[llvm] d657206 - [X86] Add inst fixup for `unpckps` -> `unpckdq`.
Noah Goldstein via llvm-commits
llvm-commits at lists.llvm.org
Sun Apr 9 22:17:36 PDT 2023
Author: Noah Goldstein
Date: 2023-04-10T00:17:00-05:00
New Revision: d65720652dd644483530ecb547365a2239a97979
URL: https://github.com/llvm/llvm-project/commit/d65720652dd644483530ecb547365a2239a97979
DIFF: https://github.com/llvm/llvm-project/commit/d65720652dd644483530ecb547365a2239a97979.diff
LOG: [X86] Add inst fixup for `unpckps` -> `unpckdq`.
`unpckps` has the same performance as `unpckpd` (only port5) wereas
`unpckdq` can run on p15 on some newer architectures.
`unpckdq` is in the integer domain, so only do the transform if the
target has no bypass delay on shuffles (SKL+).
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D147729
Added:
Modified:
llvm/lib/Target/X86/X86FixupInstTuning.cpp
llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll
llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll
Removed:
################################################################################
diff --git a/llvm/lib/Target/X86/X86FixupInstTuning.cpp b/llvm/lib/Target/X86/X86FixupInstTuning.cpp
index 6f63cca8c7d4..ab0dd72fd796 100644
--- a/llvm/lib/Target/X86/X86FixupInstTuning.cpp
+++ b/llvm/lib/Target/X86/X86FixupInstTuning.cpp
@@ -166,7 +166,9 @@ bool X86FixupInstTuningPass::processInstruction(
// -> `vunpck{l|h}qdq`
// 2) If `vshufpd` faster than `vunpck{l|h}pd`
// -> `vshufpd`
- auto ProcessUNPCKPD = [&](unsigned NewOpc, unsigned MaskImm) -> bool {
+ //
+ // `vunpcklps` -> `vunpckldq` (for all operand types if no bypass delay)
+ auto ProcessUNPCK = [&](unsigned NewOpc, unsigned MaskImm) -> bool {
if (!NewOpcPreferable(NewOpc, /*ReplaceInTie*/ false))
return false;
@@ -175,7 +177,7 @@ bool X86FixupInstTuningPass::processInstruction(
return true;
};
- auto ProcessUNPCKPDToIntDomain = [&](unsigned NewOpc) -> bool {
+ auto ProcessUNPCKToIntDomain = [&](unsigned NewOpc) -> bool {
// TODO it may be worth it to set ReplaceInTie to `true` as there is no real
// downside to the integer unpck, but if someone doesn't specify exact
// target we won't find it faster.
@@ -188,19 +190,23 @@ bool X86FixupInstTuningPass::processInstruction(
auto ProcessUNPCKLPDrr = [&](unsigned NewOpcIntDomain,
unsigned NewOpc) -> bool {
- if (ProcessUNPCKPDToIntDomain(NewOpcIntDomain))
+ if (ProcessUNPCKToIntDomain(NewOpcIntDomain))
return true;
- return ProcessUNPCKPD(NewOpc, 0x00);
+ return ProcessUNPCK(NewOpc, 0x00);
};
auto ProcessUNPCKHPDrr = [&](unsigned NewOpcIntDomain,
unsigned NewOpc) -> bool {
- if (ProcessUNPCKPDToIntDomain(NewOpcIntDomain))
+ if (ProcessUNPCKToIntDomain(NewOpcIntDomain))
return true;
- return ProcessUNPCKPD(NewOpc, 0xff);
+ return ProcessUNPCK(NewOpc, 0xff);
};
auto ProcessUNPCKPDrm = [&](unsigned NewOpcIntDomain) -> bool {
- return ProcessUNPCKPDToIntDomain(NewOpcIntDomain);
+ return ProcessUNPCKToIntDomain(NewOpcIntDomain);
+ };
+
+ auto ProcessUNPCKPS = [&](unsigned NewOpc) -> bool {
+ return ProcessUNPCKToIntDomain(NewOpc);
};
switch (Opc) {
@@ -351,6 +357,103 @@ bool X86FixupInstTuningPass::processInstruction(
return ProcessUNPCKPDrm(X86::VPUNPCKHQDQZ256rmkz);
case X86::VUNPCKHPDZrmkz:
return ProcessUNPCKPDrm(X86::VPUNPCKHQDQZrmkz);
+
+ case X86::UNPCKLPSrr:
+ return ProcessUNPCKPS(X86::PUNPCKLDQrr);
+ case X86::VUNPCKLPSrr:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQrr);
+ case X86::VUNPCKLPSYrr:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQYrr);
+ case X86::VUNPCKLPSZ128rr:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rr);
+ case X86::VUNPCKLPSZ256rr:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rr);
+ case X86::VUNPCKLPSZrr:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrr);
+ case X86::VUNPCKLPSZ128rrk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rrk);
+ case X86::VUNPCKLPSZ256rrk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rrk);
+ case X86::VUNPCKLPSZrrk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrrk);
+ case X86::VUNPCKLPSZ128rrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rrkz);
+ case X86::VUNPCKLPSZ256rrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rrkz);
+ case X86::VUNPCKLPSZrrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrrkz);
+ case X86::UNPCKHPSrr:
+ return ProcessUNPCKPS(X86::PUNPCKHDQrr);
+ case X86::VUNPCKHPSrr:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQrr);
