diff options
Diffstat (limited to 'llvm/test/CodeGen/X86')
4 files changed, 985 insertions, 1 deletions
diff --git a/llvm/test/CodeGen/X86/avx512vlvp2intersect-intrinsics.ll b/llvm/test/CodeGen/X86/avx512vlvp2intersect-intrinsics.ll new file mode 100644 index 00000000000..3dc76ec0692 --- /dev/null +++ b/llvm/test/CodeGen/X86/avx512vlvp2intersect-intrinsics.ll @@ -0,0 +1,593 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+avx512vp2intersect,+avx512vl --show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X86 +; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vp2intersect,+avx512vl --show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X64 + +define void @test_mm256_2intersect_epi32(<4 x i64> %a, <4 x i64> %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi32: +; X86: # %bb.0: # %entry +; X86-NEXT: movl 4(%esp), %eax # encoding: [0x8b,0x44,0x24,0x04] +; X86-NEXT: vp2intersectd %ymm1, %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x28,0x68,0xc1] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 8(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi32: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectd %ymm1, %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x28,0x68,0xc1] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X64-NEXT: movb %cl, (%rdi) # encoding: [0x88,0x0f] +; X64-NEXT: movb %al, (%rsi) # encoding: [0x88,0x06] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <4 x i64> %a to <8 x i32> + %1 = bitcast <4 x i64> %b to <8 x i32> + %2 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.d.256(<8 x i32> %0, <8 x i32> %1) + %3 = extractvalue { <8 x i1>, <8 x i1> } %2, 0 + %4 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %3, <8 x i1>* %4, align 8 + %5 = extractvalue { <8 x i1>, <8 x i1> } %2, 1 + %6 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %5, <8 x i1>* %6, align 8 + ret void +} + +define void @test_mm256_2intersect_epi64(<4 x i64> %a, <4 x i64> %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi64: +; X86: # %bb.0: # %entry +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x04] +; X86-NEXT: vp2intersectq %ymm1, %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x28,0x68,0xc1] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi64: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectq %ymm1, %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x28,0x68,0xc1] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdi) # encoding: [0x88,0x07] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rsi) # encoding: [0x88,0x06] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.q.256(<4 x i64> %a, <4 x i64> %b) + %1 = extractvalue { <4 x i1>, <4 x i1> } %0, 0 + %2 = shufflevector <4 x i1> %1, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %3 = bitcast <8 x i1> %2 to i8 + store i8 %3, i8* %m0, align 1 + %4 = extractvalue { <4 x i1>, <4 x i1> } %0, 1 + %5 = shufflevector <4 x i1> %4, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %6 = bitcast <8 x i1> %5 to i8 + store i8 %6, i8* %m1, align 1 + ret void +} + +define void @test_mm256_2intersect_epi32_p(<4 x i64>* nocapture readonly %a, <4 x i64>* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi32_p: +; X86: # %bb.0: # %entry +; X86-NEXT: movl 12(%esp), %eax # encoding: [0x8b,0x44,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x08] +; X86-NEXT: movl 4(%esp), %edx # encoding: [0x8b,0x54,0x24,0x04] +; X86-NEXT: vmovaps (%edx), %ymm0 # EVEX TO VEX Compression encoding: [0xc5,0xfc,0x28,0x02] +; X86-NEXT: vp2intersectd (%ecx), %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x28,0x68,0x01] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 16(%esp), %eax # encoding: [0x8b,0x44,0x24,0x10] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi32_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %ymm0 # EVEX TO VEX Compression encoding: [0xc5,0xfc,0x28,0x07] +; X64-NEXT: vp2intersectd (%rsi), %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x28,0x68,0x06] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %esi # encoding: [0xc5,0xf8,0x93,0xf0] +; X64-NEXT: movb %sil, (%rdx) # encoding: [0x40,0x88,0x32] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <4 x i64>* %a to <8 x i32>* + %1 = load <8 x i32>, <8 x i32>* %0, align 32 + %2 = bitcast <4 x i64>* %b to <8 x i32>* + %3 = load <8 x i32>, <8 x i32>* %2, align 32 + %4 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.d.256(<8 x i32> %1, <8 x i32> %3) + %5 = extractvalue { <8 x i1>, <8 x i1> } %4, 0 + %6 