; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE41
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2,AVX2-SLOW
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=CHECK,AVX,AVX2,AVX2-FAST-ALL
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=CHECK,AVX,AVX2,AVX2-FAST-PERLANE
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
; fold (sra 0, x) -> 0
define <4 x i32> @combine_vec_ashr_zero(<4 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_zero:
; SSE: # %bb.0:
; SSE-NEXT: xorps %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_zero:
; AVX: # %bb.0:
; AVX-NEXT: vxorps %xmm0, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> zeroinitializer, %x
ret <4 x i32> %1
}
; fold (sra -1, x) -> -1
define <4 x i32> @combine_vec_ashr_allones(<4 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_allones:
; SSE: # %bb.0:
; SSE-NEXT: pcmpeqd %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_allones:
; AVX: # %bb.0:
; AVX-NEXT: vpcmpeqd %xmm0, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, %x
ret <4 x i32> %1
}
; fold (sra x, c >= size(x)) -> undef
define <4 x i32> @combine_vec_ashr_outofrange0(<4 x i32> %x) {
; CHECK-LABEL: combine_vec_ashr_outofrange0:
; CHECK: # %bb.0:
; CHECK-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 33, i32 33, i32 33, i32 33>
ret <4 x i32> %1
}
define <4 x i32> @combine_vec_ashr_outofrange1(<4 x i32> %x) {
; CHECK-LABEL: combine_vec_ashr_outofrange1:
; CHECK: # %bb.0:
; CHECK-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 33, i32 34, i32 35, i32 36>
ret <4 x i32> %1
}
define <4 x i32> @combine_vec_ashr_outofrange2(<4 x i32> %x) {
; CHECK-LABEL: combine_vec_ashr_outofrange2:
; CHECK: # %bb.0:
; CHECK-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 33, i32 34, i32 35, i32 undef>
ret <4 x i32> %1
}
; fold (sra x, 0) -> x
define <4 x i32> @combine_vec_ashr_by_zero(<4 x i32> %x) {
; CHECK-LABEL: combine_vec_ashr_by_zero:
; CHECK: # %bb.0:
; CHECK-NEXT: retq
%1 = ashr <4 x i32> %x, zeroinitializer
ret <4 x i32> %1
}
; fold (sra (sra x, c1), c2) -> (sra x, (add c1, c2))
define <4 x i32> @combine_vec_ashr_ashr0(<4 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_ashr0:
; SSE: # %bb.0:
; SSE-NEXT: psrad $6, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_ashr0:
; AVX: # %bb.0:
; AVX-NEXT: vpsrad $6, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 2, i32 2, i32 2, i32 2>
%2 = ashr <4 x i32> %1, <i32 4, i32 4, i32 4, i32 4>
ret <4 x i32> %2
}
define <4 x i32> @combine_vec_ashr_ashr1(<4 x i32> %x) {
; SSE2-LABEL: combine_vec_ashr_ashr1:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa %xmm0, %xmm1
; SSE2-NEXT: psrad $10, %xmm1
; SSE2-NEXT: movdqa %xmm0, %xmm2
; SSE2-NEXT: psrad $8, %xmm2
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm2 = xmm2[1],xmm1[1]
; SSE2-NEXT: movdqa %xmm0, %xmm1
; SSE2-NEXT: psrad $6, %xmm1
; SSE2-NEXT: psrad $4, %xmm0
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,3],xmm2[0,3]
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_ashr1:
; SSE41: # %bb.0:
; SSE41-NEXT: movdqa %xmm0, %xmm1
; SSE41-NEXT: psrad $10, %xmm1
; SSE41-NEXT: movdqa %xmm0, %xmm2
; SSE41-NEXT: psrad $6, %xmm2
; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm2[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm1
; SSE41-NEXT: psrad $8, %xmm1
; SSE41-NEXT: psrad $4, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm2[2,3],xmm0[4,5],xmm2[6,7]
; SSE41-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_ashr1:
; AVX: # %bb.0:
; AVX-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 0, i32 1, i32 2, i32 3>
%2 = ashr <4 x i32> %1, <i32 4, i32 5, i32 6, i32 7>
ret <4 x i32> %2
}
define <4 x i32> @combine_vec_ashr_ashr2(<4 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_ashr2:
; SSE: # %bb.0:
; SSE-NEXT: psrad $31, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_ashr2:
; AVX: # %bb.0:
; AVX-NEXT: vpsrad $31, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 17, i32 18, i32 19, i32 20>
