: @test_vdupd_lane_u64
+uint64_t test_vdupd_lane_u64(uint64x1_t src) {
+  return vdupd_lane_u64(src, 0);
+  // CHECK: extractelement <1 x i64> %src, i32 0
+}
+
+// CHECK-LABEL: @test_vdupd_lane_f64
+float64_t test_vdupd_lane_f64(float64x1_t src) {
+  return vdupd_lane_f64(src, 0);
+  // CHECK: extractelement <1 x double> %src, i32 0
+}
diff --git a/test/CodeGen/arm64-scalar-test.c b/test/CodeGen/arm64-scalar-test.c
new file mode 100644
index 0000000..1e1087d
--- /dev/null
+++ b/test/CodeGen/arm64-scalar-test.c
@@ -0,0 +1,535 @@
+// REQUIRES: arm64-registered-target
+// RUN: %clang_cc1 -triple arm64-apple-ios7.0  \
+// RUN:   -S -O1 -o - -ffreestanding %s | FileCheck %s
+
+// We're explicitly using arm_neon.h here: some types probably don't match
+// the ACLE definitions, but we want to check current codegen.
+#include <arm_neon.h>
+
+float test_vrsqrtss_f32(float a, float b) {
+// CHECK: test_vrsqrtss_f32
+  return vrsqrtss_f32(a, b);
+// CHECK: frsqrts {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+double test_vrsqrtsd_f64(double a, double b) {
+// CHECK: test_vrsqrtsd_f64
+  return vrsqrtsd_f64(a, b);
+// CHECK: frsqrts {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+int64x1_t test_vrshl_s64(int64x1_t a, int64x1_t b) {
+// CHECK: test_vrshl_s64
+  return vrshl_s64(a, b);
+// CHECK: srshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+uint64x1_t test_vrshl_u64(uint64x1_t a, int64x1_t b) {
+// CHECK: test_vrshl_u64
+  return vrshl_u64(a, b);
+// CHECK: urshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vrshld_s64
+int64_t test_vrshld_s64(int64_t a, int64_t b) {
+  return vrshld_s64(a, b);
+// CHECK: srshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vrshld_u64
+uint64_t test_vrshld_u64(uint64_t a, uint64_t b) {
+  return vrshld_u64(a, b);
+// CHECK: urshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqrshlb_s8
+int8_t test_vqrshlb_s8(int8_t a, int8_t b) {
+  return vqrshlb_s8(a, b);
+// CHECK: sqrshl.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqrshlh_s16
+int16_t test_vqrshlh_s16(int16_t a, int16_t b) {
+  return vqrshlh_s16(a, b);
+// CHECK: sqrshl.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqrshls_s32
+int32_t test_vqrshls_s32(int32_t a, int32_t b) {
+  return vqrshls_s32(a, b);
+// CHECK: sqrshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqrshld_s64
+int64_t test_vqrshld_s64(int64_t a, int64_t b) {
+  return vqrshld_s64(a, b);
+// CHECK: sqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqrshlb_u8
+uint8_t test_vqrshlb_u8(uint8_t a, uint8_t b) {
+  return vqrshlb_u8(a, b);
+// CHECK: uqrshl.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqrshlh_u16
+uint16_t test_vqrshlh_u16(uint16_t a, uint16_t b) {
+  return vqrshlh_u16(a, b);
+// CHECK: uqrshl.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqrshls_u32
+uint32_t test_vqrshls_u32(uint32_t a, uint32_t b) {
+  return vqrshls_u32(a, b);
+// CHECK: uqrshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqrshld_u64
+uint64_t test_vqrshld_u64(uint64_t a, uint64_t b) {
+  return vqrshld_u64(a, b);
+// CHECK: uqrshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqshlb_s8
+int8_t test_vqshlb_s8(int8_t a, int8_t b) {
+  return vqshlb_s8(a, b);
+// CHECK: sqshl.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqshlh_s16
+int16_t test_vqshlh_s16(int16_t a, int16_t b) {
+  return vqshlh_s16(a, b);
+// CHECK: sqshl.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqshls_s32
+int32_t test_vqshls_s32(int32_t a, int32_t b) {
+  return vqshls_s32(a, b);
+// CHECK: sqshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqshld_s64
+int64_t test_vqshld_s64(int64_t a, int64_t b) {
+  return vqshld_s64(a, b);
+// CHECK: sqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqshlb_u8
+uint8_t test_vqshlb_u8(uint8_t a, uint8_t b) {
+  return vqshlb_u8(a, b);
+// CHECK: uqshl.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqshlh_u16
+uint16_t test_vqshlh_u16(uint16_t a, uint16_t b) {
+  return vqshlh_u16(a, b);
+// CHECK: uqshl.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqshls_u32
+uint32_t test_vqshls_u32(uint32_t a, uint32_t b) {
+  return vqshls_u32(a, b);
+// CHECK: uqshl {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqshld_u64
+uint64_t test_vqshld_u64(uint64_t a, uint64_t b) {
+  return vqshld_u64(a, b);
+// CHECK: uqshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vshld_u64
+uint64_t test_vshld_u64(uint64_t a, uint64_t b) {
+  return vshld_u64(a, b);
+// CHECK: ushl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vshld_s64
+int64_t test_vshld_s64(int64_t a, int64_t b) {
+  return vshld_s64(a, b);
+// CHECK: sshl {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqdmullh_s16
+int32_t test_vqdmullh_s16(int16_t a, int16_t b) {
+  return vqdmullh_s16(a, b);
+// CHECK: sqdmull.4s {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqdmulls_s32
+int64_t test_vqdmulls_s32(int32_t a, int32_t b) {
+  return vqdmulls_s32(a, b);
+// CHECK: sqdmull {{d[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqaddb_s8
+int8_t test_vqaddb_s8(int8_t a, int8_t b) {
+  return vqaddb_s8(a, b);
+// CHECK: sqadd.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqaddh_s16
+int16_t test_vqaddh_s16(int16_t a, int16_t b) {
+  return vqaddh_s16(a, b);
+// CHECK: sqadd.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqadds_s32
+int32_t test_vqadds_s32(int32_t a, int32_t b) {
+  return vqadds_s32(a, b);
+// CHECK: sqadd {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqaddd_s64
+int64_t test_vqaddd_s64(int64_t a, int64_t b) {
+  return vqaddd_s64(a, b);
+// CHECK: sqadd {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqaddb_u8
+uint8_t test_vqaddb_u8(uint8_t a, uint8_t b) {
+  return vqaddb_u8(a, b);
+// CHECK: uqadd.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqaddh_u16
+uint16_t test_vqaddh_u16(uint16_t a, uint16_t b) {
+  return vqaddh_u16(a, b);
+// CHECK: uqadd.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqadds_u32
+uint32_t test_vqadds_u32(uint32_t a, uint32_t b) {
+  return vqadds_u32(a, b);
+// CHECK: uqadd {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqaddd_u64
+uint64_t test_vqaddd_u64(uint64_t a, uint64_t b) {
+  return vqaddd_u64(a, b);
+// CHECK: uqadd {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqsubb_s8
+int8_t test_vqsubb_s8(int8_t a, int8_t b) {
+  return vqsubb_s8(a, b);
+// CHECK: sqsub.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqsubh_s16
+int16_t test_vqsubh_s16(int16_t a, int16_t b) {
+  return vqsubh_s16(a, b);
+// CHECK: sqsub.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqsubs_s32
+int32_t test_vqsubs_s32(int32_t a, int32_t b) {
+  return vqsubs_s32(a, b);
+// CHECK: sqsub {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqsubd_s64
+int64_t test_vqsubd_s64(int64_t a, int64_t b) {
+  return vqsubd_s64(a, b);
+// CHECK: sqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqsubb_u8
+uint8_t test_vqsubb_u8(uint8_t a, uint8_t b) {
+  return vqsubb_u8(a, b);
+// CHECK: uqsub.8b {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqsubh_u16
+uint16_t test_vqsubh_u16(uint16_t a, uint16_t b) {
+  return vqsubh_u16(a, b);
+// CHECK: uqsub.4h {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqsubs_u32
+uint32_t test_vqsubs_u32(uint32_t a, uint32_t b) {
+  return vqsubs_u32(a, b);
+// CHECK: uqsub {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqsubd_u64
+uint64_t test_vqsubd_u64(uint64_t a, uint64_t b) {
+  return vqsubd_u64(a, b);
+// CHECK: uqsub {{d[0-9]+}}, {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqmovnh_s16
+int8_t test_vqmovnh_s16(int16_t a) {
+  return vqmovnh_s16(a);
+// CHECK: sqxtn.8b {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovnh_u16
+uint8_t test_vqmovnh_u16(uint16_t a) {
+  return vqmovnh_u16(a);
+// CHECK: uqxtn.8b {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovns_s32
+int16_t test_vqmovns_s32(int32_t a) {
+  return vqmovns_s32(a);
+// CHECK: sqxtn.4h {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovns_u32
+uint16_t test_vqmovns_u32(uint32_t a) {
+  return vqmovns_u32(a);
+// CHECK: uqxtn.4h {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovnd_s64
+int32_t test_vqmovnd_s64(int64_t a) {
+  return vqmovnd_s64(a);
+// CHECK: sqxtn {{s[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqmovnd_u64
+uint32_t test_vqmovnd_u64(uint64_t a) {
+  return vqmovnd_u64(a);
+// CHECK: uqxtn {{s[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqmovunh_s16
+int8_t test_vqmovunh_s16(int16_t a) {
+  return vqmovunh_s16(a);
+// CHECK: sqxtun.8b {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovuns_s32
+int16_t test_vqmovuns_s32(int32_t a) {
+  return vqmovuns_s32(a);
+// CHECK: sqxtun.4h {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqmovund_s64
+int32_t test_vqmovund_s64(int64_t a) {
+  return vqmovund_s64(a);
+// CHECK: sqxtun {{s[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqabsb_s8
+int8_t test_vqabsb_s8(int8_t a) {
+  return vqabsb_s8(a);
+// CHECK: sqabs.8b {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqabsh_s16
+int16_t test_vqabsh_s16(int16_t a) {
+  return vqabsh_s16(a);
+// CHECK: sqabs.4h {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqabss_s32
+int32_t test_vqabss_s32(int32_t a) {
+  return vqabss_s32(a);
+// CHECK: sqabs {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqabsd_s64
+int64_t test_vqabsd_s64(int64_t a) {
+  return vqabsd_s64(a);
+// CHECK: sqabs {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vqnegb_s8
+int8_t test_vqnegb_s8(int8_t a) {
+  return vqnegb_s8(a);
+// CHECK: sqneg.8b {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqnegh_s16
+int16_t test_vqnegh_s16(int16_t a) {
+  return vqnegh_s16(a);
+// CHECK: sqneg.4h {{v[0-9]+}}, {{v[0-9]+}}
+}
+
+// CHECK: test_vqnegs_s32
+int32_t test_vqnegs_s32(int32_t a) {
+  return vqnegs_s32(a);
+// CHECK: sqneg {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vqnegd_s64
+int64_t test_vqnegd_s64(int64_t a) {
+  return vqnegd_s64(a);
+// CHECK: sqneg {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvts_n_f32_s32
+float32_t test_vcvts_n_f32_s32(int32_t a) {
+  return vcvts_n_f32_s32(a, 3);
+// CHECK: scvtf {{s[0-9]+}}, {{s[0-9]+}}, #3
+}
+
+// CHECK: test_vcvts_n_f32_u32
+float32_t test_vcvts_n_f32_u32(uint32_t a) {
+  return vcvts_n_f32_u32(a, 3);
+// CHECK: ucvtf {{s[0-9]+}}, {{s[0-9]+}}, #3
+}
+
+// CHECK: test_vcvtd_n_f64_s64
+float64_t test_vcvtd_n_f64_s64(int64_t a) {
+  return vcvtd_n_f64_s64(a, 3);
+// CHECK: scvtf {{d[0-9]+}}, {{d[0-9]+}}, #3
+}
+
+// CHECK: test_vcvtd_n_f64_u64
+float64_t test_vcvtd_n_f64_u64(uint64_t a) {
+  return vcvtd_n_f64_u64(a, 3);
+// CHECK: ucvtf {{d[0-9]+}}, {{d[0-9]+}}, #3
+}
+
+// CHECK: test_vcvts_n_s32_f32
+int32_t test_vcvts_n_s32_f32(float32_t a) {
+  return vcvts_n_s32_f32(a, 3);
+// CHECK: fcvtzs {{s[0-9]+}}, {{s[0-9]+}}, #3
+}
+
+// CHECK: test_vcvts_n_u32_f32
+uint32_t test_vcvts_n_u32_f32(float32_t a) {
+  return vcvts_n_u32_f32(a, 3);
+// CHECK: fcvtzu {{s[0-9]+}}, {{s[0-9]+}}, #3
+}
+
+// CHECK: test_vcvtd_n_s64_f64
+int64_t test_vcvtd_n_s64_f64(float64_t a) {
+  return vcvtd_n_s64_f64(a, 3);
+// CHECK: fcvtzs {{d[0-9]+}}, {{d[0-9]+}}, #3
+}
+
+// CHECK: test_vcvtd_n_u64_f64
+uint64_t test_vcvtd_n_u64_f64(float64_t a) {
+  return vcvtd_n_u64_f64(a, 3);
+// CHECK: fcvtzu {{d[0-9]+}}, {{d[0-9]+}}, #3
+}
+
+// CHECK: test_vcvtas_s32_f32
+int32_t test_vcvtas_s32_f32(float32_t a) {
+  return vcvtas_s32_f32(a);
+// CHECK: fcvtas {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtas_u32_f32
+uint32_t test_vcvtas_u32_f32(float32_t a) {
+  return vcvtas_u32_f32(a);
+// CHECK: fcvtau {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtad_s64_f64
+int64_t test_vcvtad_s64_f64(float64_t a) {
+  return vcvtad_s64_f64(a);
+// CHECK: fcvtas {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtad_u64_f64
+uint64_t test_vcvtad_u64_f64(float64_t a) {
+  return vcvtad_u64_f64(a);
+// CHECK: fcvtau {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtms_s32_f32
+int32_t test_vcvtms_s32_f32(float32_t a) {
+  return vcvtms_s32_f32(a);
+// CHECK: fcvtms {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtms_u32_f32
+uint32_t test_vcvtms_u32_f32(float32_t a) {
+  return vcvtms_u32_f32(a);
+// CHECK: fcvtmu {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtmd_s64_f64
+int64_t test_vcvtmd_s64_f64(float64_t a) {
+  return vcvtmd_s64_f64(a);
+// CHECK: fcvtms {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtmd_u64_f64
+uint64_t test_vcvtmd_u64_f64(float64_t a) {
+  return vcvtmd_u64_f64(a);
+// CHECK: fcvtmu {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtns_s32_f32
+int32_t test_vcvtns_s32_f32(float32_t a) {
+  return vcvtns_s32_f32(a);
+// CHECK: fcvtns {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtns_u32_f32
+uint32_t test_vcvtns_u32_f32(float32_t a) {
+  return vcvtns_u32_f32(a);
+// CHECK: fcvtnu {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtnd_s64_f64
+int64_t test_vcvtnd_s64_f64(float64_t a) {
+  return vcvtnd_s64_f64(a);
+// CHECK: fcvtns {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtnd_u64_f64
+uint64_t test_vcvtnd_u64_f64(float64_t a) {
+  return vcvtnd_u64_f64(a);
+// CHECK: fcvtnu {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtps_s32_f32
+int32_t test_vcvtps_s32_f32(float32_t a) {
+  return vcvtps_s32_f32(a);
+// CHECK: fcvtps {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtps_u32_f32
+uint32_t test_vcvtps_u32_f32(float32_t a) {
+  return vcvtps_u32_f32(a);
+// CHECK: fcvtpu {{w[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vcvtpd_s64_f64
+int64_t test_vcvtpd_s64_f64(float64_t a) {
+  return vcvtpd_s64_f64(a);
+// CHECK: fcvtps {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtpd_u64_f64
+uint64_t test_vcvtpd_u64_f64(float64_t a) {
+  return vcvtpd_u64_f64(a);
+// CHECK: fcvtpu {{x[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vcvtxd_f32_f64
+float32_t test_vcvtxd_f32_f64(float64_t a) {
+  return vcvtxd_f32_f64(a);
+// CHECK: fcvtxn {{s[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vabds_f32
+float32_t test_vabds_f32(float32_t a, float32_t b) {
+  return vabds_f32(a, b);
+  // CHECK: fabd {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vabdd_f64
+float64_t test_vabdd_f64(float64_t a, float64_t b) {
+  return vabdd_f64(a, b);
+  // CHECK: fabd {{d[0-9]+}}, {{d[0-9]+}}
+}
+
+// CHECK: test_vmulxs_f32
+float32_t test_vmulxs_f32(float32_t a, float32_t b) {
+  return vmulxs_f32(a, b);
+  // CHECK: fmulx {{s[0-9]+}}, {{s[0-9]+}}
+}
+
+// CHECK: test_vmulxd_f64
+float64_t test_vmulxd_f64(float64_t a, float64_t b) {
+  return vmulxd_f64(a, b);
+  // CHECK: fmulx {{d[0-9]+}}, {{d[0-9]+}}
+}
diff --git a/test/CodeGen/arm64-vrnd.c b/test/CodeGen/arm64-vrnd.c
new file mode 100644
index 0000000..4de2ec7
--- /dev/null
+++ b/test/CodeGen/arm64-vrnd.c
@@ -0,0 +1,51 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7 -ffreestanding -emit-llvm -o - %s | FileCheck %s
+
+#include <arm_neon.h>
+
+int32x2_t rnd1(float32x2_t a) { return vrnd_f32(a); }
+// CHECK: call <2 x float> @llvm.trunc.v2f32(<2 x float>
+int32x4_t rnd3(float32x4_t a) { return vrndq_f32(a); }
+// CHECK: call <4 x float> @llvm.trunc.v4f32(<4 x float>
+int64x2_t rnd5(float64x2_t a) { return vrndq_f64(a); }
+// CHECK: call <2 x double> @llvm.trunc.v2f64(<2 x double>
+
+
+int32x2_t rnd7(float32x2_t a) { return vrndn_f32(a); }
+// CHECK: call <2 x float> @llvm.arm64.neon.frintn.v2f32(<2 x float>
+int32x4_t rnd8(float32x4_t a) { return vrndnq_f32(a); }
+// CHECK: call <4 x float> @llvm.arm64.neon.frintn.v4f32(<4 x float>
+int64x2_t rnd9(float64x2_t a) { return vrndnq_f64(a); }
+// CHECK: call <2 x double> @llvm.arm64.neon.frintn.v2f64(<2 x double>
+int64x2_t rnd10(float64x2_t a) { return vrndnq_f64(a); }
+// CHECK: call <2 x double> @llvm.arm64.neon.frintn.v2f64(<2 x double>
+
+int32x2_t rnd11(float32x2_t a) { return vrndm_f32(a); }
+// CHECK: call <2 x float> @llvm.floor.v2f32(<2 x float>
+int32x4_t rnd12(float32x4_t a) { return vrndmq_f32(a); }
+// CHECK: call <4 x float> @llvm.floor.v4f32(<4 x float>
+int64x2_t rnd13(float64x2_t a) { return vrndmq_f64(a); }
+// CHECK: call <2 x double> @llvm.floor.v2f64(<2 x double>
+int64x2_t rnd14(float64x2_t a) { return vrndmq_f64(a); }
+// CHECK: call <2 x double> @llvm.floor.v2f64(<2 x double>
+
+int32x2_t rnd15(float32x2_t a) { return vrndp_f32(a); }
+// CHECK: call <2 x float> @llvm.ceil.v2f32(<2 x float>
+int32x4_t rnd16(float32x4_t a) { return vrndpq_f32(a); }
+// CHECK: call <4 x float> @llvm.ceil.v4f32(<4 x float>
+int64x2_t rnd18(float64x2_t a) { return vrndpq_f64(a); }
+// CHECK: call <2 x double> @llvm.ceil.v2f64(<2 x double>
+
+int32x2_t rnd19(float32x2_t a) { return vrnda_f32(a); }
+// CHECK: call <2 x float> @llvm.round.v2f32(<2 x float>
+int32x4_t rnd20(float32x4_t a) { return vrndaq_f32(a); }
+// CHECK: call <4 x float> @llvm.round.v4f32(<4 x float>
+int64x2_t rnd22(float64x2_t a) { return vrndaq_f64(a); }
+// CHECK: call <2 x double> @llvm.round.v2f64(<2 x double>
+
+int32x2_t rnd23(float32x2_t a) { return vrndx_f32(a); }
+// CHECK: call <2 x float> @llvm.rint.v2f32(<2 x float>
+int32x4_t rnd24(float32x4_t a) { return vrndxq_f32(a); }
+// CHECK: call <4 x float> @llvm.rint.v4f32(<4 x float>
+int64x2_t rnd25(float64x2_t a) { return vrndxq_f64(a); }
+// CHECK: call <2 x double> @llvm.rint.v2f64(<2 x double>
+
diff --git a/test/CodeGen/arm64-vrsqrt.c b/test/CodeGen/arm64-vrsqrt.c
new file mode 100644
index 0000000..45a536c
--- /dev/null
+++ b/test/CodeGen/arm64-vrsqrt.c
@@ -0,0 +1,43 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7.0 -ffreestanding -emit-llvm -O1 -o - %s | FileCheck %s
+
+#include <arm_neon.h>
+
+uint32x2_t test_vrsqrte_u32(uint32x2_t in) {
+  // CHECK-LABEL: @test_vrsqrte_u32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.ursqrte.v2i32(<2 x i32> %in)
+  return vrsqrte_u32(in);
+}
+
+float32x2_t test_vrsqrte_f32(float32x2_t in) {
+  // CHECK-LABEL: @test_vrsqrte_f32
+  // CHECK: call <2 x float> @llvm.arm64.neon.frsqrte.v2f32(<2 x float> %in)
+  return vrsqrte_f32(in);
+}
+
+
+uint32x4_t test_vrsqrteq_u32(uint32x4_t in) {
+  // CHECK-LABEL: @test_vrsqrteq_u32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.ursqrte.v4i32(<4 x i32> %in)
+  return vrsqrteq_u32(in);
+}
+
+float32x4_t test_vrsqrteq_f32(float32x4_t in) {
+  // CHECK-LABEL: @test_vrsqrteq_f32
+  // CHECK: call <4 x float> @llvm.arm64.neon.frsqrte.v4f32(<4 x float> %in)
+  return vrsqrteq_f32(in);
+}
+
+
+float32x2_t test_vrsqrts_f32(float32x2_t est, float32x2_t val) {
+  // CHECK-LABEL: @test_vrsqrts_f32
+  // CHECK: call <2 x float> @llvm.arm64.neon.frsqrts.v2f32(<2 x float> %est, <2 x float> %val)
+  return vrsqrts_f32(est, val);
+}
+
+
+float32x4_t test_vrsqrtsq_f32(float32x4_t est, float32x4_t val) {
+  // CHECK-LABEL: @test_vrsqrtsq_f32
+  // CHECK: call <4 x float> @llvm.arm64.neon.frsqrts.v4f32(<4 x float> %est, <4 x float> %val)
+  return vrsqrtsq_f32(est, val);
+}
+
diff --git a/test/CodeGen/arm64_crypto.c b/test/CodeGen/arm64_crypto.c
new file mode 100644
index 0000000..7150c56
--- /dev/null
+++ b/test/CodeGen/arm64_crypto.c
@@ -0,0 +1,93 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7.0 -ffreestanding -Os -S -o - %s | FileCheck %s
+// REQUIRES: arm64-registered-target
+
+#include <arm_neon.h>
+
+uint8x16_t test_aese(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: test_aese:
+  // CHECK: aese.16b v0, v1
+  return vaeseq_u8(data, key);
+}
+
+uint8x16_t test_aesd(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: test_aesd:
+  // CHECK: aesd.16b v0, v1
+  return vaesdq_u8(data, key);
+}
+
+uint8x16_t test_aesmc(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: test_aesmc:
+  // CHECK: aesmc.16b v0, v0
+  return vaesmcq_u8(data);
+}
+
+uint8x16_t test_aesimc(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: test_aesimc:
+  // CHECK: aesimc.16b v0, v0
+  return vaesimcq_u8(data);
+}
+
+uint32x4_t test_sha1c(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: test_sha1c:
+  // CHECK: fmov [[HASH_E:s[0-9]+]], w0
+  // CHECK: sha1c.4s q0, [[HASH_E]], v1
+  return vsha1cq_u32(hash_abcd, hash_e, wk);
+}
+
+uint32x4_t test_sha1p(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: test_sha1p:
+  // CHECK: fmov [[HASH_E:s[0-9]+]], w0
+  // CHECK: sha1p.4s q0, [[HASH_E]], v1
+  return vsha1pq_u32(hash_abcd, hash_e, wk);
+}
+
+uint32x4_t test_sha1m(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: test_sha1m:
+  // CHECK: fmov [[HASH_E:s[0-9]+]], w0
+  // CHECK: sha1m.4s q0, [[HASH_E]], v1
+  return vsha1mq_u32(hash_abcd, hash_e, wk);
+}
+
+uint32_t test_sha1h(uint32_t hash_e) {
+  // CHECK-LABEL: test_sha1h:
+  // CHECK: fmov [[HASH_E:s[0-9]+]], w0
+  // CHECK: sha1h [[RES:s[0-9]+]], [[HASH_E]]
+  // CHECK: fmov w0, [[RES]]
+  return vsha1h_u32(hash_e);
+}
+
+uint32x4_t test_sha1su0(uint32x4_t wk0_3, uint32x4_t wk4_7, uint32x4_t wk8_11) {
+  // CHECK-LABEL: test_sha1su0:
+  // CHECK: sha1su0.4s v0, v1, v2
+  return vsha1su0q_u32(wk0_3, wk4_7, wk8_11);
+}
+
+uint32x4_t test_sha1su1(uint32x4_t wk0_3, uint32x4_t wk12_15) {
+  // CHECK-LABEL: test_sha1su1:
+  // CHECK: sha1su1.4s v0, v1
+  return vsha1su1q_u32(wk0_3, wk12_15);
+}
+
+uint32x4_t test_sha256h(uint32x4_t hash_abcd, uint32x4_t hash_efgh, uint32x4_t wk) {
+  // CHECK-LABEL: test_sha256h:
+  // CHECK: sha256h.4s q0, q1, v2
+  return vsha256hq_u32(hash_abcd, hash_efgh, wk);
+}
+
+uint32x4_t test_sha256h2(uint32x4_t hash_efgh, uint32x4_t hash_abcd, uint32x4_t wk) {
+  // CHECK-LABEL: test_sha256h2:
+  // CHECK: sha256h2.4s q0, q1, v2
+  return vsha256h2q_u32(hash_efgh, hash_abcd, wk);
+}
+
+uint32x4_t test_sha256su0(uint32x4_t w0_3, uint32x4_t w4_7) {
+  // CHECK-LABEL: test_sha256su0:
+  // CHECK: sha256su0.4s v0, v1
+  return vsha256su0q_u32(w0_3, w4_7);
+}
+
+uint32x4_t test_sha256su1(uint32x4_t w0_3, uint32x4_t w8_11, uint32x4_t w12_15) {
+  // CHECK-LABEL: test_sha256su1:
+  // CHECK: sha256su1.4s v0, v1, v2
+  return vsha256su1q_u32(w0_3, w8_11, w12_15);
+}
diff --git a/test/CodeGen/arm64_neon_high_half.c b/test/CodeGen/arm64_neon_high_half.c
new file mode 100644
index 0000000..920dded
--- /dev/null
+++ b/test/CodeGen/arm64_neon_high_half.c
@@ -0,0 +1,559 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7.0 -ffreestanding -Os -S -o - %s | FileCheck %s
+// REQUIRES: arm64-registered-target
+
+#include <arm_neon.h>
+
+int16x8_t test_vaddw_high_s8(int16x8_t lhs, int8x16_t rhs) {
+  // CHECK: saddw2.8h
+  return vaddw_high_s8(lhs, rhs);
+}
+
+int32x4_t test_vaddw_high_s16(int32x4_t lhs, int16x8_t rhs) {
+  // CHECK: saddw2.4s
+  return vaddw_high_s16(lhs, rhs);
+}
+
+int64x2_t test_vaddw_high_s32(int64x2_t lhs, int32x4_t rhs) {
+  // CHECK: saddw2.2d
+  return vaddw_high_s32(lhs, rhs);
+}
+
+uint16x8_t test_vaddw_high_u8(uint16x8_t lhs, uint8x16_t rhs) {
+  // CHECK: uaddw2.8h
+  return vaddw_high_u8(lhs, rhs);
+}
+
+uint32x4_t test_vaddw_high_u16(uint32x4_t lhs, uint16x8_t rhs) {
+  // CHECK: uaddw2.4s
+  return vaddw_high_u16(lhs, rhs);
+}
+
+uint64x2_t test_vaddw_high_u32(uint64x2_t lhs, uint32x4_t rhs) {
+  // CHECK: uaddw2.2d
+  return vaddw_high_u32(lhs, rhs);
+}
+
+int16x8_t test_vsubw_high_s8(int16x8_t lhs, int8x16_t rhs) {
+  // CHECK: ssubw2.8h
+  return vsubw_high_s8(lhs, rhs);
+}
+
+int32x4_t test_vsubw_high_s16(int32x4_t lhs, int16x8_t rhs) {
+  // CHECK: ssubw2.4s
+  return vsubw_high_s16(lhs, rhs);
+}
+
+int64x2_t test_vsubw_high_s32(int64x2_t lhs, int32x4_t rhs) {
+  // CHECK: ssubw2.2d
+  return vsubw_high_s32(lhs, rhs);
+}
+
+uint16x8_t test_vsubw_high_u8(uint16x8_t lhs, uint8x16_t rhs) {
+  // CHECK: usubw2.8h
+  return vsubw_high_u8(lhs, rhs);
+}
+
+uint32x4_t test_vsubw_high_u16(uint32x4_t lhs, uint16x8_t rhs) {
+  // CHECK: usubw2.4s
+  return vsubw_high_u16(lhs, rhs);
+}
+
+uint64x2_t test_vsubw_high_u32(uint64x2_t lhs, uint32x4_t rhs) {
+  // CHECK: usubw2.2d
+  return vsubw_high_u32(lhs, rhs);
+}
+
+int16x8_t test_vabdl_high_s8(int8x16_t lhs, int8x16_t rhs) {
+  // CHECK: sabdl2.8h
+  return vabdl_high_s8(lhs, rhs);
+}
+
+int32x4_t test_vabdl_high_s16(int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: sabdl2.4s
+  return vabdl_high_s16(lhs, rhs);
+}
+
+int64x2_t test_vabdl_high_s32(int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: sabdl2.2d
+  return vabdl_high_s32(lhs, rhs);
+}
+
+uint16x8_t test_vabdl_high_u8(uint8x16_t lhs, uint8x16_t rhs) {
+  // CHECK: uabdl2.8h
+  return vabdl_high_u8(lhs, rhs);
+}
+
+uint32x4_t test_vabdl_high_u16(uint16x8_t lhs, uint16x8_t rhs) {
+  // CHECK: uabdl2.4s
+  return vabdl_high_u16(lhs, rhs);
+}
+
+uint64x2_t test_vabdl_high_u32(uint32x4_t lhs, uint32x4_t rhs) {
+  // CHECK: uabdl2.2d
+  return vabdl_high_u32(lhs, rhs);
+}
+
+int16x8_t test_vabal_high_s8(int16x8_t accum, int8x16_t lhs, int8x16_t rhs) {
+  // CHECK: sabal2.8h
+  return vabal_high_s8(accum, lhs, rhs);
+}
+
+int32x4_t test_vabal_high_s16(int32x4_t accum, int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: sabal2.4s
+  return vabal_high_s16(accum, lhs, rhs);
+}
+
+int64x2_t test_vabal_high_s32(int64x2_t accum, int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: sabal2.2d
+  return vabal_high_s32(accum, lhs, rhs);
+}
+
+uint16x8_t test_vabal_high_u8(uint16x8_t accum, uint8x16_t lhs, uint8x16_t rhs) {
+  // CHECK: uabal2.8h
+  return vabal_high_u8(accum, lhs, rhs);
+}
+
+uint32x4_t test_vabal_high_u16(uint32x4_t accum, uint16x8_t lhs, uint16x8_t rhs) {
+  // CHECK: uabal2.4s
+  return vabal_high_u16(accum, lhs, rhs);
+}
+
+uint64x2_t test_vabal_high_u32(uint64x2_t accum, uint32x4_t lhs, uint32x4_t rhs) {
+  // CHECK: uabal2.2d
+  return vabal_high_u32(accum, lhs, rhs);
+}
+
+int32x4_t test_vqdmlal_high_s16(int32x4_t accum, int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: sqdmlal2.4s
+  return vqdmlal_high_s16(accum, lhs, rhs);
+}
+
+int64x2_t test_vqdmlal_high_s32(int64x2_t accum, int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: sqdmlal2.2d
+  return vqdmlal_high_s32(accum, lhs, rhs);
+}
+
+int32x4_t test_vqdmlsl_high_s16(int32x4_t accum, int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: sqdmlsl2.4s
+  return vqdmlsl_high_s16(accum, lhs, rhs);
+}
+
+int64x2_t test_vqdmlsl_high_s32(int64x2_t accum, int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: sqdmlsl2.2d
+  return vqdmlsl_high_s32(accum, lhs, rhs);
+}
+
+int32x4_t test_vqdmull_high_s16(int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: sqdmull2.4s
+  return vqdmull_high_s16(lhs, rhs);
+}
+
+int64x2_t test_vqdmull_high_s32(int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: sqdmull2.2d
+  return vqdmull_high_s32(lhs, rhs);
+}
+
+int16x8_t test_vshll_high_n_s8(int8x16_t in) {
+  // CHECK: sshll2.8h
+  return vshll_high_n_s8(in, 7);
+}
+
+int32x4_t test_vshll_high_n_s16(int16x8_t in) {
+  // CHECK: sshll2.4s
+  return vshll_high_n_s16(in, 15);
+}
+
+int64x2_t test_vshll_high_n_s32(int32x4_t in) {
+  // CHECK: sshll2.2d
+  return vshll_high_n_s32(in, 31);
+}
+
+int16x8_t test_vshll_high_n_u8(int8x16_t in) {
+  // CHECK: ushll2.8h
+  return vshll_high_n_u8(in, 7);
+}
+
+int32x4_t test_vshll_high_n_u16(int16x8_t in) {
+  // CHECK: ushll2.4s
+  return vshll_high_n_u16(in, 15);
+}
+
+int64x2_t test_vshll_high_n_u32(int32x4_t in) {
+  // CHECK: ushll2.2d
+  return vshll_high_n_u32(in, 31);
+}
+
+int16x8_t test_vshll_high_n_s8_max(int8x16_t in) {
+  // CHECK: shll2.8h
+  return vshll_high_n_s8(in, 8);
+}
+
+int32x4_t test_vshll_high_n_s16_max(int16x8_t in) {
+  // CHECK: shll2.4s
+  return vshll_high_n_s16(in, 16);
+}
+
+int64x2_t test_vshll_high_n_s32_max(int32x4_t in) {
+  // CHECK: shll2.2d
+  return vshll_high_n_s32(in, 32);
+}
+
+int16x8_t test_vshll_high_n_u8_max(int8x16_t in) {
+  // CHECK: shll2.8h
+  return vshll_high_n_u8(in, 8);
+}
+
+int32x4_t test_vshll_high_n_u16_max(int16x8_t in) {
+  // CHECK: shll2.4s
+  return vshll_high_n_u16(in, 16);
+}
+
+int64x2_t test_vshll_high_n_u32_max(int32x4_t in) {
+  // CHECK: shll2.2d
+  return vshll_high_n_u32(in, 32);
+}
+
+int16x8_t test_vsubl_high_s8(int8x16_t lhs, int8x16_t rhs) {
+  // CHECK: ssubl2.8h
+  return vsubl_high_s8(lhs, rhs);
+}
+
+int32x4_t test_vsubl_high_s16(int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: ssubl2.4s
+  return vsubl_high_s16(lhs, rhs);
+}
+
+int64x2_t test_vsubl_high_s32(int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: ssubl2.2d
+  return vsubl_high_s32(lhs, rhs);
+}
+
+uint16x8_t test_vsubl_high_u8(uint8x16_t lhs, uint8x16_t rhs) {
+  // CHECK: usubl2.8h
+  return vsubl_high_u8(lhs, rhs);
+}
+
+uint32x4_t test_vsubl_high_u16(uint16x8_t lhs, uint16x8_t rhs) {
+  // CHECK: usubl2.4s
+  return vsubl_high_u16(lhs, rhs);
+}
+
+uint64x2_t test_vsubl_high_u32(uint32x4_t lhs, uint32x4_t rhs) {
+  // CHECK: usubl2.2d
+  return vsubl_high_u32(lhs, rhs);
+}
+
+int8x16_t test_vrshrn_high_n_s16(int8x8_t lowpart, int16x8_t input) {
+  // CHECK: rshrn2.16b
+  return vrshrn_high_n_s16(lowpart, input, 2);
+}
+
+int16x8_t test_vrshrn_high_n_s32(int16x4_t lowpart, int32x4_t input) {
+  // CHECK: rshrn2.8h
+  return vrshrn_high_n_s32(lowpart, input, 2);
+}
+
+int32x4_t test_vrshrn_high_n_s64(int32x2_t lowpart, int64x2_t input) {
+  // CHECK: shrn2.4s
+  return vrshrn_high_n_s64(lowpart, input, 2);
+}
+
+uint8x16_t test_vrshrn_high_n_u16(uint8x8_t lowpart, uint16x8_t input) {
+  // CHECK: rshrn2.16b
+  return vrshrn_high_n_u16(lowpart, input, 2);
+}
+
+uint16x8_t test_vrshrn_high_n_u32(uint16x4_t lowpart, uint32x4_t input) {
+  // CHECK: rshrn2.8h
+  return vrshrn_high_n_u32(lowpart, input, 2);
+}
+
+uint32x4_t test_vrshrn_high_n_u64(uint32x2_t lowpart, uint64x2_t input) {
+  // CHECK: rshrn2.4s
+  return vrshrn_high_n_u64(lowpart, input, 2);
+}
+
+int8x16_t test_vshrn_high_n_s16(int8x8_t lowpart, int16x8_t input) {
+  // CHECK: shrn2.16b
+  return vshrn_high_n_s16(lowpart, input, 2);
+}
+
+int16x8_t test_vshrn_high_n_s32(int16x4_t lowpart, int32x4_t input) {
+  // CHECK: shrn2.8h
+  return vshrn_high_n_s32(lowpart, input, 2);
+}
+
+int32x4_t test_vshrn_high_n_s64(int32x2_t lowpart, int64x2_t input) {
+  // CHECK: shrn2.4s
+  return vshrn_high_n_s64(lowpart, input, 2);
+}
+
+uint8x16_t test_vshrn_high_n_u16(uint8x8_t lowpart, uint16x8_t input) {
+  // CHECK: shrn2.16b
+  return vshrn_high_n_u16(lowpart, input, 2);
+}
+
+uint16x8_t test_vshrn_high_n_u32(uint16x4_t lowpart, uint32x4_t input) {
+  // CHECK: shrn2.8h
+  return vshrn_high_n_u32(lowpart, input, 2);
+}
+
+uint32x4_t test_vshrn_high_n_u64(uint32x2_t lowpart, uint64x2_t input) {
+  // CHECK: shrn2.4s
+  return vshrn_high_n_u64(lowpart, input, 2);
+}
+
+uint8x16_t test_vqshrun_high_n_s16(uint8x8_t lowpart, int16x8_t input) {
+  // CHECK: sqshrun2.16b
+  return vqshrun_high_n_s16(lowpart, input, 2);
+}
+
+uint16x8_t test_vqshrun_high_n_s32(uint16x4_t lowpart, int32x4_t input) {
+  // CHECK: sqshrun2.8h
+  return vqshrun_high_n_s32(lowpart, input, 2);
+}
+
+uint32x4_t test_vqshrun_high_n_s64(uint32x2_t lowpart, int64x2_t input) {
+  // CHECK: sqshrun2.4s
+  return vqshrun_high_n_s64(lowpart, input, 2);
+}
+
+uint8x16_t test_vqrshrun_high_n_s16(uint8x8_t lowpart, int16x8_t input) {
+  // CHECK: sqrshrun2.16b
+  return vqrshrun_high_n_s16(lowpart, input, 2);
+}
+
+uint16x8_t test_vqrshrun_high_n_s32(uint16x4_t lowpart, int32x4_t input) {
+  // CHECK: sqrshrun2.8h
+  return vqrshrun_high_n_s32(lowpart, input, 2);
+}
+
+uint32x4_t test_vqrshrun_high_n_s64(uint32x2_t lowpart, int64x2_t input) {
+  // CHECK: sqrshrun2.4s
+  return vqrshrun_high_n_s64(lowpart, input, 2);
+}
+
+int8x16_t test_vqshrn_high_n_s16(int8x8_t lowpart, int16x8_t input) {
+  // CHECK: sqshrn2.16b
+  return vqshrn_high_n_s16(lowpart, input, 2);
+}
+
+int16x8_t test_vqshrn_high_n_s32(int16x4_t lowpart, int32x4_t input) {
+  // CHECK: sqshrn2.8h
+  return vqshrn_high_n_s32(lowpart, input, 2);
+}
+
+int32x4_t test_vqshrn_high_n_s64(int32x2_t lowpart, int64x2_t input) {
+  // CHECK: sqshrn2.4s
+  return vqshrn_high_n_s64(lowpart, input, 2);
+}
+
+uint8x16_t test_vqshrn_high_n_u16(uint8x8_t lowpart, uint16x8_t input) {
+  // CHECK: uqshrn2.16b
+  return vqshrn_high_n_u16(lowpart, input, 2);
+}
+
+uint16x8_t test_vqshrn_high_n_u32(uint16x4_t lowpart, uint32x4_t input) {
+  // CHECK: uqshrn2.8h
+  return vqshrn_high_n_u32(lowpart, input, 2);
+}
+
+uint32x4_t test_vqshrn_high_n_u64(uint32x2_t lowpart, uint64x2_t input) {
+  // CHECK: uqshrn2.4s
+  return vqshrn_high_n_u64(lowpart, input, 2);
+}
+
+int8x16_t test_vqrshrn_high_n_s16(int8x8_t lowpart, int16x8_t input) {
+  // CHECK: sqrshrn2.16b
+  return vqrshrn_high_n_s16(lowpart, input, 2);
+}
+
+int16x8_t test_vqrshrn_high_n_s32(int16x4_t lowpart, int32x4_t input) {
+  // CHECK: sqrshrn2.8h
+  return vqrshrn_high_n_s32(lowpart, input, 2);
+}
+
+int32x4_t test_vqrshrn_high_n_s64(int32x2_t lowpart, int64x2_t input) {
+  // CHECK: sqrshrn2.4s
+  return vqrshrn_high_n_s64(lowpart, input, 2);
+}
+
+uint8x16_t test_vqrshrn_high_n_u16(uint8x8_t lowpart, uint16x8_t input) {
+  // CHECK: uqrshrn2.16b
+  return vqrshrn_high_n_u16(lowpart, input, 2);
+}
+
+uint16x8_t test_vqrshrn_high_n_u32(uint16x4_t lowpart, uint32x4_t input) {
+  // CHECK: uqrshrn2.8h
+  return vqrshrn_high_n_u32(lowpart, input, 2);
+}
+
+uint32x4_t test_vqrshrn_high_n_u64(uint32x2_t lowpart, uint64x2_t input) {
+  // CHECK: uqrshrn2.4s
+  return vqrshrn_high_n_u64(lowpart, input, 2);
+}
+
+int8x16_t test_vaddhn_high_s16(int8x8_t lowpart, int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: addhn2.16b v0, v1, v2
+  return vaddhn_high_s16(lowpart, lhs, rhs);
+}
+
+int16x8_t test_vaddhn_high_s32(int16x4_t lowpart, int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: addhn2.8h v0, v1, v2
+  return vaddhn_high_s32(lowpart, lhs, rhs);
+}
+
+int32x4_t test_vaddhn_high_s64(int32x2_t lowpart, int64x2_t lhs, int64x2_t rhs) {
+  // CHECK: addhn2.4s v0, v1, v2
+  return vaddhn_high_s64(lowpart, lhs, rhs);
+}
+
+uint8x16_t test_vaddhn_high_u16(uint8x8_t lowpart, uint16x8_t lhs, uint16x8_t rhs) {
+  // CHECK: addhn2.16b v0, v1, v2
+  return vaddhn_high_s16(lowpart, lhs, rhs);
+}
+
+uint16x8_t test_vaddhn_high_u32(uint16x4_t lowpart, uint32x4_t lhs, uint32x4_t rhs) {
+  // CHECK: addhn2.8h v0, v1, v2
+  return vaddhn_high_s32(lowpart, lhs, rhs);
+}
+
+uint32x4_t test_vaddhn_high_u64(uint32x2_t lowpart, uint64x2_t lhs, uint64x2_t rhs) {
+  // CHECK: addhn2.4s v0, v1, v2
+  return vaddhn_high_s64(lowpart, lhs, rhs);
+}
+
+int8x16_t test_vraddhn_high_s16(int8x8_t lowpart, int16x8_t lhs, int16x8_t rhs) {
+  // CHECK: raddhn2.16b v0, v1, v2
+  return vraddhn_high_s16(lowpart, lhs, rhs);
+}
+
+int16x8_t test_vraddhn_high_s32(int16x4_t lowpart, int32x4_t lhs, int32x4_t rhs) {
+  // CHECK: raddhn2.8h v0, v1, v2
+  return vraddhn_high_s32(lowpart, lhs, rhs);
+}
+
+int32x4_t test_vraddhn_high_s64(int32x2_t lowpart, int64x2_t lhs, int64x2_t rhs) {
+  // CHECK: raddhn2.4s v0, v1, v2
+  return vraddhn_high_s64(lowpart, lhs, rhs);
+}
+
+uint8x16_t test_vraddhn_high_u16(uint8x8_t lowpart, uint16x8_t lhs, uint16x8_t rhs) {
+  // CHECK: raddhn2.16b v0, v1, v2
+  return vraddhn_high_s16(lowpart, lhs, rhs);
+}
+
+uint16x8_t test_vraddhn_high_u32(uint16x4_t lowpart, uint32x4_t lhs, uint32x4_t rhs) {
+  // CHECK: raddhn2.8h v0, v1, v2
+  return vraddhn_high_s32(lowpart, lhs, rhs);
+}
+
+uint32x4_t test_vraddhn_high_u64(uint32x2_t lowpart, uint64x2_t lhs, uint64x2_t rhs) {
+  // CHECK: raddhn2.4s v0, v1, v2
+  return vraddhn_high_s64(lowpart, lhs, rhs);
+}
+
+int8x16_t test_vmovn_high_s16(int8x8_t lowpart, int16x8_t wide) {
+  // CHECK: xtn2.16b v0, v1
+  return vmovn_high_s16(lowpart, wide);
+}
+
+int16x8_t test_vmovn_high_s32(int16x4_t lowpart, int32x4_t wide) {
+  // CHECK: xtn2.8h v0, v1
+  return vmovn_high_s32(lowpart, wide);
+}
+
+int32x4_t test_vmovn_high_s64(int32x2_t lowpart, int64x2_t wide) {
+  // CHECK: xtn2.4s v0, v1
+  return vmovn_high_s64(lowpart, wide);
+}
+
+uint8x16_t test_vmovn_high_u16(uint8x8_t lowpart, uint16x8_t wide) {
+  // CHECK: xtn2.16b v0, v1
+  return vmovn_high_u16(lowpart, wide);
+}
+
+uint16x8_t test_vmovn_high_u32(uint16x4_t lowpart, uint32x4_t wide) {
+  // CHECK: xtn2.8h v0, v1
+  return vmovn_high_u32(lowpart, wide);
+}
+
+uint32x4_t test_vmovn_high_u64(uint32x2_t lowpart, uint64x2_t wide) {
+  // CHECK: xtn2.4s v0, v1
+  return vmovn_high_u64(lowpart, wide);
+}
+
+int8x16_t test_vqmovn_high_s16(int8x8_t lowpart, int16x8_t wide) {
+  // CHECK: sqxtn2.16b v0, v1
+  return vqmovn_high_s16(lowpart, wide);
+}
+
+int16x8_t test_vqmovn_high_s32(int16x4_t lowpart, int32x4_t wide) {
+  // CHECK: sqxtn2.8h v0, v1
+  return vqmovn_high_s32(lowpart, wide);
+}
+
+int32x4_t test_vqmovn_high_s64(int32x2_t lowpart, int64x2_t wide) {
+  // CHECK: sqxtn2.4s v0, v1
+  return vqmovn_high_s64(lowpart, wide);
+}
+
+uint8x16_t test_vqmovn_high_u16(uint8x8_t lowpart, int16x8_t wide) {
+  // CHECK: uqxtn2.16b v0, v1
+  return vqmovn_high_u16(lowpart, wide);
+}
+
+uint16x8_t test_vqmovn_high_u32(uint16x4_t lowpart, int32x4_t wide) {
+  // CHECK: uqxtn2.8h v0, v1
+  return vqmovn_high_u32(lowpart, wide);
+}
+
+uint32x4_t test_vqmovn_high_u64(uint32x2_t lowpart, int64x2_t wide) {
+  // CHECK: uqxtn2.4s v0, v1
+  return vqmovn_high_u64(lowpart, wide);
+}
+
+uint8x16_t test_vqmovun_high_s16(uint8x8_t lowpart, int16x8_t wide) {
+  // CHECK: sqxtun2.16b v0, v1
+  return vqmovun_high_s16(lowpart, wide);
+}
+
