summaryrefslogtreecommitdiffstats
path: root/llvm/test
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/test')
-rw-r--r--llvm/test/CodeGen/X86/avx-splat.ll5
-rw-r--r--llvm/test/CodeGen/X86/avx2-conversions.ll4
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll3
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll4
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll18
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll129
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-combining.ll11
-rw-r--r--llvm/test/CodeGen/X86/vector-trunc.ll35
8 files changed, 90 insertions, 119 deletions
diff --git a/llvm/test/CodeGen/X86/avx-splat.ll b/llvm/test/CodeGen/X86/avx-splat.ll
index 34b0e5ab516..c29bb0c952f 100644
--- a/llvm/test/CodeGen/X86/avx-splat.ll
+++ b/llvm/test/CodeGen/X86/avx-splat.ll
@@ -124,9 +124,8 @@ entry:
define <8 x float> @funcH(<8 x float> %a) nounwind uwtable readnone ssp {
; CHECK-LABEL: funcH:
; CHECK: ## BB#0: ## %entry
-; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm0
-; CHECK-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[1,1,1,1]
-; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[1,1,1,1,5,5,5,5]
+; CHECK-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; CHECK-NEXT: retq
entry:
%shuffle = shufflevector <8 x float> %a, <8 x float> undef, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
diff --git a/llvm/test/CodeGen/X86/avx2-conversions.ll b/llvm/test/CodeGen/X86/avx2-conversions.ll
index 17629276825..6567b933ad7 100644
--- a/llvm/test/CodeGen/X86/avx2-conversions.ll
+++ b/llvm/test/CodeGen/X86/avx2-conversions.ll
@@ -4,8 +4,8 @@
define <4 x i32> @trunc4(<4 x i64> %A) nounwind {
; CHECK-LABEL: trunc4:
; CHECK: ## BB#0:
-; CHECK-NEXT: vmovdqa {{.*#+}} ymm1 = <0,2,4,6,u,u,u,u>
-; CHECK-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; CHECK-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; CHECK-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: retq
%B = trunc <4 x i64> %A to <4 x i32>
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
index 0fc34657160..3d4e0ae6e3c 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
@@ -718,9 +718,8 @@ define <16 x i16> @shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_1
;
; AVX2-LABEL: shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_16:
; AVX2: # BB#0:
-; AVX2-NEXT: vpbroadcastw %xmm1, %ymm1
+; AVX2-NEXT: vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
; AVX2-NEXT: vpbroadcastd %xmm0, %ymm0
-; AVX2-NEXT: vpblendw {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7],ymm0[8],ymm1[9],ymm0[10],ymm1[11],ymm0[12],ymm1[13],ymm0[14],ymm1[15]
; AVX2-NEXT: retq
%shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16>
ret <16 x i16> %shuffle
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
index a63ff465d6e..aa122d14f13 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
@@ -977,10 +977,8 @@ define <32 x i8> @shuffle_v32i8_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_
;
; AVX2-LABEL: shuffle_v32i8_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32:
; AVX2: # BB#0:
+; AVX2-NEXT: vpunpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
; AVX2-NEXT: vpbroadcastw %xmm0, %ymm0
-; AVX2-NEXT: vpbroadcastb %xmm1, %ymm1
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0,255,0]
-; AVX2-NEXT: vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
; AVX2-NEXT: retq
%shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32, i32 0, i32 32>
ret <32 x i8> %shuffle
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
index 29dc59c8d85..a5a4cb5faed 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
@@ -112,8 +112,8 @@ define <4 x double> @shuffle_v4f64_1000(<4 x double> %a, <4 x double> %b) {
define <4 x double> @shuffle_v4f64_2200(<4 x double> %a, <4 x double> %b) {
; AVX1-LABEL: shuffle_v4f64_2200:
; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4f64_2200:
@@ -153,8 +153,8 @@ define <4 x double> @shuffle_v4f64_3330(<4 x double> %a, <4 x double> %b) {