%2 = ashr <4 x i32> %1, <i32 25, i32 26, i32 27, i32 28>
ret <4 x i32> %2
}
define <4 x i32> @combine_vec_ashr_ashr3(<4 x i32> %x) {
; SSE2-LABEL: combine_vec_ashr_ashr3:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa %xmm0, %xmm2
; SSE2-NEXT: psrad $27, %xmm2
; SSE2-NEXT: movdqa %xmm0, %xmm1
; SSE2-NEXT: psrad $31, %xmm1
; SSE2-NEXT: movdqa %xmm1, %xmm3
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm3 = xmm3[1],xmm2[1]
; SSE2-NEXT: psrad $15, %xmm0
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,3],xmm3[0,3]
; SSE2-NEXT: movaps %xmm1, %xmm0
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_ashr3:
; SSE41: # %bb.0:
; SSE41-NEXT: movdqa %xmm0, %xmm1
; SSE41-NEXT: psrad $27, %xmm1
; SSE41-NEXT: movdqa %xmm0, %xmm2
; SSE41-NEXT: psrad $15, %xmm2
; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm2[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: psrad $31, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm2[2,3],xmm0[4,5],xmm2[6,7]
; SSE41-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_ashr3:
; AVX: # %bb.0:
; AVX-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX-NEXT: retq
%1 = ashr <4 x i32> %x, <i32 1, i32 5, i32 50, i32 27>
%2 = ashr <4 x i32> %1, <i32 33, i32 10, i32 33, i32 0>
ret <4 x i32> %2
}
; fold (sra x, (trunc (and y, c))) -> (sra x, (and (trunc y), (trunc c))).
define <4 x i32> @combine_vec_ashr_trunc_and(<4 x i32> %x, <4 x i64> %y) {
; SSE2-LABEL: combine_vec_ashr_trunc_and:
; SSE2: # %bb.0:
; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; SSE2-NEXT: andps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm1[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm3
; SSE2-NEXT: psrad %xmm2, %xmm3
; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm1[0,1,1,1,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm2
; SSE2-NEXT: psrad %xmm4, %xmm2
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,2,3]
; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm1[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm4
; SSE2-NEXT: psrad %xmm3, %xmm4
; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,1,1,1,4,5,6,7]
; SSE2-NEXT: psrad %xmm1, %xmm0
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm0 = xmm0[1],xmm4[1]
; SSE2-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,3],xmm0[0,3]
; SSE2-NEXT: movaps %xmm2, %xmm0
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_trunc_and:
; SSE41: # %bb.0:
; SSE41-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; SSE41-NEXT: andps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
; SSE41-NEXT: pshuflw {{.*#+}} xmm2 = xmm1[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrad %xmm2, %xmm3
; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,3,2,3]
; SSE41-NEXT: pshuflw {{.*#+}} xmm4 = xmm2[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm5
; SSE41-NEXT: psrad %xmm4, %xmm5
; SSE41-NEXT: pblendw {{.*#+}} xmm5 = xmm3[0,1,2,3],xmm5[4,5,6,7]
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,1,1,1,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrad %xmm1, %xmm3
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm2[0,1,1,1,4,5,6,7]
; SSE41-NEXT: psrad %xmm1, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm3[0,1,2,3],xmm0[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm5[2,3],xmm0[4,5],xmm5[6,7]
; SSE41-NEXT: retq
;
; AVX2-SLOW-LABEL: combine_vec_ashr_trunc_and:
; AVX2-SLOW: # %bb.0:
; AVX2-SLOW-NEXT: vextractf128 $1, %ymm1, %xmm2
; AVX2-SLOW-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; AVX2-SLOW-NEXT: vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; AVX2-SLOW-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; AVX2-SLOW-NEXT: vzeroupper
; AVX2-SLOW-NEXT: retq
;
; AVX2-FAST-ALL-LABEL: combine_vec_ashr_trunc_and:
; AVX2-FAST-ALL: # %bb.0:
; AVX2-FAST-ALL-NEXT: vpmovsxbd {{.*#+}} ymm2 = [0,2,4,6,0,0,0,0]
; AVX2-FAST-ALL-NEXT: vpermd %ymm1, %ymm2, %ymm1
