[llvm] [X86] bitreverse.ll - add additional i128/i256/i512 GFNI test coverage for #187502 (PR #187552)
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Thu Mar 19 14:52:15 PDT 2026
https://github.com/RKSimon updated https://github.com/llvm/llvm-project/pull/187552
>From 6651b4ccebde9b6031060f07c198d487435a61b6 Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev at redking.me.uk>
Date: Thu, 19 Mar 2026 18:29:38 +0000
Subject: [PATCH] [X86] bitreverse.ll - add additional i128/i256/i512 GFNI test
coverage for #187502
---
llvm/test/CodeGen/X86/bitreverse.ll | 1193 +++++++++++++++++++++++++++
1 file changed, 1193 insertions(+)
diff --git a/llvm/test/CodeGen/X86/bitreverse.ll b/llvm/test/CodeGen/X86/bitreverse.ll
index d92e1a1e7b9d4..674df2540d38d 100644
--- a/llvm/test/CodeGen/X86/bitreverse.ll
+++ b/llvm/test/CodeGen/X86/bitreverse.ll
@@ -690,6 +690,1199 @@ define <2 x i16> @undef_v2i16() {
ret <2 x i16> %b
}
+; Very wide types that can be performed on the FPU.
+
+define i128 @test_bitreverse_i128(i128 %a) nounwind {
+; X86-LABEL: test_bitreverse_i128:
+; X86: # %bb.0:
+; X86-NEXT: pushl %ebp
+; X86-NEXT: movl %esp, %ebp
+; X86-NEXT: pushl %edi
+; X86-NEXT: pushl %esi
+; X86-NEXT: andl $-16, %esp
+; X86-NEXT: movl 24(%ebp), %eax
+; X86-NEXT: movl 28(%ebp), %edx
+; X86-NEXT: movl 32(%ebp), %esi
+; X86-NEXT: movl 36(%ebp), %ecx
+; X86-NEXT: bswapl %ecx
+; X86-NEXT: movl %ecx, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edi
+; X86-NEXT: shrl $4, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: orl %edi, %ecx
+; X86-NEXT: movl %ecx, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: shrl $2, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: leal (%ecx,%edi,4), %ecx
+; X86-NEXT: movl %ecx, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%edi,2), %ecx
+; X86-NEXT: bswapl %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edi
+; X86-NEXT: shrl $4, %esi
+; X86-NEXT: andl $252645135, %esi # imm = 0xF0F0F0F
+; X86-NEXT: orl %edi, %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: shrl $2, %esi
+; X86-NEXT: andl $858993459, %esi # imm = 0x33333333
+; X86-NEXT: leal (%esi,%edi,4), %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %esi
+; X86-NEXT: andl $1431655765, %esi # imm = 0x55555555
+; X86-NEXT: leal (%esi,%edi,2), %esi
+; X86-NEXT: bswapl %edx
+; X86-NEXT: movl %edx, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edi
+; X86-NEXT: shrl $4, %edx
+; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F
+; X86-NEXT: orl %edi, %edx
+; X86-NEXT: movl %edx, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: shrl $2, %edx
+; X86-NEXT: andl $858993459, %edx # imm = 0x33333333
+; X86-NEXT: leal (%edx,%edi,4), %edx
+; X86-NEXT: movl %edx, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %edx
+; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555
+; X86-NEXT: leal (%edx,%edi,2), %edx
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edi
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %edi, %eax
+; X86-NEXT: movl %eax, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%edi,4), %eax
+; X86-NEXT: movl %eax, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%edi,2), %edi
+; X86-NEXT: movl 8(%ebp), %eax
+; X86-NEXT: movl %edi, 12(%eax)
+; X86-NEXT: movl %edx, 8(%eax)
+; X86-NEXT: movl %esi, 4(%eax)
+; X86-NEXT: movl %ecx, (%eax)
+; X86-NEXT: leal -8(%ebp), %esp
+; X86-NEXT: popl %esi
+; X86-NEXT: popl %edi
+; X86-NEXT: popl %ebp
+; X86-NEXT: retl $4
+;
+; X64-LABEL: test_bitreverse_i128:
+; X64: # %bb.0:
+; X64-NEXT: bswapq %rsi
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: shrq $4, %rax
+; X64-NEXT: movabsq $1085102592571150095, %rdx # imm = 0xF0F0F0F0F0F0F0F
+; X64-NEXT: andq %rdx, %rax
+; X64-NEXT: andq %rdx, %rsi
+; X64-NEXT: shlq $4, %rsi
+; X64-NEXT: orq %rax, %rsi
