diff --git a/test/CodeGen/ARM/fabss.ll b/test/CodeGen/ARM/fabss.ll
index 46f136b..e5b5791 100644
--- a/test/CodeGen/ARM/fabss.ll
+++ b/test/CodeGen/ARM/fabss.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vabs.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vabs.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vabs.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vabs.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vabs.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %a, float %b) {
 entry:
@@ -13,3 +13,16 @@
 }
 
 declare float @fabsf(float)
+
+; VFP2: test:
+; VFP2: 	vabs.f32	s1, s1
+
+; NFP1: test:
+; NFP1: 	vabs.f32	d1, d1
+; NFP0: test:
+; NFP0: 	vabs.f32	s1, s1
+
+; CORTEXA8: test:
+; CORTEXA8: 	vabs.f32	d1, d1
+; CORTEXA9: test:
+; CORTEXA9: 	vabs.f32	s1, s1
diff --git a/test/CodeGen/ARM/fadds.ll b/test/CodeGen/ARM/fadds.ll
index 1426a2d..db18a86 100644
--- a/test/CodeGen/ARM/fadds.ll
+++ b/test/CodeGen/ARM/fadds.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vadd.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vadd.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vadd.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vadd.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vadd.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %a, float %b) {
 entry:
@@ -10,3 +10,15 @@
 	ret float %0
 }
 
+; VFP2: test:
+; VFP2: 	vadd.f32	s0, s1, s0
+
+; NFP1: test:
+; NFP1: 	vadd.f32	d0, d1, d0
+; NFP0: test:
+; NFP0: 	vadd.f32	s0, s1, s0
+
+; CORTEXA8: test:
+; CORTEXA8: 	vadd.f32	d0, d1, d0
+; CORTEXA9: test:
+; CORTEXA9: 	vadd.f32	s0, s1, s0
diff --git a/test/CodeGen/ARM/fdivs.ll b/test/CodeGen/ARM/fdivs.ll
index 45803f6..a5c86bf 100644
--- a/test/CodeGen/ARM/fdivs.ll
+++ b/test/CodeGen/ARM/fdivs.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vdiv.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vdiv.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vdiv.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vdiv.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vdiv.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %a, float %b) {
 entry:
@@ -10,3 +10,15 @@
 	ret float %0
 }
 
+; VFP2: test:
+; VFP2: 	vdiv.f32	s0, s1, s0
+
+; NFP1: test:
+; NFP1: 	vdiv.f32	s0, s1, s0
+; NFP0: test:
+; NFP0: 	vdiv.f32	s0, s1, s0
+
+; CORTEXA8: test:
+; CORTEXA8: 	vdiv.f32	s0, s1, s0
+; CORTEXA9: test:
+; CORTEXA9: 	vdiv.f32	s0, s1, s0
diff --git a/test/CodeGen/ARM/fmacs.ll b/test/CodeGen/ARM/fmacs.ll
index 57efa82..904a587 100644
--- a/test/CodeGen/ARM/fmacs.ll
+++ b/test/CodeGen/ARM/fmacs.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vmla.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vmul.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vmla.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vmul.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vmla.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %acc, float %a, float %b) {
 entry:
@@ -11,3 +11,15 @@
 	ret float %1
 }
 
+; VFP2: test:
+; VFP2: 	vmla.f32	s2, s1, s0
+
+; NFP1: test:
+; NFP1: 	vmul.f32	d0, d1, d0
+; NFP0: test:
+; NFP0: 	vmla.f32	s2, s1, s0
+
+; CORTEXA8: test:
+; CORTEXA8: 	vmul.f32	d0, d1, d0
+; CORTEXA9: test:
+; CORTEXA9: 	vmla.f32	s2, s1, s0
diff --git a/test/CodeGen/ARM/fmscs.ll b/test/CodeGen/ARM/fmscs.ll
index 31b5c52..7b9e029 100644
--- a/test/CodeGen/ARM/fmscs.ll
+++ b/test/CodeGen/ARM/fmscs.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vnmls.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vnmls.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vnmls.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vnmls.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vnmls.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %acc, float %a, float %b) {
 entry:
@@ -11,3 +11,15 @@
 	ret float %1
 }
 