; AVX2-FAST-ALL-NEXT: vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; AVX2-FAST-ALL-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; AVX2-FAST-ALL-NEXT: vzeroupper
; AVX2-FAST-ALL-NEXT: retq
;
; AVX2-FAST-PERLANE-LABEL: combine_vec_ashr_trunc_and:
; AVX2-FAST-PERLANE: # %bb.0:
; AVX2-FAST-PERLANE-NEXT: vextractf128 $1, %ymm1, %xmm2
; AVX2-FAST-PERLANE-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; AVX2-FAST-PERLANE-NEXT: vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; AVX2-FAST-PERLANE-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; AVX2-FAST-PERLANE-NEXT: vzeroupper
; AVX2-FAST-PERLANE-NEXT: retq
;
; AVX512-LABEL: combine_vec_ashr_trunc_and:
; AVX512: # %bb.0:
; AVX512-NEXT: vpmovqd %ymm1, %xmm1
; AVX512-NEXT: vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; AVX512-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = and <4 x i64> %y, <i64 15, i64 255, i64 4095, i64 65535>
%2 = trunc <4 x i64> %1 to <4 x i32>
%3 = ashr <4 x i32> %x, %2
ret <4 x i32> %3
}
; fold (sra (trunc (srl x, c1)), c2) -> (trunc (sra x, c1 + c2))
; if c1 is equal to the number of bits the trunc removes
define <4 x i32> @combine_vec_ashr_trunc_lshr(<4 x i64> %x) {
; SSE2-LABEL: combine_vec_ashr_trunc_lshr:
; SSE2: # %bb.0:
; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; SSE2-NEXT: movaps %xmm0, %xmm1
; SSE2-NEXT: psrad $3, %xmm1
; SSE2-NEXT: movaps %xmm0, %xmm2
; SSE2-NEXT: psrad $2, %xmm2
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm2 = xmm2[1],xmm1[1]
; SSE2-NEXT: movaps %xmm0, %xmm1
; SSE2-NEXT: psrad $1, %xmm1
; SSE2-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,3],xmm2[0,3]
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_trunc_lshr:
; SSE41: # %bb.0:
; SSE41-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; SSE41-NEXT: movaps %xmm0, %xmm2
; SSE41-NEXT: psrad $2, %xmm2
; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm0[0,1,2,3],xmm2[4,5,6,7]
; SSE41-NEXT: psrad $1, %xmm0
; SSE41-NEXT: psrad $3, %xmm1
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm2[0,1],xmm0[2,3],xmm2[4,5],xmm0[6,7]
; SSE41-NEXT: retq
;
; AVX2-SLOW-LABEL: combine_vec_ashr_trunc_lshr:
; AVX2-SLOW: # %bb.0:
; AVX2-SLOW-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX2-SLOW-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX2-SLOW-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-SLOW-NEXT: vzeroupper
; AVX2-SLOW-NEXT: retq
;
; AVX2-FAST-ALL-LABEL: combine_vec_ashr_trunc_lshr:
; AVX2-FAST-ALL: # %bb.0:
; AVX2-FAST-ALL-NEXT: vpmovsxbd {{.*#+}} xmm1 = [1,3,5,7]
; AVX2-FAST-ALL-NEXT: vpermd %ymm0, %ymm1, %ymm0
; AVX2-FAST-ALL-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-FAST-ALL-NEXT: vzeroupper
; AVX2-FAST-ALL-NEXT: retq
;
; AVX2-FAST-PERLANE-LABEL: combine_vec_ashr_trunc_lshr:
; AVX2-FAST-PERLANE: # %bb.0:
; AVX2-FAST-PERLANE-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX2-FAST-PERLANE-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX2-FAST-PERLANE-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-FAST-PERLANE-NEXT: vzeroupper
; AVX2-FAST-PERLANE-NEXT: retq
;
; AVX512-LABEL: combine_vec_ashr_trunc_lshr:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsrlq $32, %ymm0, %ymm0
; AVX512-NEXT: vpmovqd %ymm0, %xmm0
; AVX512-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = lshr <4 x i64> %x, <i64 32, i64 32, i64 32, i64 32>
%2 = trunc <4 x i64> %1 to <4 x i32>
%3 = ashr <4 x i32> %2, <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %3
}
define <16 x i8> @combine_vec_ashr_trunc_lshr_splat(<16 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_trunc_lshr_splat:
; SSE: # %bb.0:
; SSE-NEXT: psrad $26, %xmm3
; SSE-NEXT: psrad $26, %xmm2
; SSE-NEXT: packssdw %xmm3, %xmm2
; SSE-NEXT: psrad $26, %xmm1
; SSE-NEXT: psrad $26, %xmm0
; SSE-NEXT: packssdw %xmm1, %xmm0
; SSE-NEXT: packsswb %xmm2, %xmm0
; SSE-NEXT: retq
;
; AVX2-LABEL: combine_vec_ashr_trunc_lshr_splat:
; AVX2: # %bb.0:
; AVX2-NEXT: vpsrad $26, %ymm1, %ymm1
; AVX2-NEXT: vpsrad $26, %ymm0, %ymm0
