| ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py |
| ; RUN: llc < %s -mtriple=i686-unknown | FileCheck %s --check-prefix=X86 |
| ; RUN: llc < %s -mtriple=x86_64-unknown | FileCheck %s --check-prefix=X64 |
| |
| ; These tests just check that the plumbing is in place for @llvm.bitreverse. The |
| ; actual output is massive at the moment as llvm.bitreverse is not yet legal. |
| |
| declare <2 x i16> @llvm.bitreverse.v2i16(<2 x i16>) readnone |
| |
| define <2 x i16> @test_bitreverse_v2i16(<2 x i16> %a) nounwind { |
| ; X86-LABEL: test_bitreverse_v2i16: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movzwl {{[0-9]+}}(%esp), %ecx |
| ; X86-NEXT: movzwl {{[0-9]+}}(%esp), %eax |
| ; X86-NEXT: rolw $8, %ax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $3855, %edx # imm = 0xF0F |
| ; X86-NEXT: shll $4, %edx |
| ; X86-NEXT: andl $61680, %eax # imm = 0xF0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %edx, %eax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $13107, %edx # imm = 0x3333 |
| ; X86-NEXT: andl $52428, %eax # imm = 0xCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%edx,4), %eax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $21845, %edx # imm = 0x5555 |
| ; X86-NEXT: andl $43690, %eax # imm = 0xAAAA |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%edx,2), %eax |
| ; X86-NEXT: rolw $8, %cx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $3855, %edx # imm = 0xF0F |
| ; X86-NEXT: shll $4, %edx |
| ; X86-NEXT: andl $61680, %ecx # imm = 0xF0F0 |
| ; X86-NEXT: shrl $4, %ecx |
| ; X86-NEXT: orl %edx, %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $13107, %edx # imm = 0x3333 |
| ; X86-NEXT: andl $52428, %ecx # imm = 0xCCCC |
| ; X86-NEXT: shrl $2, %ecx |
| ; X86-NEXT: leal (%ecx,%edx,4), %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $21845, %edx # imm = 0x5555 |
| ; X86-NEXT: andl $43690, %ecx # imm = 0xAAAA |
| ; X86-NEXT: shrl %ecx |
| ; X86-NEXT: leal (%ecx,%edx,2), %edx |
| ; X86-NEXT: # kill: def $ax killed $ax killed $eax |
| ; X86-NEXT: # kill: def $dx killed $dx killed $edx |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_v2i16: |
| ; X64: # %bb.0: |
| ; X64-NEXT: pxor %xmm1, %xmm1 |
| ; X64-NEXT: movdqa %xmm0, %xmm2 |
| ; X64-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] |
| ; X64-NEXT: pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1] |
| ; X64-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[3,2,1,0,4,5,6,7] |
| ; X64-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,7,6,5,4] |
| ; X64-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] |
| ; X64-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1] |
| ; X64-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,2,1,0,4,5,6,7] |
| ; X64-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,6,5,4] |
| ; X64-NEXT: packuswb %xmm2, %xmm0 |
| ; X64-NEXT: movdqa {{.*#+}} xmm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15] |
| ; X64-NEXT: movdqa %xmm0, %xmm2 |
| ; X64-NEXT: pand %xmm1, %xmm2 |
| ; X64-NEXT: psllw $4, %xmm2 |
| ; X64-NEXT: movdqa {{.*#+}} xmm3 = [240,240,240,240,240,240,240,240,240,240,240,240,240,240,240,240] |
| ; X64-NEXT: pand %xmm3, %xmm2 |
| ; X64-NEXT: pand %xmm3, %xmm0 |
| ; X64-NEXT: psrlw $4, %xmm0 |
| ; X64-NEXT: pand %xmm1, %xmm0 |
| ; X64-NEXT: por %xmm2, %xmm0 |
| ; X64-NEXT: movdqa {{.*#+}} xmm1 = [51,51,51,51,51,51,51,51,51,51,51,51,51,51,51,51] |
| ; X64-NEXT: pand %xmm0, %xmm1 |
| ; X64-NEXT: psllw $2, %xmm1 |
| ; X64-NEXT: pand {{.*}}(%rip), %xmm1 |
| ; X64-NEXT: pand {{.*}}(%rip), %xmm0 |
| ; X64-NEXT: psrlw $2, %xmm0 |
| ; X64-NEXT: pand {{.*}}(%rip), %xmm0 |
| ; X64-NEXT: por %xmm1, %xmm0 |
| ; X64-NEXT: movdqa {{.*#+}} xmm1 = [85,85,85,85,85,85,85,85,85,85,85,85,85,85,85,85] |