+uint16x8_t test_vqmovun_high_s32(uint16x4_t lowpart, int32x4_t wide) {
+  // CHECK: sqxtun2.8h v0, v1
+  return vqmovun_high_s32(lowpart, wide);
+}
+
+uint32x4_t test_vqmovun_high_s64(uint32x2_t lowpart, int64x2_t wide) {
+  // CHECK: sqxtun2.4s v0, v1
+  return vqmovun_high_s64(lowpart, wide);
+}
+
+float32x4_t test_vcvtx_high_f32_f64(float32x2_t lowpart, float64x2_t wide) {
+  // CHECK: fcvtxn2 v0.4s, v1.2d
+  return vcvtx_high_f32_f64(lowpart, wide);
+}
+
+float64x2_t test_vcvt_f64_f32(float32x2_t x) {
+  // CHECK: fcvtl v0.2d, v0.2s
+  return vcvt_f64_f32(x);
+}
+
+float64x2_t test_vcvt_high_f64_f32(float32x4_t x) {
+  // CHECK: fcvtl2 v0.2d, v0.4s
+  return vcvt_high_f64_f32(x);
+}
+
+float32x2_t test_vcvt_f32_f64(float64x2_t v) {
+  // CHECK: fcvtn v0.2s, v0.2d
+  return vcvt_f32_f64(v);
+}
+
+float32x4_t test_vcvt_high_f32_f64(float32x2_t x, float64x2_t v) {
+  // CHECK: fcvtn2 v0.4s, v1.2d
+  return vcvt_high_f32_f64(x, v);
+}
+
+float32x2_t test_vcvtx_f32_f64(float64x2_t v) {
+  // CHECK: fcvtxn v0.2s, v0.2d
+  return vcvtx_f32_f64(v);
+}
diff --git a/test/CodeGen/arm64_vCMP.c b/test/CodeGen/arm64_vCMP.c
new file mode 100644
index 0000000..5fb1d79
--- /dev/null
+++ b/test/CodeGen/arm64_vCMP.c
@@ -0,0 +1,108 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+
+// Test ARM64 SIMD fused multiply add intrinsics
+
+#include <arm_neon.h>
+
+int64x2_t test_vabsq_s64(int64x2_t a1) {
+  // CHECK: test_vabsq_s64
+  return vabsq_s64(a1);
+  // CHECK: llvm.arm64.neon.abs.v2i64
+  // CHECK-NEXT: ret
+}
+
+int64_t test_vceqd_s64(int64_t a1, int64_t a2) {
+  // CHECK: test_vceqd_s64
+  return vceqd_s64(a1, a2);
+  // CHECK: [[BIT:%[0-9a-zA-Z.]+]] = icmp eq i64 %a1, %a2
+  // CHECK: sext i1 [[BIT]] to i64
+}
+
+int64_t test_vceqd_f64(float64_t a1, float64_t a2) {
+  // CHECK: test_vceqd_f64
+  return vceqd_f64(a1, a2);
+  // CHECK: [[BIT:%[0-9a-zA-Z.]+]] = fcmp oeq double %a1, %a2
+  // CHECK: sext i1 [[BIT]] to i64
+}
+
+uint64_t test_vcgtd_u64(uint64_t a1, uint64_t a2) {
+  // CHECK: test_vcgtd_u64
+  return vcgtd_u64(a1, a2);
+  // CHECK: [[BIT:%[0-9a-zA-Z.]+]] = icmp ugt i64 %a1, %a2
+  // CHECK: sext i1 [[BIT]] to i64
+}
+
+uint64_t test_vcled_u64(uint64_t a1, uint64_t a2) {
+  // CHECK: test_vcled_u64
+  return vcled_u64(a1, a2);
+  // CHECK: [[BIT:%[0-9a-zA-Z.]+]] = icmp ule i64 %a1, %a2
+  // CHECK: sext i1 [[BIT]] to i64
+}
+
+int64_t test_vceqzd_s64(int64_t a1) {
+  // CHECK: test_vceqzd_s64
+  return vceqzd_s64(a1);
+  // CHECK: [[BIT:%[0-9a-zA-Z.]+]] = icmp eq i64 %a1, 0
+  // CHECK: sext i1 [[BIT]] to i64
+}
+
+uint64x2_t test_vceqq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vceqq_u64
+  return vceqq_u64(a1, a2);
+  // CHECK:  icmp eq <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcgeq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vcgeq_s64
+  return vcgeq_s64(a1, a2);
+  // CHECK:  icmp sge <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcgeq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vcgeq_u64
+  return vcgeq_u64(a1, a2);
+  // CHECK:  icmp uge <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcgtq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vcgtq_s64
+  return vcgtq_s64(a1, a2);
+  // CHECK: icmp sgt <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcgtq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vcgtq_u64
+  return vcgtq_u64(a1, a2);
+  // CHECK: icmp ugt <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcleq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vcleq_s64
+  return vcleq_s64(a1, a2);
+  // CHECK: icmp sle <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcleq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vcleq_u64
+  return vcleq_u64(a1, a2);
+  // CHECK: icmp ule <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcltq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vcltq_s64
+  return vcltq_s64(a1, a2);
+  // CHECK: icmp slt <2 x i64> %a1, %a2
+}
+
+uint64x2_t test_vcltq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vcltq_u64
+  return vcltq_u64(a1, a2);
+  // CHECK: icmp ult <2 x i64> %a1, %a2
+}
+
+int64x2_t test_vqabsq_s64(int64x2_t a1) {
+  // CHECK: test_vqabsq_s64
+  return vqabsq_s64(a1);
+  // CHECK: llvm.arm64.neon.sqabs.v2i64(<2 x i64> %a1)
+  // CHECK-NEXT: ret
+}
diff --git a/test/CodeGen/arm64_vLdStNum_lane.c b/test/CodeGen/arm64_vLdStNum_lane.c
new file mode 100644
index 0000000..e36cd82
--- /dev/null
+++ b/test/CodeGen/arm64_vLdStNum_lane.c
@@ -0,0 +1,141 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD load and stores of an N-element structure  intrinsics
+
+#include <arm_neon.h>
+
+int64x2x2_t test_vld2q_lane_s64(const void * a1, int64x2x2_t a2) {
+  // CHECK: test_vld2q_lane_s64
+  return vld2q_lane_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.ld2lane.v2i64.p0i8
+}
+
+uint64x2x2_t test_vld2q_lane_u64(const void * a1, uint64x2x2_t a2) {
+  // CHECK: test_vld2q_lane_u64
+  return vld2q_lane_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.ld2lane.v2i64.p0i8
+}
+
+int64x1x2_t test_vld2_lane_s64(const void * a1, int64x1x2_t a2) {
+  // CHECK: test_vld2_lane_s64
+  return vld2_lane_s64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld2lane.v1i64.p0i8
+}
+
+uint64x1x2_t test_vld2_lane_u64(const void * a1, uint64x1x2_t a2) {
+  // CHECK: test_vld2_lane_u64
+  return vld2_lane_u64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld2lane.v1i64.p0i8
+}
+
+poly8x16x2_t test_vld2q_lane_p8(const void * a1, poly8x16x2_t a2) {
+  // CHECK: test_vld2q_lane_p8
+  return vld2q_lane_p8(a1, a2, 0);
+  // CHECK: extractvalue {{.*}} 0{{ *$}}
+  // CHECK: extractvalue {{.*}} 1{{ *$}}
+}
+
+uint8x16x2_t test_vld2q_lane_u8(const void * a1, uint8x16x2_t a2) {
+  // CHECK: test_vld2q_lane_u8
+  return vld2q_lane_u8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld2lane.v16i8.p0i8
+}
+
+int64x2x3_t test_vld3q_lane_s64(const void * a1, int64x2x3_t a2) {
+  // CHECK: test_vld3q_lane_s64
+  return vld3q_lane_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.ld3lane.v2i64.p0i8
+}
+
+uint64x2x3_t test_vld3q_lane_u64(const void * a1, uint64x2x3_t a2) {
+  // CHECK: test_vld3q_lane_u64
+  return vld3q_lane_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.ld3lane.v2i64.p0i8
+}
+
+int64x1x3_t test_vld3_lane_s64(const void * a1, int64x1x3_t a2) {
+  // CHECK: test_vld3_lane_s64
+  return vld3_lane_s64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld3lane.v1i64.p0i8
+}
+
+uint64x1x3_t test_vld3_lane_u64(const void * a1, uint64x1x3_t a2) {
+  // CHECK: test_vld3_lane_u64
+  return vld3_lane_u64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld3lane.v1i64.p0i8
+}
+
+int8x8x3_t test_vld3_lane_s8(const void * a1, int8x8x3_t a2) {
+  // CHECK: test_vld3_lane_s8
+  return vld3_lane_s8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld3lane.v8i8.p0i8
+}
+
+poly8x16x3_t test_vld3q_lane_p8(const void * a1, poly8x16x3_t a2) {
+  // CHECK: test_vld3q_lane_p8
+  return vld3q_lane_p8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld3lane.v16i8.p0i8
+}
+
+uint8x16x3_t test_vld3q_lane_u8(const void * a1, uint8x16x3_t a2) {
+  // CHECK: test_vld3q_lane_u8
+  return vld3q_lane_u8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld3lane.v16i8.p0i8
+}
+
+int64x2x4_t test_vld4q_lane_s64(const void * a1, int64x2x4_t a2) {
+  // CHECK: test_vld4q_lane_s64
+  return vld4q_lane_s64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v2i64.p0i8
+}
+
+uint64x2x4_t test_vld4q_lane_u64(const void * a1, uint64x2x4_t a2) {
+  // CHECK: test_vld4q_lane_u64
+  return vld4q_lane_u64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v2i64.p0i8
+}
+
+int64x1x4_t test_vld4_lane_s64(const void * a1, int64x1x4_t a2) {
+  // CHECK: test_vld4_lane_s64
+  return vld4_lane_s64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v1i64.p0i8
+}
+
+uint64x1x4_t test_vld4_lane_u64(const void * a1, uint64x1x4_t a2) {
+  // CHECK: test_vld4_lane_u64
+  return vld4_lane_u64(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v1i64.p0i8
+}
+
+int8x8x4_t test_vld4_lane_s8(const void * a1, int8x8x4_t a2) {
+  // CHECK: test_vld4_lane_s8
+  return vld4_lane_s8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v8i8.p0i8
+}
+
+uint8x8x4_t test_vld4_lane_u8(const void * a1, uint8x8x4_t a2) {
+  // CHECK: test_vld4_lane_u8
+  return vld4_lane_u8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v8i8.p0i8
+}
+
+poly8x16x4_t test_vld4q_lane_p8(const void * a1, poly8x16x4_t a2) {
+  // CHECK: test_vld4q_lane_p8
+  return vld4q_lane_p8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v16i8.p0i8
+}
+
+int8x16x4_t test_vld4q_lane_s8(const void * a1, int8x16x4_t a2) {
+  // CHECK: test_vld4q_lane_s8
+  return vld4q_lane_s8(a1, a2, 0);
+  // CHECK: extractvalue {{.*}} 0{{ *$}}
+  // CHECK: extractvalue {{.*}} 1{{ *$}}
+  // CHECK: extractvalue {{.*}} 2{{ *$}}
+  // CHECK: extractvalue {{.*}} 3{{ *$}}
+}
+
+uint8x16x4_t test_vld4q_lane_u8(const void * a1, uint8x16x4_t a2) {
+  // CHECK: test_vld4q_lane_u8
+  return vld4q_lane_u8(a1, a2, 0);
+  // CHECK: llvm.arm64.neon.ld4lane.v16i8.p0i8
+}
+
diff --git a/test/CodeGen/arm64_vMaxMin.c b/test/CodeGen/arm64_vMaxMin.c
new file mode 100644
index 0000000..d53c0bb
--- /dev/null
+++ b/test/CodeGen/arm64_vMaxMin.c
@@ -0,0 +1,207 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | FileCheck -check-prefix=CHECK-CODEGEN %s
+// REQUIRES: arm64-registered-target
+// Test ARM64 SIMD max/min intrinsics
+
+#include <arm_neon.h>
+
+// Test a represntative sample of 8 and 16, signed and unsigned, 64 and 128 bit reduction
+int8_t test_vmaxv_s8(int8x8_t a1) {
+  // CHECK: test_vmaxv_s8
+  return vmaxv_s8(a1);
+  // CHECK @llvm.arm64.neon.smaxv.i32.v8i8
+}
+
+uint16_t test_vminvq_u16(uint16x8_t a1) {
+  // CHECK: test_vminvq_u16
+  return vminvq_u16(a1);
+  // CHECK llvm.arm64.neon.uminv.i16.v8i16
+}
+
+// Test a represntative sample of 8 and 16, signed and unsigned, 64 and 128 bit pairwise
+uint8x8_t test_vmin_u8(uint8x8_t a1, uint8x8_t a2) {
+  // CHECK: test_vmin_u8
+  return vmin_u8(a1, a2);
+  // CHECK llvm.arm64.neon.umin.v8i8
+}
+
+uint8x16_t test_vminq_u8(uint8x16_t a1, uint8x16_t a2) {
+  // CHECK: test_vminq_u8
+  return vminq_u8(a1, a2);
+  // CHECK llvm.arm64.neon.umin.v16i8
+}
+
+int16x8_t test_vmaxq_s16(int16x8_t a1, int16x8_t a2) {
+  // CHECK: test_vmaxq_s16
+  return vmaxq_s16(a1, a2);
+  // CHECK llvm.arm64.neon.smax.v8i16
+}
+
+// Test the more complicated cases of [suf]32 and f64
+float64x2_t test_vmaxq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vmaxq_f64
+  return vmaxq_f64(a1, a2);
+  // CHECK llvm.arm64.neon.fmax.v2f64
+}
+
+float32x4_t test_vmaxq_f32(float32x4_t a1, float32x4_t a2) {
+  // CHECK: test_vmaxq_f32
+  return vmaxq_f32(a1, a2);
+  // CHECK llvm.arm64.neon.fmax.v4f32
+}
+
+float64x2_t test_vminq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vminq_f64
+  return vminq_f64(a1, a2);
+  // CHECK llvm.arm64.neon.fmin.v2f64
+}
+
+float32x2_t test_vmax_f32(float32x2_t a1, float32x2_t a2) {
+  // CHECK: test_vmax_f32
+  return vmax_f32(a1, a2);
+  // CHECK llvm.arm64.neon.fmax.v2f32
+}
+
+int32x2_t test_vmax_s32(int32x2_t a1, int32x2_t a2) {
+  // CHECK: test_vmax_s32
+  return vmax_s32(a1, a2);
+  // CHECK llvm.arm64.neon.smax.v2i32
+}
+
+uint32x2_t test_vmin_u32(uint32x2_t a1, uint32x2_t a2) {
+  // CHECK: test_vmin_u32
+  return vmin_u32(a1, a2);
+  // CHECK llvm.arm64.neon.umin.v2i32
+}
+
+float32_t test_vmaxnmv_f32(float32x2_t a1) {
+  // CHECK: test_vmaxnmv_f32
+  return vmaxnmv_f32(a1);
+  // CHECK: llvm.arm64.neon.fmaxnmv.f32.v2f32
+  // CHECK-NEXT: ret
+}
+
+// this doesn't translate into a valid instruction, regardless of what the
+// ARM doc says.
+#if 0
+float64_t test_vmaxnmvq_f64(float64x2_t a1) {
+  // CHECK@ test_vmaxnmvq_f64
+  return vmaxnmvq_f64(a1);
+  // CHECK@ llvm.arm64.neon.saddlv.i64.v2i32
+  // CHECK-NEXT@ ret
+}
+#endif
+
+float32_t test_vmaxnmvq_f32(float32x4_t a1) {
+  // CHECK: test_vmaxnmvq_f32
+  return vmaxnmvq_f32(a1);
+  // CHECK: llvm.arm64.neon.fmaxnmv.f32.v4f32
+  // CHECK-NEXT: ret
+}
+
+float32_t test_vmaxv_f32(float32x2_t a1) {
+  // CHECK: test_vmaxv_f32
+  return vmaxv_f32(a1);
+  // CHECK: llvm.arm64.neon.fmaxv.f32.v2f32
+  // FIXME check that the 2nd and 3rd arguments are the same V register below
+  // CHECK-CODEGEN: fmaxp.2s
+  // CHECK-NEXT: ret
+}
+
+int32_t test_vmaxv_s32(int32x2_t a1) {
+  // CHECK: test_vmaxv_s32
+  return vmaxv_s32(a1);
+  // CHECK: llvm.arm64.neon.smaxv.i32.v2i32
+  // FIXME check that the 2nd and 3rd arguments are the same V register below
+  // CHECK-CODEGEN: smaxp.2s
+  // CHECK-NEXT: ret
+}
+
+uint32_t test_vmaxv_u32(uint32x2_t a1) {
+  // CHECK: test_vmaxv_u32
+  return vmaxv_u32(a1);
+  // CHECK: llvm.arm64.neon.umaxv.i32.v2i32
+  // FIXME check that the 2nd and 3rd arguments are the same V register below
+  // CHECK-CODEGEN: umaxp.2s
+  // CHECK-NEXT: ret
+}
+
+// FIXME punt on this for now; don't forget to fix CHECKs
+#if 0
+float64_t test_vmaxvq_f64(float64x2_t a1) {
+  // CHECK@ test_vmaxvq_f64
+  return vmaxvq_f64(a1);
+  // CHECK@ llvm.arm64.neon.fmaxv.i64.v2f64
+  // CHECK-NEXT@ ret
+}
+#endif
+
+float32_t test_vmaxvq_f32(float32x4_t a1) {
+  // CHECK: test_vmaxvq_f32
+  return vmaxvq_f32(a1);
+  // CHECK: llvm.arm64.neon.fmaxv.f32.v4f32
+  // CHECK-NEXT: ret
+}
+
+float32_t test_vminnmv_f32(float32x2_t a1) {
+  // CHECK: test_vminnmv_f32
+  return vminnmv_f32(a1);
+  // CHECK: llvm.arm64.neon.fminnmv.f32.v2f32
+  // CHECK-NEXT: ret
+}
+
+float32_t test_vminvq_f32(float32x4_t a1) {
+  // CHECK: test_vminvq_f32
+  return vminvq_f32(a1);
+  // CHECK: llvm.arm64.neon.fminv.f32.v4f32
+  // CHECK-NEXT: ret
+}
+
+// this doesn't translate into a valid instruction, regardless of what the ARM
+// doc says.
+#if 0
+float64_t test_vminnmvq_f64(float64x2_t a1) {
+  // CHECK@ test_vminnmvq_f64
+  return vminnmvq_f64(a1);
+  // CHECK@ llvm.arm64.neon.saddlv.i64.v2i32
+  // CHECK-NEXT@ ret
+}
+#endif
+
+float32_t test_vminnmvq_f32(float32x4_t a1) {
+  // CHECK: test_vminnmvq_f32
+  return vminnmvq_f32(a1);
+  // CHECK: llvm.arm64.neon.fminnmv.f32.v4f32
+  // CHECK-NEXT: ret
+}
+
+float32_t test_vminv_f32(float32x2_t a1) {
+  // CHECK: test_vminv_f32
+  return vminv_f32(a1);
+  // CHECK: llvm.arm64.neon.fminv.f32.v2f32
+  // CHECK-NEXT: ret
+}
+
+int32_t test_vminv_s32(int32x2_t a1) {
+  // CHECK: test_vminv_s32
+  return vminv_s32(a1);
+  // CHECK: llvm.arm64.neon.sminv.i32.v2i32
+  // CHECK-CODEGEN: sminp.2s
+  // CHECK-NEXT: ret
+}
+
+uint32_t test_vminv_u32(uint32x2_t a1) {
+  // CHECK: test_vminv_u32
+  return vminv_u32(a1);
+  // CHECK: llvm.arm64.neon.fminv.f32.v2f32
+}
+
+// FIXME punt on this for now; don't forget to fix CHECKs
+#if 0
+float64_t test_vminvq_f64(float64x2_t a1) {
+  // CHECK@ test_vminvq_f64
+  return vminvq_f64(a1);
+  // CHECK@ llvm.arm64.neon.saddlv.i64.v2i32
+  // CHECK-NEXT@ ret
+}
+#endif
diff --git a/test/CodeGen/arm64_vadd.c b/test/CodeGen/arm64_vadd.c
new file mode 100644
index 0000000..83fb3e7
--- /dev/null
+++ b/test/CodeGen/arm64_vadd.c
@@ -0,0 +1,102 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD add intrinsics
+
+#include <arm_neon.h>
+int64_t test_vaddlv_s32(int32x2_t a1) {
+  // CHECK: test_vaddlv_s32
+  return vaddlv_s32(a1);
+  // CHECK: llvm.arm64.neon.saddlv.i64.v2i32
+  // CHECK-NEXT: ret
+}
+
+uint64_t test_vaddlv_u32(uint32x2_t a1) {
+  // CHECK: test_vaddlv_u32
+  return vaddlv_u32(a1);
+  // CHECK: llvm.arm64.neon.uaddlv.i64.v2i32
+  // CHECK-NEXT: ret
+}
+
+int8_t test_vaddv_s8(int8x8_t a1) {
+  // CHECK: test_vaddv_s8
+  return vaddv_s8(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v8i8
+  // don't check for return here (there's a trunc?)
+}
+
+int16_t test_vaddv_s16(int16x4_t a1) {
+  // CHECK: test_vaddv_s16
+  return vaddv_s16(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v4i16
+  // don't check for return here (there's a trunc?)
+}
+
+int32_t test_vaddv_s32(int32x2_t a1) {
+  // CHECK: test_vaddv_s32
+  return vaddv_s32(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v2i32
+  // CHECK-NEXT: ret
+}
+
+uint8_t test_vaddv_u8(int8x8_t a1) {
+  // CHECK: test_vaddv_u8
+  return vaddv_u8(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v8i8
+  // don't check for return here (there's a trunc?)
+}
+
+uint16_t test_vaddv_u16(int16x4_t a1) {
+  // CHECK: test_vaddv_u16
+  return vaddv_u16(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v4i16
+  // don't check for return here (there's a trunc?)
+}
+
+uint32_t test_vaddv_u32(int32x2_t a1) {
+  // CHECK: test_vaddv_u32
+  return vaddv_u32(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v2i32
+  // CHECK-NEXT: ret
+}
+
+int8_t test_vaddvq_s8(int8x16_t a1) {
+  // CHECK: test_vaddvq_s8
+  return vaddvq_s8(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v16i8
+  // don't check for return here (there's a trunc?)
+}
+
+int16_t test_vaddvq_s16(int16x8_t a1) {
+  // CHECK: test_vaddvq_s16
+  return vaddvq_s16(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v8i16
+  // don't check for return here (there's a trunc?)
+}
+
+int32_t test_vaddvq_s32(int32x4_t a1) {
+  // CHECK: test_vaddvq_s32
+  return vaddvq_s32(a1);
+  // CHECK: llvm.arm64.neon.saddv.i32.v4i32
+  // CHECK-NEXT: ret
+}
+
+uint8_t test_vaddvq_u8(int8x16_t a1) {
+  // CHECK: test_vaddvq_u8
+  return vaddvq_u8(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v16i8
+  // don't check for return here (there's a trunc?)
+}
+
+uint16_t test_vaddvq_u16(int16x8_t a1) {
+  // CHECK: test_vaddvq_u16
+  return vaddvq_u16(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v8i16
+  // don't check for return here (there's a trunc?)
+}
+
+uint32_t test_vaddvq_u32(int32x4_t a1) {
+  // CHECK: test_vaddvq_u32
+  return vaddvq_u32(a1);
+  // CHECK: llvm.arm64.neon.uaddv.i32.v4i32
+  // CHECK-NEXT: ret
+}
+
diff --git a/test/CodeGen/arm64_vca.c b/test/CodeGen/arm64_vca.c
new file mode 100644
index 0000000..0acbe45
--- /dev/null
+++ b/test/CodeGen/arm64_vca.c
@@ -0,0 +1,59 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 vector compare absolute intrinsics
+
+#include <arm_neon.h>
+
+uint32x2_t test_vcale_f32(float32x2_t a1, float32x2_t a2) {
+  // CHECK: test_vcale_f32
+  return vcale_f32(a1, a2);
+  // CHECK: llvm.arm64.neon.facge.v2i32.v2f32
+  // no check for ret here, as there is a bitcast
+}
+
+uint32x4_t test_vcaleq_f32(float32x4_t a1, float32x4_t a2) {
+  // CHECK: test_vcaleq_f32
+  return vcaleq_f32(a1, a2);
+  // CHECK: llvm.arm64.neon.facge.v4i32.v4f32{{.*a2,.*a1}}
+  // no check for ret here, as there is a bitcast
+}
+
+uint32x2_t test_vcalt_f32(float32x2_t a1, float32x2_t a2) {
+  // CHECK: test_vcalt_f32
+  return vcalt_f32(a1, a2);
+  // CHECK: llvm.arm64.neon.facgt.v2i32.v2f32{{.*a2,.*a1}}
+  // no check for ret here, as there is a bitcast
+}
+
+uint32x4_t test_vcaltq_f32(float32x4_t a1, float32x4_t a2) {
+  // CHECK: test_vcaltq_f32
+  return vcaltq_f32(a1, a2);
+  // CHECK: llvm.arm64.neon.facgt.v4i32.v4f32{{.*a2,.*a1}}
+}
+
+uint64x2_t test_vcagtq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vcagtq_f64
+  return vcagtq_f64(a1, a2);
+  // CHECK: llvm.arm64.neon.facgt.v2i64.v2f64{{.*a1,.*a2}}
+  // no check for ret here, as there is a bitcast
+}
+
+uint64x2_t test_vcaltq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vcaltq_f64
+  return vcaltq_f64(a1, a2);
+  // CHECK: llvm.arm64.neon.facgt.v2i64.v2f64{{.*a2,.*a1}}
+  // no check for ret here, as there is a bitcast
+}
+
+uint64x2_t test_vcageq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vcageq_f64
+  return vcageq_f64(a1, a2);
+  // CHECK: llvm.arm64.neon.facge.v2i64.v2f64{{.*a1,.*a2}}
+  // no check for ret here, as there is a bitcast
+}
+
+uint64x2_t test_vcaleq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK: test_vcaleq_f64
+  return vcaleq_f64(a1, a2);
+  // CHECK: llvm.arm64.neon.facge.v2i64.v2f64{{.*a2,.*a1}}
+  // no check for ret here, as there is a bitcast
+}
diff --git a/test/CodeGen/arm64_vcopy.c b/test/CodeGen/arm64_vcopy.c
new file mode 100644
index 0000000..7283909
--- /dev/null
+++ b/test/CodeGen/arm64_vcopy.c
@@ -0,0 +1,69 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+
+// Test ARM64 SIMD copy vector element to vector element: vcopyq_lane*
+
+#include <arm_neon.h>
+
+int8x16_t test_vcopyq_laneq_s8(int8x16_t a1, int8x16_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_s8
+  return vcopyq_laneq_s8(a1, (int64_t) 3, a2, (int64_t) 13);
+  // CHECK: shufflevector <16 x i8> %a1, <16 x i8> %a2, <16 x i32> <i32 0, i32 1, i32 2, i32 29, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+}
+
+uint8x16_t test_vcopyq_laneq_u8(uint8x16_t a1, uint8x16_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_u8
+  return vcopyq_laneq_u8(a1, (int64_t) 3, a2, (int64_t) 13);
+  // CHECK: shufflevector <16 x i8> %a1, <16 x i8> %a2, <16 x i32> <i32 0, i32 1, i32 2, i32 29, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+
+}
+
+int16x8_t test_vcopyq_laneq_s16(int16x8_t a1, int16x8_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_s16
+  return vcopyq_laneq_s16(a1, (int64_t) 3, a2, (int64_t) 7);
+  // CHECK: shufflevector <8 x i16> %a1, <8 x i16> %a2, <8 x i32> <i32 0, i32 1, i32 2, i32 15, i32 4, i32 5, i32 6, i32 7>
+
+}
+
+uint16x8_t test_vcopyq_laneq_u16(uint16x8_t a1, uint16x8_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_u16
+  return vcopyq_laneq_u16(a1, (int64_t) 3, a2, (int64_t) 7);
+  // CHECK: shufflevector <8 x i16> %a1, <8 x i16> %a2, <8 x i32> <i32 0, i32 1, i32 2, i32 15, i32 4, i32 5, i32 6, i32 7>
+
+}
+
+int32x4_t test_vcopyq_laneq_s32(int32x4_t a1, int32x4_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_s32
+  return vcopyq_laneq_s32(a1, (int64_t) 3, a2, (int64_t) 3);
+  // CHECK: shufflevector <4 x i32> %a1, <4 x i32> %a2, <4 x i32> <i32 0, i32 1, i32 2, i32 7>
+}
+
+uint32x4_t test_vcopyq_laneq_u32(uint32x4_t a1, uint32x4_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_u32
+  return vcopyq_laneq_u32(a1, (int64_t) 3, a2, (int64_t) 3);
+  // CHECK: shufflevector <4 x i32> %a1, <4 x i32> %a2, <4 x i32> <i32 0, i32 1, i32 2, i32 7>
+}
+
+int64x2_t test_vcopyq_laneq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_s64
+  return vcopyq_laneq_s64(a1, (int64_t) 0, a2, (int64_t) 1);
+  // CHECK: shufflevector <2 x i64> %a1, <2 x i64> %a2, <2 x i32> <i32 3, i32 1>
+}
+
+uint64x2_t test_vcopyq_laneq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_u64
+  return vcopyq_laneq_u64(a1, (int64_t) 0, a2, (int64_t) 1);
+  // CHECK: shufflevector <2 x i64> %a1, <2 x i64> %a2, <2 x i32> <i32 3, i32 1>
+}
+
+float32x4_t test_vcopyq_laneq_f32(float32x4_t a1, float32x4_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_f32
+  return vcopyq_laneq_f32(a1, 0, a2, 3);
+  // CHECK: shufflevector <4 x float> %a1, <4 x float> %a2, <4 x i32> <i32 7, i32 1, i32 2, i32 3>
+}
+
+float64x2_t test_vcopyq_laneq_f64(float64x2_t a1, float64x2_t a2) {
+  // CHECK-LABEL: test_vcopyq_laneq_f64
+  return vcopyq_laneq_f64(a1, 0, a2, 1);
+  // CHECK: shufflevector <2 x double> %a1, <2 x double> %a2, <2 x i32> <i32 3, i32 1>
+}
+
diff --git a/test/CodeGen/arm64_vcreate.c b/test/CodeGen/arm64_vcreate.c
new file mode 100644
index 0000000..ae1e431
--- /dev/null
+++ b/test/CodeGen/arm64_vcreate.c
@@ -0,0 +1,23 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD vcreate intrinsics
+
+/*#include <arm_neon.h>*/
+#include <arm_neon.h>
+
+float32x2_t test_vcreate_f32(uint64_t a1) {
+  // CHECK: test_vcreate_f32
+  return vcreate_f32(a1);
+  // CHECK: bitcast {{.*}} to <2 x float>
+  // CHECK-NEXT: ret
+}
+
+// FIXME enable when scalar_to_vector in backend is fixed.  Also, change
+// CHECK@ to CHECK<colon> and CHECK-NEXT@ to CHECK-NEXT<colon>
+/*
+float64x1_t test_vcreate_f64(uint64_t a1) {
+  // CHECK@ test_vcreate_f64
+  return vcreate_f64(a1);
+  // CHECK@ llvm.arm64.neon.saddlv.i64.v2i32
+  // CHECK-NEXT@ ret
+}
+*/
diff --git a/test/CodeGen/arm64_vcvtfp.c b/test/CodeGen/arm64_vcvtfp.c
new file mode 100644
index 0000000..a98a175
--- /dev/null
+++ b/test/CodeGen/arm64_vcvtfp.c
@@ -0,0 +1,48 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+
+#include <arm_neon.h>
+
+float64x2_t test_vcvt_f64_f32(float32x2_t x) {
+  // CHECK-LABEL: test_vcvt_f64_f32
+  return vcvt_f64_f32(x);
+  // CHECK: fpext <2 x float> {{%.*}} to <2 x double>
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vcvt_high_f64_f32(float32x4_t x) {
+  // CHECK-LABEL: test_vcvt_high_f64_f32
+  return vcvt_high_f64_f32(x);
+  // CHECK: [[HIGH:%.*]] = shufflevector <4 x float> {{%.*}}, <4 x float> undef, <2 x i32> <i32 2, i32 3>
+  // CHECK-NEXT: fpext <2 x float> [[HIGH]] to <2 x double>
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vcvt_f32_f64(float64x2_t v) {
+  // CHECK: test_vcvt_f32_f64
+  return vcvt_f32_f64(v);
+  // CHECK: fptrunc <2 x double> {{%.*}} to <2 x float>
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vcvt_high_f32_f64(float32x2_t x, float64x2_t v) {
+  // CHECK: test_vcvt_high_f32_f64
+  return vcvt_high_f32_f64(x, v);
+  // CHECK: [[TRUNC:%.*]] = fptrunc <2 x double> {{.*}} to <2 x float>
+  // CHECK-NEXT: shufflevector <2 x float> {{.*}}, <2 x float> [[TRUNC]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vcvtx_f32_f64(float64x2_t v) {
+  // CHECK: test_vcvtx_f32_f64
+  return vcvtx_f32_f64(v);
+  // CHECK: llvm.arm64.neon.fcvtxn.v2f32.v2f64
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vcvtx_high_f32_f64(float32x2_t x, float64x2_t v) {
+  // CHECK: test_vcvtx_high_f32_f64
+  return vcvtx_high_f32_f64(x, v);
+  // CHECK: llvm.arm64.neon.fcvtxn.v2f32.v2f64
+  // CHECK: shufflevector
+  // CHECK-NEXT: ret
+}
diff --git a/test/CodeGen/arm64_vdup.c b/test/CodeGen/arm64_vdup.c
new file mode 100644
index 0000000..8476af2
--- /dev/null
+++ b/test/CodeGen/arm64_vdup.c
@@ -0,0 +1,42 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD duplicate lane and n intrinsics
+
+#include <arm_neon.h>
+
+void test_vdup_lane_s64(int64x1_t a1) {
+  // CHECK-LABEL: test_vdup_lane_s64
+  vdup_lane_s64(a1, 0);
+  // CHECK: shufflevector
+}
+
+void test_vdup_lane_u64(uint64x1_t a1) {
+  // CHECK-LABEL: test_vdup_lane_u64
+  vdup_lane_u64(a1, 0);
+  // CHECK: shufflevector
+}
+
+// uncomment out the following code once scalar_to_vector in the backend
+// works (for 64 bit?).  Change the "CHECK@" to "CHECK<colon>"
+/*
+float64x1_t test_vdup_n_f64(float64_t a1) {
+  // CHECK-LABEL@ test_vdup_n_f64
+  return vdup_n_f64(a1);
+  // match that an element is inserted into part 0
+  // CHECK@ insertelement {{.*, i32 0 *$}}
+}
+*/
+
+float16x8_t test_vdupq_n_f16(float16_t *a1) {
+  // CHECK-LABEL: test_vdupq_n_f16
+  return vdupq_n_f16(*a1);
+  // match that an element is inserted into parts 0-7.  The backend better
+  // turn that into a single dup intruction
+  // CHECK: insertelement {{.*, i32 0 *$}}
+  // CHECK: insertelement {{.*, i32 1 *$}}
+  // CHECK: insertelement {{.*, i32 2 *$}}
+  // CHECK: insertelement {{.*, i32 3 *$}}
+  // CHECK: insertelement {{.*, i32 4 *$}}
+  // CHECK: insertelement {{.*, i32 5 *$}}
+  // CHECK: insertelement {{.*, i32 6 *$}}
+  // CHECK: insertelement {{.*, i32 7 *$}}
+}
diff --git a/test/CodeGen/arm64_vdupq_n_f64.c b/test/CodeGen/arm64_vdupq_n_f64.c
new file mode 100644
index 0000000..5e8834f
--- /dev/null
+++ b/test/CodeGen/arm64_vdupq_n_f64.c
@@ -0,0 +1,88 @@
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | FileCheck %s
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | \
+// RUN:   FileCheck -check-prefix=CHECK-IR %s
+// REQUIRES: arm64-registered-target
+
+/// Test vdupq_n_f64 and vmovq_nf64 ARM64 intrinsics
+// <rdar://problem/11778405> ARM64: vdupq_n_f64 and vdupq_lane_f64 intrinsics
+// missing
+
+
+#include <arm_neon.h>
+
+// vdupq_n_f64 -> dup.2d v0, v0[0]
+//
+float64x2_t test_vdupq_n_f64(float64_t w)
+{
+    return vdupq_n_f64(w);
+  // CHECK-LABEL: test_vdupq_n_f64:
+  // CHECK: dup.2d v0, v0[0]
+  // CHECK-NEXT: ret
+}
+
+// might as well test this while we're here
+// vdupq_n_f32 -> dup.4s v0, v0[0]
+float32x4_t test_vdupq_n_f32(float32_t w)
+{
+    return vdupq_n_f32(w);
+  // CHECK-LABEL: test_vdupq_n_f32:
+  // CHECK: dup.4s v0, v0[0]
+  // CHECK-NEXT: ret
+}
+
+// vdupq_lane_f64 -> dup.2d v0, v0[0]
+// this was in <rdar://problem/11778405>, but had already been implemented,
+// test anyway
+float64x2_t test_vdupq_lane_f64(float64x1_t V)
+{
+    return vdupq_lane_f64(V, 0);
+  // CHECK-LABEL: test_vdupq_lane_f64:
+  // CHECK: dup.2d v0, v0[0]
+  // CHECK-NEXT: ret
+}
+
+// vmovq_n_f64 -> dup Vd.2d,X0
+// this wasn't in <rdar://problem/11778405>, but it was between the vdups
+float64x2_t test_vmovq_n_f64(float64_t w)
+{
+  return vmovq_n_f64(w);
+  // CHECK-LABEL: test_vmovq_n_f64:
+  // CHECK: dup.2d v0, v0[0]
+  // CHECK-NEXT: ret
+}
+
+float16x4_t test_vmov_n_f16(float16_t *a1)
+{
+  // CHECK-IR-LABEL: test_vmov_n_f16
+  return vmov_n_f16(*a1);
+  // CHECK-IR: insertelement {{.*}} i32 0{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 1{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 2{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 3{{ *$}}
+}
+
+// Disable until scalar problem in backend is fixed. Change CHECK-IR@ to
+// CHECK-IR<colon>
+/*
+float64x1_t test_vmov_n_f64(float64_t a1)
+{
+  // CHECK-IR@ test_vmov_n_f64
+  return vmov_n_f64(a1);
+  // CHECK-IR@ insertelement {{.*}} i32 0{{ *$}}
+}
+*/
+
+float16x8_t test_vmovq_n_f16(float16_t *a1)
+{
+  // CHECK-IR-LABEL: test_vmovq_n_f16
+  return vmovq_n_f16(*a1);
+  // CHECK-IR: insertelement {{.*}} i32 0{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 1{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 2{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 3{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 4{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 5{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 6{{ *$}}
+  // CHECK-IR: insertelement {{.*}} i32 7{{ *$}}
+}
+
diff --git a/test/CodeGen/arm64_vecCmpBr.c b/test/CodeGen/arm64_vecCmpBr.c
new file mode 100644
index 0000000..ba67496
--- /dev/null
+++ b/test/CodeGen/arm64_vecCmpBr.c
@@ -0,0 +1,111 @@
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -S -ffreestanding %s -o - | FileCheck %s
+// REQUIRES: arm64-registered-target
+// test code generation for <rdar://problem/11487757>
+#include <arm_neon.h>
+
+unsigned bar();
+
+// Branch if any lane of V0 is zero; 64 bit => !min
+unsigned anyZero64(uint16x4_t a) {
+// CHECK: anyZero64:
+// CHECK: uminv.8b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: b {{_bar|bar}}
+  if (!vminv_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if any lane of V0 is zero; 128 bit => !min
+unsigned anyZero128(uint16x8_t a) {
+// CHECK: anyZero128:
+// CHECK: uminv.16b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: b {{_bar|bar}}
+  if (!vminvq_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if any lane of V0 is non-zero; 64 bit => max
+unsigned anyNonZero64(uint16x4_t a) {
+// CHECK: anyNonZero64:
+// CHECK: umaxv.8b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: movz w0, #0
+  if (vmaxv_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if any lane of V0 is non-zero; 128 bit => max
+unsigned anyNonZero128(uint16x8_t a) {
+// CHECK: anyNonZero128:
+// CHECK: umaxv.16b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: movz w0, #0
+  if (vmaxvq_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if all lanes of V0 are zero; 64 bit => !max
+unsigned allZero64(uint16x4_t a) {
+// CHECK: allZero64:
+// CHECK: umaxv.8b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: b {{_bar|bar}}
+  if (!vmaxv_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if all lanes of V0 are zero; 128 bit => !max
+unsigned allZero128(uint16x8_t a) {
+// CHECK: allZero128:
+// CHECK: umaxv.16b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: b {{_bar|bar}}
+  if (!vmaxvq_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if all lanes of V0 are non-zero; 64 bit => min
+unsigned allNonZero64(uint16x4_t a) {
+// CHECK: allNonZero64:
+// CHECK: uminv.8b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: movz w0, #0
+  if (vminv_u8(a))
+    return bar();
+  return 0;
+}
+
+// Branch if all lanes of V0 are non-zero; 128 bit => min
+unsigned allNonZero128(uint16x8_t a) {
+// CHECK: allNonZero128:
+// CHECK: uminv.16b b[[REGNO1:[0-9]+]], v0
+// CHECK-NEXT: fmov w[[REGNO2:[0-9]+]], s[[REGNO1]]
+// CHECK-NEXT: cbz w[[REGNO2]], [[LABEL:[.A-Z_0-9]+]]
+// CHECK: [[LABEL]]:
+// CHECK-NEXT: movz w0, #0
+  if (vminvq_u8(a))
+    return bar();
+  return 0;
+}
+
diff --git a/test/CodeGen/arm64_vext.c b/test/CodeGen/arm64_vext.c
new file mode 100644
index 0000000..50ff48c
--- /dev/null
+++ b/test/CodeGen/arm64_vext.c
@@ -0,0 +1,239 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+
+// Test ARM64 extract intrinsics
+// can use as back end test by adding a run line with
+// -check-prefix=CHECK-CODEGEN on the FileCheck
+
+#include <arm_neon.h>
+
+void test_vext_s8()
+{
+  // CHECK: test_vext_s8
+  int8x8_t xS8x8;
+  xS8x8 = vext_s8(xS8x8, xS8x8, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_s8:
+  // CHECK-CODEGEN: {{ext.8.*#1}}
+}
+
+void test_vext_u8()
+{
+  // CHECK: test_vext_u8
+  uint8x8_t xU8x8;
+  xU8x8 = vext_u8(xU8x8, xU8x8, 2);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_u8:
+  // CHECK-CODEGEN: {{ext.8.*#2}}
+}
+
+void test_vext_p8()
+{
+  // CHECK: test_vext_p8
+  poly8x8_t xP8x8;
+  xP8x8 = vext_p8(xP8x8, xP8x8, 3);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_p8:
+  // CHECK-CODEGEN: {{ext.8.*#3}}
+}
+
+void test_vext_s16()
+{
+  // CHECK: test_vext_s16
+  int16x4_t xS16x4;
+  xS16x4 = vext_s16(xS16x4, xS16x4, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_s16:
+  // CHECK-CODEGEN: {{ext.8.*#2}}
+}
+
+void test_vext_u16()
+{
+  // CHECK: test_vext_u16
+  uint16x4_t xU16x4;
+  xU16x4 = vext_u16(xU16x4, xU16x4, 2);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_u16:
+  // CHECK-CODEGEN: {{ext.8.*#4}}
+}
+
+void test_vext_p16()
+{
+  // CHECK: test_vext_p16
+  poly16x4_t xP16x4;
+  xP16x4 = vext_p16(xP16x4, xP16x4, 3);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_p16:
+  // CHECK-CODEGEN: {{ext.8.*#6}}
+}
+
+void test_vext_s32()
+{
+  // CHECK: test_vext_s32
+  int32x2_t xS32x2;
+  xS32x2 = vext_s32(xS32x2, xS32x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_s32:
+  // CHECK-CODEGEN: {{ext.8.*#4}}
+}
+
+void test_vext_u32()
+{
+  // CHECK: test_vext_u32
+  uint32x2_t xU32x2;
+  xU32x2 = vext_u32(xU32x2, xU32x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_u32:
+  // CHECK-CODEGEN: {{ext.8.*#4}}
+}
+
+void test_vext_f32()
+{
+  // CHECK: test_vext_f32
+  float32x2_t xF32x2;
+  xF32x2 = vext_f32(xF32x2, xF32x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_f32:
+  // CHECK-CODEGEN: {{ext.8.*#4}}
+}
+
+void test_vext_s64()
+{
+  // CHECK: test_vext_s64
+  int64x1_t xS64x1;
+  // FIXME don't use 1 as index or check for now, clang has a bug?
+  xS64x1 = vext_s64(xS64x1, xS64x1, /*1*/0);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_s64:
+  // CHECK_FIXME: {{ext.8.*#0}}
+}
+
+void test_vext_u64()
+{
+  // CHECK: test_vext_u64
+  uint64x1_t xU64x1;
+  // FIXME don't use 1 as index or check for now, clang has a bug?
+  xU64x1 = vext_u64(xU64x1, xU64x1, /*1*/0);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vext_u64:
+  // CHECK_FIXME: {{ext.8.*#0}}
+}
+
+void test_vextq_s8()
+{
+  // CHECK: test_vextq_s8
+  int8x16_t xS8x16;
+  xS8x16 = vextq_s8(xS8x16, xS8x16, 4);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_s8:
+  // CHECK-CODEGEN: {{ext.16.*#4}}
+}
+
+void test_vextq_u8()
+{
+  // CHECK: test_vextq_u8
+  uint8x16_t xU8x16;
+  xU8x16 = vextq_u8(xU8x16, xU8x16, 5);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_u8:
+  // CHECK-CODEGEN: {{ext.16.*#5}}
+}
+
+void test_vextq_p8()
+{
+  // CHECK: test_vextq_p8
+  poly8x16_t xP8x16;
+  xP8x16 = vextq_p8(xP8x16, xP8x16, 6);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_p8:
+  // CHECK-CODEGEN: {{ext.16.*#6}}
+}
+
+void test_vextq_s16()
+{
+  // CHECK: test_vextq_s16
+  int16x8_t xS16x8;
+  xS16x8 = vextq_s16(xS16x8, xS16x8, 7);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_s16:
+  // CHECK-CODEGEN: {{ext.16.*#14}}
+}
+
+void test_vextq_u16()
+{
+  // CHECK: test_vextq_u16
+  uint16x8_t xU16x8;
+  xU16x8 = vextq_u16(xU16x8, xU16x8, 4);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_u16:
+  // CHECK-CODEGEN: {{ext.16.*#8}}
+}
+
+void test_vextq_p16()
+{
+  // CHECK: test_vextq_p16
+  poly16x8_t xP16x8;
+  xP16x8 = vextq_p16(xP16x8, xP16x8, 5);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_p16:
+  // CHECK-CODEGEN: {{ext.16.*#10}}
+}
+
+void test_vextq_s32()
+{
+  // CHECK: test_vextq_s32
+  int32x4_t xS32x4;
+  xS32x4 = vextq_s32(xS32x4, xS32x4, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_s32:
+  // CHECK-CODEGEN: {{ext.16.*#4}}
+}
+
+void test_vextq_u32()
+{
+  // CHECK: test_vextq_u32
+  uint32x4_t xU32x4;
+  xU32x4 = vextq_u32(xU32x4, xU32x4, 2);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_u32:
+  // CHECK-CODEGEN: {{ext.16.*#8}}
+}
+
+void test_vextq_f32()
+{
+  // CHECK: test_vextq_f32
+  float32x4_t xF32x4;
+  xF32x4 = vextq_f32(xF32x4, xF32x4, 3);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_f32:
+  // CHECK-CODEGEN: {{ext.16.*#12}}
+}
+
+void test_vextq_s64()
+{
+  // CHECK: test_vextq_s64