; AVX2-NEXT: vpackssdw %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpacksswb %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[0,2,1,3]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: combine_vec_ashr_trunc_lshr_splat:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsrad $26, %zmm0, %zmm0
; AVX512-NEXT: vpmovdb %zmm0, %xmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = lshr <16 x i32> %x, <i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24>
%2 = trunc <16 x i32> %1 to <16 x i8>
%3 = ashr <16 x i8> %2, <i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2>
ret <16 x i8> %3
}
; fold (sra (trunc (sra x, c1)), c2) -> (trunc (sra x, c1 + c2))
; if c1 is equal to the number of bits the trunc removes
define <4 x i32> @combine_vec_ashr_trunc_ashr(<4 x i64> %x) {
; SSE2-LABEL: combine_vec_ashr_trunc_ashr:
; SSE2: # %bb.0:
; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; SSE2-NEXT: movaps %xmm0, %xmm1
; SSE2-NEXT: psrad $3, %xmm1
; SSE2-NEXT: movaps %xmm0, %xmm2
; SSE2-NEXT: psrad $2, %xmm2
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm2 = xmm2[1],xmm1[1]
; SSE2-NEXT: movaps %xmm0, %xmm1
; SSE2-NEXT: psrad $1, %xmm1
; SSE2-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,3],xmm2[0,3]
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_trunc_ashr:
; SSE41: # %bb.0:
; SSE41-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; SSE41-NEXT: movaps %xmm0, %xmm2
; SSE41-NEXT: psrad $2, %xmm2
; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm0[0,1,2,3],xmm2[4,5,6,7]
; SSE41-NEXT: psrad $1, %xmm0
; SSE41-NEXT: psrad $3, %xmm1
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm2[0,1],xmm0[2,3],xmm2[4,5],xmm0[6,7]
; SSE41-NEXT: retq
;
; AVX2-SLOW-LABEL: combine_vec_ashr_trunc_ashr:
; AVX2-SLOW: # %bb.0:
; AVX2-SLOW-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX2-SLOW-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX2-SLOW-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-SLOW-NEXT: vzeroupper
; AVX2-SLOW-NEXT: retq
;
; AVX2-FAST-ALL-LABEL: combine_vec_ashr_trunc_ashr:
; AVX2-FAST-ALL: # %bb.0:
; AVX2-FAST-ALL-NEXT: vpmovsxbd {{.*#+}} xmm1 = [1,3,5,7]
; AVX2-FAST-ALL-NEXT: vpermd %ymm0, %ymm1, %ymm0
; AVX2-FAST-ALL-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-FAST-ALL-NEXT: vzeroupper
; AVX2-FAST-ALL-NEXT: retq
;
; AVX2-FAST-PERLANE-LABEL: combine_vec_ashr_trunc_ashr:
; AVX2-FAST-PERLANE: # %bb.0:
; AVX2-FAST-PERLANE-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX2-FAST-PERLANE-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX2-FAST-PERLANE-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX2-FAST-PERLANE-NEXT: vzeroupper
; AVX2-FAST-PERLANE-NEXT: retq
;
; AVX512-LABEL: combine_vec_ashr_trunc_ashr:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsrlq $32, %ymm0, %ymm0
; AVX512-NEXT: vpmovqd %ymm0, %xmm0
; AVX512-NEXT: vpsravd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = ashr <4 x i64> %x, <i64 32, i64 32, i64 32, i64 32>
%2 = trunc <4 x i64> %1 to <4 x i32>
%3 = ashr <4 x i32> %2, <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %3
}
define <8 x i16> @combine_vec_ashr_trunc_ashr_splat(<8 x i32> %x) {
; SSE-LABEL: combine_vec_ashr_trunc_ashr_splat:
; SSE: # %bb.0:
; SSE-NEXT: psrad $19, %xmm1
; SSE-NEXT: psrad $19, %xmm0
; SSE-NEXT: packssdw %xmm1, %xmm0
; SSE-NEXT: retq
;
; AVX2-LABEL: combine_vec_ashr_trunc_ashr_splat:
; AVX2: # %bb.0:
; AVX2-NEXT: vpsrad $19, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: combine_vec_ashr_trunc_ashr_splat:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsrad $19, %ymm0, %ymm0
; AVX512-NEXT: vpmovdw %ymm0, %xmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = ashr <8 x i32> %x, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
%2 = trunc <8 x i32> %1 to <8 x i16>
%3 = ashr <8 x i16> %2, <i16 3, i16 3, i16 3, i16 3, i16 3, i16 3, i16 3, i16 3>
ret <8 x i16> %3
}
; If the sign bit is known to be zero, switch this to a SRL.