define <4 x double> @shuffle_v4f64_3210(<4 x double> %a, <4 x double> %b) {
; AVX1-LABEL: shuffle_v4f64_3210:
; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4f64_3210:
@@ -517,9 +517,8 @@ define <4 x double> @shuffle_v4f64_22uu(<4 x double> %a, <4 x double> %b) {
define <4 x double> @shuffle_v4f64_3333(<4 x double> %a, <4 x double> %b) {
; AVX1-LABEL: shuffle_v4f64_3333:
; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vmovhlps {{.*#+}} xmm0 = xmm0[1,1]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,1,3,3]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4f64_3333:
@@ -663,8 +662,8 @@ define <4 x i64> @shuffle_v4i64_1000(<4 x i64> %a, <4 x i64> %b) {
define <4 x i64> @shuffle_v4i64_2200(<4 x i64> %a, <4 x i64> %b) {
; AVX1-LABEL: shuffle_v4i64_2200:
; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4i64_2200:
@@ -704,8 +703,8 @@ define <4 x i64> @shuffle_v4i64_3330(<4 x i64> %a, <4 x i64> %b) {
define <4 x i64> @shuffle_v4i64_3210(<4 x i64> %a, <4 x i64> %b) {
; AVX1-LABEL: shuffle_v4i64_3210:
; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4i64_3210:
@@ -1172,9 +1171,8 @@ define <4 x i64> @shuffle_v4i64_22uu(<4 x i64> %a, <4 x i64> %b) {
define <4 x i64> @shuffle_v4i64_3333(<4 x i64> %a, <4 x i64> %b) {
; AVX1-LABEL: shuffle_v4i64_3333:
; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,1]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,1,3,3]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v4i64_3333:
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
index 1908a5ae207..dfd03d32453 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
@@ -195,18 +195,15 @@ define <8 x float> @shuffle_v8f32_81a3c5e7(<8 x float> %a, <8 x float> %b) {
define <8 x float> @shuffle_v8f32_08080808(<8 x float> %a, <8 x float> %b) {
; AVX1-LABEL: shuffle_v8f32_08080808:
; AVX1: # BB#0:
-; AVX1-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,0,2,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
-; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,0]
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,2,1,3]
; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; AVX1-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8f32_08080808:
; AVX2: # BB#0:
-; AVX2-NEXT: vbroadcastss %xmm1, %ymm1
+; AVX2-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0],xmm1[0],zero,zero
; AVX2-NEXT: vbroadcastsd %xmm0, %ymm0
-; AVX2-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
ret <8 x float> %shuffle
@@ -647,10 +644,10 @@ define <8 x float> @shuffle_v8f32_c348cda0(<8 x float> %a, <8 x float> %b) {
;
; AVX2-LABEL: shuffle_v8f32_c348cda0:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <u,3,4,u,u,u,u,0>
-; AVX2-NEXT: vpermps %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <4,u,u,0,4,5,2,u>
-; AVX2-NEXT: vpermps %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[0,3,0,0,4,7,4,4]
+; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,3,2,1]
+; AVX2-NEXT: vpermilps {{.*#+}} ymm1 = ymm1[0,1,2,0,4,5,6,4]
+; AVX2-NEXT: vpermpd {{.*#+}} ymm1 = ymm1[2,1,2,1]
; AVX2-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0],ymm0[1,2],ymm1[3,4,5,6],ymm0[7]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 12, i32 3, i32 4, i32 8, i32 12, i32 13, i32 10, i32 0>
@@ -660,21 +657,21 @@ define <8 x float> @shuffle_v8f32_c348cda0(<8 x float> %a, <8 x float> %b) {
define <8 x float> @shuffle_v8f32_f511235a(<8 x float> %a, <8 x float> %b) {
; AVX1-LABEL: shuffle_v8f32_f511235a:
; AVX1: # BB#0:
+; AVX1-NEXT: vpermilps {{.*#+}} ymm1 = ymm1[3,1,2,2,7,5,6,6]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm1[2,3,0,1]
; AVX1-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3,0,1]
; AVX1-NEXT: vpermilpd {{.*#+}} ymm2 = ymm2[0,0,3,2]
; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[0,1,1,1,4,5,5,5]
; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm2[0],ymm0[1],ymm2[2],ymm0[3]
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm1[2,3,0,1]
-; AVX1-NEXT: vpermilps {{.*#+}} ymm1 = ymm1[3,1,2,2,7,5,6,6]