+; X64-NEXT: movabsq $3689348814741910323, %rcx # imm = 0x3333333333333333
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: andq %rcx, %rax
+; X64-NEXT: shrq $2, %rsi
+; X64-NEXT: andq %rcx, %rsi
+; X64-NEXT: leaq (%rsi,%rax,4), %rax
+; X64-NEXT: movabsq $6148914691236517205, %rsi # imm = 0x5555555555555555
+; X64-NEXT: movq %rax, %r8
+; X64-NEXT: andq %rsi, %r8
+; X64-NEXT: shrq %rax
+; X64-NEXT: andq %rsi, %rax
+; X64-NEXT: leaq (%rax,%r8,2), %rax
+; X64-NEXT: bswapq %rdi
+; X64-NEXT: movq %rdi, %r8
+; X64-NEXT: shrq $4, %r8
+; X64-NEXT: andq %rdx, %r8
+; X64-NEXT: andq %rdx, %rdi
+; X64-NEXT: shlq $4, %rdi
+; X64-NEXT: orq %r8, %rdi
+; X64-NEXT: movq %rdi, %rdx
+; X64-NEXT: andq %rcx, %rdx
+; X64-NEXT: shrq $2, %rdi
+; X64-NEXT: andq %rcx, %rdi
+; X64-NEXT: leaq (%rdi,%rdx,4), %rcx
+; X64-NEXT: movq %rcx, %rdx
+; X64-NEXT: andq %rsi, %rdx
+; X64-NEXT: shrq %rcx
+; X64-NEXT: andq %rsi, %rcx
+; X64-NEXT: leaq (%rcx,%rdx,2), %rdx
+; X64-NEXT: retq
+;
+; X86XOP-LABEL: test_bitreverse_i128:
+; X86XOP: # %bb.0:
+; X86XOP-NEXT: pushl %ebp
+; X86XOP-NEXT: movl %esp, %ebp
+; X86XOP-NEXT: andl $-16, %esp
+; X86XOP-NEXT: subl $16, %esp
+; X86XOP-NEXT: movl 8(%ebp), %eax
+; X86XOP-NEXT: vmovdqa {{.*#+}} xmm0 = [87,86,85,84,83,82,81,80,95,94,93,92,91,90,89,88]
+; X86XOP-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm1, %xmm0, %xmm1
+; X86XOP-NEXT: vmovq {{.*#+}} xmm2 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm2, %xmm0, %xmm0
+; X86XOP-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; X86XOP-NEXT: vmovdqa %xmm0, (%eax)
+; X86XOP-NEXT: movl %ebp, %esp
+; X86XOP-NEXT: popl %ebp
+; X86XOP-NEXT: retl $4
+;
+; X86GFNI-LABEL: test_bitreverse_i128:
+; X86GFNI: # %bb.0:
+; X86GFNI-NEXT: pushl %ebp
+; X86GFNI-NEXT: movl %esp, %ebp
+; X86GFNI-NEXT: pushl %edi
+; X86GFNI-NEXT: pushl %esi
+; X86GFNI-NEXT: andl $-16, %esp
+; X86GFNI-NEXT: movl 8(%ebp), %eax
+; X86GFNI-NEXT: vpbroadcastq {{.*#+}} xmm0 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: vmovd %xmm1, %edx
+; X86GFNI-NEXT: bswapl %edx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm0
+; X86GFNI-NEXT: vpextrd $1, %xmm0, %esi
+; X86GFNI-NEXT: bswapl %esi
+; X86GFNI-NEXT: vmovd %xmm0, %edi
+; X86GFNI-NEXT: bswapl %edi
+; X86GFNI-NEXT: movl %edi, 12(%eax)
+; X86GFNI-NEXT: movl %esi, 8(%eax)
+; X86GFNI-NEXT: movl %edx, 4(%eax)
+; X86GFNI-NEXT: movl %ecx, (%eax)
+; X86GFNI-NEXT: leal -8(%ebp), %esp
+; X86GFNI-NEXT: popl %esi
+; X86GFNI-NEXT: popl %edi
+; X86GFNI-NEXT: popl %ebp
+; X86GFNI-NEXT: retl $4
+;
+; X64GFNI-LABEL: test_bitreverse_i128:
+; X64GFNI: # %bb.0:
+; X64GFNI-NEXT: vmovq %rsi, %xmm0
+; X64GFNI-NEXT: vpbroadcastq {{.*#+}} xmm1 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rax
+; X64GFNI-NEXT: bswapq %rax
+; X64GFNI-NEXT: vmovq %rdi, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rdx
+; X64GFNI-NEXT: bswapq %rdx
+; X64GFNI-NEXT: retq
+ %b = call i128 @llvm.bitreverse.i128(i128 %a)
+ ret i128 %b
+}
+
+define i256 @test_bitreverse_i256(i256 %a) nounwind {
+; X86-LABEL: test_bitreverse_i256:
+; X86: # %bb.0:
+; X86-NEXT: pushl %ebp
+; X86-NEXT: pushl %ebx
+; X86-NEXT: pushl %edi
+; X86-NEXT: pushl %esi
+; X86-NEXT: subl $8, %esp
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: movl {{[0-9]+}}(%esp), %edi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %ebx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT: bswapl %ecx
+; X86-NEXT: movl %ecx, %ebp
+; X86-NEXT: andl $252645135, %ebp # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ebp
+; X86-NEXT: shrl $4, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: orl %ebp, %ecx
+; X86-NEXT: movl %ecx, %ebp
+; X86-NEXT: andl $858993459, %ebp # imm = 0x33333333
+; X86-NEXT: shrl $2, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: leal (%ecx,%ebp,4), %ecx
+; X86-NEXT: movl %ecx, %ebp