define <4 x i32> @combine_vec_ashr_positive(<4 x i32> %x, <4 x i32> %y) {
; SSE2-LABEL: combine_vec_ashr_positive:
; SSE2: # %bb.0:
; SSE2-NEXT: pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm1[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm3
; SSE2-NEXT: psrld %xmm2, %xmm3
; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm1[0,1,1,1,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm2
; SSE2-NEXT: psrld %xmm4, %xmm2
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,2,3]
; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm1[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm4
; SSE2-NEXT: psrld %xmm3, %xmm4
; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,1,1,1,4,5,6,7]
; SSE2-NEXT: psrld %xmm1, %xmm0
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm0 = xmm0[1],xmm4[1]
; SSE2-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,3],xmm0[0,3]
; SSE2-NEXT: movaps %xmm2, %xmm0
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec_ashr_positive:
; SSE41: # %bb.0:
; SSE41-NEXT: pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE41-NEXT: pshuflw {{.*#+}} xmm2 = xmm1[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrld %xmm2, %xmm3
; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,3,2,3]
; SSE41-NEXT: pshuflw {{.*#+}} xmm4 = xmm2[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm5
; SSE41-NEXT: psrld %xmm4, %xmm5
; SSE41-NEXT: pblendw {{.*#+}} xmm5 = xmm3[0,1,2,3],xmm5[4,5,6,7]
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,1,1,1,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrld %xmm1, %xmm3
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm2[0,1,1,1,4,5,6,7]
; SSE41-NEXT: psrld %xmm1, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm3[0,1,2,3],xmm0[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm5[2,3],xmm0[4,5],xmm5[6,7]
; SSE41-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_positive:
; AVX: # %bb.0:
; AVX-NEXT: vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX-NEXT: vpsrlvd %xmm1, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = and <4 x i32> %x, <i32 15, i32 255, i32 4095, i32 65535>
%2 = ashr <4 x i32> %1, %y
ret <4 x i32> %2
}
define <4 x i32> @combine_vec_ashr_positive_splat(<4 x i32> %x, <4 x i32> %y) {
; SSE-LABEL: combine_vec_ashr_positive_splat:
; SSE: # %bb.0:
; SSE-NEXT: xorps %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: combine_vec_ashr_positive_splat:
; AVX: # %bb.0:
; AVX-NEXT: vxorps %xmm0, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = and <4 x i32> %x, <i32 1023, i32 1023, i32 1023, i32 1023>
%2 = ashr <4 x i32> %1, <i32 10, i32 10, i32 10, i32 10>
ret <4 x i32> %2
}
define <8 x i16> @combine_vec8i16_ashr_clamped(<8 x i16> %x, <8 x i16> %y) {
; SSE2-LABEL: combine_vec8i16_ashr_clamped:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa %xmm1, %xmm2
; SSE2-NEXT: psubusw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2
; SSE2-NEXT: psubw %xmm2, %xmm1
; SSE2-NEXT: psllw $12, %xmm1
; SSE2-NEXT: movdqa %xmm1, %xmm2
; SSE2-NEXT: psraw $15, %xmm2
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: pandn %xmm0, %xmm3
; SSE2-NEXT: psraw $8, %xmm0
; SSE2-NEXT: pand %xmm2, %xmm0
; SSE2-NEXT: por %xmm3, %xmm0
; SSE2-NEXT: paddw %xmm1, %xmm1
; SSE2-NEXT: movdqa %xmm1, %xmm2
; SSE2-NEXT: psraw $15, %xmm2