| ; X64-NEXT: pand %xmm0, %xmm1 |
| ; X64-NEXT: paddb %xmm1, %xmm1 |
| ; X64-NEXT: pand {{.*}}(%rip), %xmm0 |
| ; X64-NEXT: psrlw $1, %xmm0 |
| ; X64-NEXT: pand {{.*}}(%rip), %xmm0 |
| ; X64-NEXT: por %xmm1, %xmm0 |
| ; X64-NEXT: psrlq $48, %xmm0 |
| ; X64-NEXT: retq |
| %b = call <2 x i16> @llvm.bitreverse.v2i16(<2 x i16> %a) |
| ret <2 x i16> %b |
| } |
| |
| declare i64 @llvm.bitreverse.i64(i64) readnone |
| |
| define i64 @test_bitreverse_i64(i64 %a) nounwind { |
| ; X86-LABEL: test_bitreverse_i64: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx |
| ; X86-NEXT: movl {{[0-9]+}}(%esp), %eax |
| ; X86-NEXT: bswapl %eax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F |
| ; X86-NEXT: shll $4, %edx |
| ; X86-NEXT: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %edx, %eax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $858993459, %edx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%edx,4), %eax |
| ; X86-NEXT: movl %eax, %edx |
| ; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%edx,2), %eax |
| ; X86-NEXT: bswapl %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $252645135, %edx # imm = 0xF0F0F0F |
| ; X86-NEXT: shll $4, %edx |
| ; X86-NEXT: andl $-252645136, %ecx # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %ecx |
| ; X86-NEXT: orl %edx, %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $858993459, %edx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %ecx # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %ecx |
| ; X86-NEXT: leal (%ecx,%edx,4), %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %ecx # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %ecx |
| ; X86-NEXT: leal (%ecx,%edx,2), %edx |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i64: |
| ; X64: # %bb.0: |
| ; X64-NEXT: bswapq %rdi |
| ; X64-NEXT: movabsq $1085102592571150095, %rax # imm = 0xF0F0F0F0F0F0F0F |
| ; X64-NEXT: andq %rdi, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: movabsq $-1085102592571150096, %rcx # imm = 0xF0F0F0F0F0F0F0F0 |
| ; X64-NEXT: andq %rdi, %rcx |
| ; X64-NEXT: shrq $4, %rcx |
| ; X64-NEXT: orq %rax, %rcx |
| ; X64-NEXT: movabsq $3689348814741910323, %rax # imm = 0x3333333333333333 |
| ; X64-NEXT: andq %rcx, %rax |
| ; X64-NEXT: movabsq $-3689348814741910324, %rdx # imm = 0xCCCCCCCCCCCCCCCC |
| ; X64-NEXT: andq %rcx, %rdx |
| ; X64-NEXT: shrq $2, %rdx |
| ; X64-NEXT: leaq (%rdx,%rax,4), %rax |
| ; X64-NEXT: movabsq $6148914691236517205, %rcx # imm = 0x5555555555555555 |
| ; X64-NEXT: andq %rax, %rcx |
| ; X64-NEXT: movabsq $-6148914691236517206, %rdx # imm = 0xAAAAAAAAAAAAAAAA |
| ; X64-NEXT: andq %rax, %rdx |
| ; X64-NEXT: shrq %rdx |
| ; X64-NEXT: leaq (%rdx,%rcx,2), %rax |
| ; X64-NEXT: retq |
| %b = call i64 @llvm.bitreverse.i64(i64 %a) |
| ret i64 %b |
| } |
| |
| declare i32 @llvm.bitreverse.i32(i32) readnone |
| |
| define i32 @test_bitreverse_i32(i32 %a) nounwind { |
| ; X86-LABEL: test_bitreverse_i32: |
| ; X86: # %bb.0: |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%ecx,2), %eax |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i32: |
| ; X64: # %bb.0: |
| ; X64-NEXT: # kill: def $edi killed $edi def $rdi |
| ; X64-NEXT: bswapl %edi |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F |
| ; X64-NEXT: shll $4, %eax |
| ; X64-NEXT: andl $-252645136, %edi # imm = 0xF0F0F0F0 |
| ; X64-NEXT: shrl $4, %edi |
| ; X64-NEXT: orl %eax, %edi |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $858993459, %eax # imm = 0x33333333 |
| ; X64-NEXT: andl $-858993460, %edi # imm = 0xCCCCCCCC |
| ; X64-NEXT: shrl $2, %edi |
| ; X64-NEXT: leal (%rdi,%rax,4), %eax |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X64-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X64-NEXT: shrl %eax |