+ case X86::VUNPCKHPSYrr:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQYrr);
+ case X86::VUNPCKHPSZ128rr:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rr);
+ case X86::VUNPCKHPSZ256rr:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rr);
+ case X86::VUNPCKHPSZrr:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrr);
+ case X86::VUNPCKHPSZ128rrk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rrk);
+ case X86::VUNPCKHPSZ256rrk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rrk);
+ case X86::VUNPCKHPSZrrk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrrk);
+ case X86::VUNPCKHPSZ128rrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rrkz);
+ case X86::VUNPCKHPSZ256rrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rrkz);
+ case X86::VUNPCKHPSZrrkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrrkz);
+ case X86::UNPCKLPSrm:
+ return ProcessUNPCKPS(X86::PUNPCKLDQrm);
+ case X86::VUNPCKLPSrm:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQrm);
+ case X86::VUNPCKLPSYrm:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQYrm);
+ case X86::VUNPCKLPSZ128rm:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rm);
+ case X86::VUNPCKLPSZ256rm:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rm);
+ case X86::VUNPCKLPSZrm:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrm);
+ case X86::VUNPCKLPSZ128rmk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rmk);
+ case X86::VUNPCKLPSZ256rmk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rmk);
+ case X86::VUNPCKLPSZrmk:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrmk);
+ case X86::VUNPCKLPSZ128rmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ128rmkz);
+ case X86::VUNPCKLPSZ256rmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZ256rmkz);
+ case X86::VUNPCKLPSZrmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKLDQZrmkz);
+ case X86::UNPCKHPSrm:
+ return ProcessUNPCKPS(X86::PUNPCKHDQrm);
+ case X86::VUNPCKHPSrm:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQrm);
+ case X86::VUNPCKHPSYrm:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQYrm);
+ case X86::VUNPCKHPSZ128rm:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rm);
+ case X86::VUNPCKHPSZ256rm:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rm);
+ case X86::VUNPCKHPSZrm:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrm);
+ case X86::VUNPCKHPSZ128rmk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rmk);
+ case X86::VUNPCKHPSZ256rmk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rmk);
+ case X86::VUNPCKHPSZrmk:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrmk);
+ case X86::VUNPCKHPSZ128rmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ128rmkz);
+ case X86::VUNPCKHPSZ256rmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZ256rmkz);
+ case X86::VUNPCKHPSZrmkz:
+ return ProcessUNPCKPS(X86::VPUNPCKHDQZrmkz);
default:
return false;
}
diff --git a/llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll b/llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll
index a53e83ac77f8..d0e3ad9b1908 100644
--- a/llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll
+++ b/llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll
@@ -25,37 +25,137 @@ define <16 x float> @transform_VUNPCKHPSZrr(<16 x float> %a, <16 x float> %b) no
}
define <8 x float> @transform_VUNPCKLPSYrr(<8 x float> %a, <8 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ZNVER4-NEXT: retq
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
ret <8 x float> %shufp
}
define <8 x float> @transform_VUNPCKHPSYrr(<8 x float> %a, <8 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ZNVER4-NEXT: retq
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
ret <8 x float> %shufp
}
define <4 x float> @transform_VUNPCKLPSrr(<4 x float> %a, <4 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrr:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrr:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrr:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrr:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ZNVER4-NEXT: retq
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
ret <4 x float> %shufp
}
define <4 x float> @transform_VUNPCKHPSrr(<4 x float> %a, <4 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrr:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrr:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrr:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrr:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ZNVER4-NEXT: retq