+; X86-NEXT: andl $1431655765, %ebp # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%ebp,2), %ecx
+; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %edx
+; X86-NEXT: movl %edx, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %edx
+; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %edx
+; X86-NEXT: movl %edx, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %edx
+; X86-NEXT: andl $858993459, %edx # imm = 0x33333333
+; X86-NEXT: leal (%edx,%ecx,4), %ecx
+; X86-NEXT: movl %ecx, %edx
+; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%edx,2), %ecx
+; X86-NEXT: movl %ecx, (%esp) # 4-byte Spill
+; X86-NEXT: bswapl %esi
+; X86-NEXT: movl %esi, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %esi
+; X86-NEXT: andl $252645135, %esi # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %esi
+; X86-NEXT: movl %esi, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %esi
+; X86-NEXT: andl $858993459, %esi # imm = 0x33333333
+; X86-NEXT: leal (%esi,%ecx,4), %ecx
+; X86-NEXT: movl %ecx, %esi
+; X86-NEXT: andl $1431655765, %esi # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%esi,2), %esi
+; X86-NEXT: bswapl %ebx
+; X86-NEXT: movl %ebx, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %ebx
+; X86-NEXT: andl $252645135, %ebx # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %ebx
+; X86-NEXT: movl %ebx, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %ebx
+; X86-NEXT: andl $858993459, %ebx # imm = 0x33333333
+; X86-NEXT: leal (%ebx,%ecx,4), %ecx
+; X86-NEXT: movl %ecx, %ebx
+; X86-NEXT: andl $1431655765, %ebx # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%ebx,2), %ebx
+; X86-NEXT: bswapl %edi
+; X86-NEXT: movl %edi, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %edi
+; X86-NEXT: movl %edi, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: leal (%edi,%ecx,4), %ecx
+; X86-NEXT: movl %ecx, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%edi,2), %edi
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %ebp
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %edx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %ecx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: movl %ecx, 28(%eax)
+; X86-NEXT: movl %edx, 24(%eax)
+; X86-NEXT: movl %ebp, 20(%eax)
+; X86-NEXT: movl %edi, 16(%eax)
+; X86-NEXT: movl %ebx, 12(%eax)
+; X86-NEXT: movl %esi, 8(%eax)
+; X86-NEXT: movl (%esp), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 4(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, (%eax)
+; X86-NEXT: addl $8, %esp
+; X86-NEXT: popl %esi
+; X86-NEXT: popl %edi
+; X86-NEXT: popl %ebx
+; X86-NEXT: popl %ebp
+; X86-NEXT: retl $4
+;
+; X64-LABEL: test_bitreverse_i256:
+; X64: # %bb.0:
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: bswapq %r8
+; X64-NEXT: movq %r8, %rdi
+; X64-NEXT: shrq $4, %rdi
+; X64-NEXT: movabsq $1085102592571150095, %r9 # imm = 0xF0F0F0F0F0F0F0F
+; X64-NEXT: andq %r9, %rdi
+; X64-NEXT: andq %r9, %r8
+; X64-NEXT: shlq $4, %r8
+; X64-NEXT: orq %rdi, %r8
+; X64-NEXT: movabsq $3689348814741910323, %rdi # imm = 0x3333333333333333
+; X64-NEXT: movq %r8, %r10
+; X64-NEXT: andq %rdi, %r10
+; X64-NEXT: shrq $2, %r8
+; X64-NEXT: andq %rdi, %r8
+; X64-NEXT: leaq (%r8,%r10,4), %r10
+; X64-NEXT: movabsq $6148914691236517205, %r8 # imm = 0x5555555555555555
+; X64-NEXT: movq %r10, %r11
+; X64-NEXT: andq %r8, %r11
+; X64-NEXT: shrq %r10
+; X64-NEXT: andq %r8, %r10
+; X64-NEXT: leaq (%r10,%r11,2), %r10
+; X64-NEXT: bswapq %rcx
+; X64-NEXT: movq %rcx, %r11
+; X64-NEXT: shrq $4, %r11
+; X64-NEXT: andq %r9, %r11
+; X64-NEXT: andq %r9, %rcx
+; X64-NEXT: shlq $4, %rcx
+; X64-NEXT: orq %r11, %rcx
+; X64-NEXT: movq %rcx, %r11
+; X64-NEXT: andq %rdi, %r11
+; X64-NEXT: shrq $2, %rcx
+; X64-NEXT: andq %rdi, %rcx
+; X64-NEXT: leaq (%rcx,%r11,4), %rcx
+; X64-NEXT: movq %rcx, %r11
+; X64-NEXT: andq %r8, %r11
+; X64-NEXT: shrq %rcx
+; X64-NEXT: andq %r8, %rcx