| ; X64-NEXT: leal (%rax,%rcx,2), %eax |
| ; X64-NEXT: retq |
| %b = call i32 @llvm.bitreverse.i32(i32 %a) |
| ret i32 %b |
| } |
| |
| declare i24 @llvm.bitreverse.i24(i24) readnone |
| |
| define i24 @test_bitreverse_i24(i24 %a) nounwind { |
| ; X86-LABEL: test_bitreverse_i24: |
| ; X86: # %bb.0: |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655680, %ecx # imm = 0x55555500 |
| ; X86-NEXT: andl $-1431655936, %eax # imm = 0xAAAAAA00 |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%ecx,2), %eax |
| ; X86-NEXT: shrl $8, %eax |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i24: |
| ; X64: # %bb.0: |
| ; X64-NEXT: # kill: def $edi killed $edi def $rdi |
| ; X64-NEXT: bswapl %edi |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $252645135, %eax # imm = 0xF0F0F0F |
| ; X64-NEXT: shll $4, %eax |
| ; X64-NEXT: andl $-252645136, %edi # imm = 0xF0F0F0F0 |
| ; X64-NEXT: shrl $4, %edi |
| ; X64-NEXT: orl %eax, %edi |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $858993459, %eax # imm = 0x33333333 |
| ; X64-NEXT: andl $-858993460, %edi # imm = 0xCCCCCCCC |
| ; X64-NEXT: shrl $2, %edi |
| ; X64-NEXT: leal (%rdi,%rax,4), %eax |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andl $1431655680, %ecx # imm = 0x55555500 |
| ; X64-NEXT: andl $-1431655936, %eax # imm = 0xAAAAAA00 |
| ; X64-NEXT: shrl %eax |
| ; X64-NEXT: leal (%rax,%rcx,2), %eax |
| ; X64-NEXT: shrl $8, %eax |
| ; X64-NEXT: retq |
| %b = call i24 @llvm.bitreverse.i24(i24 %a) |
| ret i24 %b |
| } |
| |
| declare i16 @llvm.bitreverse.i16(i16) readnone |
| |
| define i16 @test_bitreverse_i16(i16 %a) nounwind { |
| ; X86-LABEL: test_bitreverse_i16: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movzwl {{[0-9]+}}(%esp), %eax |
| ; X86-NEXT: rolw $8, %ax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $3855, %ecx # imm = 0xF0F |
| ; X86-NEXT: shll $4, %ecx |
| ; X86-NEXT: andl $61680, %eax # imm = 0xF0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $13107, %ecx # imm = 0x3333 |
| ; X86-NEXT: andl $52428, %eax # imm = 0xCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $21845, %ecx # imm = 0x5555 |
| ; X86-NEXT: andl $43690, %eax # imm = 0xAAAA |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%ecx,2), %eax |
| ; X86-NEXT: # kill: def $ax killed $ax killed $eax |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i16: |
| ; X64: # %bb.0: |
| ; X64-NEXT: # kill: def $edi killed $edi def $rdi |
| ; X64-NEXT: rolw $8, %di |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $3855, %eax # imm = 0xF0F |
| ; X64-NEXT: shll $4, %eax |
| ; X64-NEXT: andl $61680, %edi # imm = 0xF0F0 |
| ; X64-NEXT: shrl $4, %edi |
| ; X64-NEXT: orl %eax, %edi |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: andl $13107, %eax # imm = 0x3333 |
| ; X64-NEXT: andl $52428, %edi # imm = 0xCCCC |
| ; X64-NEXT: shrl $2, %edi |
| ; X64-NEXT: leal (%rdi,%rax,4), %eax |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andl $21845, %ecx # imm = 0x5555 |
| ; X64-NEXT: andl $43690, %eax # imm = 0xAAAA |
| ; X64-NEXT: shrl %eax |
| ; X64-NEXT: leal (%rax,%rcx,2), %eax |
| ; X64-NEXT: # kill: def $ax killed $ax killed $eax |
| ; X64-NEXT: retq |
| %b = call i16 @llvm.bitreverse.i16(i16 %a) |
| ret i16 %b |
| } |
| |
| declare i8 @llvm.bitreverse.i8(i8) readnone |
| |
| define i8 @test_bitreverse_i8(i8 %a) { |
| ; X86-LABEL: test_bitreverse_i8: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movb {{[0-9]+}}(%esp), %al |
| ; X86-NEXT: rolb $4, %al |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andb $51, %cl |
| ; X86-NEXT: shlb $2, %cl |
| ; X86-NEXT: andb $-52, %al |
| ; X86-NEXT: shrb $2, %al |
| ; X86-NEXT: orb %cl, %al |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andb $85, %cl |