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: pandn %xmm0, %xmm3
; SSE2-NEXT: psraw $4, %xmm0
; SSE2-NEXT: pand %xmm2, %xmm0
; SSE2-NEXT: por %xmm3, %xmm0
; SSE2-NEXT: paddw %xmm1, %xmm1
; SSE2-NEXT: movdqa %xmm1, %xmm2
; SSE2-NEXT: psraw $15, %xmm2
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: pandn %xmm0, %xmm3
; SSE2-NEXT: psraw $2, %xmm0
; SSE2-NEXT: pand %xmm2, %xmm0
; SSE2-NEXT: por %xmm3, %xmm0
; SSE2-NEXT: paddw %xmm1, %xmm1
; SSE2-NEXT: psraw $15, %xmm1
; SSE2-NEXT: movdqa %xmm1, %xmm2
; SSE2-NEXT: pandn %xmm0, %xmm2
; SSE2-NEXT: psraw $1, %xmm0
; SSE2-NEXT: pand %xmm1, %xmm0
; SSE2-NEXT: por %xmm2, %xmm0
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec8i16_ashr_clamped:
; SSE41: # %bb.0:
; SSE41-NEXT: movdqa %xmm0, %xmm2
; SSE41-NEXT: pminuw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
; SSE41-NEXT: movdqa %xmm1, %xmm0
; SSE41-NEXT: psllw $12, %xmm0
; SSE41-NEXT: psllw $4, %xmm1
; SSE41-NEXT: por %xmm1, %xmm0
; SSE41-NEXT: movdqa %xmm0, %xmm1
; SSE41-NEXT: paddw %xmm0, %xmm1
; SSE41-NEXT: movdqa %xmm2, %xmm3
; SSE41-NEXT: psraw $8, %xmm3
; SSE41-NEXT: pblendvb %xmm0, %xmm3, %xmm2
; SSE41-NEXT: movdqa %xmm2, %xmm3
; SSE41-NEXT: psraw $4, %xmm3
; SSE41-NEXT: movdqa %xmm1, %xmm0
; SSE41-NEXT: pblendvb %xmm0, %xmm3, %xmm2
; SSE41-NEXT: movdqa %xmm2, %xmm3
; SSE41-NEXT: psraw $2, %xmm3
; SSE41-NEXT: paddw %xmm1, %xmm1
; SSE41-NEXT: movdqa %xmm1, %xmm0
; SSE41-NEXT: pblendvb %xmm0, %xmm3, %xmm2
; SSE41-NEXT: movdqa %xmm2, %xmm3
; SSE41-NEXT: psraw $1, %xmm3
; SSE41-NEXT: paddw %xmm1, %xmm1
; SSE41-NEXT: movdqa %xmm1, %xmm0
; SSE41-NEXT: pblendvb %xmm0, %xmm3, %xmm2
; SSE41-NEXT: movdqa %xmm2, %xmm0
; SSE41-NEXT: retq
;
; AVX2-LABEL: combine_vec8i16_ashr_clamped:
; AVX2: # %bb.0:
; AVX2-NEXT: vpminuw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; AVX2-NEXT: vpmovsxwd %xmm0, %ymm0
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero
; AVX2-NEXT: vpsravd %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: combine_vec8i16_ashr_clamped:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsravw %xmm1, %xmm0, %xmm0
; AVX512-NEXT: retq
%1 = tail call <8 x i16> @llvm.umin.v8i16(<8 x i16> %y, <8 x i16> <i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15>)
%2 = ashr <8 x i16> %x, %1
ret <8 x i16> %2
}
define <4 x i32> @combine_vec4i32_ashr_clamped(<4 x i32> %x, <4 x i32> %y) {
; SSE2-LABEL: combine_vec4i32_ashr_clamped:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [2147483648,2147483648,2147483648,2147483648]
; SSE2-NEXT: pxor %xmm1, %xmm2
; SSE2-NEXT: pcmpgtd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: pandn %xmm1, %xmm3
; SSE2-NEXT: psrld $27, %xmm2
; SSE2-NEXT: por %xmm3, %xmm2
; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm2[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm3
; SSE2-NEXT: psrad %xmm1, %xmm3
; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm2[0,1,1,1,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm1
; SSE2-NEXT: psrad %xmm4, %xmm1
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[2,3,2,3]
; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm2[2,3,3,3,4,5,6,7]
; SSE2-NEXT: movdqa %xmm0, %xmm4
; SSE2-NEXT: psrad %xmm3, %xmm4
; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[0,1,1,1,4,5,6,7]
; SSE2-NEXT: psrad %xmm2, %xmm0
; SSE2-NEXT: punpckhqdq {{.*#+}} xmm0 = xmm0[1],xmm4[1]
; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,3],xmm0[0,3]
; SSE2-NEXT: movaps %xmm1, %xmm0
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec4i32_ashr_clamped:
; SSE41: # %bb.0:
; SSE41-NEXT: pminud {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
; SSE41-NEXT: pshuflw {{.*#+}} xmm2 = xmm1[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrad %xmm2, %xmm3
; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,3,2,3]
; SSE41-NEXT: pshuflw {{.*#+}} xmm4 = xmm2[2,3,3,3,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm5
; SSE41-NEXT: psrad %xmm4, %xmm5
; SSE41-NEXT: pblendw {{.*#+}} xmm5 = xmm3[0,1,2,3],xmm5[4,5,6,7]
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,1,1,1,4,5,6,7]
; SSE41-NEXT: movdqa %xmm0, %xmm3
; SSE41-NEXT: psrad %xmm1, %xmm3
; SSE41-NEXT: pshuflw {{.*#+}} xmm1 = xmm2[0,1,1,1,4,5,6,7]
; SSE41-NEXT: psrad %xmm1, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm3[0,1,2,3],xmm0[4,5,6,7]
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm5[2,3],xmm0[4,5],xmm5[6,7]
; SSE41-NEXT: retq
;
; AVX-LABEL: combine_vec4i32_ashr_clamped:
; AVX: # %bb.0:
; AVX-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; AVX-NEXT: retq
%1 = tail call <4 x i32> @llvm.umin.v4i32(<4 x i32> %y, <4 x i32> <i32 31, i32 31, i32 31, i32 31>)
%2 = ashr <4 x i32> %x, %1
ret <4 x i32> %2
}
define <4 x i64> @combine_vec4i64_ashr_clamped(<4 x i64> %x, <4 x i64> %y) {
; SSE2-LABEL: combine_vec4i64_ashr_clamped:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa {{.*#+}} xmm5 = [9223372039002259456,9223372039002259456]
; SSE2-NEXT: movdqa %xmm3, %xmm4
; SSE2-NEXT: pxor %xmm5, %xmm4
; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
; SSE2-NEXT: movdqa {{.*#+}} xmm7 = [2147483711,2147483711,2147483711,2147483711]
; SSE2-NEXT: movdqa %xmm7, %xmm8
; SSE2-NEXT: pcmpgtd %xmm6, %xmm8
; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
; SSE2-NEXT: pcmpeqd %xmm5, %xmm4
; SSE2-NEXT: pand %xmm8, %xmm4
; SSE2-NEXT: movdqa {{.*#+}} xmm6 = [63,63]
; SSE2-NEXT: pand %xmm4, %xmm3
; SSE2-NEXT: pandn %xmm6, %xmm4
; SSE2-NEXT: por %xmm3, %xmm4
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: pxor %xmm5, %xmm3
; SSE2-NEXT: pshufd {{.*#+}} xmm8 = xmm3[0,0,2,2]
; SSE2-NEXT: pcmpgtd %xmm8, %xmm7
; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
; SSE2-NEXT: pcmpeqd %xmm5, %xmm3
; SSE2-NEXT: pand %xmm7, %xmm3
; SSE2-NEXT: pand %xmm3, %xmm2
; SSE2-NEXT: pandn %xmm6, %xmm3
; SSE2-NEXT: por %xmm2, %xmm3
; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [9223372036854775808,9223372036854775808]
; SSE2-NEXT: movdqa %xmm2, %xmm5
; SSE2-NEXT: psrlq %xmm3, %xmm5
; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm3[2,3,2,3]
; SSE2-NEXT: movdqa %xmm2, %xmm7
; SSE2-NEXT: psrlq %xmm6, %xmm7
; SSE2-NEXT: movsd {{.*#+}} xmm7 = xmm5[0],xmm7[1]
; SSE2-NEXT: movdqa %xmm0, %xmm5
; SSE2-NEXT: psrlq %xmm3, %xmm5
; SSE2-NEXT: psrlq %xmm6, %xmm0
; SSE2-NEXT: movsd {{.*#+}} xmm0 = xmm5[0],xmm0[1]
; SSE2-NEXT: xorpd %xmm7, %xmm0
; SSE2-NEXT: psubq %xmm7, %xmm0
; SSE2-NEXT: movdqa %xmm2, %xmm3
; SSE2-NEXT: psrlq %xmm4, %xmm3
; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[2,3,2,3]
; SSE2-NEXT: psrlq %xmm5, %xmm2
; SSE2-NEXT: movsd {{.*#+}} xmm2 = xmm3[0],xmm2[1]
; SSE2-NEXT: movdqa %xmm1, %xmm3
; SSE2-NEXT: psrlq %xmm4, %xmm3
