aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--llvm/lib/Target/X86/X86ISelLowering.cpp9
-rw-r--r--llvm/test/CodeGen/X86/vector-bitreverse.ll388
2 files changed, 313 insertions, 84 deletions
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index d91ea1ea..6281124 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -1323,11 +1323,15 @@ X86TargetLowering::X86TargetLowering(const X86TargetMachine &TM,
setOperationAction(ISD::STRICT_FDIV, MVT::v2f64, Legal);
}
- if (Subtarget.hasGFNI()) {
+ if (!Subtarget.useSoftFloat() && Subtarget.hasGFNI()) {
setOperationAction(ISD::BITREVERSE, MVT::i8, Custom);
setOperationAction(ISD::BITREVERSE, MVT::i16, Custom);
setOperationAction(ISD::BITREVERSE, MVT::i32, Custom);
setOperationAction(ISD::BITREVERSE, MVT::i64, Custom);
+
+ for (auto VT : {MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v2i64}) {
+ setOperationAction(ISD::BITREVERSE, VT, Custom);
+ }
}
if (!Subtarget.useSoftFloat() && Subtarget.hasSSSE3()) {
@@ -32694,7 +32698,8 @@ static SDValue LowerBITREVERSE(SDValue Op, const X86Subtarget &Subtarget,
if (Subtarget.hasXOP() && !VT.is512BitVector())
return LowerBITREVERSE_XOP(Op, DAG);
- assert(Subtarget.hasSSSE3() && "SSSE3 required for BITREVERSE");
+ assert((Subtarget.hasSSSE3() || Subtarget.hasGFNI()) &&
+ "SSSE3 or GFNI required for BITREVERSE");
SDValue In = Op.getOperand(0);
SDLoc DL(Op);
diff --git a/llvm/test/CodeGen/X86/vector-bitreverse.ll b/llvm/test/CodeGen/X86/vector-bitreverse.ll
index 5dcf190..834dfd6 100644
--- a/llvm/test/CodeGen/X86/vector-bitreverse.ll
+++ b/llvm/test/CodeGen/X86/vector-bitreverse.ll
@@ -8,7 +8,8 @@
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=ALL,AVX,AVX512,AVX512BW
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+xop,+avx | FileCheck %s --check-prefixes=ALL,XOP,XOPAVX1
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+xop,+avx2 | FileCheck %s --check-prefixes=ALL,XOP,XOPAVX2
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+ssse3,+gfni | FileCheck %s --check-prefixes=ALL,GFNISSE
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+gfni | FileCheck %s --check-prefixes=ALL,GFNISSE,GFNISSE2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+ssse3,+gfni | FileCheck %s --check-prefixes=ALL,GFNISSE,GFNISSSE3
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx,+gfni | FileCheck %s --check-prefixes=ALL,GFNIAVX,GFNIAVX1
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2,+gfni | FileCheck %s --check-prefixes=ALL,GFNIAVX,GFNIAVX2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+gfni | FileCheck %s --check-prefixes=ALL,GFNIAVX,GFNIAVX512,GFNIAVX512F
@@ -492,11 +493,20 @@ define <8 x i16> @test_bitreverse_v8i16(<8 x i16> %a) nounwind {
; XOP-NEXT: vpperm {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0, %xmm0
; XOP-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v8i16:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: pshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
-; GFNISSE-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v8i16:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm1
+; GFNISSE2-NEXT: psrlw $8, %xmm1
+; GFNISSE2-NEXT: psllw $8, %xmm0
+; GFNISSE2-NEXT: por %xmm1, %xmm0
+; GFNISSE2-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v8i16:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX-LABEL: test_bitreverse_v8i16:
; GFNIAVX: # %bb.0:
@@ -605,11 +615,25 @@ define <4 x i32> @test_bitreverse_v4i32(<4 x i32> %a) nounwind {
; XOP-NEXT: vpperm {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0, %xmm0
; XOP-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v4i32:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: pshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
-; GFNISSE-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v4i32:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm1, %xmm1
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm2
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm1[8],xmm2[9],xmm1[9],xmm2[10],xmm1[10],xmm2[11],xmm1[11],xmm2[12],xmm1[12],xmm2[13],xmm1[13],xmm2[14],xmm1[14],xmm2[15],xmm1[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} 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]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm2, %xmm0
+; GFNISSE2-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v4i32:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX-LABEL: test_bitreverse_v4i32:
; GFNIAVX: # %bb.0:
@@ -720,11 +744,27 @@ define <2 x i64> @test_bitreverse_v2i64(<2 x i64> %a) nounwind {
; XOP-NEXT: vpperm {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0, %xmm0
; XOP-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v2i64:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: pshufb {{.*#+}} xmm0 = xmm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
-; GFNISSE-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v2i64:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm1, %xmm1
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm2
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm1[8],xmm2[9],xmm1[9],xmm2[10],xmm1[10],xmm2[11],xmm1[11],xmm2[12],xmm1[12],xmm2[13],xmm1[13],xmm2[14],xmm1[14],xmm2[15],xmm1[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} 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]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm2, %xmm0
+; GFNISSE2-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v2i64:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX-LABEL: test_bitreverse_v2i64:
; GFNIAVX: # %bb.0:
@@ -1042,15 +1082,30 @@ define <16 x i16> @test_bitreverse_v16i16(<16 x i16> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v16i16:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm2 = [1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