+; X64-NEXT: leaq (%rcx,%r11,2), %rcx
+; X64-NEXT: bswapq %rdx
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: shrq $4, %r11
+; X64-NEXT: andq %r9, %r11
+; X64-NEXT: andq %r9, %rdx
+; X64-NEXT: shlq $4, %rdx
+; X64-NEXT: orq %r11, %rdx
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: andq %rdi, %r11
+; X64-NEXT: shrq $2, %rdx
+; X64-NEXT: andq %rdi, %rdx
+; X64-NEXT: leaq (%rdx,%r11,4), %rdx
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: andq %r8, %r11
+; X64-NEXT: shrq %rdx
+; X64-NEXT: andq %r8, %rdx
+; X64-NEXT: leaq (%rdx,%r11,2), %rdx
+; X64-NEXT: bswapq %rsi
+; X64-NEXT: movq %rsi, %r11
+; X64-NEXT: shrq $4, %r11
+; X64-NEXT: andq %r9, %r11
+; X64-NEXT: andq %r9, %rsi
+; X64-NEXT: shlq $4, %rsi
+; X64-NEXT: orq %r11, %rsi
+; X64-NEXT: movq %rsi, %r9
+; X64-NEXT: andq %rdi, %r9
+; X64-NEXT: shrq $2, %rsi
+; X64-NEXT: andq %rdi, %rsi
+; X64-NEXT: leaq (%rsi,%r9,4), %rsi
+; X64-NEXT: movq %rsi, %rdi
+; X64-NEXT: andq %r8, %rdi
+; X64-NEXT: shrq %rsi
+; X64-NEXT: andq %r8, %rsi
+; X64-NEXT: leaq (%rsi,%rdi,2), %rsi
+; X64-NEXT: movq %rsi, 24(%rax)
+; X64-NEXT: movq %rdx, 16(%rax)
+; X64-NEXT: movq %rcx, 8(%rax)
+; X64-NEXT: movq %r10, (%rax)
+; X64-NEXT: retq
+;
+; X86XOP-LABEL: test_bitreverse_i256:
+; X86XOP: # %bb.0:
+; X86XOP-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86XOP-NEXT: vmovdqa {{.*#+}} xmm0 = [87,86,85,84,83,82,81,80,95,94,93,92,91,90,89,88]
+; X86XOP-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm1, %xmm0, %xmm1
+; X86XOP-NEXT: vmovq {{.*#+}} xmm2 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm2, %xmm0, %xmm2
+; X86XOP-NEXT: vmovq {{.*#+}} xmm3 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm3, %xmm0, %xmm3
+; X86XOP-NEXT: vmovq {{.*#+}} xmm4 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm0, %xmm4, %xmm0, %xmm0
+; X86XOP-NEXT: vinsertf128 $1, %xmm3, %ymm0, %ymm0
+; X86XOP-NEXT: vinsertf128 $1, %xmm1, %ymm2, %ymm1
+; X86XOP-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
+; X86XOP-NEXT: vmovups %ymm0, (%eax)
+; X86XOP-NEXT: vzeroupper
+; X86XOP-NEXT: retl $4
+;
+; X86GFNI-LABEL: test_bitreverse_i256:
+; X86GFNI: # %bb.0:
+; X86GFNI-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86GFNI-NEXT: vpbroadcastq {{.*#+}} xmm0 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 28(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 24(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 20(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 16(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 12(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm0
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 8(%eax)
+; X86GFNI-NEXT: vmovd %xmm0, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 4(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm0, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, (%eax)
+; X86GFNI-NEXT: retl $4
+;
+; X64GFNI-LABEL: test_bitreverse_i256:
+; X64GFNI: # %bb.0:
+; X64GFNI-NEXT: movq %rdi, %rax
+; X64GFNI-NEXT: vmovq %r8, %xmm0
+; X64GFNI-NEXT: vpbroadcastq {{.*#+}} xmm1 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rdi
+; X64GFNI-NEXT: bswapq %rdi
+; X64GFNI-NEXT: vmovq %rcx, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rcx
+; X64GFNI-NEXT: bswapq %rcx
+; X64GFNI-NEXT: vmovq %rdx, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rdx
+; X64GFNI-NEXT: bswapq %rdx
+; X64GFNI-NEXT: vmovq %rsi, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rsi
+; X64GFNI-NEXT: bswapq %rsi
+; X64GFNI-NEXT: movq %rsi, 24(%rax)
+; X64GFNI-NEXT: movq %rdx, 16(%rax)
+; X64GFNI-NEXT: movq %rcx, 8(%rax)
+; X64GFNI-NEXT: movq %rdi, (%rax)
+; X64GFNI-NEXT: retq
+ %b = call i256 @llvm.bitreverse.i256(i256 %a)
+ ret i256 %b
+}
+
+define i512 @test_bitreverse_i512(i512 %a) nounwind {
+; X86-LABEL: test_bitreverse_i512:
+; X86: # %bb.0:
+; X86-NEXT: pushl %ebp
+; X86-NEXT: pushl %ebx
+; X86-NEXT: pushl %edi
+; X86-NEXT: pushl %esi
+; X86-NEXT: subl $40, %esp
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %edi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %ebx
+; X86-NEXT: bswapl %ebx
+; X86-NEXT: movl %ebx, %ebp
+; X86-NEXT: andl $252645135, %ebp # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ebp
+; X86-NEXT: shrl $4, %ebx
+; X86-NEXT: andl $252645135, %ebx # imm = 0xF0F0F0F
+; X86-NEXT: orl %ebp, %ebx
+; X86-NEXT: movl %ebx, %ebp
+; X86-NEXT: andl $858993459, %ebp # imm = 0x33333333
+; X86-NEXT: shrl $2, %ebx
+; X86-NEXT: andl $858993459, %ebx # imm = 0x33333333
+; X86-NEXT: leal (%ebx,%ebp,4), %ebx
+; X86-NEXT: movl %ebx, %ebp
+; X86-NEXT: andl $1431655765, %ebp # imm = 0x55555555
+; X86-NEXT: shrl %ebx
+; X86-NEXT: andl $1431655765, %ebx # imm = 0x55555555
+; X86-NEXT: leal (%ebx,%ebp,2), %ebx
+; X86-NEXT: movl %ebx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %edi
+; X86-NEXT: movl %edi, %ebx
+; X86-NEXT: andl $252645135, %ebx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ebx
+; X86-NEXT: shrl $4, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: orl %ebx, %edi
+; X86-NEXT: movl %edi, %ebx
+; X86-NEXT: andl $858993459, %ebx # imm = 0x33333333
+; X86-NEXT: shrl $2, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: leal (%edi,%ebx,4), %edi
+; X86-NEXT: movl %edi, %ebx
+; X86-NEXT: andl $1431655765, %ebx # imm = 0x55555555
+; X86-NEXT: shrl %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: leal (%edi,%ebx,2), %edi
+; X86-NEXT: movl %edi, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $252645135, %edi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edi
+; X86-NEXT: shrl $4, %esi
+; X86-NEXT: andl $252645135, %esi # imm = 0xF0F0F0F
+; X86-NEXT: orl %edi, %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $858993459, %edi # imm = 0x33333333
+; X86-NEXT: shrl $2, %esi
+; X86-NEXT: andl $858993459, %esi # imm = 0x33333333
+; X86-NEXT: leal (%esi,%edi,4), %esi
+; X86-NEXT: movl %esi, %edi
+; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555
+; X86-NEXT: shrl %esi
+; X86-NEXT: andl $1431655765, %esi # imm = 0x55555555
+; X86-NEXT: leal (%esi,%edi,2), %esi
+; X86-NEXT: movl %esi, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %edx
+; X86-NEXT: movl %edx, %esi
+; X86-NEXT: andl $252645135, %esi # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %esi
+; X86-NEXT: shrl $4, %edx
+; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F
+; X86-NEXT: orl %esi, %edx
+; X86-NEXT: movl %edx, %esi
+; X86-NEXT: andl $858993459, %esi # imm = 0x33333333
+; X86-NEXT: shrl $2, %edx
+; X86-NEXT: andl $858993459, %edx # imm = 0x33333333
+; X86-NEXT: leal (%edx,%esi,4), %edx
+; X86-NEXT: movl %edx, %esi
+; X86-NEXT: andl $1431655765, %esi # imm = 0x55555555
+; X86-NEXT: shrl %edx
+; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555
+; X86-NEXT: leal (%edx,%esi,2), %edx
+; X86-NEXT: movl %edx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %ecx
+; X86-NEXT: movl %ecx, %edx
+; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %edx
+; X86-NEXT: shrl $4, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: orl %edx, %ecx
+; X86-NEXT: movl %ecx, %edx
+; X86-NEXT: andl $858993459, %edx # imm = 0x33333333
+; X86-NEXT: shrl $2, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: leal (%ecx,%edx,4), %ecx
+; X86-NEXT: movl %ecx, %edx
+; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555
+; X86-NEXT: shrl %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: leal (%ecx,%edx,2), %ecx
+; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %eax
+; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %eax
+; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %eax
+; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %eax
+; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %eax
+; X86-NEXT: movl %eax, (%esp) # 4-byte Spill
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %ebp
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %ebx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %edi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %esi
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %edx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: bswapl %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $252645135, %ecx # imm = 0xF0F0F0F
+; X86-NEXT: shll $4, %ecx
+; X86-NEXT: shrl $4, %eax
+; X86-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F
+; X86-NEXT: orl %ecx, %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333
+; X86-NEXT: shrl $2, %eax
+; X86-NEXT: andl $858993459, %eax # imm = 0x33333333
+; X86-NEXT: leal (%eax,%ecx,4), %eax
+; X86-NEXT: movl %eax, %ecx
+; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555
+; X86-NEXT: shrl %eax
+; X86-NEXT: andl $1431655765, %eax # imm = 0x55555555
+; X86-NEXT: leal (%eax,%ecx,2), %ecx
+; X86-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT: movl %ecx, 60(%eax)
+; X86-NEXT: movl %edx, 56(%eax)
+; X86-NEXT: movl %esi, 52(%eax)
+; X86-NEXT: movl %edi, 48(%eax)
+; X86-NEXT: movl %ebx, 44(%eax)
+; X86-NEXT: movl %ebp, 40(%eax)
+; X86-NEXT: movl (%esp), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 36(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 32(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 28(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 24(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 20(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 16(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 12(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 8(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, 4(%eax)
+; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NEXT: movl %ecx, (%eax)
+; X86-NEXT: addl $40, %esp
+; X86-NEXT: popl %esi
+; X86-NEXT: popl %edi
+; X86-NEXT: popl %ebx
+; X86-NEXT: popl %ebp
+; X86-NEXT: retl $4
+;
+; X64-LABEL: test_bitreverse_i512:
+; X64: # %bb.0:
+; X64-NEXT: pushq %r15
+; X64-NEXT: pushq %r14
+; X64-NEXT: pushq %r12
+; X64-NEXT: pushq %rbx
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r15
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi
+; X64-NEXT: bswapq %rdi
+; X64-NEXT: movq %rdi, %r10
+; X64-NEXT: shrq $4, %r10
+; X64-NEXT: movabsq $1085102592571150095, %r11 # imm = 0xF0F0F0F0F0F0F0F
+; X64-NEXT: andq %r11, %r10
+; X64-NEXT: andq %r11, %rdi
+; X64-NEXT: shlq $4, %rdi
+; X64-NEXT: orq %r10, %rdi
+; X64-NEXT: movabsq $3689348814741910323, %r10 # imm = 0x3333333333333333
+; X64-NEXT: movq %rdi, %rbx
+; X64-NEXT: andq %r10, %rbx
+; X64-NEXT: shrq $2, %rdi
+; X64-NEXT: andq %r10, %rdi
+; X64-NEXT: leaq (%rdi,%rbx,4), %rdi
+; X64-NEXT: movabsq $6148914691236517205, %rbx # imm = 0x5555555555555555
+; X64-NEXT: movq %rdi, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %rdi
+; X64-NEXT: andq %rbx, %rdi
+; X64-NEXT: leaq (%rdi,%r12,2), %rdi
+; X64-NEXT: bswapq %r14
+; X64-NEXT: movq %r14, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %r14
+; X64-NEXT: shlq $4, %r14
+; X64-NEXT: orq %r12, %r14
+; X64-NEXT: movq %r14, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %r14
+; X64-NEXT: andq %r10, %r14
+; X64-NEXT: leaq (%r14,%r12,4), %r14
+; X64-NEXT: movq %r14, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %r14
+; X64-NEXT: andq %rbx, %r14
+; X64-NEXT: leaq (%r14,%r12,2), %r14
+; X64-NEXT: bswapq %r15
+; X64-NEXT: movq %r15, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %r15
+; X64-NEXT: shlq $4, %r15
+; X64-NEXT: orq %r12, %r15
+; X64-NEXT: movq %r15, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %r15
+; X64-NEXT: andq %r10, %r15
+; X64-NEXT: leaq (%r15,%r12,4), %r15