; AVX1-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0],ymm0[1,2,3,4,5,6],ymm1[7]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8f32_f511235a:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <7,u,u,u,u,u,u,2>
-; AVX2-NEXT: vpermps %ymm1, %ymm2, %ymm1
; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <u,5,1,1,2,3,5,u>
; AVX2-NEXT: vpermps %ymm0, %ymm2, %ymm0
+; AVX2-NEXT: vpermilps {{.*#+}} ymm1 = ymm1[3,1,2,2,7,5,6,6]
+; AVX2-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm1[2,3,0,1]
; AVX2-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0],ymm0[1,2,3,4,5,6],ymm1[7]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 15, i32 5, i32 1, i32 1, i32 2, i32 3, i32 5, i32 10>
@@ -682,50 +679,31 @@ define <8 x float> @shuffle_v8f32_f511235a(<8 x float> %a, <8 x float> %b) {
}
define <8 x float> @shuffle_v8f32_32103210(<8 x float> %a, <8 x float> %b) {
-; AVX1-LABEL: shuffle_v8f32_32103210:
-; AVX1: # BB#0:
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; AVX1-NEXT: retq
-;
-; AVX2-LABEL: shuffle_v8f32_32103210:
-; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm1 = [3,2,1,0,3,2,1,0]
-; AVX2-NEXT: vpermps %ymm0, %ymm1, %ymm0
-; AVX2-NEXT: retq
+; ALL-LABEL: shuffle_v8f32_32103210:
+; ALL: # BB#0:
+; ALL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
+; ALL-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; ALL-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 3, i32 2, i32 1, i32 0>
ret <8 x float> %shuffle
}
define <8 x float> @shuffle_v8f32_76547654(<8 x float> %a, <8 x float> %b) {
-; AVX1-LABEL: shuffle_v8f32_76547654:
-; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; AVX1-NEXT: retq
-;
-; AVX2-LABEL: shuffle_v8f32_76547654:
-; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm1 = [7,6,5,4,7,6,5,4]
-; AVX2-NEXT: vpermps %ymm0, %ymm1, %ymm0
-; AVX2-NEXT: retq
+; ALL-LABEL: shuffle_v8f32_76547654:
+; ALL: # BB#0:
+; ALL-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; ALL-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
+; ALL-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 7, i32 6, i32 5, i32 4>
ret <8 x float> %shuffle
}
define <8 x float> @shuffle_v8f32_76543210(<8 x float> %a, <8 x float> %b) {
-; AVX1-LABEL: shuffle_v8f32_76543210:
-; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
-; AVX1-NEXT: retq
-;
-; AVX2-LABEL: shuffle_v8f32_76543210:
-; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm1 = [7,6,5,4,3,2,1,0]
-; AVX2-NEXT: vpermps %ymm0, %ymm1, %ymm0
-; AVX2-NEXT: retq
+; ALL-LABEL: shuffle_v8f32_76543210:
+; ALL: # BB#0:
+; ALL-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; ALL-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
+; ALL-NEXT: retq
%shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
ret <8 x float> %shuffle
}
@@ -783,10 +761,10 @@ define <8 x float> @PR21138(<8 x float> %truc, <8 x float> %tchose) {
;
; AVX2-LABEL: PR21138:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <u,u,u,u,1,3,5,7>
-; AVX2-NEXT: vpermps %ymm1, %ymm2, %ymm1
-; AVX2-NEXT: vmovaps {{.*#+}} ymm2 = <1,3,5,7,u,u,u,u>
-; AVX2-NEXT: vpermps %ymm0, %ymm2, %ymm0
+; AVX2-NEXT: vpermilps {{.*#+}} ymm1 = ymm1[1,3,1,3,5,7,5,7]
+; AVX2-NEXT: vpermpd {{.*#+}} ymm1 = ymm1[0,1,0,3]
+; AVX2-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[1,3,1,3,5,7,5,7]
+; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,3,2,3]
; AVX2-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x float> %truc, <8 x float> %tchose, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
@@ -844,9 +822,8 @@ define <8 x float> @shuffle_v8f32_uuuu1111(<8 x float> %a, <8 x float> %b) {
define <8 x float> @shuffle_v8f32_44444444(<8 x float> %a, <8 x float> %b) {
; AVX1-LABEL: shuffle_v8f32_44444444:
; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[0,0,0,0,4,4,4,4]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8f32_44444444:
@@ -1107,18 +1084,15 @@ define <8 x i32> @shuffle_v8i32_81a3c5e7(<8 x i32> %a, <8 x i32> %b) {
