diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2018-02-07 15:48:44 +0000 |
|---|---|---|
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | 2018-02-07 15:48:44 +0000 |
| commit | b4e789e8f6aa62b628a41a73275aff95c9dd4389 (patch) | |
| tree | 30988128d144c2fc16261a378841f28e3fbb36ca /llvm/test/CodeGen/X86/vector-trunc-ssat.ll | |
| parent | b419108203ab65a3fae3b593e52f88bee011edee (diff) | |
| download | bcm5719-llvm-b4e789e8f6aa62b628a41a73275aff95c9dd4389.tar.gz bcm5719-llvm-b4e789e8f6aa62b628a41a73275aff95c9dd4389.zip | |
[X86][AVX] Add PACKSSDW/PACKUSDW support for truncation of clamped values
SSE and shorter vector sizes will have to wait until we can add support for general SMIN/SMAX matching.
llvm-svn: 324485
Diffstat (limited to 'llvm/test/CodeGen/X86/vector-trunc-ssat.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/vector-trunc-ssat.ll | 35 |
1 files changed, 7 insertions, 28 deletions
diff --git a/llvm/test/CodeGen/X86/vector-trunc-ssat.ll b/llvm/test/CodeGen/X86/vector-trunc-ssat.ll index 2dbfe76a6e8..c47bfe908b4 100644 --- a/llvm/test/CodeGen/X86/vector-trunc-ssat.ll +++ b/llvm/test/CodeGen/X86/vector-trunc-ssat.ll @@ -1256,28 +1256,14 @@ define <8 x i16> @trunc_ssat_v8i32_v8i16(<8 x i32> %a0) { ; AVX1-LABEL: trunc_ssat_v8i32_v8i16: ; AVX1: # %bb.0: ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1 -; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [32767,32767,32767,32767] -; AVX1-NEXT: vpminsd %xmm2, %xmm1, %xmm1 -; AVX1-NEXT: vpminsd %xmm2, %xmm0, %xmm0 -; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [4294934528,4294934528,4294934528,4294934528] -; AVX1-NEXT: vpmaxsd %xmm2, %xmm0, %xmm0 -; AVX1-NEXT: vpmaxsd %xmm2, %xmm1, %xmm1 -; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15] -; AVX1-NEXT: vpshufb %xmm2, %xmm1, %xmm1 -; AVX1-NEXT: vpshufb %xmm2, %xmm0, %xmm0 -; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0] +; AVX1-NEXT: vpackssdw %xmm1, %xmm0, %xmm0 ; AVX1-NEXT: vzeroupper ; AVX1-NEXT: retq ; ; AVX2-LABEL: trunc_ssat_v8i32_v8i16: ; AVX2: # %bb.0: -; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm1 = [32767,32767,32767,32767,32767,32767,32767,32767] -; AVX2-NEXT: vpminsd %ymm1, %ymm0, %ymm0 -; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm1 = [4294934528,4294934528,4294934528,4294934528,4294934528,4294934528,4294934528,4294934528] -; AVX2-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0 -; AVX2-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31] -; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3] -; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0 +; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1 +; AVX2-NEXT: vpackssdw %xmm1, %xmm0, %xmm0 ; AVX2-NEXT: vzeroupper ; AVX2-NEXT: retq ; @@ -1483,17 +1469,10 @@ define <16 x i16> @trunc_ssat_v16i32_v16i16(<16 x i32> %a0) { ; ; AVX2-LABEL: trunc_ssat_v16i32_v16i16: ; AVX2: # %bb.0: -; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm2 = [32767,32767,32767,32767,32767,32767,32767,32767] -; AVX2-NEXT: vpminsd %ymm2, %ymm0, %ymm0 -; AVX2-NEXT: vpminsd %ymm2, %ymm1, %ymm1 -; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm2 = [4294934528,4294934528,4294934528,4294934528,4294934528,4294934528,4294934528,4294934528] -; AVX2-NEXT: vpmaxsd %ymm2, %ymm1, %ymm1 -; AVX2-NEXT: vpmaxsd %ymm2, %ymm0, %ymm0 -; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31] -; AVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0 -; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3] -; AVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1 -; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3] +; AVX2-NEXT: vextracti128 $1, %ymm1, %xmm2 +; AVX2-NEXT: vpackssdw %xmm2, %xmm1, %xmm1 +; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm2 +; AVX2-NEXT: vpackssdw %xmm2, %xmm0, %xmm0 ; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0 ; AVX2-NEXT: retq ; |