| ; X86-NEXT: addb %cl, %cl |
| ; X86-NEXT: andb $-86, %al |
| ; X86-NEXT: shrb %al |
| ; X86-NEXT: orb %cl, %al |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i8: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: rolb $4, %al |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andb $51, %cl |
| ; X64-NEXT: shlb $2, %cl |
| ; X64-NEXT: andb $-52, %al |
| ; X64-NEXT: shrb $2, %al |
| ; X64-NEXT: orb %cl, %al |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andb $85, %cl |
| ; X64-NEXT: addb %cl, %cl |
| ; X64-NEXT: andb $-86, %al |
| ; X64-NEXT: shrb %al |
| ; X64-NEXT: orb %cl, %al |
| ; X64-NEXT: # kill: def $al killed $al killed $eax |
| ; X64-NEXT: retq |
| %b = call i8 @llvm.bitreverse.i8(i8 %a) |
| ret i8 %b |
| } |
| |
| declare i4 @llvm.bitreverse.i4(i4) readnone |
| |
| define i4 @test_bitreverse_i4(i4 %a) { |
| ; X86-LABEL: test_bitreverse_i4: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movb {{[0-9]+}}(%esp), %al |
| ; X86-NEXT: rolb $4, %al |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andb $51, %cl |
| ; X86-NEXT: shlb $2, %cl |
| ; X86-NEXT: andb $-52, %al |
| ; X86-NEXT: shrb $2, %al |
| ; X86-NEXT: orb %cl, %al |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andb $80, %cl |
| ; X86-NEXT: addb %cl, %cl |
| ; X86-NEXT: andb $-96, %al |
| ; X86-NEXT: shrb %al |
| ; X86-NEXT: orb %cl, %al |
| ; X86-NEXT: shrb $4, %al |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: test_bitreverse_i4: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: rolb $4, %al |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andb $51, %cl |
| ; X64-NEXT: shlb $2, %cl |
| ; X64-NEXT: andb $-52, %al |
| ; X64-NEXT: shrb $2, %al |
| ; X64-NEXT: orb %cl, %al |
| ; X64-NEXT: movl %eax, %ecx |
| ; X64-NEXT: andb $80, %cl |
| ; X64-NEXT: addb %cl, %cl |
| ; X64-NEXT: andb $-96, %al |
| ; X64-NEXT: shrb %al |
| ; X64-NEXT: orb %cl, %al |
| ; X64-NEXT: shrb $4, %al |
| ; X64-NEXT: # kill: def $al killed $al killed $eax |
| ; X64-NEXT: retq |
| %b = call i4 @llvm.bitreverse.i4(i4 %a) |
| ret i4 %b |
| } |
| |
| ; These tests check that bitreverse(constant) calls are folded |
| |
| define <2 x i16> @fold_v2i16() { |
| ; X86-LABEL: fold_v2i16: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movw $-4096, %ax # imm = 0xF000 |
| ; X86-NEXT: movw $240, %dx |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: fold_v2i16: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movaps {{.*#+}} xmm0 = [61440,240] |
| ; X64-NEXT: retq |
| %b = call <2 x i16> @llvm.bitreverse.v2i16(<2 x i16> <i16 15, i16 3840>) |
| ret <2 x i16> %b |
| } |
| |
| define i24 @fold_i24() { |
| ; X86-LABEL: fold_i24: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movl $2048, %eax # imm = 0x800 |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: fold_i24: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movl $2048, %eax # imm = 0x800 |
| ; X64-NEXT: retq |
| %b = call i24 @llvm.bitreverse.i24(i24 4096) |
| ret i24 %b |
| } |
| |
| define i8 @fold_i8() { |
| ; X86-LABEL: fold_i8: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movb $-16, %al |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: fold_i8: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movb $-16, %al |
| ; X64-NEXT: retq |
| %b = call i8 @llvm.bitreverse.i8(i8 15) |
| ret i8 %b |
| } |
| |
| define i4 @fold_i4() { |
| ; X86-LABEL: fold_i4: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movb $1, %al |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: fold_i4: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movb $1, %al |
| ; X64-NEXT: retq |
| %b = call i4 @llvm.bitreverse.i4(i4 8) |
| ret i4 %b |
| } |
| |
| ; These tests check that bitreverse(bitreverse()) calls are removed |