+  int64x2_t xS64x2;
+  xS64x2 = vextq_s64(xS64x2, xS64x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_s64:
+  // CHECK-CODEGEN: {{ext.16.*#8}}
+}
+
+void test_vextq_u64()
+{
+  // CHECK: test_vextq_u64
+  uint64x2_t xU64x2;
+  xU64x2 = vextq_u64(xU64x2, xU64x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_u64:
+  // CHECK-CODEGEN: {{ext.16.*#8}}
+}
+
+void test_vextq_f64()
+{
+  // CHECK: test_vextq_f64
+  float64x2_t xF64x2;
+  xF64x2 = vextq_f64(xF64x2, xF64x2, 1);
+  // CHECK: shufflevector
+  // CHECK-CODEGEN: test_vextq_u64:
+  // CHECK-CODEGEN: {{ext.16.*#8}}
+}
diff --git a/test/CodeGen/arm64_vfma.c b/test/CodeGen/arm64_vfma.c
new file mode 100644
index 0000000..e57161b
--- /dev/null
+++ b/test/CodeGen/arm64_vfma.c
@@ -0,0 +1,136 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD fused multiply add intrinsics
+
+#include <arm_neon.h>
+
+float32x2_t test_vfma_f32(float32x2_t a1, float32x2_t a2, float32x2_t a3) {
+  // CHECK: test_vfma_f32
+  return vfma_f32(a1, a2, a3);
+  // CHECK: llvm.fma.v2f32({{.*a2, .*a3, .*a1}})
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vfmaq_f32(float32x4_t a1, float32x4_t a2, float32x4_t a3) {
+  // CHECK: test_vfmaq_f32
+  return vfmaq_f32(a1, a2, a3);
+  // CHECK: llvm.fma.v4f32({{.*a2, .*a3, .*a1}})
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vfmaq_f64(float64x2_t a1, float64x2_t a2, float64x2_t a3) {
+  // CHECK: test_vfmaq_f64
+  return vfmaq_f64(a1, a2, a3);
+  // CHECK: llvm.fma.v2f64({{.*a2, .*a3, .*a1}})
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vfma_lane_f32(float32x2_t a1, float32x2_t a2, float32x2_t a3) {
+  // CHECK: test_vfma_lane_f32
+  return vfma_lane_f32(a1, a2, a3, 1);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: llvm.fma.v2f32(<2 x float> %a2, <2 x float> {{.*}}, <2 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vfmaq_lane_f32(float32x4_t a1, float32x4_t a2, float32x2_t a3) {
+  // CHECK: test_vfmaq_lane_f32
+  return vfmaq_lane_f32(a1, a2, a3, 1);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: llvm.fma.v4f32(<4 x float> %a2, <4 x float> {{.*}}, <4 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vfmaq_lane_f64(float64x2_t a1, float64x2_t a2, float64x1_t a3) {
+  // CHECK: test_vfmaq_lane_f64
+  return vfmaq_lane_f64(a1, a2, a3, 0);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: llvm.fma.v2f64(<2 x double> %a2, <2 x double> {{.*}}, <2 x double> %a1)
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vfma_n_f32(float32x2_t a1, float32x2_t a2, float32_t a3) {
+  // CHECK: test_vfma_n_f32
+  return vfma_n_f32(a1, a2, a3);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 0 (usually two insertelements)
+  // CHECK: llvm.fma.v2f32
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vfmaq_n_f32(float32x4_t a1, float32x4_t a2, float32_t a3) {
+  // CHECK: test_vfmaq_n_f32
+  return vfmaq_n_f32(a1, a2, a3);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 0 (usually four insertelements)
+  // CHECK: llvm.fma.v4f32
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vfmaq_n_f64(float64x2_t a1, float64x2_t a2, float64_t a3) {
+  // CHECK: test_vfmaq_n_f64
+  return vfmaq_n_f64(a1, a2, a3);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 0 (usually two insertelements)
+  // CHECK: llvm.fma.v2f64
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vfms_f32(float32x2_t a1, float32x2_t a2, float32x2_t a3) {
+  // CHECK: test_vfms_f32
+  return vfms_f32(a1, a2, a3);
+  // CHECK: [[NEG:%.*]] = fsub <2 x float> {{.*}}, %a2
+  // CHECK: llvm.fma.v2f32(<2 x float> %a3, <2 x float> [[NEG]], <2 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vfmsq_f32(float32x4_t a1, float32x4_t a2, float32x4_t a3) {
+  // CHECK: test_vfmsq_f32
+  return vfmsq_f32(a1, a2, a3);
+  // CHECK: [[NEG:%.*]] = fsub <4 x float> {{.*}}, %a2
+  // CHECK: llvm.fma.v4f32(<4 x float> %a3, <4 x float> [[NEG]], <4 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vfmsq_f64(float64x2_t a1, float64x2_t a2, float64x2_t a3) {
+  // CHECK: test_vfmsq_f64
+  return vfmsq_f64(a1, a2, a3);
+  // CHECK: [[NEG:%.*]] = fsub <2 x double> {{.*}}, %a2
+  // CHECK: llvm.fma.v2f64(<2 x double> %a3, <2 x double> [[NEG]], <2 x double> %a1)
+  // CHECK-NEXT: ret
+}
+
+float32x2_t test_vfms_lane_f32(float32x2_t a1, float32x2_t a2, float32x2_t a3) {
+  // CHECK: test_vfms_lane_f32
+  return vfms_lane_f32(a1, a2, a3, 1);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: [[NEG:%.*]] = fsub <2 x float> {{.*}}, %a3
+  // CHECK: [[LANE:%.*]] = shufflevector <2 x float> [[NEG]]
+  // CHECK: llvm.fma.v2f32(<2 x float> {{.*}}, <2 x float> [[LANE]], <2 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float32x4_t test_vfmsq_lane_f32(float32x4_t a1, float32x4_t a2, float32x2_t a3) {
+  // CHECK: test_vfmsq_lane_f32
+  return vfmsq_lane_f32(a1, a2, a3, 1);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: [[NEG:%.*]] = fsub <2 x float> {{.*}}, %a3
+  // CHECK: [[LANE:%.*]] = shufflevector <2 x float> [[NEG]]
+  // CHECK: llvm.fma.v4f32(<4 x float> {{.*}}, <4 x float> [[LANE]], <4 x float> %a1)
+  // CHECK-NEXT: ret
+}
+
+float64x2_t test_vfmsq_lane_f64(float64x2_t a1, float64x2_t a2, float64x1_t a3) {
+  // CHECK: test_vfmsq_lane_f64
+  return vfmsq_lane_f64(a1, a2, a3, 0);
+  // NB: the test below is deliberately lose, so that we don't depend too much
+  // upon the exact IR used to select lane 1 (usually a shufflevector)
+  // CHECK: [[NEG:%.*]] = fsub <1 x double> {{.*}}, %a3
+  // CHECK: [[LANE:%.*]] = shufflevector <1 x double> [[NEG]]
+  // CHECK: llvm.fma.v2f64(<2 x double> {{.*}}, <2 x double> [[LANE]], <2 x double> %a1)
+  // CHECK-NEXT: ret
+}
diff --git a/test/CodeGen/arm64_vget.c b/test/CodeGen/arm64_vget.c
new file mode 100644
index 0000000..44c7da5
--- /dev/null
+++ b/test/CodeGen/arm64_vget.c
@@ -0,0 +1,13 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD vget intrinsics
+
+#include <arm_neon.h>
+
+float64_t test_vget_lane_f64(float64x1_t a1) {
+  // CHECK: test_vget_lane_f64
+  // why isn't 1 allowed as second argument?
+  return vget_lane_f64(a1, 0);
+  // CHECK: extractelement {{.*}} i32 0
+  // CHECK-NEXT: ret
+}
+
diff --git a/test/CodeGen/arm64_vneg.c b/test/CodeGen/arm64_vneg.c
new file mode 100644
index 0000000..96b9e1f
--- /dev/null
+++ b/test/CodeGen/arm64_vneg.c
@@ -0,0 +1,18 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD negate and saturating negate intrinsics
+
+#include <arm_neon.h>
+
+int64x2_t test_vnegq_s64(int64x2_t a1) {
+  // CHECK: test_vnegq_s64
+  return vnegq_s64(a1);
+  // CHECK: sub <2 x i64> zeroinitializer, %a1
+  // CHECK-NEXT: ret
+}
+
+int64x2_t test_vqnegq_s64(int64x2_t a1) {
+  // CHECK: test_vqnegq_s64
+  return vqnegq_s64(a1);
+  // CHECK: llvm.arm64.neon.sqneg.v2i64
+  // CHECK-NEXT: ret
+}
diff --git a/test/CodeGen/arm64_vqmov.c b/test/CodeGen/arm64_vqmov.c
new file mode 100644
index 0000000..e8c0605
--- /dev/null
+++ b/test/CodeGen/arm64_vqmov.c
@@ -0,0 +1,77 @@
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | FileCheck %s
+// REQUIRES: arm64-registered-target
+/// Test vqmov[u]n_high_<su>{16,32,64) ARM64 intrinsics
+
+#include <arm_neon.h>
+
+// vqmovn_high_s16 -> UQXTN2 Vd.16b,Vn.8h
+int8x16_t test_vqmovn_high_s16(int8x8_t Vdlow, int16x8_t Vn)
+{
+    return vqmovn_high_s16(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_s16:
+  // CHECK: sqxtn2.16b {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovun_high_s16 -> UQXTN2 Vd.16b,Vn.8h
+uint8x16_t test_vqmovun_high_s16(uint8x8_t Vdlow, uint16x8_t Vn)
+{
+    return vqmovun_high_s16(Vdlow, Vn);
+  // CHECK: test_vqmovun_high_s16:
+  // CHECK: sqxtun2.16b {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovn_high_s32 -> SQXTN2 Vd.8h,Vn.4s
+int16x8_t test_vqmovn_high_s32(int16x4_t Vdlow, int32x4_t Vn)
+{
+    return vqmovn_high_s32(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_s32:
+  // CHECK: sqxtn2.8h {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovn_high_u32 -> UQXTN2 Vd.8h,Vn.4s
+uint16x8_t test_vqmovn_high_u32(uint16x4_t Vdlow, uint32x4_t Vn)
+{
+    return vqmovn_high_u32(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_u32:
+  // CHECK: uqxtn2.8h {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovn_high_s64 -> SQXTN2 Vd.4s,Vn.2d
+int32x4_t test_vqmovn_high_s64(int32x2_t Vdlow, int64x2_t Vn)
+{
+    return vqmovn_high_s64(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_s64:
+  // CHECK: sqxtn2.4s {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovn_high_u64 -> UQXTN2 Vd.4s,Vn.2d
+uint32x4_t test_vqmovn_high_u64(uint32x2_t Vdlow, uint64x2_t Vn)
+{
+    return vqmovn_high_u64(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_u64:
+  // CHECK: uqxtn2.4s {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovn_high_u16 -> UQXTN2 Vd.16b,Vn.8h
+uint8x16_t test_vqmovn_high_u16(uint8x8_t Vdlow, uint16x8_t Vn)
+{
+    return vqmovn_high_u16(Vdlow, Vn);
+  // CHECK: test_vqmovn_high_u16:
+  // CHECK: uqxtn2.16b {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovun_high_s32 -> SQXTUN2 Vd.8h,Vn.4s
+uint16x8_t test_vqmovun_high_s32(uint16x4_t Vdlow, uint32x4_t Vn)
+{
+    return vqmovun_high_s32(Vdlow, Vn);
+  // CHECK: test_vqmovun_high_s32:
+  // CHECK: sqxtun2.8h {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
+
+// vqmovun_high_s64 -> SQXTUN2  Vd.4s,Vn.2d
+uint32x4_t test_vqmovun_high_s64(uint32x2_t Vdlow, uint64x2_t Vn)
+{
+    return vqmovun_high_s64(Vdlow, Vn);
+  // CHECK: test_vqmovun_high_s64:
+  // CHECK: sqxtun2.4s {{v[0-9][0-9]*}}, {{v[0-9][0-9]*}}
+}
diff --git a/test/CodeGen/arm64_vrecps.c b/test/CodeGen/arm64_vrecps.c
new file mode 100644
index 0000000..66f06ce
--- /dev/null
+++ b/test/CodeGen/arm64_vrecps.c
@@ -0,0 +1,26 @@
+// RUN: %clang_cc1 -O3 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | FileCheck %s
+// REQUIRES: arm64-registered-target
+/// Test vrecpss_f32, vrecpsd_f64 ARM64 intrinsics
+
+
+#include <arm_neon.h>
+
+// vrecpss_f32 -> FRECPS Sd,Sa,Sb
+//
+float32_t test_vrecpss_f32(float32_t Vdlow, float32_t Vn)
+{
+    return vrecpss_f32(Vdlow, Vn);
+  // CHECK: test_vrecpss_f32:
+  // CHECK: frecps  s0, s0, s1
+  // CHECK-NEXT: ret
+}
+
+// vrecpsd_f64 -> FRECPS Dd,Da,Db
+//
+float64_t test_vrecpsd_f64(float64_t Vdlow, float64_t Vn)
+{
+    return vrecpsd_f64(Vdlow, Vn);
+  // CHECK: test_vrecpsd_f64:
+  // CHECK: frecps d0, d0, d1
+  // CHECK-NEXT: ret
+}
diff --git a/test/CodeGen/arm64_vset_lane.c b/test/CodeGen/arm64_vset_lane.c
new file mode 100644
index 0000000..6997eb8
--- /dev/null
+++ b/test/CodeGen/arm64_vset_lane.c
@@ -0,0 +1,31 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD set lane intrinsics INCOMPLETE
+
+#include <arm_neon.h>
+
+float16x4_t test_vset_lane_f16(float16_t *a1, float16x4_t a2) {
+  // CHECK-LABEL: test_vset_lane_f16
+  return vset_lane_f16(*a1, a2, 1);
+  // CHECK insertelement <4 x i16> %a2, i16 %a1, i32 1
+}
+
+float16x8_t test_vsetq_lane_f16(float16_t *a1, float16x8_t a2) {
+  // CHECK-LABEL: test_vsetq_lane_f16
+  return vsetq_lane_f16(*a1, a2, 4);
+  // CHECK insertelement <8 x i16> %a2, i16 %a1, i32 4
+}
+
+// problem with scalar_to_vector in backend.  Punt for now
+#if 0
+float64x1_t test_vset_lane_f64(float64_t a1, float64x1_t a2) {
+  // CHECK-LABEL@ test_vset_lane_f64
+  return vset_lane_f64(a1, a2, 0);
+  // CHECK@ @llvm.arm64.neon.smaxv.i32.v8i8
+}
+#endif
+
+float64x2_t test_vsetq_lane_f64(float64_t a1, float64x2_t a2) {
+  // CHECK-LABEL: test_vsetq_lane_f64
+  return vsetq_lane_f64(a1, a2, 0);
+  // CHECK insertelement <2 x double> %a2, double %a1, i32 0
+}
diff --git a/test/CodeGen/arm64_vshift.c b/test/CodeGen/arm64_vshift.c
new file mode 100644
index 0000000..cac4d12
--- /dev/null
+++ b/test/CodeGen/arm64_vshift.c
@@ -0,0 +1,357 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios7.0 -ffreestanding -emit-llvm -o - -O1 %s | FileCheck %s
+#include <arm_neon.h>
+
+int8x8_t test_vqshl_n_s8(int8x8_t in) {
+  // CHECK-LABEL: @test_vqshl_n_s8
+  // CHECK: call <8 x i8> @llvm.arm64.neon.sqshl.v8i8(<8 x i8> %in, <8 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshl_n_s8(in, 1);
+}
+
+int16x4_t test_vqshl_n_s16(int16x4_t in) {
+  // CHECK-LABEL: @test_vqshl_n_s16
+  // CHECK: call <4 x i16> @llvm.arm64.neon.sqshl.v4i16(<4 x i16> %in, <4 x i16> <i16 1, i16 1, i16 1, i16 1>)
+  return vqshl_n_s16(in, 1);
+}
+
+int32x2_t test_vqshl_n_s32(int32x2_t in) {
+  // CHECK-LABEL: @test_vqshl_n_s32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.sqshl.v2i32(<2 x i32> %in, <2 x i32> <i32 1, i32 1>)
+  return vqshl_n_s32(in, 1);
+}
+
+int64x1_t test_vqshl_n_s64(int64x1_t in) {
+  // CHECK-LABEL: @test_vqshl_n_s64
+  // CHECK: call <1 x i64> @llvm.arm64.neon.sqshl.v1i64(<1 x i64> %in, <1 x i64> <i64 1>)
+  return vqshl_n_s64(in, 1);
+}
+
+
+int8x16_t test_vqshlq_n_s8(int8x16_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_s8
+  // CHECK: call <16 x i8> @llvm.arm64.neon.sqshl.v16i8(<16 x i8> %in, <16 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshlq_n_s8(in, 1);
+}
+
+int16x8_t test_vqshlq_n_s16(int16x8_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_s16
+  // CHECK: call <8 x i16> @llvm.arm64.neon.sqshl.v8i16(<8 x i16> %in, <8 x i16> <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>)
+  return vqshlq_n_s16(in, 1);
+}
+
+int32x4_t test_vqshlq_n_s32(int32x4_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_s32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.sqshl.v4i32(<4 x i32> %in, <4 x i32> <i32 1, i32 1, i32 1, i32 1>)
+  return vqshlq_n_s32(in, 1);
+}
+
+int64x2_t test_vqshlq_n_s64(int64x2_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_s64
+  // CHECK: call <2 x i64> @llvm.arm64.neon.sqshl.v2i64(<2 x i64> %in, <2 x i64> <i64 1, i64 1>
+  return vqshlq_n_s64(in, 1);
+}
+
+uint8x8_t test_vqshl_n_u8(uint8x8_t in) {
+  // CHECK-LABEL: @test_vqshl_n_u8
+  // CHECK: call <8 x i8> @llvm.arm64.neon.uqshl.v8i8(<8 x i8> %in, <8 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshl_n_u8(in, 1);
+}
+
+uint16x4_t test_vqshl_n_u16(uint16x4_t in) {
+  // CHECK-LABEL: @test_vqshl_n_u16
+  // CHECK: call <4 x i16> @llvm.arm64.neon.uqshl.v4i16(<4 x i16> %in, <4 x i16> <i16 1, i16 1, i16 1, i16 1>)
+  return vqshl_n_u16(in, 1);
+}
+
+uint32x2_t test_vqshl_n_u32(uint32x2_t in) {
+  // CHECK-LABEL: @test_vqshl_n_u32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.uqshl.v2i32(<2 x i32> %in, <2 x i32> <i32 1, i32 1>)
+  return vqshl_n_u32(in, 1);
+}
+
+uint64x1_t test_vqshl_n_u64(uint64x1_t in) {
+  // CHECK-LABEL: @test_vqshl_n_u64
+  // CHECK: call <1 x i64> @llvm.arm64.neon.uqshl.v1i64(<1 x i64> %in, <1 x i64> <i64 1>)
+  return vqshl_n_u64(in, 1);
+}
+
+uint8x16_t test_vqshlq_n_u8(uint8x16_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_u8
+  // CHECK: call <16 x i8> @llvm.arm64.neon.uqshl.v16i8(<16 x i8> %in, <16 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshlq_n_u8(in, 1);
+}
+
+uint16x8_t test_vqshlq_n_u16(uint16x8_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_u16
+  // CHECK: call <8 x i16> @llvm.arm64.neon.uqshl.v8i16(<8 x i16> %in, <8 x i16> <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>)
+  return vqshlq_n_u16(in, 1);
+}
+
+uint32x4_t test_vqshlq_n_u32(uint32x4_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_u32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.uqshl.v4i32(<4 x i32> %in, <4 x i32> <i32 1, i32 1, i32 1, i32 1>)
+  return vqshlq_n_u32(in, 1);
+}
+
+uint64x2_t test_vqshlq_n_u64(uint64x2_t in) {
+  // CHECK-LABEL: @test_vqshlq_n_u64
+  // CHECK: call <2 x i64> @llvm.arm64.neon.uqshl.v2i64(<2 x i64> %in, <2 x i64> <i64 1, i64 1>
+  return vqshlq_n_u64(in, 1);
+}
+
+int8x8_t test_vrshr_n_s8(int8x8_t in) {
+  // CHECK-LABEL: @test_vrshr_n_s8
+  // CHECK: call <8 x i8> @llvm.arm64.neon.srshl.v8i8(<8 x i8> %in, <8 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  return vrshr_n_s8(in, 1);
+}
+
+int16x4_t test_vrshr_n_s16(int16x4_t in) {
+  // CHECK-LABEL: @test_vrshr_n_s16
+  // CHECK: call <4 x i16> @llvm.arm64.neon.srshl.v4i16(<4 x i16> %in, <4 x i16> <i16 -1, i16 -1, i16 -1, i16 -1>)
+  return vrshr_n_s16(in, 1);
+}
+
+int32x2_t test_vrshr_n_s32(int32x2_t in) {
+  // CHECK-LABEL: @test_vrshr_n_s32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.srshl.v2i32(<2 x i32> %in, <2 x i32> <i32 -1, i32 -1>)
+  return vrshr_n_s32(in, 1);
+}
+
+int64x1_t test_vrshr_n_s64(int64x1_t in) {
+  // CHECK-LABEL: @test_vrshr_n_s64
+  // CHECK: call <1 x i64> @llvm.arm64.neon.srshl.v1i64(<1 x i64> %in, <1 x i64> <i64 -1>)
+  return vrshr_n_s64(in, 1);
+}
+
+
+int8x16_t test_vrshrq_n_s8(int8x16_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_s8
+  // CHECK: call <16 x i8> @llvm.arm64.neon.srshl.v16i8(<16 x i8> %in, <16 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  return vrshrq_n_s8(in, 1);
+}
+
+int16x8_t test_vrshrq_n_s16(int16x8_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_s16
+  // CHECK: call <8 x i16> @llvm.arm64.neon.srshl.v8i16(<8 x i16> %in, <8 x i16> <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>)
+  return vrshrq_n_s16(in, 1);
+}
+
+int32x4_t test_vrshrq_n_s32(int32x4_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_s32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.srshl.v4i32(<4 x i32> %in, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>)
+  return vrshrq_n_s32(in, 1);
+}
+
+int64x2_t test_vrshrq_n_s64(int64x2_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_s64
+  // CHECK: call <2 x i64> @llvm.arm64.neon.srshl.v2i64(<2 x i64> %in, <2 x i64> <i64 -1, i64 -1>
+  return vrshrq_n_s64(in, 1);
+}
+
+uint8x8_t test_vrshr_n_u8(uint8x8_t in) {
+  // CHECK-LABEL: @test_vrshr_n_u8
+  // CHECK: call <8 x i8> @llvm.arm64.neon.urshl.v8i8(<8 x i8> %in, <8 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  return vrshr_n_u8(in, 1);
+}
+
+uint16x4_t test_vrshr_n_u16(uint16x4_t in) {
+  // CHECK-LABEL: @test_vrshr_n_u16
+  // CHECK: call <4 x i16> @llvm.arm64.neon.urshl.v4i16(<4 x i16> %in, <4 x i16> <i16 -1, i16 -1, i16 -1, i16 -1>)
+  return vrshr_n_u16(in, 1);
+}
+
+uint32x2_t test_vrshr_n_u32(uint32x2_t in) {
+  // CHECK-LABEL: @test_vrshr_n_u32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.urshl.v2i32(<2 x i32> %in, <2 x i32> <i32 -1, i32 -1>)
+  return vrshr_n_u32(in, 1);
+}
+
+uint64x1_t test_vrshr_n_u64(uint64x1_t in) {
+  // CHECK-LABEL: @test_vrshr_n_u64
+  // CHECK: call <1 x i64> @llvm.arm64.neon.urshl.v1i64(<1 x i64> %in, <1 x i64> <i64 -1>)
+  return vrshr_n_u64(in, 1);
+}
+
+uint8x16_t test_vrshrq_n_u8(uint8x16_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_u8
+  // CHECK: call <16 x i8> @llvm.arm64.neon.urshl.v16i8(<16 x i8> %in, <16 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  return vrshrq_n_u8(in, 1);
+}
+
+uint16x8_t test_vrshrq_n_u16(uint16x8_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_u16
+  // CHECK: call <8 x i16> @llvm.arm64.neon.urshl.v8i16(<8 x i16> %in, <8 x i16> <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>)
+  return vrshrq_n_u16(in, 1);
+}
+
+uint32x4_t test_vrshrq_n_u32(uint32x4_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_u32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.urshl.v4i32(<4 x i32> %in, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>)
+  return vrshrq_n_u32(in, 1);
+}
+
+uint64x2_t test_vrshrq_n_u64(uint64x2_t in) {
+  // CHECK-LABEL: @test_vrshrq_n_u64
+  // CHECK: call <2 x i64> @llvm.arm64.neon.urshl.v2i64(<2 x i64> %in, <2 x i64> <i64 -1, i64 -1>
+  return vrshrq_n_u64(in, 1);
+}
+
+int8x8_t test_vqshlu_n_s8(int8x8_t in) {
+  // CHECK-LABEL: @test_vqshlu_n_s8
+  // CHECK: call <8 x i8> @llvm.arm64.neon.sqshlu.v8i8(<8 x i8> %in, <8 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshlu_n_s8(in, 1);
+}
+
+int16x4_t test_vqshlu_n_s16(int16x4_t in) {
+  // CHECK-LABEL: @test_vqshlu_n_s16
+  // CHECK: call <4 x i16> @llvm.arm64.neon.sqshlu.v4i16(<4 x i16> %in, <4 x i16> <i16 1, i16 1, i16 1, i16 1>)
+  return vqshlu_n_s16(in, 1);
+}
+
+int32x2_t test_vqshlu_n_s32(int32x2_t in) {
+  // CHECK-LABEL: @test_vqshlu_n_s32
+  // CHECK: call <2 x i32> @llvm.arm64.neon.sqshlu.v2i32(<2 x i32> %in, <2 x i32> <i32 1, i32 1>)
+  return vqshlu_n_s32(in, 1);
+}
+
+int64x1_t test_vqshlu_n_s64(int64x1_t in) {
+  // CHECK-LABEL: @test_vqshlu_n_s64
+  // CHECK: call <1 x i64> @llvm.arm64.neon.sqshlu.v1i64(<1 x i64> %in, <1 x i64> <i64 1>)
+  return vqshlu_n_s64(in, 1);
+}
+
+
+int8x16_t test_vqshluq_n_s8(int8x16_t in) {
+  // CHECK-LABEL: @test_vqshluq_n_s8
+  // CHECK: call <16 x i8> @llvm.arm64.neon.sqshlu.v16i8(<16 x i8> %in, <16 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>)
+  return vqshluq_n_s8(in, 1);
+}
+
+int16x8_t test_vqshluq_n_s16(int16x8_t in) {
+  // CHECK-LABEL: @test_vqshluq_n_s16
+  // CHECK: call <8 x i16> @llvm.arm64.neon.sqshlu.v8i16(<8 x i16> %in, <8 x i16> <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>)
+  return vqshluq_n_s16(in, 1);
+}
+
+int32x4_t test_vqshluq_n_s32(int32x4_t in) {
+  // CHECK-LABEL: @test_vqshluq_n_s32
+  // CHECK: call <4 x i32> @llvm.arm64.neon.sqshlu.v4i32(<4 x i32> %in, <4 x i32> <i32 1, i32 1, i32 1, i32 1>)
+  return vqshluq_n_s32(in, 1);
+}
+
+int64x2_t test_vqshluq_n_s64(int64x2_t in) {
+  // CHECK-LABEL: @test_vqshluq_n_s64
+  // CHECK: call <2 x i64> @llvm.arm64.neon.sqshlu.v2i64(<2 x i64> %in, <2 x i64> <i64 1, i64 1>
+  return vqshluq_n_s64(in, 1);
+}
+
+int8x8_t test_vrsra_n_s8(int8x8_t acc, int8x8_t in) {
+  // CHECK-LABEL: @test_vrsra_n_s8
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <8 x i8> @llvm.arm64.neon.srshl.v8i8(<8 x i8> %in, <8 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  // CHECK: add <8 x i8> [[TMP]], %acc
+  return vrsra_n_s8(acc, in, 1);
+}
+
+int16x4_t test_vrsra_n_s16(int16x4_t acc, int16x4_t in) {
+  // CHECK-LABEL: @test_vrsra_n_s16
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <4 x i16> @llvm.arm64.neon.srshl.v4i16(<4 x i16> %in, <4 x i16> <i16 -1, i16 -1, i16 -1, i16 -1>)
+  // CHECK: add <4 x i16> [[TMP]], %acc
+  return vrsra_n_s16(acc, in, 1);
+}
+
+int32x2_t test_vrsra_n_s32(int32x2_t acc, int32x2_t in) {
+  // CHECK-LABEL: @test_vrsra_n_s32
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <2 x i32> @llvm.arm64.neon.srshl.v2i32(<2 x i32> %in, <2 x i32> <i32 -1, i32 -1>)
+  // CHECK: add <2 x i32> [[TMP]], %acc
+  return vrsra_n_s32(acc, in, 1);
+}
+
+int64x1_t test_vrsra_n_s64(int64x1_t acc, int64x1_t in) {
+  // CHECK-LABEL: @test_vrsra_n_s64
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <1 x i64> @llvm.arm64.neon.srshl.v1i64(<1 x i64> %in, <1 x i64> <i64 -1>)
+  // CHECK: add <1 x i64> [[TMP]], %acc
+  return vrsra_n_s64(acc, in, 1);
+}
+
+int8x16_t test_vrsraq_n_s8(int8x16_t acc, int8x16_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_s8
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <16 x i8> @llvm.arm64.neon.srshl.v16i8(<16 x i8> %in, <16 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  // CHECK: add <16 x i8> [[TMP]], %acc
+  return vrsraq_n_s8(acc, in, 1);
+}
+
+int16x8_t test_vrsraq_n_s16(int16x8_t acc, int16x8_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_s16
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <8 x i16> @llvm.arm64.neon.srshl.v8i16(<8 x i16> %in, <8 x i16> <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>)
+  // CHECK: add <8 x i16> [[TMP]], %acc
+  return vrsraq_n_s16(acc, in, 1);
+}
+
+int32x4_t test_vrsraq_n_s32(int32x4_t acc, int32x4_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_s32
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <4 x i32> @llvm.arm64.neon.srshl.v4i32(<4 x i32> %in, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>)
+  // CHECK: add <4 x i32> [[TMP]], %acc
+  return vrsraq_n_s32(acc, in, 1);
+}
+
+int64x2_t test_vrsraq_n_s64(int64x2_t acc, int64x2_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_s64
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <2 x i64> @llvm.arm64.neon.srshl.v2i64(<2 x i64> %in, <2 x i64> <i64 -1, i64 -1>)
+  // CHECK: add <2 x i64> [[TMP]], %acc
+  return vrsraq_n_s64(acc, in, 1);
+}
+
+uint8x8_t test_vrsra_n_u8(uint8x8_t acc, uint8x8_t in) {
+  // CHECK-LABEL: @test_vrsra_n_u8
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <8 x i8> @llvm.arm64.neon.urshl.v8i8(<8 x i8> %in, <8 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  // CHECK: add <8 x i8> [[TMP]], %acc
+  return vrsra_n_u8(acc, in, 1);
+}
+
+uint16x4_t test_vrsra_n_u16(uint16x4_t acc, uint16x4_t in) {
+  // CHECK-LABEL: @test_vrsra_n_u16
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <4 x i16> @llvm.arm64.neon.urshl.v4i16(<4 x i16> %in, <4 x i16> <i16 -1, i16 -1, i16 -1, i16 -1>)
+  // CHECK: add <4 x i16> [[TMP]], %acc
+  return vrsra_n_u16(acc, in, 1);
+}
+
+uint32x2_t test_vrsra_n_u32(uint32x2_t acc, uint32x2_t in) {
+  // CHECK-LABEL: @test_vrsra_n_u32
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <2 x i32> @llvm.arm64.neon.urshl.v2i32(<2 x i32> %in, <2 x i32> <i32 -1, i32 -1>)
+  // CHECK: add <2 x i32> [[TMP]], %acc
+  return vrsra_n_u32(acc, in, 1);
+}
+
+uint64x1_t test_vrsra_n_u64(uint64x1_t acc, uint64x1_t in) {
+  // CHECK-LABEL: @test_vrsra_n_u64
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <1 x i64> @llvm.arm64.neon.urshl.v1i64(<1 x i64> %in, <1 x i64> <i64 -1>)
+  // CHECK: add <1 x i64> [[TMP]], %acc
+  return vrsra_n_u64(acc, in, 1);
+}
+
+uint8x16_t test_vrsraq_n_u8(uint8x16_t acc, uint8x16_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_u8
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <16 x i8> @llvm.arm64.neon.urshl.v16i8(<16 x i8> %in, <16 x i8> <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>)
+  // CHECK: add <16 x i8> [[TMP]], %acc
+  return vrsraq_n_u8(acc, in, 1);
+}
+
+uint16x8_t test_vrsraq_n_u16(uint16x8_t acc, uint16x8_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_u16
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <8 x i16> @llvm.arm64.neon.urshl.v8i16(<8 x i16> %in, <8 x i16> <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>)
+  // CHECK: add <8 x i16> [[TMP]], %acc
+  return vrsraq_n_u16(acc, in, 1);
+}
+
+uint32x4_t test_vrsraq_n_u32(uint32x4_t acc, uint32x4_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_u32
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <4 x i32> @llvm.arm64.neon.urshl.v4i32(<4 x i32> %in, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>)
+  // CHECK: add <4 x i32> [[TMP]], %acc
+  return vrsraq_n_u32(acc, in, 1);
+}
+
+uint64x2_t test_vrsraq_n_u64(uint64x2_t acc, uint64x2_t in) {
+  // CHECK-LABEL: @test_vrsraq_n_u64
+  // CHECK: [[TMP:%[0-9a-zA-Z._]+]] = tail call <2 x i64> @llvm.arm64.neon.urshl.v2i64(<2 x i64> %in, <2 x i64> <i64 -1, i64 -1>)
+  // CHECK: add <2 x i64> [[TMP]], %acc
+  return vrsraq_n_u64(acc, in, 1);
+}
diff --git a/test/CodeGen/arm64_vsli.c b/test/CodeGen/arm64_vsli.c
new file mode 100644
index 0000000..bce1446
--- /dev/null
+++ b/test/CodeGen/arm64_vsli.c
@@ -0,0 +1,148 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | \
+// RUN:   FileCheck -check-prefix=CHECK_CODEGEN %s
+// REQUIRES: arm64-registered-target
+// Test
+
+#include <arm_neon.h>
+
+int8x8_t test_vsli_n_s8(int8x8_t a1, int8x8_t a2) {
+  // CHECK: test_vsli_n_s8
+  return vsli_n_s8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v8i8
+  // CHECK_CODEGEN: sli.8b  v0, v1, #3
+}
+
+int16x4_t test_vsli_n_s16(int16x4_t a1, int16x4_t a2) {
+  // CHECK: test_vsli_n_s16
+  return vsli_n_s16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v4i16
+  // CHECK_CODEGEN: sli.4h  v0, v1, #3
+}
+
+int32x2_t test_vsli_n_s32(int32x2_t a1, int32x2_t a2) {
+  // CHECK: test_vsli_n_s32
+  return vsli_n_s32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v2i32
+  // CHECK_CODEGEN: sli.2s  v0, v1, #1
+}
+
+int64x1_t test_vsli_n_s64(int64x1_t a1, int64x1_t a2) {
+  // CHECK: test_vsli_n_s64
+  return vsli_n_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v1i64
+  // CHECK_CODEGEN: sli     d0, d1, #1
+}
+
+uint8x8_t test_vsli_n_u8(uint8x8_t a1, uint8x8_t a2) {
+  // CHECK: test_vsli_n_u8
+  return vsli_n_u8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v8i8
+  // CHECK_CODEGEN: sli.8b  v0, v1, #3
+}
+
+uint16x4_t test_vsli_n_u16(uint16x4_t a1, uint16x4_t a2) {
+  // CHECK: test_vsli_n_u16
+  return vsli_n_u16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v4i16
+  // CHECK_CODEGEN: sli.4h  v0, v1, #3
+}
+
+uint32x2_t test_vsli_n_u32(uint32x2_t a1, uint32x2_t a2) {
+  // CHECK: test_vsli_n_u32
+  return vsli_n_u32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v2i32
+  // CHECK_CODEGEN: sli.2s  v0, v1, #1
+}
+
+uint64x1_t test_vsli_n_u64(uint64x1_t a1, uint64x1_t a2) {
+  // CHECK: test_vsli_n_u64
+  return vsli_n_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v1i64
+  // CHECK_CODEGEN: sli     d0, d1, #1
+}
+
+poly8x8_t test_vsli_n_p8(poly8x8_t a1, poly8x8_t a2) {
+  // CHECK: test_vsli_n_p8
+  return vsli_n_p8(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v8i8
+  // CHECK_CODEGEN: sli.8b  v0, v1, #1
+}
+
+poly16x4_t test_vsli_n_p16(poly16x4_t a1, poly16x4_t a2) {
+  // CHECK: test_vsli_n_p16
+  return vsli_n_p16(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v4i16
+  // CHECK_CODEGEN: sli.4h  v0, v1, #1
+}
+
+int8x16_t test_vsliq_n_s8(int8x16_t a1, int8x16_t a2) {
+  // CHECK: test_vsliq_n_s8
+  return vsliq_n_s8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v16i8
+  // CHECK_CODEGEN: sli.16b v0, v1, #3
+}
+
+int16x8_t test_vsliq_n_s16(int16x8_t a1, int16x8_t a2) {
+  // CHECK: test_vsliq_n_s16
+  return vsliq_n_s16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v8i16
+  // CHECK_CODEGEN: sli.8h  v0, v1, #3
+}
+
+int32x4_t test_vsliq_n_s32(int32x4_t a1, int32x4_t a2) {
+  // CHECK: test_vsliq_n_s32
+  return vsliq_n_s32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v4i32
+  // CHECK_CODEGEN: sli.4s  v0, v1, #1
+}
+
+int64x2_t test_vsliq_n_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vsliq_n_s64
+  return vsliq_n_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v2i64
+  // CHECK_CODEGEN: sli.2d  v0, v1, #1
+}
+
+uint8x16_t test_vsliq_n_u8(uint8x16_t a1, uint8x16_t a2) {
+  // CHECK: test_vsliq_n_u8
+  return vsliq_n_u8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v16i8
+  // CHECK_CODEGEN: sli.16b v0, v1, #3
+}
+
+uint16x8_t test_vsliq_n_u16(uint16x8_t a1, uint16x8_t a2) {
+  // CHECK: test_vsliq_n_u16
+  return vsliq_n_u16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsli.v8i16
+  // CHECK_CODEGEN: sli.8h  v0, v1, #3
+}
+
+uint32x4_t test_vsliq_n_u32(uint32x4_t a1, uint32x4_t a2) {
+  // CHECK: test_vsliq_n_u32
+  return vsliq_n_u32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v4i32
+  // CHECK_CODEGEN: sli.4s  v0, v1, #1
+}
+
+uint64x2_t test_vsliq_n_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vsliq_n_u64
+  return vsliq_n_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v2i64
+  // CHECK_CODEGEN: sli.2d  v0, v1, #1
+}
+
+poly8x16_t test_vsliq_n_p8(poly8x16_t a1, poly8x16_t a2) {
+  // CHECK: test_vsliq_n_p8
+  return vsliq_n_p8(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v16i8
+  // CHECK_CODEGEN: sli.16b v0, v1, #1
+}
+
+poly16x8_t test_vsliq_n_p16(poly16x8_t a1, poly16x8_t a2) {
+  // CHECK: test_vsliq_n_p16
+  return vsliq_n_p16(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsli.v8i16
+  // CHECK_CODEGEN: sli.8h  v0, v1, #1
+}
+
diff --git a/test/CodeGen/arm64_vsri.c b/test/CodeGen/arm64_vsri.c
new file mode 100644
index 0000000..f561339
--- /dev/null
+++ b/test/CodeGen/arm64_vsri.c
@@ -0,0 +1,149 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - %s | \
+// RUN:   FileCheck -check-prefix=CHECK_CODEGEN %s
+// REQUIRES: arm64-registered-target
+
+// Test ARM64 SIMD vector shift right and insert: vsri[q]_n_*
+
+#include <arm_neon.h>
+
+int8x8_t test_vsri_n_s8(int8x8_t a1, int8x8_t a2) {
+  // CHECK: test_vsri_n_s8
+  return vsri_n_s8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v8i8
+  // CHECK_CODEGEN: sri.8b  v0, v1, #3
+}
+
+int16x4_t test_vsri_n_s16(int16x4_t a1, int16x4_t a2) {
+  // CHECK: test_vsri_n_s16
+  return vsri_n_s16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v4i16
+  // CHECK_CODEGEN: sri.4h  v0, v1, #3
+}
+
+int32x2_t test_vsri_n_s32(int32x2_t a1, int32x2_t a2) {
+  // CHECK: test_vsri_n_s32
+  return vsri_n_s32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v2i32
+  // CHECK_CODEGEN: sri.2s  v0, v1, #1
+}
+
+int64x1_t test_vsri_n_s64(int64x1_t a1, int64x1_t a2) {
+  // CHECK: test_vsri_n_s64
+  return vsri_n_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v1i64
+  // CHECK_CODEGEN: sri     d0, d1, #1
+}
+
+uint8x8_t test_vsri_n_u8(uint8x8_t a1, uint8x8_t a2) {
+  // CHECK: test_vsri_n_u8
+  return vsri_n_u8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v8i8
+  // CHECK_CODEGEN: sri.8b  v0, v1, #3
+}
+
+uint16x4_t test_vsri_n_u16(uint16x4_t a1, uint16x4_t a2) {
+  // CHECK: test_vsri_n_u16
+  return vsri_n_u16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v4i16
+  // CHECK_CODEGEN: sri.4h  v0, v1, #3
+}
+
+uint32x2_t test_vsri_n_u32(uint32x2_t a1, uint32x2_t a2) {
+  // CHECK: test_vsri_n_u32
+  return vsri_n_u32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v2i32
+  // CHECK_CODEGEN: sri.2s  v0, v1, #1
+}
+
+uint64x1_t test_vsri_n_u64(uint64x1_t a1, uint64x1_t a2) {
+  // CHECK: test_vsri_n_u64
+  return vsri_n_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v1i64
+  // CHECK_CODEGEN: sri     d0, d1, #1
+}
+
+poly8x8_t test_vsri_n_p8(poly8x8_t a1, poly8x8_t a2) {
+  // CHECK: test_vsri_n_p8
+  return vsri_n_p8(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v8i8
+  // CHECK_CODEGEN: sri.8b  v0, v1, #1
+}
+
+poly16x4_t test_vsri_n_p16(poly16x4_t a1, poly16x4_t a2) {
+  // CHECK: test_vsri_n_p16
+  return vsri_n_p16(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v4i16
+  // CHECK_CODEGEN: sri.4h  v0, v1, #1
+}
+
+int8x16_t test_vsriq_n_s8(int8x16_t a1, int8x16_t a2) {
+  // CHECK: test_vsriq_n_s8
+  return vsriq_n_s8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v16i8
+  // CHECK_CODEGEN: sri.16b v0, v1, #3
+}
+
+int16x8_t test_vsriq_n_s16(int16x8_t a1, int16x8_t a2) {
+  // CHECK: test_vsriq_n_s16
+  return vsriq_n_s16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v8i16
+  // CHECK_CODEGEN: sri.8h  v0, v1, #3
+}
+
+int32x4_t test_vsriq_n_s32(int32x4_t a1, int32x4_t a2) {
+  // CHECK: test_vsriq_n_s32
+  return vsriq_n_s32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v4i32
+  // CHECK_CODEGEN: sri.4s  v0, v1, #1
+}
+
+int64x2_t test_vsriq_n_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vsriq_n_s64
+  return vsriq_n_s64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v2i64
+  // CHECK_CODEGEN: sri.2d  v0, v1, #1
+}
+
+uint8x16_t test_vsriq_n_u8(uint8x16_t a1, uint8x16_t a2) {
+  // CHECK: test_vsriq_n_u8
+  return vsriq_n_u8(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v16i8
+  // CHECK_CODEGEN: sri.16b v0, v1, #3
+}
+
+uint16x8_t test_vsriq_n_u16(uint16x8_t a1, uint16x8_t a2) {
+  // CHECK: test_vsriq_n_u16
+  return vsriq_n_u16(a1, a2, 3);
+  // CHECK: llvm.arm64.neon.vsri.v8i16
+  // CHECK_CODEGEN: sri.8h  v0, v1, #3
+}
+
+uint32x4_t test_vsriq_n_u32(uint32x4_t a1, uint32x4_t a2) {
+  // CHECK: test_vsriq_n_u32
+  return vsriq_n_u32(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v4i32
+  // CHECK_CODEGEN: sri.4s  v0, v1, #1
+}
+
+uint64x2_t test_vsriq_n_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vsriq_n_u64
+  return vsriq_n_u64(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v2i64
+  // CHECK_CODEGEN: sri.2d  v0, v1, #1
+}
+
+poly8x16_t test_vsriq_n_p8(poly8x16_t a1, poly8x16_t a2) {
+  // CHECK: test_vsriq_n_p8
+  return vsriq_n_p8(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v16i8
+  // CHECK_CODEGEN: sri.16b v0, v1, #1
+}
+
+poly16x8_t test_vsriq_n_p16(poly16x8_t a1, poly16x8_t a2) {
+  // CHECK: test_vsriq_n_p16
+  return vsriq_n_p16(a1, a2, 1);
+  // CHECK: llvm.arm64.neon.vsri.v8i16
+  // CHECK_CODEGEN: sri.8h  v0, v1, #1
+}
+
diff --git a/test/CodeGen/arm64_vtst.c b/test/CodeGen/arm64_vtst.c
new file mode 100644
index 0000000..c30b394
--- /dev/null
+++ b/test/CodeGen/arm64_vtst.c
@@ -0,0 +1,22 @@
+// RUN: %clang_cc1 -O1 -triple arm64-apple-ios7 -ffreestanding -S -o - -emit-llvm %s | FileCheck %s
+// Test ARM64 SIMD comparison test intrinsics
+
+#include <arm_neon.h>
+
+uint64x2_t test_vtstq_s64(int64x2_t a1, int64x2_t a2) {
+  // CHECK: test_vtstq_s64
+  return vtstq_s64(a1, a2);
+  // CHECK: [[COMMONBITS:%[A-Za-z0-9.]+]] = and <2 x i64> %a1, %a2
+  // CHECK: [[MASK:%[A-Za-z0-9.]+]] = icmp ne <2 x i64> [[COMMONBITS]], zeroinitializer
+  // CHECK: [[RES:%[A-Za-z0-9.]+]] = sext <2 x i1> [[MASK]] to <2 x i64>
+  // CHECK: ret <2 x i64> [[RES]]
+}
+
+uint64x2_t test_vtstq_u64(uint64x2_t a1, uint64x2_t a2) {
+  // CHECK: test_vtstq_u64
+  return vtstq_u64(a1, a2);
+  // CHECK: [[COMMONBITS:%[A-Za-z0-9.]+]] = and <2 x i64> %a1, %a2
+  // CHECK: [[MASK:%[A-Za-z0-9.]+]] = icmp ne <2 x i64> [[COMMONBITS]], zeroinitializer
+  // CHECK: [[RES:%[A-Za-z0-9.]+]] = sext <2 x i1> [[MASK]] to <2 x i64>
+  // CHECK: ret <2 x i64> [[RES]]
+}
diff --git a/test/CodeGen/arm_neon_intrinsics.c b/test/CodeGen/arm_neon_intrinsics.c
index 1d76e8a..9ec3451 100644
--- a/test/CodeGen/arm_neon_intrinsics.c
+++ b/test/CodeGen/arm_neon_intrinsics.c
@@ -1887,6 +1887,12 @@
   return vdup_n_p16(a);
 }
 