define <8 x i32> @shuffle_v8i32_08080808(<8 x i32> %a, <8 x i32> %b) {
; AVX1-LABEL: shuffle_v8i32_08080808:
; AVX1: # BB#0:
-; AVX1-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,0,2,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
-; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,0]
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,2,1,3]
; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; AVX1-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8i32_08080808:
; AVX2: # BB#0:
-; AVX2-NEXT: vpbroadcastd %xmm1, %ymm1
+; AVX2-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; AVX2-NEXT: vpbroadcastq %xmm0, %ymm0
-; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
ret <8 x i32> %shuffle
@@ -1762,19 +1736,18 @@ define <8 x i32> @shuffle_v8i32_uuu3uu66(<8 x i32> %a, <8 x i32> %b) {
define <8 x i32> @shuffle_v8i32_6caa87e5(<8 x i32> %a, <8 x i32> %b) {
; AVX1-LABEL: shuffle_v8i32_6caa87e5:
; AVX1: # BB#0:
+; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm1[2,3,0,1]
; AVX1-NEXT: vshufps {{.*#+}} ymm1 = ymm2[0,0],ymm1[2,2],ymm2[4,4],ymm1[6,6]
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
; AVX1-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3,4],ymm0[5],ymm1[6],ymm0[7]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8i32_6caa87e5:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <u,4,2,2,0,u,6,u>
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[3,1,3,2]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,0,2,2,4,4,6,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[2,1,0,3]
; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3,4],ymm0[5],ymm1[6],ymm0[7]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 6, i32 12, i32 10, i32 10, i32 8, i32 7, i32 14, i32 5>
@@ -1790,8 +1763,8 @@ define <8 x i32> @shuffle_v8i32_32103210(<8 x i32> %a, <8 x i32> %b) {
;
; AVX2-LABEL: shuffle_v8i32_32103210:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm1 = [3,2,1,0,3,2,1,0]
-; AVX2-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; AVX2-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[3,2,1,0]
+; AVX2-NEXT: vinserti128 $1, %xmm0, %ymm0, %ymm0
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 3, i32 2, i32 1, i32 0>
ret <8 x i32> %shuffle
@@ -1800,15 +1773,14 @@ define <8 x i32> @shuffle_v8i32_32103210(<8 x i32> %a, <8 x i32> %b) {
define <8 x i32> @shuffle_v8i32_76547654(<8 x i32> %a, <8 x i32> %b) {
; AVX1-LABEL: shuffle_v8i32_76547654:
; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8i32_76547654:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm1 = [7,6,5,4,7,6,5,4]
-; AVX2-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX2-NEXT: vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 7, i32 6, i32 5, i32 4>
ret <8 x i32> %shuffle
@@ -1817,14 +1789,14 @@ define <8 x i32> @shuffle_v8i32_76547654(<8 x i32> %a, <8 x i32> %b) {
define <8 x i32> @shuffle_v8i32_76543210(<8 x i32> %a, <8 x i32> %b) {
; AVX1-LABEL: shuffle_v8i32_76543210:
; AVX1: # BB#0:
-; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8i32_76543210:
; AVX2: # BB#0:
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm1 = [7,6,5,4,3,2,1,0]
-; AVX2-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX2-NEXT: vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
; AVX2-NEXT: retq
%shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
ret <8 x i32> %shuffle
@@ -2008,9 +1980,8 @@ define <8 x i32> @shuffle_v8i32_2A3Buuuu(<8 x i32> %a, <8 x i32> %b) {
define <8 x i32> @shuffle_v8i32_44444444(<8 x i32> %a, <8 x i32> %b) {
; AVX1-LABEL: shuffle_v8i32_44444444:
; AVX1: # BB#0:
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
-; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[0,0,0,0,4,4,4,4]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: shuffle_v8i32_44444444:
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-combining.ll b/llvm/test/CodeGen/X86/vector-shuffle-combining.ll
