[AArch64] Implement lowering of the X constraint on AArch64

Summary:
This implements the lowering of the X constraint on
AArch64.

The default behaviour of the X constraint lowering is to
restrict it to "f". This is a problem because the "f"
constraint is not implemented on AArch64 and would be too
restrictive anyway. Therefore, the AArch64 hook will
lower this to "w" (if the operand is a floating point or
vector) or "r" otherwise.

The implementation is similar with the one added for
ARM (r267411).

This is the AArch64 side of the fix for http://llvm.org/PR26493

Reviewers: rengolin

Subscribers: aemerson, rengolin, llvm-commits, t.p.northover

Differential Revision: http://reviews.llvm.org/D19967

llvm-svn: 268907
diff --git a/llvm/test/CodeGen/AArch64/inlineasm-X-allocation.ll b/llvm/test/CodeGen/AArch64/inlineasm-X-allocation.ll
new file mode 100644
index 0000000..1d7a24e
--- /dev/null
+++ b/llvm/test/CodeGen/AArch64/inlineasm-X-allocation.ll
@@ -0,0 +1,17 @@
+; RUN: llc -mtriple=aarch64-none-eabi -mattr=-fp-armv8 %s -o - | FileCheck %s  -check-prefix=nofp
+
+; In the novfp case, the compiler is forced to assign a core register,
+; even if the input is a float.
+
+; nofp-LABEL: f1
+; nofp-CHECK: ldr x0, [sp]
+
+; This can be generated by a function such as:
+;  void f1(float f) {asm volatile ("ldr $0, [sp]" : : "X" (f));}
+
+define void @f1(float %f) {
+entry:
+  call void asm sideeffect "ldr $0, [sp]", "X" (float %f) nounwind
+
+  ret void
+}