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %5, <8 x i1>* %6, align 8 + %7 = extractvalue { <8 x i1>, <8 x i1> } %4, 1 + %8 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %7, <8 x i1>* %8, align 8 + ret void +} + +define void @test_mm256_2intersect_epi64_p(<4 x i64>* nocapture readonly %a, <4 x i64>* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi64_p: +; X86: .cfi_startproc +; X86-NEXT: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl 20(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl 16(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl 12(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vmovaps (%esi), %ymm0 # EVEX TO VEX Compression encoding: [0xc5,0xfc,0x28,0x06] +; X86-NEXT: vp2intersectq (%edx), %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x28,0x68,0x02] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi64_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %ymm0 # EVEX TO VEX Compression encoding: [0xc5,0xfc,0x28,0x07] +; X64-NEXT: vp2intersectq (%rsi), %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x28,0x68,0x06] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load <4 x i64>, <4 x i64>* %a, align 32 + %1 = load <4 x i64>, <4 x i64>* %b, align 32 + %2 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.q.256(<4 x i64> %0, <4 x i64> %1) + %3 = extractvalue { <4 x i1>, <4 x i1> } %2, 0 + %4 = shufflevector <4 x i1> %3, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <4 x i1>, <4 x i1> } %2, 1 + %7 = shufflevector <4 x i1> %6, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +define void @test_mm256_2intersect_epi32_b(i32* nocapture readonly %a, i32* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi32_b: +; X86: .cfi_startproc +; X86-NEXT: # %bb.0: # %entry +; X86-NEXT: movl 12(%esp), %eax # encoding: [0x8b,0x44,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x08] +; X86-NEXT: movl 4(%esp), %edx # encoding: [0x8b,0x54,0x24,0x04] +; X86-NEXT: vbroadcastss (%edx), %ymm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x18,0x02] +; X86-NEXT: vp2intersectd (%ecx){1to8}, %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x38,0x68,0x01] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 16(%esp), %eax # encoding: [0x8b,0x44,0x24,0x10] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi32_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vbroadcastss (%rdi), %ymm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x18,0x07] +; X64-NEXT: vp2intersectd (%rsi){1to8}, %ymm0, %k0 # encoding: [0x62,0xf2,0x7f,0x38,0x68,0x06] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %esi # encoding: [0xc5,0xf8,0x93,0xf0] +; X64-NEXT: movb %sil, (%rdx) # encoding: [0x40,0x88,0x32] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i32, i32* %a, align 4 + %vecinit.i.i = insertelement <8 x i32> undef, i32 %0, i32 0 + %vecinit7.i.i = shufflevector <8 x i32> %vecinit.i.i, <8 x i32> undef, <8 x i32> zeroinitializer + %1 = load i32, i32* %b, align 4 + %vecinit.i.i2 = insertelement <8 x i32> undef, i32 %1, i32 0 + %vecinit7.i.i3 = shufflevector <8 x i32> %vecinit.i.i2, <8 x i32> undef, <8 x i32> zeroinitializer + %2 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.d.256(<8 x i32> %vecinit7.i.i, <8 x i32> %vecinit7.i.i3) + %3 = extractvalue { <8 x i1>, <8 x i1> } %2, 0 + %4 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %3, <8 x i1>* %4, align 8 + %5 = extractvalue { <8 x i1>, <8 x i1> } %2, 1 + %6 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %5, <8 x i1>* %6, align 8 + ret void +} + +define void @test_mm256_2intersect_epi64_b(i64* nocapture readonly %a, i64* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm256_2intersect_epi64_b: +; X86: .cfi_startproc +; X86-NEXT: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl 20(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl 16(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl 12(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vbroadcastsd (%esi), %ymm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x19,0x06] +; X86-NEXT: vbroadcastsd (%edx), %ymm1 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x19,0x0a] +; X86-NEXT: vp2intersectq %ymm1, %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x28,0x68,0xc1] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm256_2intersect_epi64_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vbroadcastsd (%rdi), %ymm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x19,0x07] +; X64-NEXT: vp2intersectq (%rsi){1to4}, %ymm0, %k0 # encoding: [0x62,0xf2,0xff,0x38,0x68,0x06] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i64, i64* %a, align 8 + %vecinit.i.i = insertelement <4 x i64> undef, i64 %0, i32 0 + %vecinit3.i.i = shufflevector <4 x i64> %vecinit.i.i, <4 x i64> undef, <4 x i32> zeroinitializer + %1 = load i64, i64* %b, align 8 + %vecinit.i.i2 = insertelement <4 x i64> undef, i64 %1, i32 0 + %vecinit3.i.i3 = shufflevector <4 x i64> %vecinit.i.i2, <4 x i64> undef, <4 x i32> zeroinitializer + %2 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.q.256(<4 x i64> %vecinit3.i.i, <4 x i64> %vecinit3.i.i3) + %3 = extractvalue { <4 x i1>, <4 x i1> } %2, 0 + %4 = shufflevector <4 x i1> %3, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <4 x i1>, <4 x i1> } %2, 1 + %7 = shufflevector <4 x i1> %6, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi32(<2 x i64> %a, <2 x i64> %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi32: +; X86: .cfi_startproc +; X86-NEXT: # %bb.0: # %entry +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x04] +; X86-NEXT: vp2intersectd %xmm1, %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x08,0x68,0xc1] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi32: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectd %xmm1, %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x08,0x68,0xc1] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdi) # encoding: [0x88,0x07] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rsi) # encoding: [0x88,0x06] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <2 x i64> %a to <4 x i32> + %1 = bitcast <2 x i64> %b to <4 x i32> + %2 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.d.128(<4 x i32> %0, <4 x i32> %1) + %3 = extractvalue { <4 x i1>, <4 x i1> } %2, 0 + %4 = shufflevector <4 x i1> %3, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <4 x i1>, <4 x i1> } %2, 1 + %7 = shufflevector <4 x i1> %6, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi64(<2 x i64> %a, <2 x i64> %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi64: +; X86: # %bb.0: # %entry +; X86-NEXT: movl 8(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movl 4(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x04] +; X86-NEXT: vp2intersectq %xmm1, %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x08,0x68,0xc1] +; X86-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X86-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X86-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi64: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectq %xmm1, %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x08,0x68,0xc1] +; X64-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X64-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdi) # encoding: [0x88,0x07] +; X64-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X64-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rsi) # encoding: [0x88,0x06] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = tail call { <2 x i1>, <2 x i1> } @llvm.x86.avx512.vp2intersect.q.128(<2 x i64> %a, <2 x i64> %b) + %1 = extractvalue { <2 x i1>, <2 x i1> } %0, 0 + %2 = shufflevector <2 x i1> %1, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %3 = bitcast <8 x i1> %2 to i8 + store i8 %3, i8* %m0, align 1 + %4 = extractvalue { <2 x i1>, <2 x i1> } %0, 1 + %5 = shufflevector <2 x i1> %4, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %6 = bitcast <8 x i1> %5 to i8 + store i8 %6, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi32_p(<2 x i64>* nocapture readonly %a, <2 x i64>* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi32_p: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl {{[0-9]+}}(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vmovaps (%esi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xf8,0x28,0x06] +; X86-NEXT: vp2intersectd (%edx), %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x08,0x68,0x02] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi32_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xf8,0x28,0x07] +; X64-NEXT: vp2intersectd (%rsi), %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x08,0x68,0x06] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <2 x i64>* %a to <4 x i32>* + %1 = load <4 x i32>, <4 x i32>* %0, align 16 + %2 = bitcast <2 x i64>* %b to <4 x i32>* + %3 = load <4 x i32>, <4 x i32>* %2, align 16 + %4 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.d.128(<4 x i32> %1, <4 x i32> %3) + %5 = extractvalue { <4 x i1>, <4 x