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
ret <4 x float> %shufp
}
@@ -85,11 +185,41 @@ define <16 x float> @transform_VUNPCKHPSZrrkz(<16 x float> %a, <16 x float> %b,
}
define <8 x float> @transform_VUNPCKLPSYrrkz(<8 x float> %a, <8 x float> %b, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrrkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrrkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
%res = select <8 x i1> %mask, <8 x float> %shufp, <8 x float> zeroinitializer
@@ -97,11 +227,41 @@ define <8 x float> @transform_VUNPCKLPSYrrkz(<8 x float> %a, <8 x float> %b, i8
}
define <8 x float> @transform_VUNPCKHPSYrrkz(<8 x float> %a, <8 x float> %b, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrrkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrrkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
%res = select <8 x i1> %mask, <8 x float> %shufp, <8 x float> zeroinitializer
@@ -109,11 +269,41 @@ define <8 x float> @transform_VUNPCKHPSYrrkz(<8 x float> %a, <8 x float> %b, i8
}
define <4 x float> @transform_VUNPCKLPSrrkz(<4 x float> %a, <4 x float> %b, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrrkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrrkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
%res = select <4 x i1> %mask, <4 x float> %shufp, <4 x float> zeroinitializer
@@ -121,11 +311,41 @@ define <4 x float> @transform_VUNPCKLPSrrkz(<4 x float> %a, <4 x float> %b, i4 %
}
define <4 x float> @transform_VUNPCKHPSrrkz(<4 x float> %a, <4 x float> %b, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrrkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrrkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
%res = select <4 x i1> %mask, <4 x float> %shufp, <4 x float> zeroinitializer
@@ -159,12 +379,47 @@ define <16 x float> @transform_VUNPCKHPSZrrk(<16 x float> %a, <16 x float> %b, <
}
define <8 x float> @transform_VUNPCKLPSYrrk(<8 x float> %a, <8 x float> %b, <8 x float> %c, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrrk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
-; CHECK-NEXT: vmovaps %ymm2, %ymm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-SKX-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-V4-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-AVX512-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrrk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm2 {%k1} = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ZNVER4-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
%res = select <8 x i1> %mask, <8 x float> %shufp, <8 x float> %c
@@ -172,12 +427,47 @@ define <8 x float> @transform_VUNPCKLPSYrrk(<8 x float> %a, <8 x float> %b, <8 x
}
define <8 x float> @transform_VUNPCKHPSYrrk(<8 x float> %a, <8 x float> %b, <8 x float> %c, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrrk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
-; CHECK-NEXT: vmovaps %ymm2, %ymm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-SKX-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-V4-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-AVX512-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrrk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm2 {%k1} = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ZNVER4-NEXT: vmovaps %ymm2, %ymm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
%res = select <8 x i1> %mask, <8 x float> %shufp, <8 x float> %c
@@ -185,12 +475,47 @@ define <8 x float> @transform_VUNPCKHPSYrrk(<8 x float> %a, <8 x float> %b, <8 x
}
define <4 x float> @transform_VUNPCKLPSrrk(<4 x float> %a, <4 x float> %b, <4 x float> %c, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrrk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; CHECK-NEXT: vmovaps %xmm2, %xmm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-SKX-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-V4-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-AVX512-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrrk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm2 {%k1} = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ZNVER4-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
%res = select <4 x i1> %mask, <4 x float> %shufp, <4 x float> %c
@@ -198,12 +523,47 @@ define <4 x float> @transform_VUNPCKLPSrrk(<4 x float> %a, <4 x float> %b, <4 x
}