+// CHECK: test_vdup_n_f16
+// CHECK: vld1.16 {{{d[0-9]+\[\]}}}
+float16x4_t test_vdup_n_f16(float16_t *a) {
+  return vdup_n_f16(*a);
+}
+
 // CHECK: test_vdup_n_f32
 // CHECK: vmov 
 float32x2_t test_vdup_n_f32(float32_t a) {
@@ -1941,6 +1947,12 @@
   return vdupq_n_p16(a);
 }
 
+// CHECK: test_vdupq_n_f16
+// CHECK: vld1.16 {{{d[0-9]+\[\], d[0-9]+\[\]}}}
+float16x8_t test_vdupq_n_f16(float16_t *a) {
+  return vdupq_n_f16(*a);
+}
+
 // CHECK: test_vdupq_n_f32
 // CHECK: vmov 
 float32x4_t test_vdupq_n_f32(float32_t a) {
@@ -4868,6 +4880,12 @@
   return vmov_n_p16(a);
 }
 
+// CHECK: test_vmov_n_f16
+// CHECK: vld1.16 {{{d[0-9]+\[\]}}}
+float16x4_t test_vmov_n_f16(float16_t *a) {
+  return vmov_n_f16(*a);
+}
+
 // CHECK: test_vmov_n_f32
 // CHECK: vmov 
 float32x2_t test_vmov_n_f32(float32_t a) {
@@ -4922,6 +4940,12 @@
   return vmovq_n_p16(a);
 }
 