| |
| define i8 @identity_i8(i8 %a) { |
| ; X86-LABEL: identity_i8: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movb {{[0-9]+}}(%esp), %al |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: identity_i8: |
| ; X64: # %bb.0: |
| ; X64-NEXT: movl %edi, %eax |
| ; X64-NEXT: # kill: def $al killed $al killed $eax |
| ; X64-NEXT: retq |
| %b = call i8 @llvm.bitreverse.i8(i8 %a) |
| %c = call i8 @llvm.bitreverse.i8(i8 %b) |
| ret i8 %c |
| } |
| |
| define <2 x i16> @identity_v2i16(<2 x i16> %a) { |
| ; X86-LABEL: identity_v2i16: |
| ; X86: # %bb.0: |
| ; X86-NEXT: movzwl {{[0-9]+}}(%esp), %eax |
| ; X86-NEXT: movzwl {{[0-9]+}}(%esp), %edx |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: identity_v2i16: |
| ; X64: # %bb.0: |
| ; X64-NEXT: retq |
| %b = call <2 x i16> @llvm.bitreverse.v2i16(<2 x i16> %a) |
| %c = call <2 x i16> @llvm.bitreverse.v2i16(<2 x i16> %b) |
| ret <2 x i16> %c |
| } |
| |
| ; These tests check that bitreverse(undef) calls are removed |
| |
| define i8 @undef_i8() { |
| ; X86-LABEL: undef_i8: |
| ; X86: # %bb.0: |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: undef_i8: |
| ; X64: # %bb.0: |
| ; X64-NEXT: retq |
| %b = call i8 @llvm.bitreverse.i8(i8 undef) |
| ret i8 %b |
| } |
| |
| define <2 x i16> @undef_v2i16() { |
| ; X86-LABEL: undef_v2i16: |
| ; X86: # %bb.0: |
| ; X86-NEXT: retl |
| ; |
| ; X64-LABEL: undef_v2i16: |
| ; X64: # %bb.0: |
| ; X64-NEXT: retq |
| %b = call <2 x i16> @llvm.bitreverse.v2i16(<2 x i16> undef) |
| ret <2 x i16> %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. |
| define i528 @large_promotion(i528 %A) nounwind { |
| ; X86-LABEL: large_promotion: |
| ; X86: # %bb.0: |
| ; X86-NEXT: pushl %ebp |
| ; X86-NEXT: pushl %ebx |
| ; X86-NEXT: pushl %edi |
| ; X86-NEXT: pushl %esi |
| ; X86-NEXT: subl $56, %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: andl $-252645136, %ebx # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %ebx |
| ; X86-NEXT: orl %ebp, %ebx |
| ; X86-NEXT: movl %ebx, %ebp |
| ; X86-NEXT: andl $858993459, %ebp # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %ebx # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %ebx |
| ; X86-NEXT: leal (%ebx,%ebp,4), %ebx |
| ; X86-NEXT: movl %ebx, %ebp |
| ; X86-NEXT: andl $1431633920, %ebp # imm = 0x55550000 |
| ; X86-NEXT: andl $-1431699456, %ebx # imm = 0xAAAA0000 |
| ; X86-NEXT: shrl %ebx |
| ; X86-NEXT: leal (%ebx,%ebp,2), %ebx |
| ; X86-NEXT: movl %ebx, (%esp) # 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: andl $-252645136, %edi # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %edi |
| ; X86-NEXT: orl %ebx, %edi |
| ; X86-NEXT: movl %edi, %ebx |
| ; X86-NEXT: andl $858993459, %ebx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %edi # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %edi |
| ; X86-NEXT: leal (%edi,%ebx,4), %edi |
| ; X86-NEXT: movl %edi, %ebx |
| ; X86-NEXT: andl $1431655765, %ebx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %edi # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %edi |
| ; 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: andl $-252645136, %esi # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %esi |
| ; X86-NEXT: orl %edi, %esi |
| ; X86-NEXT: movl %esi, %edi |
| ; X86-NEXT: andl $858993459, %edi # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %esi # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %esi |
| ; X86-NEXT: leal (%esi,%edi,4), %esi |
| ; X86-NEXT: movl %esi, %edi |
| ; X86-NEXT: andl $1431655765, %edi # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %esi # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %esi |
| ; X86-NEXT: leal (%esi,%edi,2), %ebx |
| ; X86-NEXT: bswapl %edx |