+; X64-NEXT: movq %r15, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %r15
+; X64-NEXT: andq %rbx, %r15
+; X64-NEXT: leaq (%r15,%r12,2), %r15
+; X64-NEXT: bswapq %r9
+; X64-NEXT: movq %r9, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %r9
+; X64-NEXT: shlq $4, %r9
+; X64-NEXT: orq %r12, %r9
+; X64-NEXT: movq %r9, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %r9
+; X64-NEXT: andq %r10, %r9
+; X64-NEXT: leaq (%r9,%r12,4), %r9
+; X64-NEXT: movq %r9, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %r9
+; X64-NEXT: andq %rbx, %r9
+; X64-NEXT: leaq (%r9,%r12,2), %r9
+; X64-NEXT: bswapq %r8
+; X64-NEXT: movq %r8, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %r8
+; X64-NEXT: shlq $4, %r8
+; X64-NEXT: orq %r12, %r8
+; X64-NEXT: movq %r8, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %r8
+; X64-NEXT: andq %r10, %r8
+; X64-NEXT: leaq (%r8,%r12,4), %r8
+; X64-NEXT: movq %r8, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %r8
+; X64-NEXT: andq %rbx, %r8
+; X64-NEXT: leaq (%r8,%r12,2), %r8
+; X64-NEXT: bswapq %rcx
+; X64-NEXT: movq %rcx, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %rcx
+; X64-NEXT: shlq $4, %rcx
+; X64-NEXT: orq %r12, %rcx
+; X64-NEXT: movq %rcx, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %rcx
+; X64-NEXT: andq %r10, %rcx
+; X64-NEXT: leaq (%rcx,%r12,4), %rcx
+; X64-NEXT: movq %rcx, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %rcx
+; X64-NEXT: andq %rbx, %rcx
+; X64-NEXT: leaq (%rcx,%r12,2), %rcx
+; X64-NEXT: bswapq %rdx
+; X64-NEXT: movq %rdx, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %rdx
+; X64-NEXT: shlq $4, %rdx
+; X64-NEXT: orq %r12, %rdx
+; X64-NEXT: movq %rdx, %r12
+; X64-NEXT: andq %r10, %r12
+; X64-NEXT: shrq $2, %rdx
+; X64-NEXT: andq %r10, %rdx
+; X64-NEXT: leaq (%rdx,%r12,4), %rdx
+; X64-NEXT: movq %rdx, %r12
+; X64-NEXT: andq %rbx, %r12
+; X64-NEXT: shrq %rdx
+; X64-NEXT: andq %rbx, %rdx
+; X64-NEXT: leaq (%rdx,%r12,2), %rdx
+; X64-NEXT: bswapq %rsi
+; X64-NEXT: movq %rsi, %r12
+; X64-NEXT: shrq $4, %r12
+; X64-NEXT: andq %r11, %r12
+; X64-NEXT: andq %r11, %rsi
+; X64-NEXT: shlq $4, %rsi
+; X64-NEXT: orq %r12, %rsi
+; X64-NEXT: movq %rsi, %r11
+; X64-NEXT: andq %r10, %r11
+; X64-NEXT: shrq $2, %rsi
+; X64-NEXT: andq %r10, %rsi
+; X64-NEXT: leaq (%rsi,%r11,4), %rsi
+; X64-NEXT: movq %rsi, %r10
+; X64-NEXT: andq %rbx, %r10
+; X64-NEXT: shrq %rsi
+; X64-NEXT: andq %rbx, %rsi
+; X64-NEXT: leaq (%rsi,%r10,2), %rsi
+; X64-NEXT: movq %rsi, 56(%rax)
+; X64-NEXT: movq %rdx, 48(%rax)
+; X64-NEXT: movq %rcx, 40(%rax)
+; X64-NEXT: movq %r8, 32(%rax)
+; X64-NEXT: movq %r9, 24(%rax)
+; X64-NEXT: movq %r15, 16(%rax)
+; X64-NEXT: movq %r14, 8(%rax)
+; X64-NEXT: movq %rdi, (%rax)
+; X64-NEXT: popq %rbx
+; X64-NEXT: popq %r12
+; X64-NEXT: popq %r14
+; X64-NEXT: popq %r15
+; X64-NEXT: retq
+;
+; X86XOP-LABEL: test_bitreverse_i512:
+; X86XOP: # %bb.0:
+; X86XOP-NEXT: subl $16, %esp
+; X86XOP-NEXT: vmovdqa {{.*#+}} xmm4 = [87,86,85,84,83,82,81,80,95,94,93,92,91,90,89,88]
+; X86XOP-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm0, %xmm0, %xmm0
+; X86XOP-NEXT: vmovdqu %xmm0, (%esp) # 16-byte Spill
+; X86XOP-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm1, %xmm0, %xmm1
+; X86XOP-NEXT: vmovq {{.*#+}} xmm2 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm2, %xmm0, %xmm2
+; X86XOP-NEXT: vmovq {{.*#+}} xmm3 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm3, %xmm0, %xmm3
+; X86XOP-NEXT: vmovq {{.*#+}} xmm5 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm5, %xmm0, %xmm5
+; X86XOP-NEXT: vmovq {{.*#+}} xmm6 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm6, %xmm0, %xmm6
+; X86XOP-NEXT: vmovq {{.*#+}} xmm7 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm7, %xmm0, %xmm7
+; X86XOP-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
+; X86XOP-NEXT: vpperm %xmm4, %xmm0, %xmm0, %xmm0
