llvm-mirror/avx-128.ll at d2e2d612c00d963e5be93e32aba6fd56a50ea78d - llvm-mirror - Git.je (Gitea)

RPCS3/llvm-mirror

mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-01 16:33:37 +01:00

Nate Begeman cb6d1c8193 Formalize the notion that AVX and SSE are non-overlapping extensions from the compiler's point of view. Per email discussion, we either want to always use VEX-prefixed instructions or never use them, and are taking "HasAVX" to mean "Always use VEX". Passing -mattr=-avx,+sse42 should serve to restore legacy SSE support when desirable.

llvm-svn: 121439

2010-12-10 00:26:57 +00:00

13 lines

310 B

LLVM

Raw Blame History

 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -march=x86 -mcpu=corei7 -mattr=avx | FileCheck %s
 @z = common global <4 x float> zeroinitializer, align 16
 define void @zero() nounwind ssp {
 entry:
   ; CHECK: vxorps
   ; CHECK: vmovaps
   store <4 x float> zeroinitializer, <4 x float>* @z, align 16
   ret void
 }