| ; X86-NEXT: movl %edx, %esi |
| ; X86-NEXT: andl $252645135, %esi # imm = 0xF0F0F0F |
| ; X86-NEXT: shll $4, %esi |
| ; X86-NEXT: andl $-252645136, %edx # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %edx |
| ; X86-NEXT: orl %esi, %edx |
| ; X86-NEXT: movl %edx, %esi |
| ; X86-NEXT: andl $858993459, %esi # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %edx # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %edx |
| ; X86-NEXT: leal (%edx,%esi,4), %edx |
| ; X86-NEXT: movl %edx, %esi |
| ; X86-NEXT: andl $1431655765, %esi # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %edx # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %edx |
| ; 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: andl $-252645136, %ecx # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %ecx |
| ; X86-NEXT: orl %edx, %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $858993459, %edx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %ecx # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %ecx |
| ; X86-NEXT: leal (%ecx,%edx,4), %ecx |
| ; X86-NEXT: movl %ecx, %edx |
| ; X86-NEXT: andl $1431655765, %edx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %ecx # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %ecx |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; 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: andl $-252645136, %eax # imm = 0xF0F0F0F0 |
| ; X86-NEXT: shrl $4, %eax |
| ; X86-NEXT: orl %ecx, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $858993459, %ecx # imm = 0x33333333 |
| ; X86-NEXT: andl $-858993460, %eax # imm = 0xCCCCCCCC |
| ; X86-NEXT: shrl $2, %eax |
| ; X86-NEXT: leal (%eax,%ecx,4), %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: andl $1431655765, %ecx # imm = 0x55555555 |
| ; X86-NEXT: andl $-1431655766, %eax # imm = 0xAAAAAAAA |
| ; X86-NEXT: shrl %eax |
| ; X86-NEXT: leal (%eax,%ecx,2), %edx |
| ; X86-NEXT: movl (%esp), %esi # 4-byte Reload |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %esi |
| ; X86-NEXT: shrdl $16, %ebx, %eax |
| ; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ecx, %ebx |
| ; X86-NEXT: movl %ebx, (%esp) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %ecx |
| ; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ecx, %eax |
| ; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %ecx |
| ; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ecx, %eax |
| ; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %ecx |
| ; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ecx, %eax |
| ; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %ecx |
| ; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ecx, %eax |
| ; X86-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ebp # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ebp, %ecx |
| ; X86-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ebx # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %ebx, %ebp |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload |
| ; X86-NEXT: shrdl $16, %eax, %ebx |
| ; X86-NEXT: shrdl $16, %edi, %eax |
| ; X86-NEXT: movl %eax, %ecx |
| ; X86-NEXT: shrdl $16, %edx, %edi |
| ; X86-NEXT: movl {{[0-9]+}}(%esp), %eax |
| ; X86-NEXT: movl %edi, 60(%eax) |
| ; X86-NEXT: movl %ecx, 56(%eax) |
| ; X86-NEXT: movl %ebx, 52(%eax) |
| ; X86-NEXT: movl %ebp, 48(%eax) |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: movl %ecx, 44(%eax) |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload |
| ; X86-NEXT: movl %ecx, 40(%eax) |
| ; X86-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %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 (%esp), %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 %esi, (%eax) |
| ; X86-NEXT: shrl $16, %edx |
| ; X86-NEXT: movw %dx, 64(%eax) |