index c1160c00c24..a7a69494196 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-combining.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-combining.ll
@@ -2636,15 +2636,16 @@ define <8 x i32> @combine_unneeded_subvector1(<8 x i32> %a) {
; AVX1: # BB#0:
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX1-NEXT: vpaddd {{.*}}(%rip), %xmm0, %xmm0
-; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX1-NEXT: retq
;
; AVX2-LABEL: combine_unneeded_subvector1:
; AVX2: # BB#0:
; AVX2-NEXT: vpaddd {{.*}}(%rip), %ymm0, %ymm0
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm1 = [7,6,5,4,7,6,5,4]
-; AVX2-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; AVX2-NEXT: vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,2,3]
; AVX2-NEXT: retq
%b = add <8 x i32> %a, <i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8>
%c = shufflevector <8 x i32> %b, <8 x i32> undef, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 7, i32 6, i32 5, i32 4>
@@ -2898,8 +2899,8 @@ define <8 x float> @PR22412(<8 x float> %a, <8 x float> %b) {
; AVX2-LABEL: PR22412:
; AVX2: # BB#0: # %entry
; AVX2-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3]
-; AVX2-NEXT: vmovaps {{.*#+}} ymm1 = [1,0,7,6,5,4,3,2]
-; AVX2-NEXT: vpermps %ymm0, %ymm1, %ymm0
+; AVX2-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[1,0,3,2,5,4,7,6]
+; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,3,2,1]
; AVX2-NEXT: retq
entry:
%s1 = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 1, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
diff --git a/llvm/test/CodeGen/X86/vector-trunc.ll b/llvm/test/CodeGen/X86/vector-trunc.ll
index 8c02c5a5433..fb6c41ab1ce 100644
--- a/llvm/test/CodeGen/X86/vector-trunc.ll
+++ b/llvm/test/CodeGen/X86/vector-trunc.ll
@@ -52,9 +52,10 @@ define <8 x i32> @trunc8i64_8i32(<8 x i64> %a) {
;
; AVX2-LABEL: trunc8i64_8i32:
; AVX2: # BB#0: # %entry
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <0,2,4,6,u,u,u,u>
-; AVX2-NEXT: vpermd %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,3,2,3]
; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; AVX2-NEXT: retq
;
@@ -135,9 +136,10 @@ define <8 x i16> @trunc8i64_8i16(<8 x i64> %a) {
;
; AVX2-LABEL: trunc8i64_8i16:
; AVX2: # BB#0: # %entry
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <0,2,4,6,u,u,u,u>
-; AVX2-NEXT: vpermd %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,3,2,3]
; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; AVX2-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13],zero,zero,zero,zero,zero,zero,zero,zero,ymm0[16,17,20,21,24,25,28,29],zero,zero,zero,zero,zero,zero,zero,zero
; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
@@ -187,9 +189,10 @@ define void @trunc8i64_8i8(<8 x i64> %a) {
;
; AVX2-LABEL: trunc8i64_8i8:
; AVX2: # BB#0: # %entry
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <0,2,4,6,u,u,u,u>
-; AVX2-NEXT: vpermd %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,3,2,3]
; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1
; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0
@@ -422,9 +425,10 @@ define <8 x i32> @trunc2x4i64_8i32(<4 x i64> %a, <4 x i64> %b) {
;
; AVX2-LABEL: trunc2x4i64_8i32:
; AVX2: # BB#0: # %entry
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <0,2,4,6,u,u,u,u>
-; AVX2-NEXT: vpermd %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,3,2,3]
; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; AVX2-NEXT: retq
;
@@ -517,9 +521,10 @@ define <8 x i16> @trunc2x4i64_8i16(<4 x i64> %a, <4 x i64> %b) {
;
; AVX2-LABEL: trunc2x4i64_8i16:
; AVX2: # BB#0: # %entry
-; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = <0,2,4,6,u,u,u,u>
-; AVX2-NEXT: vpermd %ymm0, %ymm2, %ymm0
-; AVX2-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,2,3]
+; AVX2-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,0,2,4,6,4,6]
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,3,2,3]
; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1
; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0
OpenPOWER on IntegriCloud