+; X86XOP-NEXT: vinsertf128 $1, %xmm7, %ymm0, %ymm0
+; X86XOP-NEXT: vinsertf128 $1, %xmm5, %ymm6, %ymm4
+; X86XOP-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm4[0],ymm0[0],ymm4[2],ymm0[2]
+; X86XOP-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86XOP-NEXT: vmovups %ymm0, 32(%eax)
+; X86XOP-NEXT: vinsertf128 $1, %xmm2, %ymm3, %ymm0
+; X86XOP-NEXT: vinsertf128 $1, (%esp), %ymm1, %ymm1 # 16-byte Folded Reload
+; X86XOP-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
+; X86XOP-NEXT: vmovups %ymm0, (%eax)
+; X86XOP-NEXT: addl $16, %esp
+; X86XOP-NEXT: vzeroupper
+; X86XOP-NEXT: retl $4
+;
+; X86GFNI-LABEL: test_bitreverse_i512:
+; X86GFNI: # %bb.0:
+; X86GFNI-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86GFNI-NEXT: vpbroadcastq {{.*#+}} xmm0 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 60(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 56(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 52(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 48(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 44(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 40(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 36(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 32(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 28(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 24(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 20(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm1
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 16(%eax)
+; X86GFNI-NEXT: vmovd %xmm1, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 12(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm1, %ecx
+; X86GFNI-NEXT: vmovq {{.*#+}} xmm1 = mem[0],zero
+; X86GFNI-NEXT: vgf2p8affineqb $0, %xmm0, %xmm1, %xmm0
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 8(%eax)
+; X86GFNI-NEXT: vmovd %xmm0, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, 4(%eax)
+; X86GFNI-NEXT: vpextrd $1, %xmm0, %ecx
+; X86GFNI-NEXT: bswapl %ecx
+; X86GFNI-NEXT: movl %ecx, (%eax)
+; X86GFNI-NEXT: retl $4
+;
+; X64GFNI-LABEL: test_bitreverse_i512:
+; X64GFNI: # %bb.0:
+; X64GFNI-NEXT: movq %rdi, %rax
+; X64GFNI-NEXT: vmovq %r9, %xmm0
+; X64GFNI-NEXT: vpbroadcastq {{.*#+}} xmm1 = [1,2,4,8,16,32,64,128,1,2,4,8,16,32,64,128]
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rdi
+; X64GFNI-NEXT: bswapq %rdi
+; X64GFNI-NEXT: vmovq %r8, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %r8
+; X64GFNI-NEXT: bswapq %r8
+; X64GFNI-NEXT: vmovq %rcx, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rcx
+; X64GFNI-NEXT: bswapq %rcx
+; X64GFNI-NEXT: vmovq %rdx, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rdx
+; X64GFNI-NEXT: bswapq %rdx
+; X64GFNI-NEXT: vmovq %rsi, %xmm0
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %rsi
+; X64GFNI-NEXT: bswapq %rsi
+; X64GFNI-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %r9
+; X64GFNI-NEXT: bswapq %r9
+; X64GFNI-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %r10
+; X64GFNI-NEXT: bswapq %r10
+; X64GFNI-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
+; X64GFNI-NEXT: vgf2p8affineqb $0, %xmm1, %xmm0, %xmm0
+; X64GFNI-NEXT: vmovq %xmm0, %r11
+; X64GFNI-NEXT: bswapq %r11
+; X64GFNI-NEXT: movq %r11, 16(%rax)
+; X64GFNI-NEXT: movq %r10, 8(%rax)
+; X64GFNI-NEXT: movq %r9, (%rax)
+; X64GFNI-NEXT: movq %rsi, 56(%rax)
+; X64GFNI-NEXT: movq %rdx, 48(%rax)
+; X64GFNI-NEXT: movq %rcx, 40(%rax)
+; X64GFNI-NEXT: movq %r8, 32(%rax)
+; X64GFNI-NEXT: movq %rdi, 24(%rax)
+; X64GFNI-NEXT: retq
+ %b = call i512 @llvm.bitreverse.i512(i512 %a)
+ ret i512 %b
+}
+
; Make sure we don't assert during type legalization promoting a large
; bitreverse due to the need for a large shift that won't fit in the i8 returned
; from getShiftAmountTy.
More information about the llvm-commits
mailing list