| ; X86-NEXT: addl $56, %esp |
| ; X86-NEXT: popl %esi |
| ; X86-NEXT: popl %edi |
| ; X86-NEXT: popl %ebx |
| ; X86-NEXT: popl %ebp |
| ; X86-NEXT: retl $4 |
| ; |
| ; X64-LABEL: large_promotion: |
| ; X64: # %bb.0: |
| ; X64-NEXT: pushq %rbp |
| ; X64-NEXT: pushq %r15 |
| ; X64-NEXT: pushq %r14 |
| ; X64-NEXT: pushq %r13 |
| ; X64-NEXT: pushq %r12 |
| ; X64-NEXT: pushq %rbx |
| ; X64-NEXT: movq %rdi, %r12 |
| ; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp |
| ; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbx |
| ; X64-NEXT: bswapq %rbx |
| ; X64-NEXT: movabsq $1085102592571150095, %r13 # imm = 0xF0F0F0F0F0F0F0F |
| ; X64-NEXT: movq %rbx, %r10 |
| ; X64-NEXT: andq %r13, %r10 |
| ; X64-NEXT: shlq $4, %r10 |
| ; X64-NEXT: movabsq $-1085102592571150096, %rax # imm = 0xF0F0F0F0F0F0F0F0 |
| ; X64-NEXT: andq %rax, %rbx |
| ; X64-NEXT: shrq $4, %rbx |
| ; X64-NEXT: orq %r10, %rbx |
| ; X64-NEXT: movabsq $3689348814741910323, %r11 # imm = 0x3333333333333333 |
| ; X64-NEXT: movq %rbx, %r10 |
| ; X64-NEXT: andq %r11, %r10 |
| ; X64-NEXT: movabsq $-3689348814741910324, %r14 # imm = 0xCCCCCCCCCCCCCCCC |
| ; X64-NEXT: andq %r14, %rbx |
| ; X64-NEXT: shrq $2, %rbx |
| ; X64-NEXT: leaq (%rbx,%r10,4), %r10 |
| ; X64-NEXT: movabsq $6148820866244280320, %rbx # imm = 0x5555000000000000 |
| ; X64-NEXT: andq %r10, %rbx |
| ; X64-NEXT: movabsq $-6149102341220990976, %rdi # imm = 0xAAAA000000000000 |
| ; X64-NEXT: andq %r10, %rdi |
| ; X64-NEXT: shrq %rdi |
| ; X64-NEXT: leaq (%rdi,%rbx,2), %rdi |
| ; X64-NEXT: movq %rdi, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill |
| ; X64-NEXT: bswapq %rbp |
| ; X64-NEXT: movq %rbp, %rdi |
| ; X64-NEXT: andq %r13, %rdi |
| ; X64-NEXT: shlq $4, %rdi |
| ; X64-NEXT: andq %rax, %rbp |
| ; X64-NEXT: shrq $4, %rbp |
| ; X64-NEXT: orq %rdi, %rbp |
| ; X64-NEXT: movq %rbp, %rdi |
| ; X64-NEXT: andq %r11, %rdi |
| ; X64-NEXT: andq %r14, %rbp |
| ; X64-NEXT: shrq $2, %rbp |
| ; X64-NEXT: leaq (%rbp,%rdi,4), %rbp |
| ; X64-NEXT: movabsq $6148914691236517205, %rbx # imm = 0x5555555555555555 |
| ; X64-NEXT: movq %rbp, %r10 |
| ; X64-NEXT: andq %rbx, %r10 |
| ; X64-NEXT: movabsq $-6148914691236517206, %rdi # imm = 0xAAAAAAAAAAAAAAAA |
| ; X64-NEXT: andq %rdi, %rbp |
| ; X64-NEXT: shrq %rbp |
| ; X64-NEXT: leaq (%rbp,%r10,2), %rbp |
| ; X64-NEXT: movq %rbp, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill |
| ; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp |
| ; X64-NEXT: bswapq %rbp |
| ; X64-NEXT: movq %rbp, %r10 |
| ; X64-NEXT: andq %r13, %r10 |
| ; X64-NEXT: shlq $4, %r10 |
| ; X64-NEXT: andq %rax, %rbp |
| ; X64-NEXT: movq %rax, %r15 |
| ; X64-NEXT: shrq $4, %rbp |
| ; X64-NEXT: orq %r10, %rbp |
| ; X64-NEXT: movq %rbp, %r10 |
| ; X64-NEXT: andq %r11, %r10 |
| ; X64-NEXT: andq %r14, %rbp |
| ; X64-NEXT: shrq $2, %rbp |
| ; X64-NEXT: leaq (%rbp,%r10,4), %rbp |
| ; X64-NEXT: movq %rbp, %r10 |
| ; X64-NEXT: andq %rbx, %r10 |
| ; X64-NEXT: andq %rdi, %rbp |
| ; X64-NEXT: shrq %rbp |
| ; X64-NEXT: leaq (%rbp,%r10,2), %rbp |
| ; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 |
| ; X64-NEXT: bswapq %r10 |
| ; X64-NEXT: movq %r10, %rax |
| ; X64-NEXT: andq %r13, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: movq %r15, %rdi |
| ; X64-NEXT: andq %r15, %r10 |
| ; X64-NEXT: shrq $4, %r10 |
| ; X64-NEXT: orq %rax, %r10 |
| ; X64-NEXT: movq %r10, %rax |
| ; X64-NEXT: andq %r11, %rax |
| ; X64-NEXT: andq %r14, %r10 |
| ; X64-NEXT: shrq $2, %r10 |
| ; X64-NEXT: leaq (%r10,%rax,4), %rax |
| ; X64-NEXT: movq %rax, %r10 |
| ; X64-NEXT: andq %rbx, %r10 |
| ; X64-NEXT: movabsq $-6148914691236517206, %r15 # imm = 0xAAAAAAAAAAAAAAAA |
| ; X64-NEXT: andq %r15, %rax |
| ; X64-NEXT: shrq %rax |
| ; X64-NEXT: leaq (%rax,%r10,2), %r10 |