; SSE2-NEXT: psrlq %xmm5, %xmm1
; SSE2-NEXT: movsd {{.*#+}} xmm1 = xmm3[0],xmm1[1]
; SSE2-NEXT: xorpd %xmm2, %xmm1
; SSE2-NEXT: psubq %xmm2, %xmm1
; SSE2-NEXT: retq
;
; SSE41-LABEL: combine_vec4i64_ashr_clamped:
; SSE41: # %bb.0:
; SSE41-NEXT: movdqa %xmm0, %xmm4
; SSE41-NEXT: movdqa {{.*#+}} xmm7 = [9223372039002259456,9223372039002259456]
; SSE41-NEXT: movdqa %xmm3, %xmm0
; SSE41-NEXT: pxor %xmm7, %xmm0
; SSE41-NEXT: movdqa {{.*#+}} xmm8 = [9223372039002259519,9223372039002259519]
; SSE41-NEXT: movdqa %xmm8, %xmm6
; SSE41-NEXT: pcmpeqd %xmm0, %xmm6
; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm0[0,0,2,2]
; SSE41-NEXT: movdqa {{.*#+}} xmm5 = [2147483711,2147483711,2147483711,2147483711]
; SSE41-NEXT: movdqa %xmm5, %xmm0
; SSE41-NEXT: pcmpgtd %xmm9, %xmm0
; SSE41-NEXT: pand %xmm6, %xmm0
; SSE41-NEXT: movapd {{.*#+}} xmm9 = [63,63]
; SSE41-NEXT: movapd %xmm9, %xmm6
; SSE41-NEXT: blendvpd %xmm0, %xmm3, %xmm6
; SSE41-NEXT: pxor %xmm2, %xmm7
; SSE41-NEXT: pcmpeqd %xmm7, %xmm8
; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm7[0,0,2,2]
; SSE41-NEXT: pcmpgtd %xmm0, %xmm5
; SSE41-NEXT: pand %xmm8, %xmm5
; SSE41-NEXT: movdqa %xmm5, %xmm0
; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm9
; SSE41-NEXT: movdqa {{.*#+}} xmm0 = [9223372036854775808,9223372036854775808]
; SSE41-NEXT: movdqa %xmm0, %xmm2
; SSE41-NEXT: psrlq %xmm9, %xmm2
; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm9[2,3,2,3]
; SSE41-NEXT: movdqa %xmm0, %xmm5
; SSE41-NEXT: psrlq %xmm3, %xmm5
; SSE41-NEXT: pblendw {{.*#+}} xmm5 = xmm2[0,1,2,3],xmm5[4,5,6,7]
; SSE41-NEXT: movdqa %xmm4, %xmm2
; SSE41-NEXT: psrlq %xmm9, %xmm2
; SSE41-NEXT: psrlq %xmm3, %xmm4
; SSE41-NEXT: pblendw {{.*#+}} xmm4 = xmm2[0,1,2,3],xmm4[4,5,6,7]
; SSE41-NEXT: pxor %xmm5, %xmm4
; SSE41-NEXT: psubq %xmm5, %xmm4
; SSE41-NEXT: movdqa %xmm0, %xmm2
; SSE41-NEXT: psrlq %xmm6, %xmm2
; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm6[2,3,2,3]
; SSE41-NEXT: psrlq %xmm3, %xmm0
; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm2[0,1,2,3],xmm0[4,5,6,7]
; SSE41-NEXT: movdqa %xmm1, %xmm2
; SSE41-NEXT: psrlq %xmm6, %xmm2
; SSE41-NEXT: psrlq %xmm3, %xmm1
; SSE41-NEXT: pblendw {{.*#+}} xmm1 = xmm2[0,1,2,3],xmm1[4,5,6,7]
; SSE41-NEXT: pxor %xmm0, %xmm1
; SSE41-NEXT: psubq %xmm0, %xmm1
; SSE41-NEXT: movdqa %xmm4, %xmm0
; SSE41-NEXT: retq
;
; AVX2-LABEL: combine_vec4i64_ashr_clamped:
; AVX2: # %bb.0:
; AVX2-NEXT: vpbroadcastq {{.*#+}} ymm2 = [9223372036854775808,9223372036854775808,9223372036854775808,9223372036854775808]
; AVX2-NEXT: vpxor %ymm2, %ymm1, %ymm3
; AVX2-NEXT: vpbroadcastq {{.*#+}} ymm4 = [9223372036854775870,9223372036854775870,9223372036854775870,9223372036854775870]
; AVX2-NEXT: vpcmpgtq %ymm4, %ymm3, %ymm3
; AVX2-NEXT: vbroadcastsd {{.*#+}} ymm4 = [63,63,63,63]
; AVX2-NEXT: vblendvpd %ymm3, %ymm4, %ymm1, %ymm1
; AVX2-NEXT: vpsrlvq %ymm1, %ymm2, %ymm2
; AVX2-NEXT: vpsrlvq %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vpxor %ymm2, %ymm0, %ymm0
; AVX2-NEXT: vpsubq %ymm2, %ymm0, %ymm0
; AVX2-NEXT: retq
;
; AVX512-LABEL: combine_vec4i64_ashr_clamped:
; AVX512: # %bb.0:
; AVX512-NEXT: vpsravq %ymm1, %ymm0, %ymm0
; AVX512-NEXT: retq
%1 = tail call <4 x i64> @llvm.umin.v4i64(<4 x i64> %y, <4 x i64> <i64 63, i64 63, i64 63, i64 63>)
%2 = ashr <4 x i64> %x, %1
ret <4 x i64> %2
}
Codebase Browser
llvm