+// CHECK: test_vmovq_n_f16
+// CHECK: vld1.16 {{{d[0-9]+\[\], d[0-9]+\[\]}}}
+float16x8_t test_vmovq_n_f16(float16_t *a) {
+  return vmovq_n_f16(*a);
+}
+
 // CHECK: test_vmovq_n_f32
 // CHECK: vmov 
 float32x4_t test_vmovq_n_f32(float32_t a) {
diff --git a/test/CodeGen/asm_arm64.c b/test/CodeGen/asm_arm64.c
new file mode 100644
index 0000000..0eaa347
--- /dev/null
+++ b/test/CodeGen/asm_arm64.c
@@ -0,0 +1,45 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios -emit-llvm -o - %s | FileCheck %s
+
+// rdar://9167275
+
+int t1()
+{
+  int x;
+  __asm__("mov %0, 7" : "=r" (x));
+  return x;
+}
+
+long t2()
+{
+  long x;
+  __asm__("mov %0, 7" : "=r" (x));
+  return x;
+}
+
+long t3()
+{
+  long x;
+  __asm__("mov %w0, 7" : "=r" (x));
+  return x;
+}
+
+// rdar://9281206
+
+void t4(long op) {
+  long x1;
+  asm ("mov x0, %1; svc #0;" : "=r"(x1) :"r"(op),"r"(x1) :"x0" );
+}
+
+// rdar://9394290
+
+float t5(float x) {
+  __asm__("fadd %0, %0, %0" : "+w" (x));
+  return x;
+}
+
+// rdar://9865712
+void t6 (void *f, int g) {
+  // CHECK: t6
+  // CHECK: call void asm "str $1, $0", "=*Q,r"
+  asm("str %1, %0" : "=Q"(f) : "r"(g));
+}
diff --git a/test/CodeGen/atomic-arm64.c b/test/CodeGen/atomic-arm64.c
new file mode 100644
index 0000000..147b570
--- /dev/null
+++ b/test/CodeGen/atomic-arm64.c
@@ -0,0 +1,73 @@
+// RUN: %clang_cc1 %s -emit-llvm -o - -triple=arm64-apple-ios7 | FileCheck %s
+
+// Memory ordering values.
+enum {
+  memory_order_relaxed = 0,
+  memory_order_consume = 1,
+  memory_order_acquire = 2,
+  memory_order_release = 3,
+  memory_order_acq_rel = 4,
+  memory_order_seq_cst = 5
+};
+
+typedef struct { void *a, *b; } pointer_pair_t;
+typedef struct { void *a, *b, *c, *d; } pointer_quad_t;
+
+// rdar://13489679
+
+extern _Atomic(_Bool) a_bool;
+extern _Atomic(float) a_float;
+extern _Atomic(void*) a_pointer;
+extern _Atomic(pointer_pair_t) a_pointer_pair;
+extern _Atomic(pointer_quad_t) a_pointer_quad;
+
+// CHECK:    define void @test0()
+// CHECK:      [[TEMP:%.*]] = alloca i8, align 1
+// CHECK-NEXT: store i8 1, i8* [[TEMP]]
+// CHECK-NEXT: [[T0:%.*]] = load i8* [[TEMP]], align 1
+// CHECK-NEXT: store atomic i8 [[T0]], i8* @a_bool seq_cst, align 1
+void test0() {
+  __c11_atomic_store(&a_bool, 1, memory_order_seq_cst);
+}
+
+// CHECK:    define void @test1()
+// CHECK:      [[TEMP:%.*]] = alloca float, align 4
+// CHECK-NEXT: store float 3.000000e+00, float* [[TEMP]]
+// CHECK-NEXT: [[T0:%.*]] = bitcast float* [[TEMP]] to i32*
+// CHECK-NEXT: [[T1:%.*]] = load i32* [[T0]], align 4
+// CHECK-NEXT: store atomic i32 [[T1]], i32* bitcast (float* @a_float to i32*) seq_cst, align 4
+void test1() {
+  __c11_atomic_store(&a_float, 3, memory_order_seq_cst);
+}
+
+// CHECK:    define void @test2()
+// CHECK:      [[TEMP:%.*]] = alloca i8*, align 8
+// CHECK-NEXT: store i8* @a_bool, i8** [[TEMP]]
+// CHECK-NEXT: [[T0:%.*]] = bitcast i8** [[TEMP]] to i64*
+// CHECK-NEXT: [[T1:%.*]] = load i64* [[T0]], align 8
+// CHECK-NEXT: store atomic i64 [[T1]], i64* bitcast (i8** @a_pointer to i64*) seq_cst, align 8
+void test2() {
+  __c11_atomic_store(&a_pointer, &a_bool, memory_order_seq_cst);
+}
+
+// CHECK:    define void @test3(
+// CHECK:      [[PAIR:%.*]] = alloca [[PAIR_T:%.*]], align 8
+// CHECK-NEXT: [[TEMP:%.*]] = alloca [[PAIR_T]], align 8
+// CHECK:      llvm.memcpy
+// CHECK-NEXT: [[T0:%.*]] = bitcast [[PAIR_T]]* [[TEMP]] to i128*
+// CHECK-NEXT: [[T1:%.*]] = load i128* [[T0]], align 16
+// CHECK-NEXT: store atomic i128 [[T1]], i128* bitcast ([[PAIR_T]]* @a_pointer_pair to i128*) seq_cst, align 16
+void test3(pointer_pair_t pair) {
+  __c11_atomic_store(&a_pointer_pair, pair, memory_order_seq_cst);
+}
+
+// CHECK:    define void @test4([[QUAD_T:%.*]]*
+// CHECK:      [[TEMP:%.*]] = alloca [[QUAD_T:%.*]], align 8
+// CHECK-NEXT: [[T0:%.*]] = bitcast [[QUAD_T]]* [[TEMP]] to i8*
+// CHECK-NEXT: [[T1:%.*]] = bitcast [[QUAD_T]]* {{%.*}} to i8*
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T0]], i8* [[T1]], i64 32, i32 8, i1 false)
+// CHECK-NEXT: [[T0:%.*]] = bitcast [[QUAD_T]]* [[TEMP]] to i8*
+// CHECK-NEXT: call void @__atomic_store(i64 32, i8* bitcast ([[QUAD_T]]* @a_pointer_quad to i8*), i8* [[T0]], i32 5)
+void test4(pointer_quad_t quad) {
+  __c11_atomic_store(&a_pointer_quad, quad, memory_order_seq_cst);
+}
diff --git a/test/CodeGen/atomic-ops-libcall.c b/test/CodeGen/atomic-ops-libcall.c
new file mode 100644
index 0000000..2a2ff5e
--- /dev/null
+++ b/test/CodeGen/atomic-ops-libcall.c
@@ -0,0 +1,37 @@
+// RUN: %clang_cc1 < %s -triple armv5e-none-linux-gnueabi -emit-llvm -O1 | FileCheck %s
+
+enum memory_order {
+  memory_order_relaxed, memory_order_consume, memory_order_acquire,
+  memory_order_release, memory_order_acq_rel, memory_order_seq_cst
+};
+
+int *test_c11_atomic_fetch_add_int_ptr(_Atomic(int *) *p) {
+  // CHECK: test_c11_atomic_fetch_add_int_ptr
+  // CHECK: {{%[^ ]*}} = tail call i32* @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 12, i32 5)
+  return __c11_atomic_fetch_add(p, 3, memory_order_seq_cst);
+}
+
+int *test_c11_atomic_fetch_sub_int_ptr(_Atomic(int *) *p) {
+  // CHECK: test_c11_atomic_fetch_sub_int_ptr
+  // CHECK: {{%[^ ]*}} = tail call i32* @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 20, i32 5)
+  return __c11_atomic_fetch_sub(p, 5, memory_order_seq_cst);
+}
+
+int test_c11_atomic_fetch_add_int(_Atomic(int) *p) {
+  // CHECK: test_c11_atomic_fetch_add_int
+  // CHECK: {{%[^ ]*}} = tail call i32 bitcast (i32* (i8*, i32, i32)* @__atomic_fetch_add_4 to i32 (i8*, i32, i32)*)(i8* {{%[0-9]+}}, i32 3, i32 5)
+  return __c11_atomic_fetch_add(p, 3, memory_order_seq_cst);
+}
+
+int test_c11_atomic_fetch_sub_int(_Atomic(int) *p) {
+  // CHECK: test_c11_atomic_fetch_sub_int
+  // CHECK: {{%[^ ]*}} = tail call i32 bitcast (i32* (i8*, i32, i32)* @__atomic_fetch_sub_4 to i32 (i8*, i32, i32)*)(i8* {{%[0-9]+}}, i32 5, i32 5)
+  return __c11_atomic_fetch_sub(p, 5, memory_order_seq_cst);
+}
+
+int *fp2a(int **p) {
+  // CHECK: @fp2a
+  // CHECK: {{%[^ ]*}} = tail call i32* @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 4, i32 0)
+  // Note, the GNU builtins do not multiply by sizeof(T)!
+  return __atomic_fetch_sub(p, 4, memory_order_relaxed);
+}
diff --git a/test/CodeGen/atomic-ops.c b/test/CodeGen/atomic-ops.c
index 830f21a..2517b67 100644
--- a/test/CodeGen/atomic-ops.c
+++ b/test/CodeGen/atomic-ops.c
@@ -91,14 +91,20 @@
 
 _Bool fi4(_Atomic(int) *i) {
   // CHECK: @fi4
-  // CHECK: cmpxchg i32*
+  // CHECK: [[OLD:%[.0-9A-Z_a-z]+]] = cmpxchg i32* [[PTR:%[.0-9A-Z_a-z]+]], i32 [[EXPECTED:%[.0-9A-Z_a-z]+]], i32 [[DESIRED:%[.0-9A-Z_a-z]+]]
+  // CHECK: [[CMP:%[.0-9A-Z_a-z]+]] = icmp eq i32 [[OLD]], [[EXPECTED]]
+  // CHECK: br i1 [[CMP]], label %[[STORE_EXPECTED:[.0-9A-Z_a-z]+]], label %[[CONTINUE:[.0-9A-Z_a-z]+]]
+  // CHECK: store i32 [[OLD]]
   int cmp = 0;
   return __c11_atomic_compare_exchange_strong(i, &cmp, 1, memory_order_acquire, memory_order_acquire);
 }
 
 _Bool fi4a(int *i) {
   // CHECK: @fi4
-  // CHECK: cmpxchg i32*
+  // CHECK: [[OLD:%[.0-9A-Z_a-z]+]] = cmpxchg i32* [[PTR:%[.0-9A-Z_a-z]+]], i32 [[EXPECTED:%[.0-9A-Z_a-z]+]], i32 [[DESIRED:%[.0-9A-Z_a-z]+]]
+  // CHECK: [[CMP:%[.0-9A-Z_a-z]+]] = icmp eq i32 [[OLD]], [[EXPECTED]]
+  // CHECK: br i1 [[CMP]], label %[[STORE_EXPECTED:[.0-9A-Z_a-z]+]], label %[[CONTINUE:[.0-9A-Z_a-z]+]]
+  // CHECK: store i32 [[OLD]]
   int cmp = 0;
   int desired = 1;
   return __atomic_compare_exchange(i, &cmp, &desired, 0, memory_order_acquire, memory_order_acquire);
@@ -106,7 +112,10 @@
 
 _Bool fi4b(int *i) {
   // CHECK: @fi4
-  // CHECK: cmpxchg i32*
+  // CHECK: [[OLD:%[.0-9A-Z_a-z]+]] = cmpxchg i32* [[PTR:%[.0-9A-Z_a-z]+]], i32 [[EXPECTED:%[.0-9A-Z_a-z]+]], i32 [[DESIRED:%[.0-9A-Z_a-z]+]]
+  // CHECK: [[CMP:%[.0-9A-Z_a-z]+]] = icmp eq i32 [[OLD]], [[EXPECTED]]
+  // CHECK: br i1 [[CMP]], label %[[STORE_EXPECTED:[.0-9A-Z_a-z]+]], label %[[CONTINUE:[.0-9A-Z_a-z]+]]
+  // CHECK: store i32 [[OLD]]
   int cmp = 0;
   return __atomic_compare_exchange_n(i, &cmp, 1, 1, memory_order_acquire, memory_order_acquire);
 }
@@ -306,13 +315,92 @@
   // CHECK: }
 }
 
-// CHECK: @invalid_atomic
-void invalid_atomic(_Atomic(int) *i) {
-  __c11_atomic_store(i, 1, memory_order_consume);
-  __c11_atomic_store(i, 1, memory_order_acquire);
-  __c11_atomic_store(i, 1, memory_order_acq_rel);
-  __c11_atomic_load(i, memory_order_release);
-  __c11_atomic_load(i, memory_order_acq_rel);
+// CHECK-LABEL: @failureOrder
+void failureOrder(_Atomic(int) *ptr, int *ptr2) {
+  __c11_atomic_compare_exchange_strong(ptr, ptr2, 43, memory_order_acquire, memory_order_relaxed);
+  // CHECK: cmpxchg i32* {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z_.]+}} acquire monotonic
+
+  __c11_atomic_compare_exchange_weak(ptr, ptr2, 43, memory_order_seq_cst, memory_order_acquire);
+  // CHECK: cmpxchg i32* {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z_.]+}} seq_cst acquire
+
+  // Unknown ordering: conservatively pick strongest valid option (for now!).
+  __atomic_compare_exchange(ptr2, ptr2, ptr2, 0, memory_order_acq_rel, *ptr2);
+  // CHECK: cmpxchg i32* {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z_.]+}} acq_rel acquire
+
+  // Undefined behaviour: don't really care what that last ordering is so leave
+  // it out:
+  __atomic_compare_exchange_n(ptr2, ptr2, 43, 1, memory_order_seq_cst, 42);
+  // CHECK: cmpxchg i32* {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z._]+}}, i32 {{%[0-9A-Za-z_.]+}} seq_cst
+}
+
+// CHECK-LABEL: @generalFailureOrder
+void generalFailureOrder(_Atomic(int) *ptr, int *ptr2, int success, int fail) {
+  __c11_atomic_compare_exchange_strong(ptr, ptr2, 42, success, fail);
+  // CHECK: switch i32 {{.*}}, label %[[MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: i32 1, label %[[ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 2, label %[[ACQUIRE]]
+  // CHECK-NEXT: i32 3, label %[[RELEASE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 4, label %[[ACQREL:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 5, label %[[SEQCST:[0-9a-zA-Z._]+]]
+
+  // CHECK: [[MONOTONIC]]
+  // CHECK: switch {{.*}}, label %[[MONOTONIC_MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: ]
+
+  // CHECK: [[ACQUIRE]]
+  // CHECK: switch {{.*}}, label %[[ACQUIRE_MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: i32 1, label %[[ACQUIRE_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 2, label %[[ACQUIRE_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: ]
+
+  // CHECK: [[RELEASE]]
+  // CHECK: switch {{.*}}, label %[[RELEASE_MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: ]
+
+  // CHECK: [[ACQREL]]
+  // CHECK: switch {{.*}}, label %[[ACQREL_MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: i32 1, label %[[ACQREL_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 2, label %[[ACQREL_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: ]
+
+  // CHECK: [[SEQCST]]
+  // CHECK: switch {{.*}}, label %[[SEQCST_MONOTONIC:[0-9a-zA-Z._]+]] [
+  // CHECK-NEXT: i32 1, label %[[SEQCST_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 2, label %[[SEQCST_ACQUIRE:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: i32 5, label %[[SEQCST_SEQCST:[0-9a-zA-Z._]+]]
+  // CHECK-NEXT: ]
+
+  // CHECK: [[MONOTONIC_MONOTONIC]]
+  // CHECK: cmpxchg {{.*}} monotonic monotonic
+  // CHECK: br
+
+  // CHECK: [[ACQUIRE_MONOTONIC]]
+  // CHECK: cmpxchg {{.*}} acquire monotonic
+  // CHECK: br
+
+  // CHECK: [[ACQUIRE_ACQUIRE]]
+  // CHECK: cmpxchg {{.*}} acquire acquire
+  // CHECK: br
+
+  // CHECK: [[ACQREL_MONOTONIC]]
+  // CHECK: cmpxchg {{.*}} acq_rel monotonic
+  // CHECK: br
+
+  // CHECK: [[ACQREL_ACQUIRE]]
+  // CHECK: cmpxchg {{.*}} acq_rel acquire
+  // CHECK: br
+
+  // CHECK: [[SEQCST_MONOTONIC]]
+  // CHECK: cmpxchg {{.*}} seq_cst monotonic
+  // CHECK: br
+
+  // CHECK: [[SEQCST_ACQUIRE]]
+  // CHECK: cmpxchg {{.*}} seq_cst acquire
+  // CHECK: br
+
+  // CHECK: [[SEQCST_SEQCST]]
+  // CHECK: cmpxchg {{.*}} seq_cst seq_cst
+  // CHECK: br
 }
 
 #endif
diff --git a/test/CodeGen/attr-optnone.c b/test/CodeGen/attr-optnone.c
new file mode 100644
index 0000000..7c4873f
--- /dev/null
+++ b/test/CodeGen/attr-optnone.c
@@ -0,0 +1,25 @@
+// RUN: %clang_cc1 -emit-llvm < %s > %t
+// RUN: FileCheck %s --check-prefix=PRESENT < %t
+// RUN: FileCheck %s --check-prefix=ABSENT  < %t
+
+__attribute__((always_inline))
+int test2() { return 0; }
+// PRESENT-DAG: @test2{{.*}}[[ATTR2:#[0-9]+]]
+
+__attribute__((optnone)) __attribute__((minsize))
+int test3() { return 0; }
+// PRESENT-DAG: @test3{{.*}}[[ATTR3:#[0-9]+]]
+
+__attribute__((optnone)) __attribute__((cold))
+int test4() { return test2(); }
+// PRESENT-DAG: @test4{{.*}}[[ATTR4:#[0-9]+]]
+// Also check that test2 is inlined into test4 (always_inline still works).
+// PRESENT-DAG-NOT: call i32 @test2
+
+// Check for both noinline and optnone on each optnone function.
+// PRESENT-DAG: attributes [[ATTR3]] = { {{.*}}noinline{{.*}}optnone{{.*}} }
+// PRESENT-DAG: attributes [[ATTR4]] = { {{.*}}noinline{{.*}}optnone{{.*}} }
+
+// Check that no 'optsize' or 'minsize' attributes appear.
+// ABSENT-NOT: optsize
+// ABSENT-NOT: minsize
diff --git a/test/CodeGen/big-atomic-ops.c b/test/CodeGen/big-atomic-ops.c
index b09aede..01a2801 100644
--- a/test/CodeGen/big-atomic-ops.c
+++ b/test/CodeGen/big-atomic-ops.c
@@ -311,13 +311,4 @@
   // CHECK: }
 }
 
-// CHECK: @invalid_atomic
-void invalid_atomic(_Atomic(int) *i) {
-  __c11_atomic_store(i, 1, memory_order_consume);
-  __c11_atomic_store(i, 1, memory_order_acquire);
-  __c11_atomic_store(i, 1, memory_order_acq_rel);
-  __c11_atomic_load(i, memory_order_release);
-  __c11_atomic_load(i, memory_order_acq_rel);
-}
-
 #endif
diff --git a/test/CodeGen/bitfield-2.c b/test/CodeGen/bitfield-2.c
index 58b17f1..c5154fc 100644
--- a/test/CodeGen/bitfield-2.c
+++ b/test/CodeGen/bitfield-2.c
@@ -11,7 +11,7 @@
 // CHECK-RECORD: *** Dumping IRgen Record Layout
 // CHECK-RECORD: Record: RecordDecl{{.*}}s0
 // CHECK-RECORD: Layout: <CGRecordLayout
-// CHECK-RECORD:   LLVMType:%struct.s0 = type <{ [3 x i8] }>
+// CHECK-RECORD:   LLVMType:%struct.s0 = type { [3 x i8] }
 // CHECK-RECORD:   IsZeroInitializable:1
 // CHECK-RECORD:   BitFields:[
 // CHECK-RECORD:     <CGBitFieldInfo Offset:0 Size:24 IsSigned:1 StorageSize:24 StorageAlignment:1>
@@ -51,7 +51,7 @@
 // CHECK-RECORD: *** Dumping IRgen Record Layout
 // CHECK-RECORD: Record: RecordDecl{{.*}}s1
 // CHECK-RECORD: Layout: <CGRecordLayout
-// CHECK-RECORD:   LLVMType:%struct.s1 = type <{ [3 x i8] }>
+// CHECK-RECORD:   LLVMType:%struct.s1 = type { [3 x i8] }
 // CHECK-RECORD:   IsZeroInitializable:1
 // CHECK-RECORD:   BitFields:[
 // CHECK-RECORD:     <CGBitFieldInfo Offset:0 Size:10 IsSigned:1 StorageSize:24 StorageAlignment:1>
@@ -99,7 +99,7 @@
 // CHECK-RECORD: *** Dumping IRgen Record Layout
 // CHECK-RECORD: Record: RecordDecl{{.*}}u2
 // CHECK-RECORD: Layout: <CGRecordLayout
-// CHECK-RECORD:   LLVMType:%union.u2 = type <{ i8 }>
+// CHECK-RECORD:   LLVMType:%union.u2 = type { i8 }
 // CHECK-RECORD:   IsZeroInitializable:1
 // CHECK-RECORD:   BitFields:[
 // CHECK-RECORD:     <CGBitFieldInfo Offset:0 Size:3 IsSigned:0 StorageSize:8 StorageAlignment:1>
@@ -271,7 +271,7 @@
 // CHECK-RECORD: *** Dumping IRgen Record Layout
 // CHECK-RECORD: Record: RecordDecl{{.*}}s7
 // CHECK-RECORD: Layout: <CGRecordLayout
-// CHECK-RECORD:   LLVMType:%struct.s7 = type { i32, i32, i32, i8, [3 x i8], [4 x i8], [12 x i8] }
+// CHECK-RECORD:   LLVMType:%struct.s7 = type { i32, i32, i32, i8, i32, [12 x i8] }
 // CHECK-RECORD:   IsZeroInitializable:1
 // CHECK-RECORD:   BitFields:[
 // CHECK-RECORD:     <CGBitFieldInfo Offset:0 Size:5 IsSigned:1 StorageSize:8 StorageAlignment:4>
diff --git a/test/CodeGen/blockstret.c b/test/CodeGen/blockstret.c
index d5dae6f..edd1218 100644
--- a/test/CodeGen/blockstret.c
+++ b/test/CodeGen/blockstret.c
@@ -1,13 +1,20 @@
 // RUN: %clang_cc1 -fblocks -triple x86_64-apple-darwin9 %s -emit-llvm -o - | FileCheck %s -check-prefix=X64
 // RUN: %clang_cc1 -fblocks -triple i686-apple-darwin9 %s -emit-llvm -o - | FileCheck %s -check-prefix=X32
+// RUN: %clang_cc1 -fblocks -triple arm64-apple-darwin %s -emit-llvm -o - | FileCheck %s -check-prefix=ARM64
 
 // X64:   internal constant {{.*}} { i8** @_NSConcreteGlobalBlock, i32 1879048192
-// X64:     store i32 1610612736, i32* %want
+// X64:   store i32 1610612736, i32* %want
 
 // X32:   @_NSConcreteGlobalBlock, i32 1879048192, i32 0,
 // X32:   store i32 1610612736, i32* %want
 
 // rdar://7677537
+
+// ARM64: @_NSConcreteGlobalBlock, i32 1342177280, i32 0,
+// ARM64: store i32 1610612736, i32* %want
+
+// rdar://9757126
+
 int printf(const char *, ...);
 void *malloc(__SIZE_TYPE__ size);
 
diff --git a/test/CodeGen/bool-convert.c b/test/CodeGen/bool-convert.c
index 8bde837..344fb6b 100644
--- a/test/CodeGen/bool-convert.c
+++ b/test/CodeGen/bool-convert.c
@@ -1,10 +1,24 @@
-// RUN: %clang_cc1 -emit-llvm < %s | grep i1 | count 1
+// RUN: %clang_cc1 -triple i686-pc-linux -emit-llvm < %s | FileCheck %s
 // All of these should uses the memory representation of _Bool
+
+// CHECK-LABEL: %struct.teststruct1 = type { i8, i8 }
+// CHECK-LABEL: @test1 = common global %struct.teststruct1
 struct teststruct1 {_Bool a, b;} test1;
+
+// CHECK-LABEL: @test2 = common global i8* null
 _Bool* test2;
+
+// CHECK-LABEL: @test3 = common global [10 x i8]
 _Bool test3[10];
+
+// CHECK-LABEL: @test4 = common global [0 x i8]* null
 _Bool (*test4)[];
+
+// CHECK-LABEL: define void @f(i32 %x)
 void f(int x) {
+  // CHECK: alloca i8, align 1
   _Bool test5;
+
+  // CHECK: alloca i8, i32 %{{.*}}, align 1
   _Bool test6[x];
 }
diff --git a/test/CodeGen/bool-init.c b/test/CodeGen/bool-init.c
index 1a8f127..09b4a87 100644
--- a/test/CodeGen/bool-init.c
+++ b/test/CodeGen/bool-init.c
@@ -1,4 +1,5 @@
-// RUN: %clang_cc1 -emit-llvm < %s | grep i1 | count 1
+// RUN: %clang_cc1 -triple x86_64-pc-linux -emit-llvm < %s | FileCheck %s
 
 // Check that the type of this global isn't i1
+// CHECK: @test = global i8 1
 _Bool test = &test;
diff --git a/test/CodeGen/bool_test.c b/test/CodeGen/bool_test.c
index a4aa669..c836b98 100644
--- a/test/CodeGen/bool_test.c
+++ b/test/CodeGen/bool_test.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc-apple-macosx10.4.0 -emit-llvm -o - %s -O2 -disable-llvm-optzns | FileCheck %s
 
 int boolsize = sizeof(_Bool);
diff --git a/test/CodeGen/branch-on-bool.c b/test/CodeGen/branch-on-bool.c
index 78dae1b..98a3845 100644
--- a/test/CodeGen/branch-on-bool.c
+++ b/test/CodeGen/branch-on-bool.c
@@ -12,11 +12,3 @@
   else
     bar();
 }
-
-void fold_for(int a, int b) {
-  // CHECK: define {{.*}} @fold_for(
-  // CHECK-NOT: = phi
-  // CHECK: }
-  for (int i = 0; a && i < b; ++i) foo();
-  for (int i = 0; a || i < b; ++i) bar();
-}
diff --git a/test/CodeGen/builtin-ms-noop.cpp b/test/CodeGen/builtin-ms-noop.cpp
index 7c5068d..b579e2d 100644
--- a/test/CodeGen/builtin-ms-noop.cpp
+++ b/test/CodeGen/builtin-ms-noop.cpp
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -fms-extensions -triple i686-pc-win32 -emit-llvm %s -o - | FileCheck %s
+// RUN: %clang_cc1 -fms-extensions -triple %itanium_abi_triple -emit-llvm %s -o - | FileCheck %s
 
 class A {
  public:
@@ -11,4 +11,3 @@
 // CHECK: ret void
   __noop(A());
 };
-
diff --git a/test/CodeGen/builtins-arm-exclusive.c b/test/CodeGen/builtins-arm-exclusive.c
index 7eccb9e..3d6cc54 100644
--- a/test/CodeGen/builtins-arm-exclusive.c
+++ b/test/CodeGen/builtins-arm-exclusive.c
@@ -1,5 +1,6 @@
 // REQUIRES: arm-registered-target
 // RUN: %clang_cc1 -Wall -Werror -triple thumbv7-linux-gnueabi -fno-signed-char -O3 -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -Wall -Werror -triple arm64-apple-ios7.0 -O3 -emit-llvm -o - %s | FileCheck %s --check-prefix=CHECK-ARM64
 
 // Make sure the canonical use works before going into smaller details:
 int atomic_inc(int *addr) {
@@ -12,46 +13,80 @@
   return OldVal;
 }
 
-// CHECK: @atomic_inc
+// CHECK-LABEL: @atomic_inc
 // CHECK:   [[OLDVAL:%.*]] = tail call i32 @llvm.arm.ldrex.p0i32(i32* %addr)
 // CHECK:   [[INC:%.*]] = add nsw i32 [[OLDVAL]], 1
 // CHECK:   [[FAILURE:%.*]] = tail call i32 @llvm.arm.strex.p0i32(i32 [[INC]], i32* %addr)
 // CHECK:   [[TST:%.*]] = icmp eq i32 [[FAILURE]], 0
 // CHECK:   br i1 [[TST]], label {{%[a-zA-Z0-9.]+}}, label {{%[a-zA-Z0-9.]+}}
 
+// CHECK-ARM64-LABEL: @atomic_inc
+// CHECK-ARM64:   [[OLDVAL:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i32(i32* %addr)
+// CHECK-ARM64:   [[INC:%.*]] = add i64 [[OLDVAL]], 1
+// CHECK-ARM64:   [[TRUNC:%.*]] = and i64 [[INC]], 4294967295
+// CHECK-ARM64:   [[FAILURE:%.*]] = tail call i32 @llvm.arm64.stxr.p0i32(i64 [[TRUNC]], i32* %addr)
+// CHECK-ARM64:   [[TST:%.*]] = icmp eq i32 [[FAILURE]], 0
+// CHECK-ARM64:   br i1 [[TST]], label {{%[a-zA-Z0-9.]+}}, label {{%[a-zA-Z0-9.]+}}
+
 struct Simple {
   char a, b;
 };
 
 int test_ldrex(char *addr, long long *addr64, float *addrfloat) {
-// CHECK: @test_ldrex
+// CHECK-LABEL: @test_ldrex
+// CHECK-ARM64-LABEL: @test_ldrex
   int sum = 0;
   sum += __builtin_arm_ldrex(addr);
 // CHECK: [[INTRES:%.*]] = tail call i32 @llvm.arm.ldrex.p0i8(i8* %addr)
 // CHECK: and i32 [[INTRES]], 255
 
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i8(i8* %addr)
+// CHECK-ARM64: [[TRUNCRES:%.*]] = trunc i64 [[INTRES]] to i32
+// CHECK-ARM64: [[SEXTTMP:%.*]] = shl i32 [[TRUNCRES]], 24
+// CHECK-ARM64: ashr exact i32 [[SEXTTMP]], 24
+
   sum += __builtin_arm_ldrex((short *)addr);
 // CHECK: [[ADDR16:%.*]] = bitcast i8* %addr to i16*
 // CHECK: [[INTRES:%.*]] = tail call i32 @llvm.arm.ldrex.p0i16(i16* [[ADDR16]])
 // CHECK: [[TMPSEXT:%.*]] = shl i32 [[INTRES]], 16
 // CHECK: ashr exact i32 [[TMPSEXT]], 16
 
+// CHECK-ARM64: [[ADDR16:%.*]] = bitcast i8* %addr to i16*
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i16(i16* [[ADDR16]])
+// CHECK-ARM64: [[TRUNCRES:%.*]] = trunc i64 [[INTRES]] to i32
+// CHECK-ARM64: [[TMPSEXT:%.*]] = shl i32 [[TRUNCRES]], 16
+// CHECK-ARM64: ashr exact i32 [[TMPSEXT]], 16
+
   sum += __builtin_arm_ldrex((int *)addr);
 // CHECK: [[ADDR32:%.*]] = bitcast i8* %addr to i32*
 // CHECK:  call i32 @llvm.arm.ldrex.p0i32(i32* [[ADDR32]])
 
+// CHECK-ARM64: [[ADDR32:%.*]] = bitcast i8* %addr to i32*
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i32(i32* [[ADDR32]])
+// CHECK-ARM64: trunc i64 [[INTRES]] to i32
+
   sum += __builtin_arm_ldrex((long long *)addr);
 // CHECK: call { i32, i32 } @llvm.arm.ldrexd(i8* %addr)
 
+// CHECK-ARM64: [[ADDR64:%.*]] = bitcast i8* %addr to i64*
+// CHECK-ARM64: call i64 @llvm.arm64.ldxr.p0i64(i64* [[ADDR64]])
+
   sum += __builtin_arm_ldrex(addr64);
 // CHECK: [[ADDR64_AS8:%.*]] = bitcast i64* %addr64 to i8*
 // CHECK: call { i32, i32 } @llvm.arm.ldrexd(i8* [[ADDR64_AS8]])
 
+// CHECK-ARM64: call i64 @llvm.arm64.ldxr.p0i64(i64* %addr64)
+
   sum += __builtin_arm_ldrex(addrfloat);
 // CHECK: [[INTADDR:%.*]] = bitcast float* %addrfloat to i32*
 // CHECK: [[INTRES:%.*]] = tail call i32 @llvm.arm.ldrex.p0i32(i32* [[INTADDR]])
 // CHECK: bitcast i32 [[INTRES]] to float
 
+// CHECK-ARM64: [[INTADDR:%.*]] = bitcast float* %addrfloat to i32*
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i32(i32* [[INTADDR]])
+// CHECK-ARM64: [[TRUNCRES:%.*]] = trunc i64 [[INTRES]] to i32
+// CHECK-ARM64: bitcast i32 [[TRUNCRES]] to float
+
   sum += __builtin_arm_ldrex((double *)addr);
 // CHECK: [[STRUCTRES:%.*]] = tail call { i32, i32 } @llvm.arm.ldrexd(i8* %addr)
 // CHECK: [[RESHI:%.*]] = extractvalue { i32, i32 } [[STRUCTRES]], 1
@@ -62,51 +97,110 @@
 // CHECK: [[INTRES:%.*]] = or i64 [[RESHIHI]], [[RESLO64]]
 // CHECK: bitcast i64 [[INTRES]] to double
 
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i64(i64* [[ADDR64]])
+// CHECK-ARM64: bitcast i64 [[INTRES]] to double
+
   sum += *__builtin_arm_ldrex((int **)addr);
 // CHECK: [[INTRES:%.*]] = tail call i32 @llvm.arm.ldrex.p0i32(i32* [[ADDR32]])
 // CHECK: inttoptr i32 [[INTRES]] to i32*
 
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i64(i64* [[ADDR64]])
+// CHECK-ARM64: inttoptr i64 [[INTRES]] to i32*
+
   sum += __builtin_arm_ldrex((struct Simple **)addr)->a;
 // CHECK: [[INTRES:%.*]] = tail call i32 @llvm.arm.ldrex.p0i32(i32* [[ADDR32]])
 // CHECK: inttoptr i32 [[INTRES]] to %struct.Simple*
 
+// CHECK-ARM64: [[INTRES:%.*]] = tail call i64 @llvm.arm64.ldxr.p0i64(i64* [[ADDR64]])
+// CHECK-ARM64: inttoptr i64 [[INTRES]] to %struct.Simple*
   return sum;
 }
 
 int test_strex(char *addr) {
-// CHECK: @test_strex
+// CHECK-LABEL: @test_strex
+// CHECK-ARM64-LABEL: @test_strex
   int res = 0;
   struct Simple var = {0};
   res |= __builtin_arm_strex(4, addr);
 // CHECK: call i32 @llvm.arm.strex.p0i8(i32 4, i8* %addr)
 
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i8(i64 4, i8* %addr)
+
   res |= __builtin_arm_strex(42, (short *)addr);
 // CHECK: [[ADDR16:%.*]] = bitcast i8* %addr to i16*
 // CHECK:  call i32 @llvm.arm.strex.p0i16(i32 42, i16* [[ADDR16]])
 
+// CHECK-ARM64: [[ADDR16:%.*]] = bitcast i8* %addr to i16*
+// CHECK-ARM64:  call i32 @llvm.arm64.stxr.p0i16(i64 42, i16* [[ADDR16]])
+
   res |= __builtin_arm_strex(42, (int *)addr);
 // CHECK: [[ADDR32:%.*]] = bitcast i8* %addr to i32*
 // CHECK: call i32 @llvm.arm.strex.p0i32(i32 42, i32* [[ADDR32]])
 