| ; X64-NEXT: bswapq %r9 |
| ; X64-NEXT: movq %r9, %rax |
| ; X64-NEXT: andq %r13, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: andq %rdi, %r9 |
| ; X64-NEXT: shrq $4, %r9 |
| ; X64-NEXT: orq %rax, %r9 |
| ; X64-NEXT: movq %r9, %rax |
| ; X64-NEXT: andq %r11, %rax |
| ; X64-NEXT: andq %r14, %r9 |
| ; X64-NEXT: shrq $2, %r9 |
| ; X64-NEXT: leaq (%r9,%rax,4), %rax |
| ; X64-NEXT: movq %rax, %r9 |
| ; X64-NEXT: andq %rbx, %r9 |
| ; X64-NEXT: andq %r15, %rax |
| ; X64-NEXT: shrq %rax |
| ; X64-NEXT: leaq (%rax,%r9,2), %r9 |
| ; X64-NEXT: bswapq %r8 |
| ; X64-NEXT: movq %r8, %rax |
| ; X64-NEXT: andq %r13, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: andq %rdi, %r8 |
| ; X64-NEXT: shrq $4, %r8 |
| ; X64-NEXT: orq %rax, %r8 |
| ; X64-NEXT: movq %r8, %rax |
| ; X64-NEXT: andq %r11, %rax |
| ; X64-NEXT: andq %r14, %r8 |
| ; X64-NEXT: shrq $2, %r8 |
| ; X64-NEXT: leaq (%r8,%rax,4), %rax |
| ; X64-NEXT: movq %rax, %r8 |
| ; X64-NEXT: andq %rbx, %r8 |
| ; X64-NEXT: andq %r15, %rax |
| ; X64-NEXT: shrq %rax |
| ; X64-NEXT: leaq (%rax,%r8,2), %r8 |
| ; X64-NEXT: bswapq %rcx |
| ; X64-NEXT: movq %rcx, %rax |
| ; X64-NEXT: andq %r13, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: andq %rdi, %rcx |
| ; X64-NEXT: shrq $4, %rcx |
| ; X64-NEXT: orq %rax, %rcx |
| ; X64-NEXT: movq %rcx, %rax |
| ; X64-NEXT: andq %r11, %rax |
| ; X64-NEXT: andq %r14, %rcx |
| ; X64-NEXT: shrq $2, %rcx |
| ; X64-NEXT: leaq (%rcx,%rax,4), %rax |
| ; X64-NEXT: movq %rax, %rcx |
| ; X64-NEXT: andq %rbx, %rcx |
| ; X64-NEXT: andq %r15, %rax |
| ; X64-NEXT: shrq %rax |
| ; X64-NEXT: leaq (%rax,%rcx,2), %rcx |
| ; X64-NEXT: bswapq %rdx |
| ; X64-NEXT: movq %rdx, %rax |
| ; X64-NEXT: andq %r13, %rax |
| ; X64-NEXT: shlq $4, %rax |
| ; X64-NEXT: andq %rdi, %rdx |
| ; X64-NEXT: shrq $4, %rdx |
| ; X64-NEXT: orq %rax, %rdx |
| ; X64-NEXT: movq %rdx, %rax |
| ; X64-NEXT: andq %r11, %rax |
| ; X64-NEXT: andq %r14, %rdx |
| ; X64-NEXT: shrq $2, %rdx |
| ; X64-NEXT: leaq (%rdx,%rax,4), %rax |
| ; X64-NEXT: movq %rax, %rdx |
| ; X64-NEXT: andq %rbx, %rdx |
| ; X64-NEXT: andq %r15, %rax |
| ; X64-NEXT: shrq %rax |
| ; X64-NEXT: leaq (%rax,%rdx,2), %rax |
| ; X64-NEXT: bswapq %rsi |
| ; X64-NEXT: andq %rsi, %r13 |
| ; X64-NEXT: andq %rdi, %rsi |
| ; X64-NEXT: shlq $4, %r13 |
| ; X64-NEXT: shrq $4, %rsi |
| ; X64-NEXT: orq %r13, %rsi |
| ; X64-NEXT: andq %rsi, %r11 |
| ; X64-NEXT: andq %r14, %rsi |
| ; X64-NEXT: shrq $2, %rsi |
| ; X64-NEXT: leaq (%rsi,%r11,4), %rdx |
| ; X64-NEXT: andq %rdx, %rbx |
| ; X64-NEXT: andq %r15, %rdx |
| ; X64-NEXT: shrq %rdx |
| ; X64-NEXT: leaq (%rdx,%rbx,2), %rdx |
| ; X64-NEXT: movq {{[-0-9]+}}(%r{{[sb]}}p), %rsi # 8-byte Reload |
| ; X64-NEXT: movq {{[-0-9]+}}(%r{{[sb]}}p), %rdi # 8-byte Reload |
| ; X64-NEXT: shrdq $48, %rdi, %rsi |
| ; X64-NEXT: shrdq $48, %rbp, %rdi |
| ; X64-NEXT: shrdq $48, %r10, %rbp |
| ; X64-NEXT: shrdq $48, %r9, %r10 |
| ; X64-NEXT: shrdq $48, %r8, %r9 |
| ; X64-NEXT: shrdq $48, %rcx, %r8 |
| ; X64-NEXT: shrdq $48, %rax, %rcx |
| ; X64-NEXT: shrdq $48, %rdx, %rax |
| ; X64-NEXT: movq %rax, 56(%r12) |
| ; X64-NEXT: movq %rcx, 48(%r12) |
| ; X64-NEXT: movq %r8, 40(%r12) |
| ; X64-NEXT: movq %r9, 32(%r12) |
| ; X64-NEXT: movq %r10, 24(%r12) |
| ; X64-NEXT: movq %rbp, 16(%r12) |
| ; X64-NEXT: movq %rdi, 8(%r12) |
| ; X64-NEXT: movq %rsi, (%r12) |
| ; X64-NEXT: shrq $48, %rdx |
| ; X64-NEXT: movw %dx, 64(%r12) |
| ; X64-NEXT: movq %r12, %rax |
| ; X64-NEXT: popq %rbx |
| ; X64-NEXT: popq %r12 |
| ; X64-NEXT: popq %r13 |
| ; X64-NEXT: popq %r14 |
| ; X64-NEXT: popq %r15 |
| ; X64-NEXT: popq %rbp |
| ; X64-NEXT: retq |
| %Z = call i528 @llvm.bitreverse.i528(i528 %A) |
| ret i528 %Z |
| } |
| declare i528 @llvm.bitreverse.i528(i528) |