+; VFP2: test:
+; VFP2: 	vnmls.f32	s2, s1, s0
+
+; NFP1: test:
+; NFP1: 	vnmls.f32	s2, s1, s0
+; NFP0: test:
+; NFP0: 	vnmls.f32	s2, s1, s0
+
+; CORTEXA8: test:
+; CORTEXA8: 	vnmls.f32	s2, s1, s0
+; CORTEXA9: test:
+; CORTEXA9: 	vnmls.f32	s2, s1, s0
diff --git a/test/CodeGen/ARM/fmuls.ll b/test/CodeGen/ARM/fmuls.ll
index 735263c..d3c9c82 100644
--- a/test/CodeGen/ARM/fmuls.ll
+++ b/test/CodeGen/ARM/fmuls.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vmul.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vmul.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vmul.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vmul.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 1
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vmul.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 1
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test(float %a, float %b) {
 entry:
@@ -10,3 +10,15 @@
 	ret float %0
 }
 
+; VFP2: test:
+; VFP2: 	vmul.f32	s0, s1, s0
+
+; NFP1: test:
+; NFP1: 	vmul.f32	d0, d1, d0
+; NFP0: test:
+; NFP0: 	vmul.f32	s0, s1, s0
+
+; CORTEXA8: test:
+; CORTEXA8: 	vmul.f32	d0, d1, d0
+; CORTEXA9: test:
+; CORTEXA9: 	vmul.f32	s0, s1, s0
diff --git a/test/CodeGen/ARM/fnegs.ll b/test/CodeGen/ARM/fnegs.ll
index bc3d42d..d6c22f1 100644
--- a/test/CodeGen/ARM/fnegs.ll
+++ b/test/CodeGen/ARM/fnegs.ll
@@ -1,8 +1,8 @@
-; RUN: llc < %s -march=arm -mattr=+vfp2 | grep -E {vneg.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 2
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | grep -E {vneg.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 2
-; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | grep -E {vneg.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 2
-; RUN: llc < %s -march=arm -mcpu=cortex-a8 | grep -E {vneg.f32\\W*d\[0-9\]+,\\W*d\[0-9\]+} | count 2
-; RUN: llc < %s -march=arm -mcpu=cortex-a9 | grep -E {vneg.f32\\W*s\[0-9\]+,\\W*s\[0-9\]+} | count 2
+; RUN: llc < %s -march=arm -mattr=+vfp2 | FileCheck %s -check-prefix=VFP2
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=1 | FileCheck %s -check-prefix=NFP1
+; RUN: llc < %s -march=arm -mattr=+neon -arm-use-neon-fp=0 | FileCheck %s -check-prefix=NFP0
+; RUN: llc < %s -march=arm -mcpu=cortex-a8 | FileCheck %s -check-prefix=CORTEXA8
+; RUN: llc < %s -march=arm -mcpu=cortex-a9 | FileCheck %s -check-prefix=CORTEXA9
 
 define float @test1(float* %a) {
 entry:
@@ -13,6 +13,20 @@
 	%retval = select i1 %3, float %1, float %0		; <float> [#uses=1]
 	ret float %retval
 }
+; VFP2: test1:
+; VFP2: 	vneg.f32	s1, s0
+
+; NFP1: test1:
+; NFP1: 	vneg.f32	d1, d0
+
+; NFP0: test1:
+; NFP0: 	vneg.f32	s1, s0
+
+; CORTEXA8: test1:
+; CORTEXA8: 	vneg.f32	d1, d0
+
+; CORTEXA9: test1:
+; CORTEXA9: 	vneg.f32	s1, s0
 
 define float @test2(float* %a) {
 entry:
@@ -23,3 +37,18 @@
 	%retval = select i1 %3, float %1, float %0		; <float> [#uses=1]
 	ret float %retval
 }
+; VFP2: test2:
+; VFP2: 	vneg.f32	s1, s0
+
+; NFP1: test2:
+; NFP1: 	vneg.f32	d1, d0
+
+; NFP0: test2:
+; NFP0: 	vneg.f32	s1, s0
+
+; CORTEXA8: test2:
+; CORTEXA8: 	vneg.f32	d1, d0
+
+; CORTEXA9: test2:
+; CORTEXA9: 	vneg.f32	s1, s0
+