mirror of
https://github.com/RPCS3/llvm.git
synced 2024-12-13 14:47:00 +00:00
49c9300c28
As discussed in D11886, this patch moves the SSE/AVX vector blend folding to instcombiner from PerformINTRINSIC_WO_CHAINCombine (which allows us to remove this completely). InstCombiner already had partial support for this, I just had to add support for zero (ConstantAggregateZero) masks and also the case where both selection inputs were the same (allowing us to ignore the mask). I also moved all the relevant combine tests into InstCombine/blend_x86.ll Differential Revision: http://reviews.llvm.org/D11934 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@244723 91177308-0d34-0410-b5e6-96231b3b80d8
61 lines
1.8 KiB
LLVM
61 lines
1.8 KiB
LLVM
; RUN: llc < %s -march=x86-64 -mcpu=corei7-avx | FileCheck %s
|
|
|
|
|
|
define <4 x double> @test_x86_avx_blend_pd_256(<4 x double> %a0) {
|
|
%1 = call <4 x double> @llvm.x86.avx.blend.pd.256(<4 x double> %a0, <4 x double> %a0, i32 7)
|
|
ret <4 x double> %1
|
|
}
|
|
; CHECK-LABEL: test_x86_avx_blend_pd_256
|
|
; CHECK-NOT: vblendpd
|
|
; CHECK: ret
|
|
|
|
|
|
define <8 x float> @test_x86_avx_blend_ps_256(<8 x float> %a0) {
|
|
%1 = call <8 x float> @llvm.x86.avx.blend.ps.256(<8 x float> %a0, <8 x float> %a0, i32 7)
|
|
ret <8 x float> %1
|
|
}
|
|
; CHECK-LABEL: test_x86_avx_blend_ps_256
|
|
; CHECK-NOT: vblendps
|
|
; CHECK: ret
|
|
|
|
|
|
define <4 x double> @test2_x86_avx_blend_pd_256(<4 x double> %a0, <4 x double> %a1) {
|
|
%1 = call <4 x double> @llvm.x86.avx.blend.pd.256(<4 x double> %a0, <4 x double> %a1, i32 0)
|
|
ret <4 x double> %1
|
|
}
|
|
; CHECK-LABEL: test2_x86_avx_blend_pd_256
|
|
; CHECK-NOT: vblendpd
|
|
; CHECK: ret
|
|
|
|
|
|
define <8 x float> @test2_x86_avx_blend_ps_256(<8 x float> %a0, <8 x float> %a1) {
|
|
%1 = call <8 x float> @llvm.x86.avx.blend.ps.256(<8 x float> %a0, <8 x float> %a1, i32 0)
|
|
ret <8 x float> %1
|
|
}
|
|
; CHECK-LABEL: test2_x86_avx_blend_ps_256
|
|
; CHECK-NOT: vblendps
|
|
; CHECK: ret
|
|
|
|
|
|
define <4 x double> @test3_x86_avx_blend_pd_256(<4 x double> %a0, <4 x double> %a1) {
|
|
%1 = call <4 x double> @llvm.x86.avx.blend.pd.256(<4 x double> %a0, <4 x double> %a1, i32 -1)
|
|
ret <4 x double> %1
|
|
}
|
|
; CHECK-LABEL: test3_x86_avx_blend_pd_256
|
|
; CHECK-NOT: vblendpd
|
|
; CHECK: ret
|
|
|
|
|
|
define <8 x float> @test3_x86_avx_blend_ps_256(<8 x float> %a0, <8 x float> %a1) {
|
|
%1 = call <8 x float> @llvm.x86.avx.blend.ps.256(<8 x float> %a0, <8 x float> %a1, i32 -1)
|
|
ret <8 x float> %1
|
|
}
|
|
; CHECK-LABEL: test3_x86_avx_blend_ps_256
|
|
; CHECK-NOT: vblendps
|
|
; CHECK: ret
|
|
|
|
|
|
declare <4 x double> @llvm.x86.avx.blend.pd.256(<4 x double>, <4 x double>, i32)
|
|
declare <8 x float> @llvm.x86.avx.blend.ps.256(<8 x float>, <8 x float>, i32)
|
|
|