i1> } %4, 0 + %6 = shufflevector <4 x i1> %5, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %7 = bitcast <8 x i1> %6 to i8 + store i8 %7, i8* %m0, align 1 + %8 = extractvalue { <4 x i1>, <4 x i1> } %4, 1 + %9 = shufflevector <4 x i1> %8, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %10 = bitcast <8 x i1> %9 to i8 + store i8 %10, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi64_p(<2 x i64>* nocapture readonly %a, <2 x i64>* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi64_p: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl {{[0-9]+}}(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vmovaps (%esi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xf8,0x28,0x06] +; X86-NEXT: vp2intersectq (%edx), %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x08,0x68,0x02] +; X86-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X86-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X86-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi64_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xf8,0x28,0x07] +; X64-NEXT: vp2intersectq (%rsi), %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x08,0x68,0x06] +; X64-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X64-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X64-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load <2 x i64>, <2 x i64>* %a, align 16 + %1 = load <2 x i64>, <2 x i64>* %b, align 16 + %2 = tail call { <2 x i1>, <2 x i1> } @llvm.x86.avx512.vp2intersect.q.128(<2 x i64> %0, <2 x i64> %1) + %3 = extractvalue { <2 x i1>, <2 x i1> } %2, 0 + %4 = shufflevector <2 x i1> %3, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <2 x i1>, <2 x i1> } %2, 1 + %7 = shufflevector <2 x i1> %6, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi32_b(i32* nocapture readonly %a, i32* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi32_b: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl {{[0-9]+}}(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vbroadcastss (%esi), %xmm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x79,0x18,0x06] +; X86-NEXT: vp2intersectd (%edx){1to4}, %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x18,0x68,0x02] +; X86-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X86-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X86-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi32_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vbroadcastss (%rdi), %xmm0 # EVEX TO VEX Compression encoding: [0xc4,0xe2,0x79,0x18,0x07] +; X64-NEXT: vp2intersectd (%rsi){1to4}, %xmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x18,0x68,0x06] +; X64-NEXT: kshiftlw $12, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0c] +; X64-NEXT: kshiftrw $12, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0c] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $12, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0c] +; X64-NEXT: kshiftrw $12, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0c] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i32, i32* %a, align 4 + %vecinit.i.i = insertelement <4 x i32> undef, i32 %0, i32 0 + %vecinit3.i.i = shufflevector <4 x i32> %vecinit.i.i, <4 x i32> undef, <4 x i32> zeroinitializer + %1 = load i32, i32* %b, align 4 + %vecinit.i.i2 = insertelement <4 x i32> undef, i32 %1, i32 0 + %vecinit3.i.i3 = shufflevector <4 x i32> %vecinit.i.i2, <4 x i32> undef, <4 x i32> zeroinitializer + %2 = tail call { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.d.128(<4 x i32> %vecinit3.i.i, <4 x i32> %vecinit3.i.i3) + %3 = extractvalue { <4 x i1>, <4 x i1> } %2, 0 + %4 = shufflevector <4 x i1> %3, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <4 x i1>, <4 x i1> } %2, 1 + %7 = shufflevector <4 x i1> %6, <4 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +define void @test_mm_2intersect_epi64_b(i64* nocapture readonly %a, i64* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm_2intersect_epi64_b: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl 20(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl 16(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl 12(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vmovddup (%esi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xfb,0x12,0x06] +; X86-NEXT: # xmm0 = mem[0,0] +; X86-NEXT: vmovddup (%edx), %xmm1 # EVEX TO VEX Compression encoding: [0xc5,0xfb,0x12,0x0a] +; X86-NEXT: # xmm1 = mem[0,0] +; X86-NEXT: vp2intersectq %xmm1, %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x08,0x68,0xc1] +; X86-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X86-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X86-NEXT: kmovw %k2, %edx # encoding: [0xc5,0xf8,0x93,0xd2] +; X86-NEXT: movb %dl, (%ecx) # encoding: [0x88,0x11] +; X86-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X86-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X86-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm_2intersect_epi64_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovddup (%rdi), %xmm0 # EVEX TO VEX Compression encoding: [0xc5,0xfb,0x12,0x07] +; X64-NEXT: # xmm0 = mem[0,0] +; X64-NEXT: vp2intersectq (%rsi){1to2}, %xmm0, %k0 # encoding: [0x62,0xf2,0xff,0x18,0x68,0x06] +; X64-NEXT: kshiftlw $14, %k0, %k2 # encoding: [0xc4,0xe3,0xf9,0x32,0xd0,0x0e] +; X64-NEXT: kshiftrw $14, %k2, %k2 # encoding: [0xc4,0xe3,0xf9,0x30,0xd2,0x0e] +; X64-NEXT: kmovw %k2, %eax # encoding: [0xc5,0xf8,0x93,0xc2] +; X64-NEXT: movb %al, (%rdx) # encoding: [0x88,0x02] +; X64-NEXT: kshiftlw $14, %k1, %k0 # encoding: [0xc4,0xe3,0xf9,0x32,0xc1,0x0e] +; X64-NEXT: kshiftrw $14, %k0, %k0 # encoding: [0xc4,0xe3,0xf9,0x30,0xc0,0x0e] +; X64-NEXT: kmovw %k0, %eax # encoding: [0xc5,0xf8,0x93,0xc0] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i64, i64* %a, align 8 + %vecinit.i.i = insertelement <2 x i64> undef, i64 %0, i32 0 + %vecinit1.i.i = shufflevector <2 x i64> %vecinit.i.i, <2 x i64> undef, <2 x i32> zeroinitializer + %1 = load i64, i64* %b, align 8 + %vecinit.i.i2 = insertelement <2 x i64> undef, i64 %1, i32 0 + %vecinit1.i.i3 = shufflevector <2 x i64> %vecinit.i.i2, <2 x i64> undef, <2 x i32> zeroinitializer + %2 = tail call { <2 x i1>, <2 x i1> } @llvm.x86.avx512.vp2intersect.q.128(<2 x i64> %vecinit1.i.i, <2 x i64> %vecinit1.i.i3) + %3 = extractvalue { <2 x i1>, <2 x i1> } %2, 0 + %4 = shufflevector <2 x i1> %3, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %5 = bitcast <8 x i1> %4 to i8 + store i8 %5, i8* %m0, align 1 + %6 = extractvalue { <2 x i1>, <2 x i1> } %2, 1 + %7 = shufflevector <2 x i1> %6, <2 x i1> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3> + %8 = bitcast <8 x i1> %7 to i8 + store i8 %8, i8* %m1, align 1 + ret void +} + +declare { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.d.256(<8 x i32>, <8 x i32>) +declare { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.q.256(<4 x i64>, <4 x i64>) +declare { <4 x i1>, <4 x i1> } @llvm.x86.avx512.vp2intersect.d.128(<4 x i32>, <4 x i32>) +declare { <2 x i1>, <2 x i1> } @llvm.x86.avx512.vp2intersect.q.128(<2 x i64>, <2 x i64>) diff --git a/llvm/test/CodeGen/X86/avx512vp2intersect-intrinsics.ll b/llvm/test/CodeGen/X86/avx512vp2intersect-intrinsics.ll new file mode 100644 index 00000000000..e2415b03c1e --- /dev/null +++ b/llvm/test/CodeGen/X86/avx512vp2intersect-intrinsics.ll @@ -0,0 +1,240 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+avx512vp2intersect --show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X86 +; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vp2intersect --show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X64 + +define void @test_mm512_2intersect_epi32(<8 x i64> %a, <8 x i64> %b, i16* nocapture %m0, i16* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi32: +; X86: # %bb.0: # %entry +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x04] +; X86-NEXT: vp2intersectd %zmm1, %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x48,0x68,0xc1] +; X86-NEXT: kmovw %k0, (%ecx) # encoding: [0xc5,0xf8,0x91,0x01] +; X86-NEXT: kmovw %k1, (%eax) # encoding: [0xc5,0xf8,0x91,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm512_2intersect_epi32: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectd %zmm1, %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x48,0x68,0xc1] +; X64-NEXT: kmovw %k0, (%rdi) # encoding: [0xc5,0xf8,0x91,0x07] +; X64-NEXT: kmovw %k1, (%rsi) # encoding: [0xc5,0xf8,0x91,0x0e] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <8 x i64> %a to <16 x i32> + %1 = bitcast <8 x i64> %b to <16 x i32> + %2 = tail call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %0, <16 x i32> %1) + %3 = extractvalue { <16 x i1>, <16 x i1> } %2, 0 + %4 = bitcast i16* %m0 to <16 x i1>* + store <16 x i1> %3, <16 x i1>* %4, align 16 + %5 = extractvalue { <16 x i1>, <16 x i1> } %2, 1 + %6 = bitcast i16* %m1 to <16 x i1>* + store <16 x i1> %5, <16 x i1>* %6, align 16 + ret void +} + +define void @test_mm512_2intersect_epi64(<8 x i64> %a, <8 x i64> %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi64: +; X86: # %bb.0: # %entry +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x04] +; X86-NEXT: vp2intersectq %zmm1, %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x48,0x68,0xc1] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 8(%esp), %eax # encoding: [0x8b,0x44,0x24,0x08] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm512_2intersect_epi64: +; X64: # %bb.0: # %entry +; X64-NEXT: vp2intersectq %zmm1, %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x48,0x68,0xc1] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %ecx # encoding: [0xc5,0xf8,0x93,0xc8] +; X64-NEXT: movb %cl, (%rdi) # encoding: [0x88,0x0f] +; X64-NEXT: movb %al, (%rsi) # encoding: [0x88,0x06] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.q.512(<8 x i64> %a, <8 x i64> %b) + %1 = extractvalue { <8 x i1>, <8 x i1> } %0, 0 + %2 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %1, <8 x i1>* %2, align 8 + %3 = extractvalue { <8 x i1>, <8 x i1> } %0, 1 + %4 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %3, <8 x i1>* %4, align 8 + ret void +} + +define void @test_mm512_2intersect_epi32_p(<8 x i64>* nocapture readonly %a, <8 x i64>* nocapture readonly %b, i16* nocapture %m0, i16* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi32_p: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl {{[0-9]+}}(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl {{[0-9]+}}(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vmovaps (%esi), %zmm0 # encoding: [0x62,0xf1,0x7c,0x48,0x28,0x06] +; X86-NEXT: vp2intersectd (%edx), %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x48,0x68,0x02] +; X86-NEXT: kmovw %k0, (%ecx) # encoding: [0xc5,0xf8,0x91,0x01] +; X86-NEXT: kmovw %k1, (%eax) # encoding: [0xc5,0xf8,0x91,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm512_2intersect_epi32_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %zmm0 # encoding: [0x62,0xf1,0x7c,0x48,0x28,0x07] +; X64-NEXT: vp2intersectd (%rsi), %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x48,0x68,0x06] +; X64-NEXT: kmovw %k0, (%rdx) # encoding: [0xc5,0xf8,0x91,0x02] +; X64-NEXT: kmovw %k1, (%rcx) # encoding: [0xc5,0xf8,0x91,0x09] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = bitcast <8 x i64>* %a to <16 x i32>* + %1 = load <16 x i32>, <16 x i32>* %0, align 64 + %2 = bitcast <8 x i64>* %b to <16 x i32>* + %3 = load <16 x i32>, <16 x i32>* %2, align 64 + %4 = tail call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %1, <16 x i32> %3) + %5 = extractvalue { <16 x i1>, <16 x i1> } %4, 0 + %6 = bitcast i16* %m0 to <16 x i1>* + store <16 x i1> %5, <16 x i1>* %6, align 16 + %7 = extractvalue { <16 x i1>, <16 x i1> } %4, 1 + %8 = bitcast i16* %m1 to <16 x i1>* + store <16 x i1> %7, <16 x i1>* %8, align 16 + ret void +} + +define void @test_mm512_2intersect_epi64_p(<8 x i64>* nocapture readonly %a, <8 x i64>* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi64_p: +; X86: # %bb.0: # %entry +; X86-NEXT: movl 12(%esp), %eax # encoding: [0x8b,0x44,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x08] +; X86-NEXT: movl 4(%esp), %edx # encoding: [0x8b,0x54,0x24,0x04] +; X86-NEXT: vmovaps (%edx), %zmm0 # encoding: [0x62,0xf1,0x7c,0x48,0x28,0x02] +; X86-NEXT: vp2intersectq (%ecx), %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x48,0x68,0x01] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 16(%esp), %eax # encoding: [0x8b,0x44,0x24,0x10] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl + +; +; X64-LABEL: test_mm512_2intersect_epi64_p: +; X64: # %bb.0: # %entry +; X64-NEXT: vmovaps (%rdi), %zmm0 # encoding: [0x62,0xf1,0x7c,0x48,0x28,0x07] +; X64-NEXT: vp2intersectq (%rsi), %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x48,0x68,0x06] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %esi # encoding: [0xc5,0xf8,0x93,0xf0] +; X64-NEXT: movb %sil, (%rdx) # encoding: [0x40,0x88,0x32] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load <8 x i64>, <8 x i64>* %a, align 64 + %1 = load <8 x i64>, <8 x i64>* %b, align 64 + %2 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.q.512(<8 x i64> %0, <8 x i64> %1) + %3 = extractvalue { <8 x i1>, <8 x i1> } %2, 0 + %4 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %3, <8 x i1>* %4, align 8 + %5 = extractvalue { <8 x i1>, <8 x i1> } %2, 1 + %6 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %5, <8 x i1>* %6, align 8 + ret void +} + +define void @test_mm512_2intersect_epi32_b(i32* nocapture readonly %a, i32* nocapture readonly %b, i16* nocapture %m0, i16* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi32_b: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %esi # encoding: [0x56] +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %esi, -8 +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax # encoding: [0x8b,0x44,0x24,0x14] +; X86-NEXT: movl {{[0-9]+}}(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x10] +; X86-NEXT: movl {{[0-9]+}}(%esp), %edx # encoding: [0x8b,0x54,0x24,0x0c] +; X86-NEXT: movl {{[0-9]+}}(%esp), %esi # encoding: [0x8b,0x74,0x24,0x08] +; X86-NEXT: vbroadcastss (%esi), %zmm0 # encoding: [0x62,0xf2,0x7d,0x48,0x18,0x06] +; X86-NEXT: vp2intersectd (%edx){1to16}, %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x58,0x68,0x02] +; X86-NEXT: kmovw %k0, (%ecx) # encoding: [0xc5,0xf8,0x91,0x01] +; X86-NEXT: kmovw %k1, (%eax) # encoding: [0xc5,0xf8,0x91,0x08] +; X86-NEXT: popl %esi # encoding: [0x5e] +; X86-NEXT: .cfi_def_cfa_offset 4 +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm512_2intersect_epi32_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vbroadcastss (%rdi), %zmm0 # encoding: [0x62,0xf2,0x7d,0x48,0x18,0x07] +; X64-NEXT: vp2intersectd (%rsi){1to16}, %zmm0, %k0 # encoding: [0x62,0xf2,0x7f,0x58,0x68,0x06] +; X64-NEXT: kmovw %k0, (%rdx) # encoding: [0xc5,0xf8,0x91,0x02] +; X64-NEXT: kmovw %k1, (%rcx) # encoding: [0xc5,0xf8,0x91,0x09] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i32, i32* %a, align 4 + %vecinit.i = insertelement <16 x i32> undef, i32 %0, i32 0 + %vecinit15.i = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer + %1 = load i32, i32* %b, align 4 + %vecinit.i2 = insertelement <16 x i32> undef, i32 %1, i32 0 + %vecinit15.i3 = shufflevector <16 x i32> %vecinit.i2, <16 x i32> undef, <16 x i32> zeroinitializer + %2 = tail call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %vecinit15.i, <16 x i32> %vecinit15.i3) + %3 = extractvalue { <16 x i1>, <16 x i1> } %2, 0 + %4 = bitcast i16* %m0 to <16 x i1>* + store <16 x i1> %3, <16 x i1>* %4, align 16 + %5 = extractvalue { <16 x i1>, <16 x i1> } %2, 1 + %6 = bitcast i16* %m1 to <16 x i1>* + store <16 x i1> %5, <16 x i1>* %6, align 16 + ret void +} + +define void @test_mm512_2intersect_epi64_b(i64* nocapture readonly %a, i64* nocapture readonly %b, i8* nocapture %m0, i8* nocapture %m1) { +; X86-LABEL: test_mm512_2intersect_epi64_b: +; X86: # %bb.0: # %entry +; X86-NEXT: movl 12(%esp), %eax # encoding: [0x8b,0x44,0x24,0x0c] +; X86-NEXT: movl 8(%esp), %ecx # encoding: [0x8b,0x4c,0x24,0x08] +; X86-NEXT: movl 4(%esp), %edx # encoding: [0x8b,0x54,0x24,0x04] +; X86-NEXT: vbroadcastsd (%edx), %zmm0 # encoding: [0x62,0xf2,0xfd,0x48,0x19,0x02] +; X86-NEXT: vbroadcastsd (%ecx), %zmm1 # encoding: [0x62,0xf2,0xfd,0x48,0x19,0x09] +; X86-NEXT: vp2intersectq %zmm1, %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x48,0x68,0xc1] +; X86-NEXT: kmovw %k1, %ecx # encoding: [0xc5,0xf8,0x93,0xc9] +; X86-NEXT: kmovw %k0, %edx # encoding: [0xc5,0xf8,0x93,0xd0] +; X86-NEXT: movb %dl, (%eax) # encoding: [0x88,0x10] +; X86-NEXT: movl 16(%esp), %eax # encoding: [0x8b,0x44,0x24,0x10] +; X86-NEXT: movb %cl, (%eax) # encoding: [0x88,0x08] +; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X86-NEXT: retl # encoding: [0xc3] +; +; X64-LABEL: test_mm512_2intersect_epi64_b: +; X64: # %bb.0: # %entry +; X64-NEXT: vbroadcastsd (%rdi), %zmm0 # encoding: [0x62,0xf2,0xfd,0x48,0x19,0x07] +; X64-NEXT: vp2intersectq (%rsi){1to8}, %zmm0, %k0 # encoding: [0x62,0xf2,0xff,0x58,0x68,0x06] +; X64-NEXT: kmovw %k1, %eax # encoding: [0xc5,0xf8,0x93,0xc1] +; X64-NEXT: kmovw %k0, %esi # encoding: [0xc5,0xf8,0x93,0xf0] +; X64-NEXT: movb %sil, (%rdx) # encoding: [0x40,0x88,0x32] +; X64-NEXT: movb %al, (%rcx) # encoding: [0x88,0x01] +; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77] +; X64-NEXT: retq # encoding: [0xc3] +entry: + %0 = load i64, i64* %a, align 8 + %vecinit.i = insertelement <8 x i64> undef, i64 %0, i32 0 + %vecinit7.i = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer + %1 = load i64, i64* %b, align 8 + %vecinit.i2 = insertelement <8 x i64> undef, i64 %1, i32 0 + %vecinit7.i3 = shufflevector <8 x i64> %vecinit.i2, <8 x i64> undef, <8 x i32> zeroinitializer + %2 = tail call { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.q.512(<8 x i64> %vecinit7.i, <8 x i64> %vecinit7.i3) + %3 = extractvalue { <8 x i1>, <8 x i1> } %2, 0 + %4 = bitcast i8* %m0 to <8 x i1>* + store <8 x i1> %3, <8 x i1>* %4, align 8 + %5 = extractvalue { <8 x i1>, <8 x i1> } %2, 1 + %6 = bitcast i8* %m1 to <8 x i1>* + store <8 x i1> %5, <8 x i1>* %6, align 8 + ret void +} + +declare { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32>, <16 x i32>) +declare { <8 x i1>, <8 x i1> } @llvm.x86.avx512.vp2intersect.q.512(<8 x i64>, <8 x i64>) diff --git a/llvm/test/CodeGen/X86/inline-asm-avx512f-x-constraint.ll b/llvm/test/CodeGen/X86/inline-asm-avx512f-x-constraint.ll index 2f711a86bc2..56e0855d955 100644 --- a/llvm/test/CodeGen/X86/inline-asm-avx512f-x-constraint.ll +++ b/llvm/test/CodeGen/X86/inline-asm-avx512f-x-constraint.ll @@ -2,7 +2,8 @@ ; CHECK: %[[REG1:.