+// CHECK-ARM64: [[ADDR32:%.*]] = bitcast i8* %addr to i32*
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i32(i64 42, i32* [[ADDR32]])
+
   res |= __builtin_arm_strex(42, (long long *)addr);
 // CHECK: call i32 @llvm.arm.strexd(i32 42, i32 0, i8* %addr)
 
+// CHECK-ARM64: [[ADDR64:%.*]] = bitcast i8* %addr to i64*
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i64(i64 42, i64* [[ADDR64]])
+
   res |= __builtin_arm_strex(2.71828f, (float *)addr);
 // CHECK: call i32 @llvm.arm.strex.p0i32(i32 1076754509, i32* [[ADDR32]])
 
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i32(i64 1076754509, i32* [[ADDR32]])
+
   res |= __builtin_arm_strex(3.14159, (double *)addr);
 // CHECK: call i32 @llvm.arm.strexd(i32 -266631570, i32 1074340345, i8* %addr)
 
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i64(i64 4614256650576692846, i64* [[ADDR64]])
+
   res |= __builtin_arm_strex(&var, (struct Simple **)addr);
 // CHECK: [[INTVAL:%.*]] = ptrtoint i16* %var to i32
 // CHECK: call i32 @llvm.arm.strex.p0i32(i32 [[INTVAL]], i32* [[ADDR32]])
 
+// CHECK-ARM64: [[INTVAL:%.*]] = ptrtoint i16* %var to i64
+// CHECK-ARM64: call i32 @llvm.arm64.stxr.p0i64(i64 [[INTVAL]], i64* [[ADDR64]])
+
   return res;
 }
 
 void test_clrex() {
-// CHECK: @test_clrex
+// CHECK-LABEL: @test_clrex
+// CHECK-ARM64-LABEL: @test_clrex
 
   __builtin_arm_clrex();
 // CHECK: call void @llvm.arm.clrex()
+// CHECK-ARM64: call void @llvm.arm64.clrex()
 }
+
+#ifdef __aarch64__
+// 128-bit tests
+
+__int128 test_ldrex_128(__int128 *addr) {
+// CHECK-ARM64-LABEL: @test_ldrex_128
+
+  return __builtin_arm_ldrex(addr);
+// CHECK-ARM64: [[ADDR8:%.*]] = bitcast i128* %addr to i8*
+// CHECK-ARM64: [[STRUCTRES:%.*]] = tail call { i64, i64 } @llvm.arm64.ldxp(i8* [[ADDR8]])
+// CHECK-ARM64: [[RESHI:%.*]] = extractvalue { i64, i64 } [[STRUCTRES]], 1
+// CHECK-ARM64: [[RESLO:%.*]] = extractvalue { i64, i64 } [[STRUCTRES]], 0
+// CHECK-ARM64: [[RESHI64:%.*]] = zext i64 [[RESHI]] to i128
+// CHECK-ARM64: [[RESLO64:%.*]] = zext i64 [[RESLO]] to i128
+// CHECK-ARM64: [[RESHIHI:%.*]] = shl nuw i128 [[RESHI64]], 64
+// CHECK-ARM64: [[INTRES:%.*]] = or i128 [[RESHIHI]], [[RESLO64]]
+// CHECK-ARM64: ret i128 [[INTRES]]
+}
+
+int test_strex_128(__int128 *addr, __int128 val) {
+// CHECK-ARM64-LABEL: @test_strex_128
+
+  return __builtin_arm_strex(val, addr);
+// CHECK-ARM64: [[VALLO:%.*]] = trunc i128 %val to i64
+// CHECK-ARM64: [[VALHI128:%.*]] = lshr i128 %val, 64
+// CHECK-ARM64: [[VALHI:%.*]] = trunc i128 [[VALHI128]] to i64
+// CHECK-ARM64: [[ADDR8:%.*]] = bitcast i128* %addr to i8*
+// CHECK-ARM64: [[RES:%.*]] = tail call i32 @llvm.arm64.stxp(i64 [[VALLO]], i64 [[VALHI]], i8* [[ADDR8]])
+}
+#endif
diff --git a/test/CodeGen/builtins-arm64.c b/test/CodeGen/builtins-arm64.c
new file mode 100644
index 0000000..fcda92e
--- /dev/null
+++ b/test/CodeGen/builtins-arm64.c
@@ -0,0 +1,6 @@
+// RUN: %clang_cc1 -triple arm64-apple-ios -O3 -emit-llvm -o - %s | FileCheck %s
+
+void f0(void *a, void *b) {
+	__clear_cache(a,b);
+// CHECK: call {{.*}} @__clear_cache
+}
diff --git a/test/CodeGen/builtins-mips-msa.c b/test/CodeGen/builtins-mips-msa.c
index 69cb8e2..38aea04 100644
--- a/test/CodeGen/builtins-mips-msa.c
+++ b/test/CodeGen/builtins-mips-msa.c
@@ -701,15 +701,15 @@
   v8i16_r = __builtin_msa_shf_h(v8i16_a, 3); // CHECK: call <8  x i16> @llvm.mips.shf.h(
   v4i32_r = __builtin_msa_shf_w(v4i32_a, 3); // CHECK: call <4  x i32> @llvm.mips.shf.w(
 
-  v16i8_r = __builtin_msa_sld_b(v16i8_a, 10); // CHECK: call <16 x i8>  @llvm.mips.sld.b(
-  v8i16_r = __builtin_msa_sld_h(v8i16_a, 10); // CHECK: call <8  x i16> @llvm.mips.sld.h(
-  v4i32_r = __builtin_msa_sld_w(v4i32_a, 10); // CHECK: call <4  x i32> @llvm.mips.sld.w(
-  v2i64_r = __builtin_msa_sld_d(v2i64_a, 10); // CHECK: call <2  x i64> @llvm.mips.sld.d(
+  v16i8_r = __builtin_msa_sld_b(v16i8_r, v16i8_a, 10); // CHECK: call <16 x i8>  @llvm.mips.sld.b(
+  v8i16_r = __builtin_msa_sld_h(v8i16_r, v8i16_a, 10); // CHECK: call <8  x i16> @llvm.mips.sld.h(
+  v4i32_r = __builtin_msa_sld_w(v4i32_r, v4i32_a, 10); // CHECK: call <4  x i32> @llvm.mips.sld.w(
+  v2i64_r = __builtin_msa_sld_d(v2i64_r, v2i64_a, 10); // CHECK: call <2  x i64> @llvm.mips.sld.d(
 
-  v16i8_r = __builtin_msa_sldi_b(v16i8_a, 3); // CHECK: call <16 x i8>  @llvm.mips.sldi.b(
-  v8i16_r = __builtin_msa_sldi_h(v8i16_a, 3); // CHECK: call <8  x i16> @llvm.mips.sldi.h(
-  v4i32_r = __builtin_msa_sldi_w(v4i32_a, 3); // CHECK: call <4  x i32> @llvm.mips.sldi.w(
-  v2i64_r = __builtin_msa_sldi_d(v2i64_a, 3); // CHECK: call <2  x i64> @llvm.mips.sldi.d(
+  v16i8_r = __builtin_msa_sldi_b(v16i8_r, v16i8_a, 3); // CHECK: call <16 x i8>  @llvm.mips.sldi.b(
+  v8i16_r = __builtin_msa_sldi_h(v8i16_r, v8i16_a, 3); // CHECK: call <8  x i16> @llvm.mips.sldi.h(
+  v4i32_r = __builtin_msa_sldi_w(v4i32_r, v4i32_a, 3); // CHECK: call <4  x i32> @llvm.mips.sldi.w(
+  v2i64_r = __builtin_msa_sldi_d(v2i64_r, v2i64_a, 3); // CHECK: call <2  x i64> @llvm.mips.sldi.d(
 
   v16i8_r = __builtin_msa_sll_b(v16i8_a, v16i8_b); // CHECK: call <16 x i8>  @llvm.mips.sll.b(
   v8i16_r = __builtin_msa_sll_h(v8i16_a, v8i16_b); // CHECK: call <8  x i16> @llvm.mips.sll.h(
diff --git a/test/CodeGen/builtins-nvptx.c b/test/CodeGen/builtins-nvptx.c
index 7deee8e..cee9061 100644
--- a/test/CodeGen/builtins-nvptx.c
+++ b/test/CodeGen/builtins-nvptx.c
@@ -1,5 +1,4 @@
 // REQUIRES: nvptx-registered-target
-// REQUIRES: nvptx64-registered-target
 // RUN: %clang_cc1 -triple nvptx-unknown-unknown -S -emit-llvm -o - %s | FileCheck %s
 // RUN: %clang_cc1 -triple nvptx64-unknown-unknown -S -emit-llvm -o - %s | FileCheck %s
 
diff --git a/test/CodeGen/builtins-ppc-altivec.c b/test/CodeGen/builtins-ppc-altivec.c
index 47a198f..5cd8d32 100644
--- a/test/CodeGen/builtins-ppc-altivec.c
+++ b/test/CodeGen/builtins-ppc-altivec.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -faltivec -triple powerpc-unknown-unknown -emit-llvm %s -o - | FileCheck %s
 
 vector bool char vbc = { 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0 };
diff --git a/test/CodeGen/builtins-ppc.c b/test/CodeGen/builtins-ppc.c
index ee27a4c..9ef5e37 100644
--- a/test/CodeGen/builtins-ppc.c
+++ b/test/CodeGen/builtins-ppc.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc-unknown-unknown -emit-llvm %s -o - | FileCheck %s
 
 void test_eh_return_data_regno()
diff --git a/test/CodeGen/c-strings.c b/test/CodeGen/c-strings.c
index ff86619..d82bc25 100644
--- a/test/CodeGen/c-strings.c
+++ b/test/CodeGen/c-strings.c
@@ -1,14 +1,18 @@
-// RUN: %clang_cc1 -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - %s | FileCheck %s --check-prefix=CHECK --check-prefix=ITANIUM
+// RUN: %clang_cc1 -triple %ms_abi_triple -emit-llvm -o - %s | FileCheck %s --check-prefix=CHECK --check-prefix=MSABI
 
 // Should be 3 hello strings, two global (of different sizes), the rest are
 // shared.
 
 // CHECK: @align = global i8 [[ALIGN:[0-9]+]]
-// CHECK: @.str = private unnamed_addr constant [6 x i8] c"hello\00"
-// CHECK: @f1.x = internal global i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0)
+// ITANIUM: @.str = private unnamed_addr constant [6 x i8] c"hello\00"
+// MSABI: @"\01??_C@_05CJBACGMB@hello?$AA@" = linkonce_odr unnamed_addr constant [6 x i8] c"hello\00", align 1
+// ITANIUM: @f1.x = internal global i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0)
+// MSABI: @f1.x = internal global i8* getelementptr inbounds ([6 x i8]* @"\01??_C@_05CJBACGMB@hello?$AA@", i32 0, i32 0)
 // CHECK: @f2.x = internal global [6 x i8] c"hello\00", align [[ALIGN]]
 // CHECK: @f3.x = internal global [8 x i8] c"hello\00\00\00", align [[ALIGN]]
-// CHECK: @f4.x = internal global %struct.s { i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0) }
+// ITANIUM: @f4.x = internal global %struct.s { i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0) }
+// MSABI: @f4.x = internal global %struct.s { i8* getelementptr inbounds ([6 x i8]* @"\01??_C@_05CJBACGMB@hello?$AA@", i32 0, i32 0) }
 // CHECK: @x = global [3 x i8] c"ola", align [[ALIGN]]
 
 #if defined(__s390x__)
@@ -22,7 +26,8 @@
 // CHECK-LABEL: define void @f0()
 void f0() {
   bar("hello");
-  // CHECK: call void @bar({{.*}} @.str
+  // ITANIUM: call void @bar({{.*}} @.str
+  // MSABI: call void @bar({{.*}} @"\01??_C@_05CJBACGMB@hello?$AA@"
 }
 
 // CHECK-LABEL: define void @f1()
diff --git a/test/CodeGen/captured-statements.c b/test/CodeGen/captured-statements.c
index c87c187..b52d115 100644
--- a/test/CodeGen/captured-statements.c
+++ b/test/CodeGen/captured-statements.c
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -emit-llvm %s -o %t
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm %s -o %t
 // RUN: FileCheck %s -input-file=%t -check-prefix=CHECK-GLOBALS
 // RUN: FileCheck %s -input-file=%t -check-prefix=CHECK-1
 // RUN: FileCheck %s -input-file=%t -check-prefix=CHECK-2
diff --git a/test/CodeGen/cfstring.c b/test/CodeGen/cfstring.c
index 9d98b56..fc86e42 100644
--- a/test/CodeGen/cfstring.c
+++ b/test/CodeGen/cfstring.c
@@ -5,9 +5,9 @@
 //
 // RUN: %clang_cc1 -fwritable-strings -emit-llvm %s -o - | FileCheck %s
 //
-// CHECK: @.str = linker_private unnamed_addr constant [14 x i8] c"Hello, World!\00", align 1
-// CHECK: @.str1 = linker_private unnamed_addr constant [7 x i8] c"yo joe\00", align 1
-// CHECK: @.str3 = linker_private unnamed_addr constant [16 x i8] c"Goodbye, World!\00", align 1
+// CHECK: @.str = private unnamed_addr constant [14 x i8] c"Hello, World!\00", section "__TEXT,__cstring,cstring_literals", align 1
+// CHECK: @.str1 = private unnamed_addr constant [7 x i8] c"yo joe\00", section "__TEXT,__cstring,cstring_literals", align 1
+// CHECK: @.str3 = private unnamed_addr constant [16 x i8] c"Goodbye, World!\00", section "__TEXT,__cstring,cstring_literals", align 1
 
 #define CFSTR __builtin___CFStringMakeConstantString
 
diff --git a/test/CodeGen/clear_cache.c b/test/CodeGen/clear_cache.c
new file mode 100644
index 0000000..f859d7f
--- /dev/null
+++ b/test/CodeGen/clear_cache.c
@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 -emit-llvm %s -o - | FileCheck %s
+
+char buffer[32] = "This is a largely unused buffer";
+
+// __builtin___clear_cache always maps to @llvm.clear_cache, but what
+// each back-end produces is different, and this is tested in LLVM
+
+int main() {
+  __builtin___clear_cache(buffer, buffer+32);
+// CHECK: @llvm.clear_cache(i8* getelementptr {{.*}}, i8* getelementptr {{.*}} (i8* getelementptr {{.*}} 32))
+  return 0;
+}
diff --git a/test/CodeGen/complex-convert.c b/test/CodeGen/complex-convert.c
index e35be9c..6ecb884 100644
--- a/test/CodeGen/complex-convert.c
+++ b/test/CodeGen/complex-convert.c
@@ -22,9 +22,14 @@
   _Complex signed long long csll1;
   _Complex unsigned long long cull1;
   // CHECK-LABEL: define void @foo(
-  // CHECK: alloca i[[CHSIZE:[0-9]+]], align [[CHALIGN:[0-9]+]]
-  // CHECK-NEXT: alloca i[[CHSIZE]], align [[CHALIGN]]
-  // CHECK-NEXT: alloca i[[LLSIZE:[0-9]+]], align [[LLALIGN:[0-9]+]]
+  // Match the prototype to pick up the size of sc and sll.
+  // CHECK: i[[CHSIZE:[0-9]+]]{{[^,]*}},
+  // CHECK: i[[CHSIZE]]{{[^,]*}},
+  // CHECK: i[[LLSIZE:[0-9]+]]
+
+  // Match against the allocas to pick up the alignments.
+  // CHECK: alloca i[[CHSIZE]], align [[CHALIGN:[0-9]+]]
+  // CHECK: alloca i[[LLSIZE]], align [[LLALIGN:[0-9]+]]
 
   sc1 = csc;
   // CHECK: %[[VAR1:[A-Za-z0-9.]+]] = getelementptr inbounds { i[[CHSIZE]], i[[CHSIZE]]  }* %[[CSC:[A-Za-z0-9.]+]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
diff --git a/test/CodeGen/cxx-default-arg.cpp b/test/CodeGen/cxx-default-arg.cpp
index 12e2666..7688e79 100644
--- a/test/CodeGen/cxx-default-arg.cpp
+++ b/test/CodeGen/cxx-default-arg.cpp
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -emit-llvm %s -o %t
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm %s -o %t
 
 // Note-LABEL: define CLANG_GENERATE_KNOWN_GOOD and compile to generate code
 // that makes all of the defaulted arguments explicit. The resulting
diff --git a/test/CodeGen/darwin-string-literals.c b/test/CodeGen/darwin-string-literals.c
index c7d9ff9..41e59cb 100644
--- a/test/CodeGen/darwin-string-literals.c
+++ b/test/CodeGen/darwin-string-literals.c
@@ -1,17 +1,17 @@
 // RUN: %clang_cc1 -triple i386-apple-darwin9 -emit-llvm %s -o - | FileCheck -check-prefix CHECK-LSB %s
 
 // CHECK-LSB: @.str = private unnamed_addr constant [8 x i8] c"string0\00"
-// CHECK-LSB: @.str1 = linker_private unnamed_addr constant [8 x i8] c"string1\00"
-// CHECK-LSB: @.str2 = internal unnamed_addr constant [18 x i16] [i16 104, i16 101, i16 108, i16 108, i16 111, i16 32, i16 8594, i16 32, i16 9731, i16 32, i16 8592, i16 32, i16 119, i16 111, i16 114, i16 108, i16 100, i16 0], align 2
-// CHECK-LSB: @.str4 = internal unnamed_addr constant [6 x i16] [i16 116, i16 101, i16 115, i16 116, i16 8482, i16 0], align 2
+// CHECK-LSB: @.str1 = private unnamed_addr constant [8 x i8] c"string1\00"
+// CHECK-LSB: @.str2 = private unnamed_addr constant [18 x i16] [i16 104, i16 101, i16 108, i16 108, i16 111, i16 32, i16 8594, i16 32, i16 9731, i16 32, i16 8592, i16 32, i16 119, i16 111, i16 114, i16 108, i16 100, i16 0], section "__TEXT,__ustring", align 2
+// CHECK-LSB: @.str4 = private unnamed_addr constant [6 x i16] [i16 116, i16 101, i16 115, i16 116, i16 8482, i16 0], section "__TEXT,__ustring", align 2
 
 
 // RUN: %clang_cc1 -triple powerpc-apple-darwin9 -emit-llvm %s -o - | FileCheck -check-prefix CHECK-MSB %s
 
 // CHECK-MSB: @.str = private unnamed_addr constant [8 x i8] c"string0\00"
-// CHECK-MSB: @.str1 = linker_private unnamed_addr constant [8 x i8] c"string1\00"
-// CHECK-MSB: @.str2 = internal unnamed_addr constant [18 x i16] [i16 104, i16 101, i16 108, i16 108, i16 111, i16 32, i16 8594, i16 32, i16 9731, i16 32, i16 8592, i16 32, i16 119, i16 111, i16 114, i16 108, i16 100, i16 0], align 2
-// CHECK-MSB: @.str4 = internal unnamed_addr constant [6 x i16] [i16 116, i16 101, i16 115, i16 116, i16 8482, i16 0], align 2
+// CHECK-MSB: @.str1 = private unnamed_addr constant [8 x i8] c"string1\00"
+// CHECK-MSB: @.str2 = private unnamed_addr constant [18 x i16] [i16 104, i16 101, i16 108, i16 108, i16 111, i16 32, i16 8594, i16 32, i16 9731, i16 32, i16 8592, i16 32, i16 119, i16 111, i16 114, i16 108, i16 100, i16 0], section "__TEXT,__ustring", align 2
+// CHECK-MSB: @.str4 = private unnamed_addr constant [6 x i16] [i16 116, i16 101, i16 115, i16 116, i16 8482, i16 0], section "__TEXT,__ustring", align 2
 
 const char *g0 = "string0";
 const void *g1 = __builtin___CFStringMakeConstantString("string1");
diff --git a/test/CodeGen/debug-dead-local-var.c b/test/CodeGen/debug-dead-local-var.c
index f9ca8f2..19cd6fe 100644
--- a/test/CodeGen/debug-dead-local-var.c
+++ b/test/CodeGen/debug-dead-local-var.c
@@ -1,4 +1,5 @@
-// RUN: %clang_cc1 -mllvm -asm-verbose -S -O2 -g %s -o - | FileCheck %s
+// FIXME: Check IR rather than asm, then triple is not needed.
+// RUN: %clang_cc1 -mllvm -asm-verbose -triple %itanium_abi_triple -S -O2 -g %s -o - | FileCheck %s
 // Radar 8122864
 
 // Code is not generated for function foo, but preserve type information of
diff --git a/test/CodeGen/debug-info-block.c b/test/CodeGen/debug-info-block.c
index 403e4aa..35ee0dd 100644
--- a/test/CodeGen/debug-info-block.c
+++ b/test/CodeGen/debug-info-block.c
@@ -1,5 +1,4 @@
 // RUN: %clang_cc1 -fblocks -g -emit-llvm -o - %s | FileCheck %s
-// APPLE LOCAL file 5939894 */
 // Verify that the desired debugging type is generated for a structure
 //  member that is a pointer to a block. 
 
diff --git a/test/CodeGen/debug-info-iv.c b/test/CodeGen/debug-info-iv.c
deleted file mode 100644
index aafd71d..0000000
--- a/test/CodeGen/debug-info-iv.c
+++ /dev/null
@@ -1,36 +0,0 @@
-// RUN: %clang_cc1 -triple x86_64-apple-darwin -Os -S -g  -o - %s | FileCheck %s
-// REQUIRES: x86-registered-target
-
-int calculate(int);
-static void test_indvars(int *Array1, int Array2[100][200]) {
-  unsigned i, j;
-  Array1[1] = Array2[3][6] = 12345;
-
-  for (i = 0; i < 100; i+=2)
-    Array1[i] = i;           /* Step by non unit amount */
-
-  for (i = 3; i < 103; i++)
-    Array1[i] = i+4;         /* Step with an offset */
-
-  for (i = 13; i < 100; i++)
-    for (j = 0; j < 100; j+=3)       /* 2d array access */
-      Array2[i][j/3] = Array2[i][i];
-}
-
-
-int main() {
-  int Array[100][200], i, j;
-  double sum = 0.0;
-
-  for (i=0; i < 100; i+=2)
-    for (j=0; j < 200; j++)
-      Array[i][j] = 0;
-  test_indvars(Array[0], Array);
-
-//CHECK:	.loc	2 31
-  for (i=0; i < 100; i+=2)
-    for (j=0; j < 200; j++)
-      sum += Array[i][j];
-
-  return calculate(sum);
-}
diff --git a/test/CodeGen/debug-info-typedef.c b/test/CodeGen/debug-info-typedef.c
new file mode 100644
index 0000000..51ebcc4
--- /dev/null
+++ b/test/CodeGen/debug-info-typedef.c
@@ -0,0 +1,11 @@
+// RUN: %clang_cc1 -emit-llvm -g -I%p %s -o - | FileCheck %s
+// Test that the location of the typedef points to the header file.
+#line 1 "a.c"
+#line 2 "b.h"
+typedef int MyType;
+#line 2 "a.c"
+
+MyType a;
+
+// CHECK: metadata ![[HEADER:[0-9]+]], null, metadata !"MyType"{{.*}} ; [ DW_TAG_typedef ] [MyType] [line 2, size 0, align 0, offset 0] [from int]
+// CHECK: ![[HEADER]] = metadata !{metadata !"b.h",
diff --git a/test/CodeGen/debug-info-var-location.c b/test/CodeGen/debug-info-var-location.c
index 12edb08..41da7d3 100644
--- a/test/CodeGen/debug-info-var-location.c
+++ b/test/CodeGen/debug-info-var-location.c
@@ -1,4 +1,5 @@
-// RUN: %clang -S -g -fverbose-asm %s -o - | FileCheck %s
+// FIXME: Check IR rather than asm, then triple is not needed.
+// RUN: %clang -Xclang -triple=%itanium_abi_triple -S -g -fverbose-asm %s -o - | FileCheck %s
 // Radar 8461032
 // CHECK: DW_AT_location
 // CHECK-NEXT: byte 145
diff --git a/test/CodeGen/dllimport-dllexport.c b/test/CodeGen/dllimport-dllexport.c
index e70ac03..b005e34 100644
--- a/test/CodeGen/dllimport-dllexport.c
+++ b/test/CodeGen/dllimport-dllexport.c
@@ -1,5 +1,10 @@
 // RUN: %clang_cc1 -triple i386-mingw32 -emit-llvm < %s | FileCheck %s
 
+__attribute__((dllexport)) int bar1 = 2;
+// CHECK-LABEL: @bar1 = dllexport global i32 2
+__attribute__((dllimport)) extern int bar2;
+// CHECK-LABEL: @bar2 = external dllimport global i32
+
 void __attribute__((dllimport)) foo1();
 void __attribute__((dllexport)) foo1(){}
 // CHECK-LABEL: define dllexport void @foo1
@@ -10,3 +15,8 @@
 __declspec(dllexport) void foo3(){}
 // CHECK-LABEL: define dllexport void @foo3
 __declspec(dllexport) void foo4();
+
+__declspec(dllimport) void foo5();
+// CHECK-LABEL: declare dllimport void @foo5
+
+int use() { foo5(); return bar2; }
diff --git a/test/CodeGen/exceptions-seh.c b/test/CodeGen/exceptions-seh.c
index eadbe15..f7d24bd 100644
--- a/test/CodeGen/exceptions-seh.c
+++ b/test/CodeGen/exceptions-seh.c
@@ -13,6 +13,6 @@
   *res = myres;
   return 1;
 }
-// CHECK-NOT error
+// CHECK-NOT: error:
 // CHECK: error: cannot compile this SEH __try yet
-// CHECK-NOT error
+// CHECK-NOT: error:
diff --git a/test/CodeGen/ffp-contract-option.c b/test/CodeGen/ffp-contract-option.c
index eb95f1e..61913b0 100644
--- a/test/CodeGen/ffp-contract-option.c
+++ b/test/CodeGen/ffp-contract-option.c
@@ -1,5 +1,5 @@
 // RUN: %clang_cc1 -O3 -ffp-contract=fast -triple=powerpc-apple-darwin10 -S -o - %s | FileCheck %s
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 
 float fma_test1(float a, float b, float c) {
 // CHECK: fmadds
diff --git a/test/CodeGen/fp-contract-pragma.cpp b/test/CodeGen/fp-contract-pragma.cpp
index afd8c43..b4e24b9 100644
--- a/test/CodeGen/fp-contract-pragma.cpp
+++ b/test/CodeGen/fp-contract-pragma.cpp
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -O3 -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -O3 -triple %itanium_abi_triple -emit-llvm -o - %s | FileCheck %s
 
 // Is FP_CONTRACT is honored in a simple case?
 float fp_contract_1(float a, float b, float c) {
diff --git a/test/CodeGen/inline.c b/test/CodeGen/inline.c
index 70ab696..96c9a86 100644
--- a/test/CodeGen/inline.c
+++ b/test/CodeGen/inline.c
@@ -53,12 +53,13 @@
 
 // RUN: echo "MS C Mode tests:"
 // RUN: %clang_cc1 %s -triple i386-unknown-unknown -O1 -disable-llvm-optzns -emit-llvm -o - -std=c99 -fms-compatibility | FileCheck %s --check-prefix=CHECK4
+// CHECK4-LABEL: define weak_odr i32 @ei()
 // CHECK4-LABEL: define i32 @bar()
+// CHECK4-NOT: unreferenced1
+// CHECK4-LABEL: define weak_odr void @unreferenced2()
 // CHECK4-LABEL: define void @gnu_inline()
 // CHECK4-LABEL: define available_externally void @gnu_ei_inline()
 // CHECK4-LABEL: define linkonce_odr i32 @foo()
-// CHECK4-NOT: unreferenced
-// CHECK4-LABEL: define linkonce_odr i32 @ei()
 
 extern __inline int ei() { return 123; }
 
diff --git a/test/CodeGen/inline2.c b/test/CodeGen/inline2.c
index 670ae20..84cd4db 100644
--- a/test/CodeGen/inline2.c
+++ b/test/CodeGen/inline2.c
@@ -39,6 +39,9 @@
 // CHECK-GNU89-LABEL: define i32 @fA()
 inline int fA(void) { return 0; }
 
+// CHECK-GNU89-LABEL: define i32 @fB()
+inline int fB() { return 0; }
+
 // CHECK-GNU89-LABEL: define available_externally i32 @f4()
 // CHECK-C99-LABEL: define i32 @f4()
 int f4(void);
@@ -56,7 +59,11 @@
 
 // CHECK-C99-LABEL: define available_externally i32 @fA()
 
+// CHECK-C99-LABEL: define i32 @fB()
+
 int test_all() { 
   return f0() + f1() + f2() + f3() + f4() + f5() + f6() + f7() + f8() + f9() 
-    + fA();
+    + fA() + fB();
 }
+
+int fB(void);
diff --git a/test/CodeGen/le32-vaarg.c b/test/CodeGen/le32-vaarg.c
new file mode 100644
index 0000000..51bbb02
--- /dev/null
+++ b/test/CodeGen/le32-vaarg.c
@@ -0,0 +1,28 @@
+// RUN: %clang_cc1 -triple le32-unknown-nacl -emit-llvm -o - %s | FileCheck %s
+#include <stdarg.h>
+
+int get_int(va_list *args) {
+  return va_arg(*args, int);
+}
+// CHECK: define i32 @get_int
+// CHECK: [[RESULT:%[a-z_0-9]+]] = va_arg {{.*}}, i32{{$}}
+// CHECK: ret i32 [[RESULT]]
+
+struct Foo {
+  int x;
+};
+
+struct Foo dest;
+
+void get_struct(va_list *args) {
+  dest = va_arg(*args, struct Foo);
+}
+// CHECK: define void @get_struct
+// CHECK: [[RESULT:%[a-z_0-9]+]] = va_arg {{.*}}, %struct.Foo{{$}}
+// CHECK: store %struct.Foo [[RESULT]], %struct.Foo* @dest
+
+void skip_struct(va_list *args) {
+  va_arg(*args, struct Foo);
+}
+// CHECK: define void @skip_struct
+// CHECK: va_arg {{.*}}, %struct.Foo{{$}}
diff --git a/test/CodeGen/libcall-declarations.c b/test/CodeGen/libcall-declarations.c
index 6442e29..345b74f 100644
--- a/test/CodeGen/libcall-declarations.c
+++ b/test/CodeGen/libcall-declarations.c
@@ -244,6 +244,16 @@
 double _Complex ctanh(double _Complex);
 float _Complex ctanhf(float _Complex);
 long double _Complex ctanhl(long double _Complex);
+
+double __sinpi(double);
+float __sinpif(float);
+double __cospi(double);
+float __cospif(float);
+double __tanpi(double);
+float __tanpif(float);
+
+double __exp10(double);
+float __exp10f(float);
 #ifdef __cplusplus
 }
 #endif
@@ -298,7 +308,8 @@
   F(crealf),     F(creall),     F(csin),       F(csinf),       F(csinl),
   F(csinh),      F(csinhf),     F(csinhl),     F(csqrt),       F(csqrtf),
   F(csqrtl),     F(ctan),       F(ctanf),      F(ctanl),       F(ctanh),
-  F(ctanhf),     F(ctanhl)
+  F(ctanhf),     F(ctanhl),     F(__sinpi),    F(__sinpif),    F(__cospi),
+  F(__cospif),   F(__tanpi),    F(__tanpif),   F(__exp10),     F(__exp10f)
 };
 
 // CHECK-NOERRNO: declare double @atan2(double, double) [[NUW:#[0-9]+]]
@@ -510,6 +521,14 @@
 // CHECK-NOERRNO: declare <2 x float> @ctanf(<2 x float>) [[NUW]]
 // CHECK-NOERRNO: declare { double, double } @ctanh(double, double) [[NUW]]
 // CHECK-NOERRNO: declare <2 x float> @ctanhf(<2 x float>) [[NUW]]
+// CHECK-NOERRNO: declare double @__sinpi(double) [[NUW]]
+// CHECK-NOERRNO: declare float @__sinpif(float) [[NUW]]
+// CHECK-NOERRNO: declare double @__cospi(double) [[NUW]]
+// CHECK-NOERRNO: declare float @__cospif(float) [[NUW]]
+// CHECK-NOERRNO: declare double @__tanpi(double) [[NUW]]
+// CHECK-NOERRNO: declare float @__tanpif(float) [[NUW]]
+// CHECK-NOERRNO: declare double @__exp10(double) [[NUW]]
+// CHECK-NOERRNO: declare float @__exp10f(float) [[NUW]]
 
 // CHECK-ERRNO: declare i32 @abs(i32) [[NUW:#[0-9]+]]
 // CHECK-ERRNO: declare i64 @labs(i64) [[NUW]]
diff --git a/test/CodeGen/libcalls.c b/test/CodeGen/libcalls.c
index 3112c87..8120b75 100644
--- a/test/CodeGen/libcalls.c
+++ b/test/CodeGen/libcalls.c
@@ -52,9 +52,9 @@
 // CHECK-YES: declare float @powf(float, float)
 // CHECK-YES: declare double @pow(double, double)
 // CHECK-YES: declare x86_fp80 @powl(x86_fp80, x86_fp80)
-// CHECK-NO: declare float @llvm.pow.f32(float, float) [[NUW_RO:#[0-9]+]]
-// CHECK-NO: declare double @llvm.pow.f64(double, double) [[NUW_RO]]
-// CHECK-NO: declare x86_fp80 @llvm.pow.f80(x86_fp80, x86_fp80) [[NUW_RO]]
+// CHECK-NO: declare float @llvm.pow.f32(float, float) [[NUW_RNI:#[0-9]+]]
+// CHECK-NO: declare double @llvm.pow.f64(double, double) [[NUW_RNI]]
+// CHECK-NO: declare x86_fp80 @llvm.pow.f80(x86_fp80, x86_fp80) [[NUW_RNI]]
 
 // CHECK-YES-LABEL: define void @test_fma
 // CHECK-NO-LABEL: define void @test_fma
@@ -81,8 +81,8 @@
 
 // Just checking to make sure these library functions are marked readnone
 void test_builtins(double d, float f, long double ld) {
-// CHEC-NO: @test_builtins
-// CHEC-YES: @test_builtins
+// CHECK-NO: @test_builtins
+// CHECK-YES: @test_builtins
   double atan_ = atan(d);
   long double atanl_ = atanl(ld);
   float atanf_ = atanf(f);
@@ -127,4 +127,4 @@
 // CHECK-YES: attributes [[NUW_RN]] = { nounwind readnone }
 