define <4 x float> @transform_VUNPCKHPSrrk(<4 x float> %a, <4 x float> %b, <4 x float> %c, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrrk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %edi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT: vmovaps %xmm2, %xmm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %edi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-SKX-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %edi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %edi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-V4-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %edi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-AVX512-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrrk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %edi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm2 {%k1} = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ZNVER4-NEXT: vmovaps %xmm2, %xmm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
%res = select <4 x i1> %mask, <4 x float> %shufp, <4 x float> %c
@@ -231,40 +591,140 @@ define <16 x float> @transform_VUNPCKHPSZrm(<16 x float> %a, ptr %pb) nounwind {
}
define <8 x float> @transform_VUNPCKLPSYrm(<8 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ZNVER4-NEXT: retq
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
ret <8 x float> %shufp
}
define <8 x float> @transform_VUNPCKHPSYrm(<8 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ZNVER4-NEXT: retq
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
ret <8 x float> %shufp
}
define <4 x float> @transform_VUNPCKLPSrm(<4 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrm:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrm:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrm:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrm:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ZNVER4-NEXT: retq
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
ret <4 x float> %shufp
}
define <4 x float> @transform_VUNPCKHPSrm(<4 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrm:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrm:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrm:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrm:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ZNVER4-NEXT: retq
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
ret <4 x float> %shufp
@@ -297,11 +757,41 @@ define <16 x float> @transform_VUNPCKHPSZrmkz(<16 x float> %a, ptr %pb, i16 %mas
}
define <8 x float> @transform_VUNPCKLPSYrmkz(<8 x float> %a, ptr %pb, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrmkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrmkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm0 {%k1} {z} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
@@ -310,11 +800,41 @@ define <8 x float> @transform_VUNPCKLPSYrmkz(<8 x float> %a, ptr %pb, i8 %mask_i
}
define <8 x float> @transform_VUNPCKHPSYrmkz(<8 x float> %a, ptr %pb, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrmkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrmkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm0 {%k1} {z} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
@@ -323,11 +843,41 @@ define <8 x float> @transform_VUNPCKHPSYrmkz(<8 x float> %a, ptr %pb, i8 %mask_i
}
define <4 x float> @transform_VUNPCKLPSrmkz(<4 x float> %a, ptr %pb, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrmkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrmkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm0 {%k1} {z} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
@@ -336,11 +886,41 @@ define <4 x float> @transform_VUNPCKLPSrmkz(<4 x float> %a, ptr %pb, i4 %mask_in
}
define <4 x float> @transform_VUNPCKHPSrmkz(<4 x float> %a, ptr %pb, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrmkz:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrmkz:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm0 {%k1} {z} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
@@ -377,12 +957,47 @@ define <16 x float> @transform_VUNPCKHPSZrmk(<16 x float> %a, ptr %pb, <16 x flo
}
define <8 x float> @transform_VUNPCKLPSYrmk(<8 x float> %a, ptr %pb, <8 x float> %c, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrmk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