-; GFNISSE-NEXT: pshufb %xmm2, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
-; GFNISSE-NEXT: pshufb %xmm2, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v16i16:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm2
+; GFNISSE2-NEXT: psrlw $8, %xmm2
+; GFNISSE2-NEXT: psllw $8, %xmm0
+; GFNISSE2-NEXT: por %xmm2, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm2 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm2, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm3
+; GFNISSE2-NEXT: psrlw $8, %xmm3
+; GFNISSE2-NEXT: psllw $8, %xmm1
+; GFNISSE2-NEXT: por %xmm3, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm2, %xmm1
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v16i16:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm2 = [1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v16i16:
; GFNIAVX1: # %bb.0:
@@ -1241,15 +1296,39 @@ define <8 x i32> @test_bitreverse_v8i32(<8 x i32> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v8i32:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm2 = [3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
-; GFNISSE-NEXT: pshufb %xmm2, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
-; GFNISSE-NEXT: pshufb %xmm2, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v8i32:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm2, %xmm2
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm3
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm2[8],xmm3[9],xmm2[9],xmm3[10],xmm2[10],xmm3[11],xmm2[11],xmm3[12],xmm2[12],xmm3[13],xmm2[13],xmm3[14],xmm2[14],xmm3[15],xmm2[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm3, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm4
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm4 = xmm4[8],xmm2[8],xmm4[9],xmm2[9],xmm4[10],xmm2[10],xmm4[11],xmm2[11],xmm4[12],xmm2[12],xmm4[13],xmm2[13],xmm4[14],xmm2[14],xmm4[15],xmm2[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm4 = xmm4[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm4, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v8i32:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm2 = [3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v8i32:
; GFNIAVX1: # %bb.0:
@@ -1444,15 +1523,43 @@ define <4 x i64> @test_bitreverse_v4i64(<4 x i64> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v4i64:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm2 = [7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
-; GFNISSE-NEXT: pshufb %xmm2, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
-; GFNISSE-NEXT: pshufb %xmm2, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v4i64:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm2, %xmm2
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm3
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm2[8],xmm3[9],xmm2[9],xmm3[10],xmm2[10],xmm3[11],xmm2[11],xmm3[12],xmm2[12],xmm3[13],xmm2[13],xmm3[14],xmm2[14],xmm3[15],xmm2[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm3, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm4
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm4 = xmm4[8],xmm2[8],xmm4[9],xmm2[9],xmm4[10],xmm2[10],xmm4[11],xmm2[11],xmm4[12],xmm2[12],xmm4[13],xmm2[13],xmm4[14],xmm2[14],xmm4[15],xmm2[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm4 = xmm4[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm4, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v4i64:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm2 = [7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm3 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm2, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm3, %xmm1
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v4i64:
; GFNIAVX1: # %bb.0:
@@ -2035,19 +2142,44 @@ define <32 x i16> @test_bitreverse_v32i16(<32 x i16> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm2, %ymm1, %ymm1
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v32i16:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm4 = [1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
-; GFNISSE-NEXT: pshufb %xmm4, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
-; GFNISSE-NEXT: pshufb %xmm4, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
-; GFNISSE-NEXT: pshufb %xmm4, %xmm2
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
-; GFNISSE-NEXT: pshufb %xmm4, %xmm3
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v32i16:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm4
+; GFNISSE2-NEXT: psrlw $8, %xmm4
+; GFNISSE2-NEXT: psllw $8, %xmm0
+; GFNISSE2-NEXT: por %xmm4, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm4 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm4, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm5
+; GFNISSE2-NEXT: psrlw $8, %xmm5
+; GFNISSE2-NEXT: psllw $8, %xmm1
+; GFNISSE2-NEXT: por %xmm5, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm4, %xmm1
+; GFNISSE2-NEXT: movdqa %xmm2, %xmm5
+; GFNISSE2-NEXT: psrlw $8, %xmm5
+; GFNISSE2-NEXT: psllw $8, %xmm2
+; GFNISSE2-NEXT: por %xmm5, %xmm2
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm4, %xmm2
+; GFNISSE2-NEXT: movdqa %xmm3, %xmm5
+; GFNISSE2-NEXT: psrlw $8, %xmm5
+; GFNISSE2-NEXT: psllw $8, %xmm3
+; GFNISSE2-NEXT: por %xmm5, %xmm3
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm4, %xmm3
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v32i16:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm4 = [1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm2
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm3
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v32i16:
; GFNIAVX1: # %bb.0:
@@ -2393,19 +2525,61 @@ define <16 x i32> @test_bitreverse_v16i32(<16 x i32> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm2, %ymm1, %ymm1
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v16i32:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm4 = [3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
-; GFNISSE-NEXT: pshufb %xmm4, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
-; GFNISSE-NEXT: pshufb %xmm4, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
-; GFNISSE-NEXT: pshufb %xmm4, %xmm2
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
-; GFNISSE-NEXT: pshufb %xmm4, %xmm3
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v16i32:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm4, %xmm4
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm5
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm5 = xmm5[8],xmm4[8],xmm5[9],xmm4[9],xmm5[10],xmm4[10],xmm5[11],xmm4[11],xmm5[12],xmm4[12],xmm5[13],xmm4[13],xmm5[14],xmm4[14],xmm5[15],xmm4[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm5 = xmm5[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm5 = xmm5[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm5, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1],xmm1[2],xmm4[2],xmm1[3],xmm4[3],xmm1[4],xmm4[4],xmm1[5],xmm4[5],xmm1[6],xmm4[6],xmm1[7],xmm4[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
+; GFNISSE2-NEXT: movdqa %xmm2, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1],xmm2[2],xmm4[2],xmm2[3],xmm4[3],xmm2[4],xmm4[4],xmm2[5],xmm4[5],xmm2[6],xmm4[6],xmm2[7],xmm4[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm2
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
+; GFNISSE2-NEXT: movdqa %xmm3, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1],xmm3[2],xmm4[2],xmm3[3],xmm4[3],xmm3[4],xmm4[4],xmm3[5],xmm4[5],xmm3[6],xmm4[6],xmm3[7],xmm4[7]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm3
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v16i32:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm4 = [3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm2
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm3
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v16i32:
; GFNIAVX1: # %bb.0:
@@ -2759,19 +2933,69 @@ define <8 x i64> @test_bitreverse_v8i64(<8 x i64> %a) nounwind {
; XOPAVX2-NEXT: vinserti128 $1, %xmm2, %ymm1, %ymm1
; XOPAVX2-NEXT: retq
;
-; GFNISSE-LABEL: test_bitreverse_v8i64:
-; GFNISSE: # %bb.0:
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm4 = [7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
-; GFNISSE-NEXT: pshufb %xmm4, %xmm0
-; GFNISSE-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
-; GFNISSE-NEXT: pshufb %xmm4, %xmm1
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
-; GFNISSE-NEXT: pshufb %xmm4, %xmm2
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
-; GFNISSE-NEXT: pshufb %xmm4, %xmm3
-; GFNISSE-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
-; GFNISSE-NEXT: retq
+; GFNISSE2-LABEL: test_bitreverse_v8i64:
+; GFNISSE2: # %bb.0:
+; GFNISSE2-NEXT: pxor %xmm4, %xmm4
+; GFNISSE2-NEXT: movdqa %xmm0, %xmm5
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm5 = xmm5[8],xmm4[8],xmm5[9],xmm4[9],xmm5[10],xmm4[10],xmm5[11],xmm4[11],xmm5[12],xmm4[12],xmm5[13],xmm4[13],xmm5[14],xmm4[14],xmm5[15],xmm4[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm5 = xmm5[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm5 = xmm5[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm5, %xmm0
+; GFNISSE2-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
+; GFNISSE2-NEXT: movdqa %xmm1, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1],xmm1[2],xmm4[2],xmm1[3],xmm4[3],xmm1[4],xmm4[4],xmm1[5],xmm4[5],xmm1[6],xmm4[6],xmm1[7],xmm4[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm1
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
+; GFNISSE2-NEXT: movdqa %xmm2, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1],xmm2[2],xmm4[2],xmm2[3],xmm4[3],xmm2[4],xmm4[4],xmm2[5],xmm4[5],xmm2[6],xmm4[6],xmm2[7],xmm4[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm2
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
+; GFNISSE2-NEXT: movdqa %xmm3, %xmm6
+; GFNISSE2-NEXT: punpckhbw {{.*#+}} xmm6 = xmm6[8],xmm4[8],xmm6[9],xmm4[9],xmm6[10],xmm4[10],xmm6[11],xmm4[11],xmm6[12],xmm4[12],xmm6[13],xmm4[13],xmm6[14],xmm4[14],xmm6[15],xmm4[15]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm6[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1],xmm3[2],xmm4[2],xmm3[3],xmm4[3],xmm3[4],xmm4[4],xmm3[5],xmm4[5],xmm3[6],xmm4[6],xmm3[7],xmm4[7]
+; GFNISSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
+; GFNISSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; GFNISSE2-NEXT: pshufhw {{.*#+}} xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; GFNISSE2-NEXT: packuswb %xmm6, %xmm3
+; GFNISSE2-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
+; GFNISSE2-NEXT: retq
+;
+; GFNISSSE3-LABEL: test_bitreverse_v8i64:
+; GFNISSSE3: # %bb.0:
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm4 = [7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm0
+; GFNISSSE3-NEXT: movdqa {{.*#+}} xmm5 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm0
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm1
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm1
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm2
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm2
+; GFNISSSE3-NEXT: pshufb %xmm4, %xmm3
+; GFNISSSE3-NEXT: gf2p8affineqb $0, %xmm5, %xmm3
+; GFNISSSE3-NEXT: retq
;
; GFNIAVX1-LABEL: test_bitreverse_v8i64:
; GFNIAVX1: # %bb.0:
generated by cgit v1.1 at 2025-08-04 02:07:18 +0000