 // CHECK-NO: attributes [[NUW_RN]] = { nounwind readnone{{.*}} }
-// CHECK-NO: attributes [[NUW_RO]] = { nounwind readonly }
+// CHECK-NO: attributes [[NUW_RNI]] = { nounwind readnone }
diff --git a/test/CodeGen/lineno-dbginfo.c b/test/CodeGen/lineno-dbginfo.c
index 72fa337..1f9b7a5 100644
--- a/test/CodeGen/lineno-dbginfo.c
+++ b/test/CodeGen/lineno-dbginfo.c
@@ -1,5 +1,5 @@
 // RUN: echo "#include <stddef.h>" > %t.h
-// RUN: %clang -S -g -include %t.h %s -emit-llvm -o %t.ll
+// RUN: %clang_cc1 -S -g -include %t.h %s -emit-llvm -o %t.ll
 // RUN: grep "i32 5" %t.ll
 // outer is at line number 5.
 int outer = 42;
diff --git a/test/CodeGen/long-double-x86-nacl.c b/test/CodeGen/long-double-x86-nacl.c
deleted file mode 100644
index cec381a..0000000
--- a/test/CodeGen/long-double-x86-nacl.c
+++ /dev/null
@@ -1,7 +0,0 @@
-// RUN: %clang_cc1 %s -emit-llvm -o - -triple=i686-unknown-nacl | FileCheck %s
-
-long double x = 0;
-int checksize[sizeof(x) == 8 ? 1 : -1];
-
-// CHECK-LABEL: define void @s1(double %a)
-void s1(long double a) {}
diff --git a/test/CodeGen/malign-double-x86-nacl.c b/test/CodeGen/malign-double-x86-nacl.c
new file mode 100644
index 0000000..d673155
--- /dev/null
+++ b/test/CodeGen/malign-double-x86-nacl.c
@@ -0,0 +1,43 @@
+// RUN: %clang_cc1 %s -emit-llvm -o - -triple=i686-unknown-nacl | FileCheck %s
+// Check that i686-nacl essentially has -malign-double, which aligns
+// double, long double, and long long to 64-bits.
+
+int checksize[sizeof(long double) == 8 ? 1 : -1];
+int checkalign[__alignof(long double) == 8 ? 1 : -1];
+
+// CHECK-LABEL: define void @s1(double %a)
+void s1(long double a) {}
+
+struct st_ld {
+  char c;
+  long double ld;
+};
+int checksize2[sizeof(struct st_ld) == 16 ? 1 : -1];
+int checkalign2[__alignof(struct st_ld) == 8 ? 1 : -1];
+
+int checksize3[sizeof(double) == 8 ? 1 : -1];
+int checkalign3[__alignof(double) == 8 ? 1 : -1];
+
+// CHECK-LABEL: define void @s2(double %a)
+void s2(double a) {}
+
+struct st_d {
+  char c;
+  double d;
+};
+int checksize4[sizeof(struct st_d) == 16 ? 1 : -1];
+int checkalign4[__alignof(struct st_d) == 8 ? 1 : -1];
+
+
+int checksize5[sizeof(long long) == 8 ? 1 : -1];
+int checkalign5[__alignof(long long) == 8 ? 1 : -1];
+
+// CHECK-LABEL: define void @s3(i64 %a)
+void s3(long long a) {}
+
+struct st_ll {
+  char c;
+  long long ll;
+};
+int checksize6[sizeof(struct st_ll) == 16 ? 1 : -1];
+int checkalign6[__alignof(struct st_ll) == 8 ? 1 : -1];
diff --git a/test/CodeGen/mangle-windows.c b/test/CodeGen/mangle-windows.c
index 6706492..37d1018 100644
--- a/test/CodeGen/mangle-windows.c
+++ b/test/CodeGen/mangle-windows.c
@@ -1,5 +1,4 @@
-// RUN: %clang_cc1 -emit-llvm %s -o - -cxx-abi microsoft \
-// RUN:     -triple=i386-pc-win32 | FileCheck %s
+// RUN: %clang_cc1 -emit-llvm %s -o - -triple=i386-pc-win32 | FileCheck %s
 // RUN: %clang_cc1 -emit-llvm %s -o - -triple=i386-mingw32 | FileCheck %s
 
 void __stdcall f1(void) {}
diff --git a/test/CodeGen/mcount.c b/test/CodeGen/mcount.c
index 1cf3d6a..5c608bc 100644
--- a/test/CodeGen/mcount.c
+++ b/test/CodeGen/mcount.c
@@ -1,4 +1,8 @@
 // RUN: %clang_cc1 -pg -triple i386-unknown-unknown -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -pg -triple powerpc-unknown-gnu-linux -emit-llvm -o - %s | FileCheck -check-prefix=CHECK-PPC %s
+// RUN: %clang_cc1 -pg -triple powerpc64-unknown-gnu-linux -emit-llvm -o - %s | FileCheck -check-prefix=CHECK-PPC %s
+// RUN: %clang_cc1 -pg -triple powerpc64le-unknown-gnu-linux -emit-llvm -o - %s | FileCheck -check-prefix=CHECK-PPC %s
 void foo(void) {
 // CHECK: call void @mcount()
+// CHECK-PPC: call void @_mcount()
 }
diff --git a/test/CodeGen/mips-target-data.c b/test/CodeGen/mips-target-data.c
deleted file mode 100644
index 6475ccb..0000000
--- a/test/CodeGen/mips-target-data.c
+++ /dev/null
@@ -1,16 +0,0 @@
-// RUN: %clang -target mipsel-linux-gnu -o - -emit-llvm -S %s |\
-// RUN: FileCheck %s -check-prefix=32EL
-// RUN: %clang -target mips-linux-gnu -o - -emit-llvm -S %s |\
-// RUN: FileCheck %s -check-prefix=32EB
-// RUN: %clang -target mips64el-linux-gnu -o - -emit-llvm -S %s |\
-// RUN: FileCheck %s -check-prefix=64EL
-// RUN: %clang -target mips64-linux-gnu -o - -emit-llvm -S %s |\
-// RUN: FileCheck %s -check-prefix=64EB
-// RUN: %clang -target mipsel-linux-gnu -o - -emit-llvm -S -mfp64 %s |\
-// RUN: FileCheck %s -check-prefix=32EL
-
-// 32EL: e-p:32:32:32-i1:8:8-i8:8:32-i16:16:32-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-n32-S64
-// 32EB: E-p:32:32:32-i1:8:8-i8:8:32-i16:16:32-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-n32-S64
-// 64EL: e-p:64:64:64-i1:8:8-i8:8:32-i16:16:32-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v64:64:64-n32:64-S128
-// 64EB: E-p:64:64:64-i1:8:8-i8:8:32-i16:16:32-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v64:64:64-n32:64-S128
-
diff --git a/test/CodeGen/mmx-builtins.c b/test/CodeGen/mmx-builtins.c
index b142684..346676c 100644
--- a/test/CodeGen/mmx-builtins.c
+++ b/test/CodeGen/mmx-builtins.c
@@ -1,4 +1,4 @@
-// REQUIRES: x86-64-registered-target
+// REQUIRES: x86-registered-target
 // RUN: %clang_cc1 %s -O3 -triple=x86_64-apple-darwin -target-feature +ssse3 -S -o - | FileCheck %s
 
 // FIXME: Disable inclusion of mm_malloc.h, our current implementation is broken
diff --git a/test/CodeGen/ms-inline-asm-64.c b/test/CodeGen/ms-inline-asm-64.c
index f667708..69066aa 100644
--- a/test/CodeGen/ms-inline-asm-64.c
+++ b/test/CodeGen/ms-inline-asm-64.c
@@ -1,4 +1,4 @@
-// REQUIRES: x86-64-registered-target
+// REQUIRES: x86-registered-target
 // RUN: %clang_cc1 %s -triple x86_64-apple-darwin10 -fasm-blocks -emit-llvm -o - | FileCheck %s
 
 void t1() {
diff --git a/test/CodeGen/ms-inline-asm.c b/test/CodeGen/ms-inline-asm.c
index c4486f6..6395e42 100644
--- a/test/CodeGen/ms-inline-asm.c
+++ b/test/CodeGen/ms-inline-asm.c
@@ -1,4 +1,4 @@
-// REQUIRES: x86-64-registered-target
+// REQUIRES: x86-registered-target
 // RUN: %clang_cc1 %s -triple i386-apple-darwin10 -fasm-blocks -emit-llvm -o - | FileCheck %s
 
 void t1() {
@@ -77,7 +77,7 @@
     pop ebx
   }
 // CHECK: t9
-// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, $$0x07\0A\09pop ebx", "~{ebx},~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, $$0x07\0A\09pop ebx", "~{ebx},~{esp},~{dirflag},~{fpsr},~{flags}"()
 }
 
 unsigned t10(void) {
@@ -211,7 +211,7 @@
     __asm pop ebx
   }
 // CHECK: t21
-// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, $$0x07\0A\09pop ebx", "~{ebx},~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, $$0x07\0A\09pop ebx", "~{ebx},~{esp},~{dirflag},~{fpsr},~{flags}"()
 }
 
 extern void t22_helper(int x);
@@ -227,7 +227,7 @@
     __asm pop ebx
   }
 // CHECK: t22
-// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, esp", "~{ebx},~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "push ebx\0A\09mov ebx, esp", "~{ebx},~{esp},~{dirflag},~{fpsr},~{flags}"()
 // CHECK: call void @t22_helper
 // CHECK: call void asm sideeffect inteldialect "mov esp, ebx\0A\09pop ebx", "~{ebx},~{esp},~{dirflag},~{fpsr},~{flags}"()
 }
@@ -268,13 +268,14 @@
   __asm __emit 0a2h
   __asm __EMIT 0a2h
   __asm popad
+// FIXME: These all need to be merged into the same asm blob.
 // CHECK: t26
-// CHECK: call void asm sideeffect inteldialect "pushad", "~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "pushad", "~{esp},~{dirflag},~{fpsr},~{flags}"()
 // CHECK: call void asm sideeffect inteldialect "mov eax, $$0", "~{eax},~{dirflag},~{fpsr},~{flags}"()
 // CHECK: call void asm sideeffect inteldialect ".byte 0fh", "~{dirflag},~{fpsr},~{flags}"()
 // CHECK: call void asm sideeffect inteldialect ".byte 0a2h", "~{dirflag},~{fpsr},~{flags}"()
 // CHECK: call void asm sideeffect inteldialect ".byte 0a2h", "~{dirflag},~{fpsr},~{flags}"()
-// CHECK: call void asm sideeffect inteldialect "popad", "~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "popad", "~{eax},~{ebp},~{ebx},~{ecx},~{edi},~{edx},~{esi},~{esp},~{dirflag},~{fpsr},~{flags}"()
 }
 
 void t27() {
@@ -323,8 +324,8 @@
   __asm pushad
   __asm popad
 // CHECK: t31
-// CHECK: call void asm sideeffect inteldialect "pushad", "~{dirflag},~{fpsr},~{flags}"()
-// CHECK: call void asm sideeffect inteldialect "popad", "~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "pushad", "~{esp},~{dirflag},~{fpsr},~{flags}"()
+// CHECK: call void asm sideeffect inteldialect "popad", "~{eax},~{ebp},~{ebx},~{ecx},~{edi},~{edx},~{esi},~{esp},~{dirflag},~{fpsr},~{flags}"()
 }
 
 void t32() {
@@ -438,3 +439,33 @@
 // CHECK: call void asm sideeffect inteldialect "mov eax, dword ptr $$8$0", "*m,~{eax},~{dirflag},~{fpsr},~{flags}"([4 x i32]* %{{.*}})
 // CHECK: call void asm sideeffect inteldialect "mov eax, dword ptr $$0$0", "*m,~{eax},~{dirflag},~{fpsr},~{flags}"([4 x i32]* %{{.*}})
 }
+
+void cpuid() {
+  __asm cpuid
+// CHECK-LABEL: define void @cpuid
+// CHECK: call void asm sideeffect inteldialect "cpuid", "~{eax},~{ebx},~{ecx},~{edx},~{dirflag},~{fpsr},~{flags}"()
+}
+
+typedef struct {
+  int a;
+  int b;
+} A;
+
+void t39() {
+  __asm mov eax, [eax].A.b
+  __asm mov eax, [eax] A.b
+  __asm mov eax, fs:[0] A.b
+  // CHECK-LABEL: define void @t39
+  // CHECK: call void asm sideeffect inteldialect "mov eax, [eax].4", "~{eax},~{dirflag},~{fpsr},~{flags}"()
+  // CHECK: call void asm sideeffect inteldialect "mov eax, [eax] .4", "~{eax},~{dirflag},~{fpsr},~{flags}"()
+  // CHECK: call void asm sideeffect inteldialect "mov eax, fs:[$$0] .4", "~{eax},~{dirflag},~{fpsr},~{flags}"()
+}
+
+void t40(float a) {
+  int i;
+  __asm fld a
+  __asm fistp i
+  // CHECK-LABEL: define void @t40
+  // CHECK: call void asm sideeffect inteldialect "fld dword ptr $0", "*m,~{dirflag},~{fpsr},~{flags}"(float* {{.*}})
+  // CHECK: call void asm sideeffect inteldialect "fistp dword ptr $0", "=*m,~{dirflag},~{fpsr},~{flags}"(i32* {{.*}})
+}
diff --git a/test/CodeGen/ms-inline-asm.cpp b/test/CodeGen/ms-inline-asm.cpp
index 64b8558..b1c13e5 100644
--- a/test/CodeGen/ms-inline-asm.cpp
+++ b/test/CodeGen/ms-inline-asm.cpp
@@ -1,4 +1,4 @@
-// REQUIRES: x86-64-registered-target
+// REQUIRES: x86-registered-target
 // RUN: %clang_cc1 -x c++ %s -triple i386-apple-darwin10 -fasm-blocks -emit-llvm -o - | FileCheck %s
 
 // rdar://13645930
@@ -97,7 +97,7 @@
   // CHECK: [[Y:%.*]] = alloca i32
   int x, y;
   __asm push y
-  // CHECK: call void asm sideeffect inteldialect "push dword ptr $0", "=*m,~{dirflag},~{fpsr},~{flags}"(i32* [[Y]])
+  // CHECK: call void asm sideeffect inteldialect "push dword ptr $0", "=*m,~{esp},~{dirflag},~{fpsr},~{flags}"(i32* [[Y]])
   __asm call T5<int>::create<float>
   // CHECK: call void asm sideeffect inteldialect "call $0", "r,~{dirflag},~{fpsr},~{flags}"(i32 (float)* @_ZN2T5IiE6createIfEEiT_)
   __asm mov x, eax
diff --git a/test/CodeGen/ms_struct-bitfield.c b/test/CodeGen/ms_struct-bitfield.c
index a8f4c91..08f2a5b 100644
--- a/test/CodeGen/ms_struct-bitfield.c
+++ b/test/CodeGen/ms_struct-bitfield.c
@@ -1,6 +1,11 @@
-// RUN: %clang_cc1 -emit-llvm-only  -triple x86_64-apple-darwin9 %s
+// RUN: %clang_cc1 -emit-llvm -o - -triple x86_64-apple-darwin9 %s | FileCheck %s
+// RUN: %clang_cc1 -emit-llvm -o - -triple thumbv7-apple-ios -target-abi apcs-gnu %s | FileCheck %s -check-prefix=CHECK-ARM
+
 // rdar://8823265
 
+// Note that we're declaring global variables with these types,
+// triggering both Sema and IRGen struct layout.
+
 #define ATTR __attribute__((__ms_struct__))
 
 struct
@@ -9,6 +14,9 @@
    long : 0;
    char bar;
 } ATTR t1;
+int s1 = sizeof(t1);
+// CHECK: @s1 = global i32 2
+// CHECK-ARM: @s1 = global i32 2
 
 struct
 {
@@ -18,6 +26,9 @@
    int : 0;
    char bar;
 } ATTR t2;
+int s2 = sizeof(t2);
+// CHECK: @s2 = global i32 2
+// CHECK-ARM: @s2 = global i32 2
 
 struct
 {
@@ -29,12 +40,18 @@
    long : 0;
    char : 0;
 } ATTR t3;
+int s3 = sizeof(t3);
+// CHECK: @s3 = global i32 2
+// CHECK-ARM: @s3 = global i32 2
 
 struct
 {
    long : 0;
    char bar;
 } ATTR t4;
+int s4 = sizeof(t4);
+// CHECK: @s4 = global i32 1
+// CHECK-ARM: @s4 = global i32 1
 
 struct
 {
@@ -43,6 +60,9 @@
    char : 0;
    char bar;
 } ATTR t5;
+int s5 = sizeof(t5);
+// CHECK: @s5 = global i32 1
+// CHECK-ARM: @s5 = global i32 1
 
 struct
 {
@@ -51,6 +71,9 @@
    char : 0;
    char bar;
 } ATTR t6;
+int s6 = sizeof(t6);
+// CHECK: @s6 = global i32 1
+// CHECK-ARM: @s6 = global i32 1
 
 struct
 {
@@ -69,6 +92,9 @@
    char bar6;
    char bar7;
 } ATTR t7;
+int s7 = sizeof(t7);
+// CHECK: @s7 = global i32 9
+// CHECK-ARM: @s7 = global i32 9
 
 struct
 {
@@ -76,6 +102,9 @@
    long : 0;
    char : 0;
 } ATTR t8;
+int s8 = sizeof(t8);
+// CHECK: @s8 = global i32 0
+// CHECK-ARM: @s8 = global i32 0
 
 struct
 {
@@ -106,6 +135,9 @@
    long : 0;
    char :4;
 } ATTR t9;
+int s9 = sizeof(t9);
+// CHECK: @s9 = global i32 28
+// CHECK-ARM: @s9 = global i32 28
 
 struct
 {
@@ -113,19 +145,35 @@
    long : 0;
    char bar;
 } ATTR t10;
+int s10 = sizeof(t10);
+// CHECK: @s10 = global i32 16
+// CHECK-ARM: @s10 = global i32 8
 
-static int arr1[(sizeof(t1) == 2) -1];
-static int arr2[(sizeof(t2) == 2) -1];
-static int arr3[(sizeof(t3) == 2) -1];
-static int arr4[(sizeof(t4) == 1) -1];
-static int arr5[(sizeof(t5) == 1) -1];
-static int arr6[(sizeof(t6) == 1) -1];
-static int arr7[(sizeof(t7) == 9) -1];
-static int arr8[(sizeof(t8) == 0) -1];
-static int arr9[(sizeof(t9) == 28) -1];
-static int arr10[(sizeof(t10) == 16) -1];
+// rdar://16041826 - ensure that ms_structs work correctly on a
+// !useBitFieldTypeAlignment() target
+struct {
+  unsigned int a : 31;
+  unsigned int b : 2;
+  unsigned int c;
+} ATTR t11;
+int s11 = sizeof(t11);
+// CHECK: @s11 = global i32 12
+// CHECK-ARM: @s11 = global i32 12
 
-int main() {
-  return 0;
-}
+struct {
+  unsigned char a : 3;
+  unsigned char b : 4;
+  unsigned short c : 6;
+} ATTR t12;
+int s12 = sizeof(t12);
+// CHECK: @s12 = global i32 4
+// CHECK-ARM: @s12 = global i32 4
 
+struct {
+  unsigned char a : 3;
+  unsigned char b : 4;
+  __attribute__((packed)) unsigned short c : 6;
+} ATTR t13;
+int s13 = sizeof(t13);
+// CHECK: @s13 = global i32 4
+// CHECK-ARM: @s13 = global i32 4
diff --git a/test/CodeGen/ms_struct-pack.c b/test/CodeGen/ms_struct-pack.c
index da94f54..6486f29 100644
--- a/test/CodeGen/ms_struct-pack.c
+++ b/test/CodeGen/ms_struct-pack.c
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -emit-llvm-only  -triple i386-apple-darwin9 %s
+// RUN: %clang_cc1 -emit-llvm-only  -triple i386-apple-darwin9 -fdump-record-layouts %s | FileCheck %s
 // rdar://8823265
 
 #pragma pack(1)
@@ -123,3 +123,22 @@
 
 static int a8[(sizeof(eight_ms) == 48) - 1];
 
+// rdar://15926990
+#pragma pack(2)
+struct test0 {
+  unsigned long a : 8;
+  unsigned long b : 8;
+  unsigned long c : 8;
+  unsigned long d : 10;
+  unsigned long e : 1;
+} __attribute__((__ms_struct__));
+
+// CHECK:      Type: struct test0
+// CHECK-NEXT: Record:
+// CHECK-NEXT: Layout:
+// CHECK-NEXT:   Size:64
+// CHECK-NEXT:   DataSize:64
+// CHECK-NEXT:   Alignment:16
+// CHECK-NEXT:   FieldOffsets: [0, 8, 16, 32, 42]>
+
+static int test0[(sizeof(struct test0) == 8) ? 1 : -1];
diff --git a/test/CodeGen/neon-crypto.c b/test/CodeGen/neon-crypto.c
new file mode 100644
index 0000000..5dcef4b
--- /dev/null
+++ b/test/CodeGen/neon-crypto.c
@@ -0,0 +1,95 @@
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -target-feature +crypto -emit-llvm -O1 -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple arm-none-linux-gnueabi -target-feature +neon \
+// RUN:  -target-feature +crypto -target-cpu cortex-a57 -emit-llvm -O1 -o - %s | FileCheck %s
+// RUN: not %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +neon \
+// RUN:   -S -O3 -o - %s 2>&1 | FileCheck --check-prefix=CHECK-NO-CRYPTO %s
+
+// Test new aarch64 intrinsics and types
+
+#include <arm_neon.h>
+
+uint8x16_t test_vaeseq_u8(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: @test_vaeseq_u8
+  // CHECK-NO-CRYPTO: warning: implicit declaration of function 'vaeseq_u8' is invalid in C99
+  return vaeseq_u8(data, key);
+  // CHECK: call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %data, <16 x i8> %key)
+}
+
+uint8x16_t test_vaesdq_u8(uint8x16_t data, uint8x16_t key) {
+  // CHECK-LABEL: @test_vaesdq_u8
+  return vaesdq_u8(data, key);
+  // CHECK: call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %data, <16 x i8> %key)
+}
+
+uint8x16_t test_vaesmcq_u8(uint8x16_t data) {
+  // CHECK-LABEL: @test_vaesmcq_u8
+  return vaesmcq_u8(data);
+  // CHECK: call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %data)
+}
+
+uint8x16_t test_vaesimcq_u8(uint8x16_t data) {
+  // CHECK-LABEL: @test_vaesimcq_u8
+  return vaesimcq_u8(data);
+  // CHECK: call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %data)
+}
+
+uint32_t test_vsha1h_u32(uint32_t hash_e) {
+  // CHECK-LABEL: @test_vsha1h_u32
+  return vsha1h_u32(hash_e);
+  // CHECK: call i32 @llvm.arm.neon.sha1h(i32 %hash_e)
+}
+
+uint32x4_t test_vsha1su1q_u32(uint32x4_t w0_3, uint32x4_t w12_15) {
+  // CHECK-LABEL: @test_vsha1su1q_u32
+  return vsha1su1q_u32(w0_3, w12_15);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha1su1(<4 x i32> %w0_3, <4 x i32> %w12_15)
+}
+
+uint32x4_t test_vsha256su0q_u32(uint32x4_t w0_3, uint32x4_t w4_7) {
+  // CHECK-LABEL: @test_vsha256su0q_u32
+  return vsha256su0q_u32(w0_3, w4_7);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha256su0(<4 x i32> %w0_3, <4 x i32> %w4_7)
+}
+
+uint32x4_t test_vsha1cq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: @test_vsha1cq_u32
+  return vsha1cq_u32(hash_abcd, hash_e, wk);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha1c(<4 x i32> %hash_abcd, i32 %hash_e, <4 x i32> %wk)
+}
+
+uint32x4_t test_vsha1pq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: @test_vsha1pq_u32
+  return vsha1pq_u32(hash_abcd, hash_e, wk);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha1p(<4 x i32> %hash_abcd, i32 %hash_e, <4 x i32> %wk)
+}
+
+uint32x4_t test_vsha1mq_u32(uint32x4_t hash_abcd, uint32_t hash_e, uint32x4_t wk) {
+  // CHECK-LABEL: @test_vsha1mq_u32
+  return vsha1mq_u32(hash_abcd, hash_e, wk);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha1m(<4 x i32> %hash_abcd, i32 %hash_e, <4 x i32> %wk)
+}
+
+uint32x4_t test_vsha1su0q_u32(uint32x4_t w0_3, uint32x4_t w4_7, uint32x4_t w8_11) {
+  // CHECK-LABEL: @test_vsha1su0q_u32
+  return vsha1su0q_u32(w0_3, w4_7, w8_11);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha1su0(<4 x i32> %w0_3, <4 x i32> %w4_7, <4 x i32> %w8_11)
+}
+
+uint32x4_t test_vsha256hq_u32(uint32x4_t hash_abcd, uint32x4_t hash_efgh, uint32x4_t wk) {
+  // CHECK-LABEL: @test_vsha256hq_u32
+  return vsha256hq_u32(hash_abcd, hash_efgh, wk);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha256h(<4 x i32> %hash_abcd, <4 x i32> %hash_efgh, <4 x i32> %wk)
+}
+
+uint32x4_t test_vsha256h2q_u32(uint32x4_t hash_efgh, uint32x4_t hash_abcd, uint32x4_t wk) {
+  // CHECK-LABEL: @test_vsha256h2q_u32
+  return vsha256h2q_u32(hash_efgh, hash_abcd, wk);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha256h2(<4 x i32> %hash_efgh, <4 x i32> %hash_abcd, <4 x i32> %wk)
+}
+
+uint32x4_t test_vsha256su1q_u32(uint32x4_t w0_3, uint32x4_t w8_11, uint32x4_t w12_15) {
+  // CHECK-LABEL: @test_vsha256su1q_u32
+  return vsha256su1q_u32(w0_3, w8_11, w12_15);
+  // CHECK: call <4 x i32> @llvm.arm.neon.sha256su1(<4 x i32> %w0_3, <4 x i32> %w8_11, <4 x i32> %w12_15)
+}
diff --git a/test/CodeGen/noduplicate-cxx11-test.cpp b/test/CodeGen/noduplicate-cxx11-test.cpp
new file mode 100644
index 0000000..0127863
--- /dev/null
+++ b/test/CodeGen/noduplicate-cxx11-test.cpp
@@ -0,0 +1,20 @@
+// RUN: %clang_cc1 -triple=i686-pc-unknown -std=c++11 %s  -emit-llvm -o - | FileCheck %s
+
+// This was a problem in Sema, but only shows up as noinline missing
+// in CodeGen.
+
+// CHECK: define i32 @_Z15noduplicatedfuni(i32 %a) [[NI:#[0-9]+]]
+
+int noduplicatedfun [[clang::noduplicate]] (int a) {
+
+  return a+1;
+
+}
+
+int main() {
+
+  return noduplicatedfun(5);
+
+}
+
+// CHECK: attributes [[NI]] = { noduplicate nounwind{{.*}} }
diff --git a/test/CodeGen/noinline.c b/test/CodeGen/noinline.c
index e64a1a5..bd5a9d8 100644
--- a/test/CodeGen/noinline.c
+++ b/test/CodeGen/noinline.c
@@ -5,10 +5,17 @@
 
 inline int dont_inline_me(int a, int b) { return(a+b); }
 
+inline __attribute__ ((__always_inline__)) int inline_me(int a, int b) { return(a*b); }
+
 volatile int *pa = (int*) 0x1000;
 void foo() {
 // NOINLINE: @foo
 // NOINLINE: dont_inline_me
 // NOINLINE-NOT: inlinehint
     pa[0] = dont_inline_me(pa[1],pa[2]);	
+// NOINLINE-NOT: inline_me
+    pa[3] = inline_me(pa[4],pa[5]);
 }
+
+// NOINLINE: Function Attrs: noinline
+// NOINLINE: @dont_inline_me
diff --git a/test/CodeGen/overloadable.c b/test/CodeGen/overloadable.c
index 1ed72b1..8b40e4d 100644
--- a/test/CodeGen/overloadable.c
+++ b/test/CodeGen/overloadable.c
@@ -1,4 +1,6 @@
-// RUN: %clang_cc1 -emit-llvm %s -o - | grep _Z1fPA10_1X
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm %s -o - | FileCheck %s
+// CHECK: _Z1fPA10_1X
+
 int __attribute__((overloadable)) f(int x) { return x; }
 float __attribute__((overloadable)) f(float x) { return x; }
 double __attribute__((overloadable)) f(double x) { return x; }
diff --git a/test/CodeGen/packed-nest-unpacked.c b/test/CodeGen/packed-nest-unpacked.c
index 3931741..ea45660 100644
--- a/test/CodeGen/packed-nest-unpacked.c
+++ b/test/CodeGen/packed-nest-unpacked.c
@@ -60,6 +60,6 @@
 
 unsigned test7() {
   // CHECK: @test7
-  // CHECK: load i32* bitcast (%struct.XBitfield* getelementptr inbounds (%struct.YBitfield* @gbitfield, i32 0, i32 1) to i32*), align 4
+  // CHECK: load i32* getelementptr inbounds (%struct.YBitfield* @gbitfield, i32 0, i32 1, i32 0), align 4
   return gbitfield.y.b2;
 }
diff --git a/test/CodeGen/packed-union.c b/test/CodeGen/packed-union.c
index 31ce614..cb49999 100644
--- a/test/CodeGen/packed-union.c
+++ b/test/CodeGen/packed-union.c
@@ -1,11 +1,11 @@
-// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -emit-llvm %s -o %t
+// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck %s
 
 typedef struct _attrs {
         unsigned file_attributes;
         unsigned char filename_length;
 } __attribute__((__packed__)) attrs;
 
-// RUN: grep "union._attr_union = type <{ i32, i8 }>" %t
+// CHECK: %union._attr_union = type <{ i32, i8 }>
 typedef union _attr_union {
   attrs file_attrs;
   unsigned owner_id;
diff --git a/test/CodeGen/powerpc_types.c b/test/CodeGen/powerpc_types.c
index b4de318..8b2b156 100644
--- a/test/CodeGen/powerpc_types.c
+++ b/test/CodeGen/powerpc_types.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc32-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc-unknown-freebsd -emit-llvm -o - %s| FileCheck -check-prefix=SVR4-CHECK %s
 
 #include <stdarg.h>
diff --git a/test/CodeGen/ppc64-align-long-double.c b/test/CodeGen/ppc64-align-long-double.c
index c4dcfa0..6d07f70 100644
--- a/test/CodeGen/ppc64-align-long-double.c
+++ b/test/CodeGen/ppc64-align-long-double.c
@@ -1,8 +1,6 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
-// CHECK: -f128:128:128-
-
 struct S {
   double a;
   long double b;
diff --git a/test/CodeGen/ppc64-complex-parms.c b/test/CodeGen/ppc64-complex-parms.c
index 92a6fa5..ec56f9f 100644
--- a/test/CodeGen/ppc64-complex-parms.c
+++ b/test/CodeGen/ppc64-complex-parms.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 float crealf(_Complex float);
diff --git a/test/CodeGen/ppc64-complex-return.c b/test/CodeGen/ppc64-complex-return.c
index b3fd549..cdb0ed6 100644
--- a/test/CodeGen/ppc64-complex-return.c
+++ b/test/CodeGen/ppc64-complex-return.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 float crealf(_Complex float);
diff --git a/test/CodeGen/ppc64-extend.c b/test/CodeGen/ppc64-extend.c
index d46b651..52e5f13 100644
--- a/test/CodeGen/ppc64-extend.c
+++ b/test/CodeGen/ppc64-extend.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 void f1(int x) { return; }
diff --git a/test/CodeGen/ppc64-inline-asm.c b/test/CodeGen/ppc64-inline-asm.c
new file mode 100644
index 0000000..552fe28
--- /dev/null
+++ b/test/CodeGen/ppc64-inline-asm.c
@@ -0,0 +1,26 @@
+// RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -O2 -emit-llvm -o - %s | FileCheck %s
+
+_Bool test_wc_i1(_Bool b1, _Bool b2) {
+  _Bool o;
+  asm("crand %0, %1, %2" : "=wc"(o) : "wc"(b1), "wc"(b2) : );
+  return o;
+// CHECK-LABEL: define zeroext i1 @test_wc_i1(i1 zeroext %b1, i1 zeroext %b2)
+// CHECK: call i8 asm "crand $0, $1, $2", "=^wc,^wc,^wc"(i1 %b1, i1 %b2)
+}
+
+int test_wc_i32(int b1, int b2) {
+  int o;
+  asm("crand %0, %1, %2" : "=wc"(o) : "wc"(b1), "wc"(b2) : );
+  return o;
+// CHECK-LABEL: signext i32 @test_wc_i32(i32 signext %b1, i32 signext %b2)
+// CHECK: call i32 asm "crand $0, $1, $2", "=^wc,^wc,^wc"(i32 %b1, i32 %b2)
+}
+
+unsigned char test_wc_i8(unsigned char b1, unsigned char b2) {
+  unsigned char o;
+  asm("crand %0, %1, %2" : "=wc"(o) : "wc"(b1), "wc"(b2) : );
+  return o;
+// CHECK-LABEL: zeroext i8 @test_wc_i8(i8 zeroext %b1, i8 zeroext %b2)
+// CHECK: call i8 asm "crand $0, $1, $2", "=^wc,^wc,^wc"(i8 %b1, i8 %b2)
+}
+
diff --git a/test/CodeGen/ppc64-struct-onefloat.c b/test/CodeGen/ppc64-struct-onefloat.c
index e26987f..11b16a4 100644
--- a/test/CodeGen/ppc64-struct-onefloat.c
+++ b/test/CodeGen/ppc64-struct-onefloat.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 typedef struct s1 { float f; } Sf;
diff --git a/test/CodeGen/ppc64-struct-onevect.c b/test/CodeGen/ppc64-struct-onevect.c
index a5a1232..34f4cc0 100644
--- a/test/CodeGen/ppc64-struct-onevect.c
+++ b/test/CodeGen/ppc64-struct-onevect.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -O2 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 typedef float v4sf __attribute__ ((vector_size (16)));
diff --git a/test/CodeGen/ppc64-varargs-complex.c b/test/CodeGen/ppc64-varargs-complex.c
index b65a773..8fc8839 100644
--- a/test/CodeGen/ppc64-varargs-complex.c
+++ b/test/CodeGen/ppc64-varargs-complex.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 #include <stdarg.h>
diff --git a/test/CodeGen/ppc64-varargs-struct.c b/test/CodeGen/ppc64-varargs-struct.c
index 61c33b0..70b242c 100644
--- a/test/CodeGen/ppc64-varargs-struct.c
+++ b/test/CodeGen/ppc64-varargs-struct.c
@@ -1,4 +1,4 @@
-// REQUIRES: ppc64-registered-target
+// REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
 
 #include <stdarg.h>
diff --git a/test/CodeGen/pr18235.c b/test/CodeGen/pr18235.c
new file mode 100644
index 0000000..d3f12ee
--- /dev/null
+++ b/test/CodeGen/pr18235.c
@@ -0,0 +1,3 @@
+// RUN: not %clang_cc1 -triple le32-unknown-nacl %s -S -o - 2>&1 | FileCheck %s
+
+// CHECK: error: unable to create target: 'No available targets are compatible with this triple, see -version for the available targets.'
diff --git a/test/CodeGen/pragma-pack-1.c b/test/CodeGen/pragma-pack-1.c
index 2a71c8f..773318c 100644
--- a/test/CodeGen/pragma-pack-1.c
+++ b/test/CodeGen/pragma-pack-1.c
@@ -53,12 +53,12 @@
   int e;
 } s4;
 
-// CHECK: [[struct_ref:%[a-zA-Z0-9_.]+]] = type <{ [[struct_ref]]* }>
+// CHECK: [[struct_ref:%[a-zA-Z0-9_.]+]] = type { [[struct_ref]]* }
 // CHECK: [[struct_S:%[a-zA-Z0-9_.]+]] = type { [3 x i8], [[struct_T:%[a-zA-Z0-9_.]+]], [[struct_T2:%[a-zA-Z0-9_.]+]] }
 // CHECK: [[struct_T]] = type <{ i8, i32 }>
 // CHECK: [[struct_T2]] = type { i8, i32 }
 
-// CHECK: %struct.S3 = type <{ [3 x i8], i8, %struct.T3, [2 x i8], %struct.T32 }>
+// CHECK: %struct.S3 = type { [3 x i8], i8, %struct.T3, %struct.T32 }
 // CHECK: %struct.T3 = type <{ i8, i8, i32 }>
 // CHECK: %struct.T32 = type { i8, i32 }
 // CHECK: %struct.S4 = type { [3 x i8], %struct.T4, i32 }
diff --git a/test/CodeGen/pragma-pack-2.c b/test/CodeGen/pragma-pack-2.c
index 1ca3bdf..8a77280 100644
--- a/test/CodeGen/pragma-pack-2.c
+++ b/test/CodeGen/pragma-pack-2.c
@@ -4,7 +4,7 @@
 
 // RUN: %clang_cc1 -triple x86_64-apple-darwin9 %s -emit-llvm -o - | FileCheck -check-prefix CHECK-X64 %s
 // CHECK-X64: %struct.s0 = type <{ i64, i64, i32, [12 x i32] }>
-// CHECK-X64: %struct.s1 = type <{ [15 x i32], %struct.s0 }>
+// CHECK-X64: %struct.s1 = type { [15 x i32], %struct.s0 }
 
 // rdar://problem/7095436
 #pragma pack(4)
@@ -20,4 +20,3 @@
   int a[15];
   struct s0 b;
 } b;
-
diff --git a/test/CodeGen/r5.c b/test/CodeGen/r5.c
deleted file mode 100644
index 30a0c0d..0000000
--- a/test/CodeGen/r5.c
+++ /dev/null
@@ -1,5 +0,0 @@
-// RUN: %clang -target armv7-none-linux-gnueabi -mcpu=cortex-r5 -emit-llvm -S %s  -o /dev/null
-
-int main() {
-  return 0;
-}
diff --git a/test/CodeGen/sparc-target-data.c b/test/CodeGen/sparc-target-data.c
deleted file mode 100644
index bb32a21..0000000
--- a/test/CodeGen/sparc-target-data.c
+++ /dev/null
@@ -1,5 +0,0 @@
-// RUN: %clang -target sparc-sun-solaris   -o - -emit-llvm -S %s | FileCheck %s -check-prefix=V8
-// RUN: %clang -target sparcv9-sun-solaris -o - -emit-llvm -S %s | FileCheck %s -check-prefix=V9
-
-// V8: E-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-n32-S64
-// V9: E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-n32:64-S128
diff --git a/test/CodeGen/sparcv9-abi.c b/test/CodeGen/sparcv9-abi.c
index 4ba4be8..d4fff81 100644
--- a/test/CodeGen/sparcv9-abi.c
+++ b/test/CodeGen/sparcv9-abi.c
@@ -18,6 +18,9 @@
 // CHECK-LABEL: define signext i8 @f_int_4(i8 signext %x)
 char f_int_4(char x) { return x; }
 