-; CHECK-NEXT: vmovaps %ymm1, %ymm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-SKX-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-V4-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-AVX512-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSYrmk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} ymm1 {%k1} = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ZNVER4-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
@@ -391,12 +1006,47 @@ define <8 x float> @transform_VUNPCKLPSYrmk(<8 x float> %a, ptr %pb, <8 x float>
}
define <8 x float> @transform_VUNPCKHPSYrmk(<8 x float> %a, ptr %pb, <8 x float> %c, i8 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrmk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
-; CHECK-NEXT: vmovaps %ymm1, %ymm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-SKX-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-V4-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-AVX512-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSYrmk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} ymm1 {%k1} = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ZNVER4-NEXT: vmovaps %ymm1, %ymm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i8 %mask_int to <8 x i1>
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
@@ -405,12 +1055,47 @@ define <8 x float> @transform_VUNPCKHPSYrmk(<8 x float> %a, ptr %pb, <8 x float>
}
define <4 x float> @transform_VUNPCKLPSrmk(<4 x float> %a, ptr %pb, <4 x float> %c, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrmk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
-; CHECK-NEXT: vmovaps %xmm1, %xmm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-SKX-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-V4-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-AVX512-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKLPSrmk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpcklps {{.*#+}} xmm1 {%k1} = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ZNVER4-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
@@ -419,12 +1104,47 @@ define <4 x float> @transform_VUNPCKLPSrmk(<4 x float> %a, ptr %pb, <4 x float>
}
define <4 x float> @transform_VUNPCKHPSrmk(<4 x float> %a, ptr %pb, <4 x float> %c, i4 %mask_int) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrmk:
-; CHECK: # %bb.0:
-; CHECK-NEXT: kmovd %esi, %k1
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
-; CHECK-NEXT: vmovaps %xmm1, %xmm0
-; CHECK-NEXT: retq
+; CHECK-SKX-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-SKX: # %bb.0:
+; CHECK-SKX-NEXT: kmovd %esi, %k1
+; CHECK-SKX-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-SKX-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-SKX-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: kmovd %esi, %k1
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-V4-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-V4: # %bb.0:
+; CHECK-V4-NEXT: kmovd %esi, %k1
+; CHECK-V4-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-V4-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-V4-NEXT: retq
+;
+; CHECK-AVX512-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-AVX512: # %bb.0:
+; CHECK-AVX512-NEXT: kmovd %esi, %k1
+; CHECK-AVX512-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-AVX512-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-AVX512-NEXT: retq
+;
+; CHECK-ZNVER4-LABEL: transform_VUNPCKHPSrmk:
+; CHECK-ZNVER4: # %bb.0:
+; CHECK-ZNVER4-NEXT: kmovd %esi, %k1
+; CHECK-ZNVER4-NEXT: vunpckhps {{.*#+}} xmm1 {%k1} = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ZNVER4-NEXT: vmovaps %xmm1, %xmm0
+; CHECK-ZNVER4-NEXT: retq
%mask = bitcast i4 %mask_int to <4 x i1>
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
@@ -432,10 +1152,4 @@ define <4 x float> @transform_VUNPCKHPSrmk(<4 x float> %a, ptr %pb, <4 x float>
ret <4 x float> %res
}
;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
-; CHECK-AVX512: {{.*}}
; CHECK-ICX: {{.*}}
-; CHECK-ICX-BYPASS-DELAY: {{.*}}
-; CHECK-ICX-NO-BYPASS-DELAY: {{.*}}
-; CHECK-SKX: {{.*}}
-; CHECK-V4: {{.*}}
-; CHECK-ZNVER4: {{.*}}
diff --git a/llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll b/llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll
index f6f69824b29f..8fa2c3dab047 100644
--- a/llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll
+++ b/llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll
@@ -8,87 +8,222 @@
define <8 x float> @transform_VUNPCKLPSYrr(<8 x float> %a, <8 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-LABEL: transform_VUNPCKLPSYrr:
+; CHECK-SNB: # %bb.0:
+; CHECK-SNB-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; CHECK-SNB-NEXT: retq
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
ret <8 x float> %shufp
}
define <8 x float> @transform_VUNPCKHPSYrr(<8 x float> %a, <8 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-LABEL: transform_VUNPCKHPSYrr:
+; CHECK-SNB: # %bb.0:
+; CHECK-SNB-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; CHECK-SNB-NEXT: retq
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
ret <8 x float> %shufp
}
define <4 x float> @transform_VUNPCKLPSrr(<4 x float> %a, <4 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKLPSrr:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-SNB-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrr:
+; CHECK-SNB-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; CHECK-SNB-BYPASS-DELAY-NEXT: retq
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
ret <4 x float> %shufp
}
define <4 x float> @transform_VUNPCKHPSrr(<4 x float> %a, <4 x float> %b) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrr:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKHPSrr:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-SNB-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrr:
+; CHECK-SNB-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-SNB-BYPASS-DELAY-NEXT: retq
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
ret <4 x float> %shufp
}
define <8 x float> @transform_VUNPCKLPSYrm(<8 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSYrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-LABEL: transform_VUNPCKLPSYrm:
+; CHECK-SNB: # %bb.0:
+; CHECK-SNB-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
+; CHECK-SNB-NEXT: retq
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
ret <8 x float> %shufp
}
define <8 x float> @transform_VUNPCKHPSYrm(<8 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSYrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-LABEL: transform_VUNPCKHPSYrm:
+; CHECK-SNB: # %bb.0:
+; CHECK-SNB-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
+; CHECK-SNB-NEXT: retq
%b = load <8 x float>, ptr %pb
%shufp = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
ret <8 x float> %shufp
}
define <4 x float> @transform_VUNPCKLPSrm(<4 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKLPSrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKLPSrm:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-NO-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-SNB-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-BYPASS-DELAY-LABEL: transform_VUNPCKLPSrm:
+; CHECK-SNB-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-BYPASS-DELAY-NEXT: vunpcklps {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-SNB-BYPASS-DELAY-NEXT: retq
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
ret <4 x float> %shufp
}
define <4 x float> @transform_VUNPCKHPSrm(<4 x float> %a, ptr %pb) nounwind {
-; CHECK-LABEL: transform_VUNPCKHPSrm:
-; CHECK: # %bb.0:
-; CHECK-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
-; CHECK-NEXT: retq
+; CHECK-AVX2-LABEL: transform_VUNPCKHPSrm:
+; CHECK-AVX2: # %bb.0:
+; CHECK-AVX2-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-AVX2-NEXT: retq
+;
+; CHECK-ICX-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-ICX-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-ICX-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-ICX-BYPASS-DELAY: # %bb.0:
+; CHECK-ICX-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-ICX-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-NO-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-SNB-NO-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: vpunpckhdq {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-SNB-NO-BYPASS-DELAY-NEXT: retq
+;
+; CHECK-SNB-BYPASS-DELAY-LABEL: transform_VUNPCKHPSrm:
+; CHECK-SNB-BYPASS-DELAY: # %bb.0:
+; CHECK-SNB-BYPASS-DELAY-NEXT: vunpckhps {{.*#+}} xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
+; CHECK-SNB-BYPASS-DELAY-NEXT: retq
%b = load <4 x float>, ptr %pb
%shufp = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
ret <4 x float> %shufp
}
;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
-; CHECK-AVX2: {{.*}}
+; CHECK: {{.*}}
; CHECK-ICX: {{.*}}
-; CHECK-ICX-BYPASS-DELAY: {{.*}}
-; CHECK-ICX-NO-BYPASS-DELAY: {{.*}}
; CHECK-SKL: {{.*}}
-; CHECK-SNB: {{.*}}
-; CHECK-SNB-BYPASS-DELAY: {{.*}}
-; CHECK-SNB-NO-BYPASS-DELAY: {{.*}}
; CHECK-V3: {{.*}}
More information about the llvm-commits
mailing list