+// CHECK-LABEL: define fp128 @f_ld(fp128 %x)
+long double f_ld(long double x) { return x; }
+
 // Small structs are passed in registers.
 struct small {
   int *a, *b;
diff --git a/test/CodeGen/sparcv9-dwarf.c b/test/CodeGen/sparcv9-dwarf.c
new file mode 100644
index 0000000..11ac28c
--- /dev/null
+++ b/test/CodeGen/sparcv9-dwarf.c
@@ -0,0 +1,99 @@
+// RUN: %clang_cc1 -triple sparcv9-unknown-unknown -emit-llvm %s -o - | FileCheck %s
+static unsigned char dwarf_reg_size_table[102+1];
+
+int test() {
+  __builtin_init_dwarf_reg_size_table(dwarf_reg_size_table);
+
+  return __builtin_dwarf_sp_column();
+}
+
+// CHECK-LABEL: define signext i32 @test()
+// CHECK:       store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 0)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 1)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 2)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 3)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 4)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 5)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 6)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 7)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 8)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 9)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 10)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 11)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 12)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 13)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 14)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 15)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 16)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 17)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 18)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 19)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 20)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 21)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 22)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 23)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 24)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 25)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 26)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 27)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 28)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 29)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 30)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 31)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 32)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 33)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 34)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 35)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 36)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 37)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 38)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 39)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 40)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 41)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 42)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 43)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 44)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 45)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 46)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 47)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 48)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 49)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 50)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 51)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 52)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 53)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 54)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 55)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 56)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 57)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 58)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 59)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 60)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 61)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 62)
+// CHECK-NEXT:  store i8 4, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 63)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 64)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 65)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 66)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 67)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 68)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 69)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 70)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 71)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 72)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 73)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 74)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 75)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 76)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 77)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 78)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 79)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 80)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 81)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 82)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 83)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 84)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 85)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 86)
+// CHECK-NEXT:  store i8 8, i8* getelementptr inbounds ([103 x i8]* @dwarf_reg_size_table, i32 0, i32 87)
+// CHECK-NEXT:  ret i32 14
diff --git a/test/CodeGen/sret.c b/test/CodeGen/sret.c
index 828bf9b..5f0d074 100644
--- a/test/CodeGen/sret.c
+++ b/test/CodeGen/sret.c
@@ -4,6 +4,8 @@
  long a;
  long b;
  long c;
+ long d;
+ long e;
 };
  
 struct abc foo1(void);
diff --git a/test/CodeGen/sret2.c b/test/CodeGen/sret2.c
index 3757462..d103d87 100644
--- a/test/CodeGen/sret2.c
+++ b/test/CodeGen/sret2.c
@@ -4,6 +4,8 @@
  long a;
  long b;
  long c;
+ long d;
+ long e;
 };
  
 struct abc foo2(){}
diff --git a/test/CodeGen/stack-protector.c b/test/CodeGen/stack-protector.c
index e47e5b3..2fb9b2c 100644
--- a/test/CodeGen/stack-protector.c
+++ b/test/CodeGen/stack-protector.c
@@ -2,7 +2,9 @@
 // NOSSP: define void @test1(i8* %msg) #0 {
 // RUN: %clang_cc1 -emit-llvm -o - %s -stack-protector 1 | FileCheck -check-prefix=WITHSSP %s
 // WITHSSP: define void @test1(i8* %msg) #0 {
-// RUN: %clang_cc1 -emit-llvm -o - %s -stack-protector 2 | FileCheck -check-prefix=SSPREQ %s
+// RUN: %clang_cc1 -emit-llvm -o - %s -stack-protector 2 | FileCheck -check-prefix=SSPSTRONG %s
+// SSPSTRONG: define void @test1(i8* %msg) #0 {
+// RUN: %clang_cc1 -emit-llvm -o - %s -stack-protector 3 | FileCheck -check-prefix=SSPREQ %s
 // SSPREQ: define void @test1(i8* %msg) #0 {
 
 typedef __SIZE_TYPE__ size_t;
@@ -21,4 +23,6 @@
 
 // WITHSSP: attributes #{{.*}} = { nounwind ssp{{.*}} }
 
+// SSPSTRONG: attributes #{{.*}} = { nounwind sspstrong{{.*}} }
+
 // SSPREQ: attributes #{{.*}} = { nounwind sspreq{{.*}} }
diff --git a/test/CodeGen/string-literal-short-wstring.c b/test/CodeGen/string-literal-short-wstring.c
index 88e4a1e..89aa6f7 100644
--- a/test/CodeGen/string-literal-short-wstring.c
+++ b/test/CodeGen/string-literal-short-wstring.c
@@ -1,4 +1,5 @@
-// RUN: %clang_cc1 -x c++ -emit-llvm -fshort-wchar %s -o - | FileCheck %s
+// RUN: %clang_cc1 -x c++ -triple %itanium_abi_triple -emit-llvm -fshort-wchar %s -o - | FileCheck %s --check-prefix=CHECK --check-prefix=ITANIUM
+// RUN: %clang_cc1 -x c++ -triple %ms_abi_triple -emit-llvm -fshort-wchar %s -o - | FileCheck %s --check-prefix=CHECK --check-prefix=MSABI
 // Runs in c++ mode so that wchar_t is available.
 
 int main() {
@@ -6,11 +7,13 @@
   // CHECK: private unnamed_addr constant [10 x i8] c"\E1\84\A0\C8\A0\F4\82\80\B0\00", align 1
   char b[10] = "\u1120\u0220\U00102030";
 
-  // CHECK: private unnamed_addr constant [3 x i16] [i16 65, i16 66, i16 0]
+  // ITANIUM: private unnamed_addr constant [3 x i16] [i16 65, i16 66, i16 0]
+  // MSABI: linkonce_odr unnamed_addr constant [3 x i16] [i16 65, i16 66, i16 0]
   const wchar_t *foo = L"AB";
 
   // This should convert to utf16.
-  // CHECK: private unnamed_addr constant [5 x i16] [i16 4384, i16 544, i16 -9272, i16 -9168, i16 0]
+  // ITANIUM: private unnamed_addr constant [5 x i16] [i16 4384, i16 544, i16 -9272, i16 -9168, i16 0]
+  // MSABI: linkonce_odr unnamed_addr constant [5 x i16] [i16 4384, i16 544, i16 -9272, i16 -9168, i16 0]
   const wchar_t *bar = L"\u1120\u0220\U00102030";
 
 
diff --git a/test/CodeGen/struct-x86-darwin.c b/test/CodeGen/struct-x86-darwin.c
index afdcb8a..5191441 100644
--- a/test/CodeGen/struct-x86-darwin.c
+++ b/test/CodeGen/struct-x86-darwin.c
@@ -1,13 +1,13 @@
-// RUN: %clang_cc1 < %s -emit-llvm > %t1 -triple=i686-apple-darwin9
-// RUN: grep "STest1 = type { i32, \[4 x i16\], double }" %t1
-// RUN: grep "STest2 = type { i16, i16, i32, i32 }" %t1
-// RUN: grep "STest3 = type { i8, i16, i32 }" %t1
-// RUN: grep "STestB1 = type { i8, i8 }" %t1
-// RUN: grep "STestB2 = type { i8, i8, i8 }" %t1
-// RUN: grep "STestB3 = type { i8, i8 }" %t1
-// RUN: grep "STestB4 = type { i8, i8, i8, i8 }" %t1
-// RUN: grep "STestB5 = type { i8, i8, \[2 x i8\], i8, i8 }" %t1
-// RUN: grep "STestB6 = type { i8, i8, \[2 x i8\] }" %t1
+// RUN: %clang_cc1 %s -emit-llvm -triple=i686-apple-darwin9 -o - | FileCheck %s
+// CHECK: STest1 = type { i32, [4 x i16], double }
+// CHECK: STest2 = type { i16, i16, i32, i32 }
+// CHECK: STest3 = type { i8, i16, i32 }
+// CHECK: STestB1 = type { i8, i8 }
+// CHECK: STestB2 = type { i8, i8, i8 }
+// CHECK: STestB3 = type { i8, i8 }
+// CHECK: STestB4 = type { i8, i8, i8, i8 }
+// CHECK: STestB5 = type { i8, i16, i8 }
+// CHECK: STestB6 = type { i8, i8, i16 }
 // Test struct layout for x86-darwin target
 
 struct STest1 {int x; short y[4]; double z; } st1;
diff --git a/test/CodeGen/target-data.c b/test/CodeGen/target-data.c
index fc8f758..0cc46b8 100644
--- a/test/CodeGen/target-data.c
+++ b/test/CodeGen/target-data.c
@@ -1,6 +1,167 @@
-// RUN: %clang_cc1 -triple i686-unknown-unknown -emit-llvm -o %t %s
-// RUN: grep 'target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32-S128"' %t
-// RUN: %clang_cc1 -triple i686-apple-darwin9 -emit-llvm -o %t %s
-// RUN: grep 'target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32-S128"' %t
-// RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm -o %t %s
-// RUN: grep 'target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"' %t
+// RUN: %clang_cc1 -triple i686-unknown-unknown -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=I686-UNKNOWN %s
+// I686-UNKNOWN: target datalayout = "e-m:e-p:32:32-f64:32:64-f80:32-n8:16:32-S128"
+
+// RUN: %clang_cc1 -triple i686-apple-darwin9 -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=I686-DARWIN %s
+// I686-DARWIN: target datalayout = "e-m:o-p:32:32-f64:32:64-f80:128-n8:16:32-S128"
+
+// RUN: %clang_cc1 -triple i686-unknown-win32 -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=I686-WIN32 %s
+// I686-WIN32: target datalayout = "e-m:w-p:32:32-i64:64-f80:32-n8:16:32-S32"
+
+// RUN: %clang_cc1 -triple i686-unknown-cygwin -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=I686-CYGWIN %s
+// I686-CYGWIN: target datalayout = "e-m:w-p:32:32-i64:64-f80:32-n8:16:32-S32"
+
+// RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=X86_64 %s
+// X86_64: target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+
+// RUN: %clang_cc1 -triple xcore-unknown-unknown -emit-llvm -o - %s | \
+// RUN:     FileCheck --check-prefix=XCORE %s
+// XCORE: target datalayout = "e-m:e-p:32:32-i1:8:32-i8:8:32-i16:16:32-i64:32-f64:32-a:0:32-n32"
+
+// RUN: %clang_cc1 -triple sparc-sun-solaris -emit-llvm -o - %s | \
+// RUN:     FileCheck %s --check-prefix=SPARC-V8
+// SPARC-V8: target datalayout = "E-m:e-p:32:32-i64:64-f128:64-n32-S64"
+
+// RUN: %clang_cc1 -triple sparcv9-sun-solaris -emit-llvm -o - %s | \
+// RUN: FileCheck %s --check-prefix=SPARC-V9
+// SPARC-V9: target datalayout = "E-m:e-i64:64-n32:64-S128"
+
+// RUN: %clang_cc1 -triple mipsel-linux-gnu -o - -emit-llvm %s |     \
+// RUN: FileCheck %s -check-prefix=MIPS-32EL
+// MIPS-32EL: target datalayout = "e-m:m-p:32:32-i8:8:32-i16:16:32-i64:64-n32-S64"
+
+// RUN: %clang_cc1 -triple mips-linux-gnu -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=MIPS-32EB
+// MIPS-32EB: target datalayout = "E-m:m-p:32:32-i8:8:32-i16:16:32-i64:64-n32-S64"
+
+// RUN: %clang_cc1 -triple mips64el-linux-gnu -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=MIPS-64EL
+// MIPS-64EL: target datalayout = "e-m:m-i8:8:32-i16:16:32-i64:64-n32:64-S128"
+
+// RUN: %clang_cc1 -triple mips64el-linux-gnu -o - -emit-llvm -target-abi n32 \
+// RUN: %s | FileCheck %s -check-prefix=MIPS-64EL-N32
+// MIPS-64EL-N32: target datalayout = "e-m:m-p:32:32-i8:8:32-i16:16:32-i64:64-n32:64-S128"
+
+// RUN: %clang_cc1 -triple mips64-linux-gnu -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=MIPS-64EB
+// MIPS-64EB: target datalayout = "E-m:m-i8:8:32-i16:16:32-i64:64-n32:64-S128"
+
+// RUN: %clang_cc1 -triple mips64-linux-gnu -o - -emit-llvm %s -target-abi n32 \
+// RUN: | FileCheck %s -check-prefix=MIPS-64EB-N32
+// MIPS-64EB-N32: target datalayout = "E-m:m-p:32:32-i8:8:32-i16:16:32-i64:64-n32:64-S128"
+
+// RUN: %clang_cc1 -triple powerpc64-lv2 -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PS3
+// PS3: target datalayout = "E-m:e-p:32:32-i64:64-n32:64"
+
+// RUN: %clang_cc1 -triple i686-nacl -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=I686-NACL
+// I686-NACL: target datalayout = "e-m:e-p:32:32-i64:64-n8:16:32-S128"
+
+// RUN: %clang_cc1 -triple x86_64-nacl -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=X86_64-NACL
+// X86_64-NACL: target datalayout = "e-m:e-p:32:32-i64:64-n8:16:32:64-S128"
+
+// RUN: %clang_cc1 -triple arm-nacl-gnueabi -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=ARM-NACL
+// ARM-NACL: target datalayout = "e-m:e-p:32:32-i64:64-v128:64:128-n32-S128"
+
+// RUN: %clang_cc1 -triple mipsel-nacl -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=MIPS-NACL
+// MIPS-NACL: target datalayout = "e-m:m-p:32:32-i8:8:32-i16:16:32-i64:64-n32-S64"
+
+// RUN: %clang_cc1 -triple le32-nacl -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=LE32-NACL
+// LE32-NACL: target datalayout = "e-p:32:32-i64:64"
+
+// RUN: %clang_cc1 -triple powerpc-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC
+// PPC: target datalayout = "E-m:e-p:32:32-i64:64-n32"
+
+// RUN: %clang_cc1 -triple powerpc64-freebsd -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC64-FREEBSD
+// PPC64-FREEBSD: target datalayout = "E-m:e-i64:64-n32:64"
+
+// RUN: %clang_cc1 -triple powerpc64-linux -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC64-LINUX
+// PPC64-LINUX: target datalayout = "E-m:e-i64:64-n32:64"
+
+// RUN: %clang_cc1 -triple powerpc64le-linux -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC64LE-LINUX
+// PPC64LE-LINUX: target datalayout = "e-m:e-i64:64-n32:64"
+
+// RUN: %clang_cc1 -triple powerpc-darwin -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC32-DARWIN
+// PPC32-DARWIN: target datalayout = "E-m:o-p:32:32-f64:32:64-n32"
+
+// RUN: %clang_cc1 -triple powerpc64-darwin -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=PPC64-DARWIN
+// PPC64-DARWIN: target datalayout = "E-m:o-i64:64-n32:64"
+
+// RUN: %clang_cc1 -triple nvptx-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=NVPTX
+// NVPTX: target datalayout = "e-p:32:32-i64:64-v16:16-v32:32-n16:32:64"
+
+// RUN: %clang_cc1 -triple nvptx64-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=NVPTX64
+// NVPTX64: target datalayout = "e-i64:64-v16:16-v32:32-n16:32:64"
+
+// RUN: %clang_cc1 -triple r600-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=R600
+// R600: target datalayout = "e-p:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64"
+
+// RUN: %clang_cc1 -triple r600-unknown -target-cpu cayman -o - -emit-llvm %s \
+// RUN: | FileCheck %s -check-prefix=R600D
+// R600D: target datalayout = "e-p:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64"
+
+// RUN: %clang_cc1 -triple r600-unknown -target-cpu hawaii -o - -emit-llvm %s \
+// RUN: | FileCheck %s -check-prefix=R600SI
+// R600SI: target datalayout = "e-p:32:32-p1:64:64-p2:64:64-p3:32:32-p4:32:32-p5:64:64-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64"
+
+// RUN: %clang_cc1 -triple aarch64-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=AARCH64
+// AARCH64: target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
+
+// RUN: %clang_cc1 -triple thumb-unknown-gnueabi -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=THUMB
+// THUMB: target datalayout = "e-m:e-p:32:32-i1:8:32-i8:8:32-i16:16:32-i64:64-v128:64:128-a:0:32-n32-S64"
+
+// RUN: %clang_cc1 -triple arm-unknown-gnueabi -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=ARM
+// ARM: target datalayout = "e-m:e-p:32:32-i64:64-v128:64:128-n32-S64"
+
+// RUN: %clang_cc1 -triple thumb-unknown -o - -emit-llvm -target-abi apcs-gnu \
+// RUN: %s | FileCheck %s -check-prefix=THUMB-GNU
+// THUMB-GNU: target datalayout = "e-m:e-p:32:32-i1:8:32-i8:8:32-i16:16:32-f64:32:64-v64:32:64-v128:32:128-a:0:32-n32-S32"
+
+// RUN: %clang_cc1 -triple arm-unknown -o - -emit-llvm -target-abi apcs-gnu \
+// RUN: %s | FileCheck %s -check-prefix=ARM-GNU
+// ARM-GNU: target datalayout = "e-m:e-p:32:32-f64:32:64-v64:32:64-v128:32:128-a:0:32-n32-S32"
+
+// RUN: %clang_cc1 -triple hexagon-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=HEXAGON
+// HEXAGON: target datalayout = "e-m:e-p:32:32-i1:32-i64:64-a:0-n32"
+
+// RUN: %clang_cc1 -triple s390x-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=SYSTEMZ
+// SYSTEMZ: target datalayout = "E-m:e-i1:8:16-i8:8:16-i64:64-f128:64-a:8:16-n32:64"
+
+// RUN: %clang_cc1 -triple msp430-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=MSP430
+// MSP430: target datalayout = "e-m:e-p:16:16-i32:16:32-n8:16"
+
+// RUN: %clang_cc1 -triple tce-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=TCE
+// TCE: target datalayout = "E-p:32:32-i8:8:32-i16:16:32-i64:32-f64:32-v64:32-v128:32-a:0:32-n32"
+
+// RUN: %clang_cc1 -triple spir-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=SPIR
+// SPIR: target datalayout = "e-p:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024"
+
+// RUN: %clang_cc1 -triple spir64-unknown -o - -emit-llvm %s | \
+// RUN: FileCheck %s -check-prefix=SPIR64
+// SPIR64: target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024"
diff --git a/test/CodeGen/tbaa-for-vptr.cpp b/test/CodeGen/tbaa-for-vptr.cpp
index 7ba058b..ded574e 100644
--- a/test/CodeGen/tbaa-for-vptr.cpp
+++ b/test/CodeGen/tbaa-for-vptr.cpp
@@ -1,12 +1,12 @@
-// RUN: %clang_cc1 -emit-llvm -o - -fsanitize=thread %s | FileCheck %s
-// RUN: %clang_cc1 -emit-llvm -o - -O1 %s | FileCheck %s
-// RUN: %clang_cc1 -emit-llvm -o - -O1  -relaxed-aliasing -fsanitize=thread %s | FileCheck %s
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - -fsanitize=thread %s | FileCheck %s
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - -O1 %s | FileCheck %s
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - -O1  -relaxed-aliasing -fsanitize=thread %s | FileCheck %s
 //
-// RUN: %clang_cc1 -emit-llvm -o - %s | FileCheck %s --check-prefix=NOTBAA
-// RUN: %clang_cc1 -emit-llvm -o - -O2  -relaxed-aliasing %s | FileCheck %s --check-prefix=NOTBAA
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - %s | FileCheck %s --check-prefix=NOTBAA
+// RUN: %clang_cc1 -triple %itanium_abi_triple -emit-llvm -o - -O2  -relaxed-aliasing %s | FileCheck %s --check-prefix=NOTBAA
 //
 // Check that we generate TBAA for vtable pointer loads and stores.
-// When -fthread-sanitizer is used TBAA should be generated at all opt levels
+// When -fsanitize=thread is used TBAA should be generated at all opt levels
 // even if -relaxed-aliasing is present.
 struct A {
   virtual int foo() const ;
@@ -17,12 +17,19 @@
   new A;
 }
 
-void CallFoo(A *a) {
+void CallFoo(A *a, int (A::*fp)() const) {
   a->foo();
+  (a->*fp)();
 }
 
+// CHECK-LABEL: @_Z7CallFoo
 // CHECK: %{{.*}} = load {{.*}} !tbaa ![[NUM:[0-9]+]]
+// CHECK: br i1
+// CHECK: load {{.*}}, !tbaa ![[NUM]]
+//
+// CHECK-LABEL: @_ZN1AC2Ev
 // CHECK: store {{.*}} !tbaa ![[NUM]]
+//
 // CHECK: [[NUM]] = metadata !{metadata [[TYPE:!.*]], metadata [[TYPE]], i64 0}
 // CHECK: [[TYPE]] = metadata !{metadata !"vtable pointer", metadata !{{.*}}
 // NOTBAA-NOT: = metadata !{metadata !"Simple C/C++ TBAA"}
diff --git a/test/CodeGen/tbaa-ms-abi.cpp b/test/CodeGen/tbaa-ms-abi.cpp
index 67390b1..2a9e47e 100644
--- a/test/CodeGen/tbaa-ms-abi.cpp
+++ b/test/CodeGen/tbaa-ms-abi.cpp
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -cxx-abi microsoft -triple i686-pc-win32 -disable-llvm-optzns -emit-llvm -o - -O1 %s | FileCheck %s
+// RUN: %clang_cc1 -triple i686-pc-win32 -disable-llvm-optzns -emit-llvm -o - -O1 %s | FileCheck %s
 //
 // Test that TBAA works in the Microsoft C++ ABI.  We used to error out while
 // attempting to mangle RTTI.
@@ -16,7 +16,7 @@
 // CHECK: store i32 42, i32* {{.*}}, !tbaa [[TAG_A_i32:!.*]]
 }
 
-// CHECK: [[TYPE_CHAR:!.*]] = metadata !{metadata !"omnipotent char", metadata
-// CHECK: [[TYPE_INT:!.*]] = metadata !{metadata !"int", metadata [[TYPE_CHAR]], i64 0}
+// CHECK: [[TYPE_INT:!.*]] = metadata !{metadata !"int", metadata [[TYPE_CHAR:!.*]], i64 0}
+// CHECK: [[TYPE_CHAR]] = metadata !{metadata !"omnipotent char", metadata
 // CHECK: [[TAG_A_i32]] = metadata !{metadata [[TYPE_A:!.*]], metadata [[TYPE_INT]], i64 0}
 // CHECK: [[TYPE_A]] = metadata !{metadata !"?AUStructA@@", metadata [[TYPE_INT]], i64 0}
diff --git a/test/CodeGen/tbaa-struct.cpp b/test/CodeGen/tbaa-struct.cpp
index f8bd124..71d3ed1 100644
--- a/test/CodeGen/tbaa-struct.cpp
+++ b/test/CodeGen/tbaa-struct.cpp
@@ -12,8 +12,7 @@
   *a = *b;
 }
 
-// CHECK: target datalayout = "{{.*}}p:[[P:64|32]]
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i[[P]](i8* %{{.*}}, i8* %{{.*}}, i[[P]] 16, i32 4, i1 false), !tbaa.struct [[TS:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 16, i32 4, i1 false), !tbaa.struct [[TS:!.*]]
 
 struct B {
   char c1;
@@ -25,7 +24,7 @@
   *a = *b;
 }
 
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i[[P]](i8* %{{.*}}, i8* %{{.*}}, i[[P]] 24, i32 4, i1 false), !tbaa.struct [[TS2:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 24, i32 4, i1 false), !tbaa.struct [[TS2:!.*]]
 
 typedef _Complex int T2;
 typedef _Complex char T5;
@@ -37,7 +36,7 @@
   *a = *b;
 }
 
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i[[P]](i8* %{{.*}}, i8* %{{.*}}, i[[P]] 12, i32 4, i1 false), !tbaa.struct [[TS3:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 12, i32 4, i1 false), !tbaa.struct [[TS3:!.*]]
 
 // Make sure that zero-length bitfield works.
 #define ATTR __attribute__ ((ms_struct))
@@ -50,7 +49,7 @@
 void copy4(struct five *a, struct five *b) {
   *a = *b;
 }
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i[[P]](i8* %{{.*}}, i8* %{{.*}}, i[[P]] 3, i32 1, i1 false), !tbaa.struct [[TS4:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 3, i32 1, i1 false), !tbaa.struct [[TS4:!.*]]
 
 struct six {
   char a;
@@ -61,7 +60,7 @@
 void copy5(struct six *a, struct six *b) {
   *a = *b;
 }
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i[[P]](i8* %{{.*}}, i8* %{{.*}}, i[[P]] 6, i32 1, i1 false), !tbaa.struct [[TS5:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 6, i32 1, i1 false), !tbaa.struct [[TS5:!.*]]
 
 // CHECK: [[TS]] = metadata !{i64 0, i64 2, metadata !{{.*}}, i64 4, i64 4, metadata !{{.*}}, i64 8, i64 1, metadata !{{.*}}, i64 12, i64 4, metadata !{{.*}}}
 // CHECK: [[CHAR:!.*]] = metadata !{metadata !"omnipotent char", metadata !{{.*}}}
diff --git a/test/CodeGen/union.c b/test/CodeGen/union.c
index 5c89e2d..7302182 100644
--- a/test/CodeGen/union.c
+++ b/test/CodeGen/union.c
@@ -44,3 +44,16 @@
 T0 t0;
 
 union { int large_bitfield: 31; char c } u2;
+
+struct dt_t_s {
+  union {
+    long long u : 56;
+  } __attribute__((packed));
+};
+struct {
+  struct {
+    struct {
+      struct dt_t_s t;
+    };
+  };
+} a;
diff --git a/test/CodeGen/utf16-cfstrings.c b/test/CodeGen/utf16-cfstrings.c
index d4f214b..dbe9bda 100644
--- a/test/CodeGen/utf16-cfstrings.c
+++ b/test/CodeGen/utf16-cfstrings.c
@@ -1,7 +1,7 @@
 // RUN: %clang_cc1 -triple x86_64-apple-darwin -emit-llvm %s -o - | FileCheck %s
 // <rdar://problem/10655949>
 
-// CHECK: @.str = internal unnamed_addr constant [9 x i16] [i16 252, i16 98, i16 101, i16 114, i16 104, i16 117, i16 110, i16 100, i16 0], align 2
+// CHECK: @.str = private unnamed_addr constant [9 x i16] [i16 252, i16 98, i16 101, i16 114, i16 104, i16 117, i16 110, i16 100, i16 0], section "__TEXT,__ustring", align 2
 
 #define CFSTR __builtin___CFStringMakeConstantString
 
diff --git a/test/CodeGen/volatile-complex.c b/test/CodeGen/volatile-complex.c
index 71e5db6..fd5e52b 100644
--- a/test/CodeGen/volatile-complex.c
+++ b/test/CodeGen/volatile-complex.c
@@ -5,16 +5,14 @@
 //
 // This test assumes that floats are 32-bit aligned and doubles are
 // 64-bit aligned, and uses x86-64 as a target that should have this
-// datalayout.
-
-// CHECK: target datalayout = "{{.*}}f32:32:32-f64:64:64{{.*}}"
+// property.
 
 volatile _Complex float cf;
 volatile _Complex double cd;
 volatile _Complex float cf32 __attribute__((aligned(32)));
 volatile _Complex double cd32 __attribute__((aligned(32)));
 
-// CHECK-LABEL-LABEL: define void @test_cf()
+// CHECK-LABEL: define void @test_cf()
 void test_cf() {
   // CHECK:      load volatile float* getelementptr inbounds ({ float, float }* @cf, i32 0, i32 0), align 4
   // CHECK-NEXT: load volatile float* getelementptr inbounds ({ float, float }* @cf, i32 0, i32 1), align 4
@@ -27,7 +25,7 @@
   // CHECK-NEXT: ret void
 }
 
-// CHECK-LABEL-LABEL: define void @test_cd()
+// CHECK-LABEL: define void @test_cd()
 void test_cd() {
   // CHECK:      load volatile double* getelementptr inbounds ({ double, double }* @cd, i32 0, i32 0), align 8
   // CHECK-NEXT: load volatile double* getelementptr inbounds ({ double, double }* @cd, i32 0, i32 1), align 8
@@ -40,7 +38,7 @@
   // CHECK-NEXT: ret void
 }
 
-// CHECK-LABEL-LABEL: define void @test_cf32()
+// CHECK-LABEL: define void @test_cf32()
 void test_cf32() {
   // CHECK:      load volatile float* getelementptr inbounds ({ float, float }* @cf32, i32 0, i32 0), align 32
   // CHECK-NEXT: load volatile float* getelementptr inbounds ({ float, float }* @cf32, i32 0, i32 1), align 4
@@ -53,7 +51,7 @@
   // CHECK-NEXT: ret void
 }
 
-// CHECK-LABEL-LABEL: define void @test_cd32()
+// CHECK-LABEL: define void @test_cd32()
 void test_cd32() {
   // CHECK:      load volatile double* getelementptr inbounds ({ double, double }* @cd32, i32 0, i32 0), align 32
   // CHECK-NEXT: load volatile double* getelementptr inbounds ({ double, double }* @cd32, i32 0, i32 1), align 8
diff --git a/test/CodeGen/volatile.c b/test/CodeGen/volatile.c
index 0dcdc15..3e891aa 100644
--- a/test/CodeGen/volatile.c
+++ b/test/CodeGen/volatile.c
@@ -1,4 +1,5 @@
-// RUN: %clang_cc1 -emit-llvm < %s | FileCheck %s
+// RUN: %clang_cc1 -triple=%itanium_abi_triple -emit-llvm < %s | FileCheck %s -check-prefix CHECK -check-prefix CHECK-IT
+// RUN: %clang_cc1 -triple=%ms_abi_triple -emit-llvm < %s | FileCheck %s -check-prefix CHECK -check-prefix CHECK-MS
 
 int S;
 volatile int vS;
@@ -83,10 +84,12 @@
 // CHECK: load volatile i32* getelementptr {{.*}} @vF3
 // CHECK: store i32 {{.*}}, i32* [[I]]
   i=BF.x;
-// CHECK: load i8* getelementptr {{.*}} @BF
+// CHECK-IT: load i8* getelementptr {{.*}} @BF
+// CHECK-MS: load i32* getelementptr {{.*}} @BF
 // CHECK: store i32 {{.*}}, i32* [[I]]
   i=vBF.x;
-// CHECK: load volatile i8* getelementptr {{.*}} @vBF
+// CHECK-IT: load volatile i8* getelementptr {{.*}} @vBF
+// CHECK-MS: load volatile i32* getelementptr {{.*}} @vBF
 // CHECK: store i32 {{.*}}, i32* [[I]]
   i=V[3];
 // CHECK: load <4 x i32>* @V
@@ -154,12 +157,16 @@
 // CHECK: store volatile i32 {{.*}}, i32* getelementptr {{.*}} @vF3
   BF.x=i;
 // CHECK: load i32* [[I]]
-// CHECK: load i8* getelementptr {{.*}} @BF
-// CHECK: store i8 {{.*}}, i8* getelementptr {{.*}} @BF
+// CHECK-IT: load i8* getelementptr {{.*}} @BF
+// CHECK-MS: load i32* getelementptr {{.*}} @BF
+// CHECK-IT: store i8 {{.*}}, i8* getelementptr {{.*}} @BF
+// CHECK-MS: store i32 {{.*}}, i32* getelementptr {{.*}} @BF
   vBF.x=i;
 // CHECK: load i32* [[I]]
-// CHECK: load volatile i8* getelementptr {{.*}} @vBF
-// CHECK: store volatile i8 {{.*}}, i8* getelementptr {{.*}} @vBF
+// CHECK-IT: load volatile i8* getelementptr {{.*}} @vBF
+// CHECK-MS: load volatile i32* getelementptr {{.*}} @vBF
+// CHECK-IT: store volatile i8 {{.*}}, i8* getelementptr {{.*}} @vBF
+// CHECK-MS: store volatile i32 {{.*}}, i32* getelementptr {{.*}} @vBF
   V[3]=i;
 // CHECK: load i32* [[I]]
 // CHECK: load <4 x i32>* @V
diff --git a/test/CodeGen/wchar-const.c b/test/CodeGen/wchar-const.c
index 2e9af53..34da249 100644
--- a/test/CodeGen/wchar-const.c
+++ b/test/CodeGen/wchar-const.c
@@ -15,7 +15,7 @@
 
 
 // CHECK-DAR: private unnamed_addr constant [18 x i32] [i32 84,
-// CHECK-WIN: private unnamed_addr constant [18 x i16] [i16 84,
+// CHECK-WIN: linkonce_odr unnamed_addr constant [18 x i16] [i16 84,
 extern void foo(const wchar_t* p);
 int main (int argc, const char * argv[])
 {
diff --git a/test/CodeGen/xcore-abi.c b/test/CodeGen/xcore-abi.c
index 10881de..6dbc44b 100644
--- a/test/CodeGen/xcore-abi.c
+++ b/test/CodeGen/xcore-abi.c
@@ -1,3 +1,4 @@
+// REQUIRES: xcore-registered-target
 // RUN: %clang_cc1 -triple xcore -verify %s
 _Static_assert(sizeof(long long) == 8, "sizeof long long is wrong");
 _Static_assert(_Alignof(long long) == 4, "alignof long long is wrong");
@@ -7,11 +8,15 @@
 
 // RUN: %clang_cc1 -triple xcore-unknown-unknown -fno-signed-char -fno-common -emit-llvm -o - %s | FileCheck %s
 
-// CHECK: target datalayout = "e-p:32:32:32-a0:0:32-n32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:32-f16:16:32-f32:32:32-f64:32:32"
 // CHECK: target triple = "xcore-unknown-unknown"
 
+// CHECK: @cgx = external constant i32, section ".cp.rodata"
+extern const int cgx;
+int fcgx() { return cgx;}
 // CHECK: @g1 = global i32 0, align 4
 int g1;
+// CHECK: @cg1 = constant i32 0, section ".cp.rodata", align 4
+const int cg1;
 
 #include <stdarg.h>
 struct x { int a[5]; };
@@ -66,7 +71,7 @@
   // CHECK:[[V2:%[a-z0-9]+]] = bitcast i64* [[V]] to i8*
   // CHECK: call void @f(i8* [[V2]])
 
-  struct x v5 = va_arg (ap, struct x);  // typical agregate type
+  struct x v5 = va_arg (ap, struct x);  // typical aggregate type
   f(&v5);
   // CHECK: [[I:%[a-z0-9]+]] = load i8** [[AP]]
   // CHECK: [[I2:%[a-z0-9]+]] = bitcast i8* [[I]] to %struct.x**
@@ -79,7 +84,7 @@
   // CHECK: [[V2:%[a-z0-9]+]] = bitcast %struct.x* [[V]] to i8*
   // CHECK: call void @f(i8* [[V2]])
 
-  int* v6 = va_arg (ap, int[4]);  // an unusual agregate type
+  int* v6 = va_arg (ap, int[4]);  // an unusual aggregate type
   f(v6);
   // CHECK: [[I:%[a-z0-9]+]] = load i8** [[AP]]
   // CHECK: [[I2:%[a-z0-9]+]] = bitcast i8* [[I]] to [4 x i32]**
@@ -113,10 +118,18 @@
   // CHECK: call i32 @llvm.xcore.getps(i32 {{%[a-z0-9]+}})
   // CHECK: call i32 @llvm.xcore.bitrev(i32 {{%[a-z0-9]+}})
   // CHECK: call void @llvm.xcore.setps(i32 {{%[a-z0-9]+}}, i32 {{%[a-z0-9]+}})
-  int i = __builtin_getid();
-  unsigned int ui = __builtin_getps(i);
+  volatile int i = __builtin_getid();
+  volatile unsigned int ui = __builtin_getps(i);
   ui = __builtin_bitrev(ui);
   __builtin_setps(i,ui);
+
+  // CHECK: store volatile i32 0, i32* {{%[a-z0-9]+}}, align 4
+  // CHECK: store volatile i32 1, i32* {{%[a-z0-9]+}}, align 4
+  // CHECK: store volatile i32 -1, i32* {{%[a-z0-9]+}}, align 4
+  volatile int res;
+  res = __builtin_eh_return_data_regno(0);
+  res = __builtin_eh_return_data_regno(1);
+  res = __builtin_eh_return_data_regno(2);
 }
 
 // CHECK-LABEL: define zeroext i8 @testchar()