*]]:vr512_0_15 = COPY %1 ; CHECK: %[[REG2:.*]]:vr512_0_15 = COPY %2 -; CHECK: INLINEASM &"vpaddq\09$3, $2, $0 {$1}", 0, 7340042, def %{{.*}}, 1179657, %{{.*}}, 7340041, %[[REG1]], 7340041, %[[REG2]], 12, implicit-def early-clobber $df, 12, implicit-def early-clobber $fpsw, 12, implicit-def early-clobber $eflags +; CHECK: INLINEASM &"vpaddq\09$3, $2, $0 {$1}", 0, {{.*}}, def %{{.*}}, {{.*}}, %{{.*}}, {{.*}}, %[[REG1]], {{.*}}, %[[REG2]], 12, implicit-def early-clobber $df, 12, implicit-def early-clobber $fpsw, 12, implicit-def early-clobber $eflags + define <8 x i64> @mask_Yk_i8(i8 signext %msk, <8 x i64> %x, <8 x i64> %y) { entry: %0 = tail call <8 x i64> asm "vpaddq\09$3, $2, $0 {$1}", "=x,^Yk,x,x,~{dirflag},~{fpsr},~{flags}"(i8 %msk, <8 x i64> %x, <8 x i64> %y) diff --git a/llvm/test/CodeGen/X86/vp2intersect_multiple_pairs.ll b/llvm/test/CodeGen/X86/vp2intersect_multiple_pairs.ll new file mode 100644 index 00000000000..a38d95b6f94 --- /dev/null +++ b/llvm/test/CodeGen/X86/vp2intersect_multiple_pairs.ll @@ -0,0 +1,150 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+avx512vp2intersect | FileCheck %s --check-prefixes=CHECK,X86 +; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vp2intersect | FileCheck %s --check-prefixes=CHECK,X64 + +; Test with more than four live mask pairs + +define void @test(<16 x i32> %a0, <16 x i32> %b0, + <16 x i32> %a1, <16 x i32> %b1, + <16 x i32> %a2, <16 x i32> %b2, + <16 x i32> %a3, <16 x i32> %b3, + <16 x i32> %a4, <16 x i32> %b4, + i16* nocapture %m0, i16* nocapture %m1) { +; X86-LABEL: test: +; X86: # %bb.0: # %entry +; X86-NEXT: pushl %ebp +; X86-NEXT: .cfi_def_cfa_offset 8 +; X86-NEXT: .cfi_offset %ebp, -8 +; X86-NEXT: movl %esp, %ebp +; X86-NEXT: .cfi_def_cfa_register %ebp +; X86-NEXT: pushl %edi +; X86-NEXT: pushl %esi +; X86-NEXT: andl $-64, %esp +; X86-NEXT: subl $64, %esp +; X86-NEXT: .cfi_offset %esi, -16 +; X86-NEXT: .cfi_offset %edi, -12 +; X86-NEXT: movl 456(%ebp), %esi +; X86-NEXT: vmovaps 328(%ebp), %zmm3 +; X86-NEXT: vmovaps 200(%ebp), %zmm4 +; X86-NEXT: vmovaps 72(%ebp), %zmm5 +; X86-NEXT: vp2intersectd %zmm1, %zmm0, %k0 +; X86-NEXT: kmovw %k0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill +; X86-NEXT: kmovw %k1, {{[0-9]+}}(%esp) +; X86-NEXT: vp2intersectd 8(%ebp), %zmm2, %k0 +; X86-NEXT: kmovw %k0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill +; X86-NEXT: kmovw %k1, {{[0-9]+}}(%esp) +; X86-NEXT: vp2intersectd 136(%ebp), %zmm5, %k0 +; X86-NEXT: kmovw %k0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill +; X86-NEXT: kmovw %k1, {{[0-9]+}}(%esp) +; X86-NEXT: vp2intersectd 264(%ebp), %zmm4, %k0 +; X86-NEXT: kmovw %k0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill +; X86-NEXT: kmovw %k1, {{[0-9]+}}(%esp) +; X86-NEXT: vp2intersectd 392(%ebp), %zmm3, %k0 +; X86-NEXT: kmovw %k0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill +; X86-NEXT: kmovw %k1, {{[0-9]+}}(%esp) +; X86-NEXT: vzeroupper +; X86-NEXT: calll dummy +; X86-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k0 # 4-byte Folded Reload +; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 +; X86-NEXT: kmovw %k0, %eax +; X86-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k0 # 4-byte Folded Reload +; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 +; X86-NEXT: kmovw %k0, %ecx +; X86-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k0 # 4-byte Folded Reload +; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 +; X86-NEXT: kmovw %k0, %edx +; X86-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k0 # 4-byte Folded Reload +; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 +; X86-NEXT: kmovw %k0, %edi +; X86-NEXT: addl %edi, %eax +; X86-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k2 # 4-byte Folded Reload +; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k3 +; X86-NEXT: kmovw %k2, %edi +; X86-NEXT: addl %ecx, %edx +; X86-NEXT: kmovw %k1, %ecx +; +; X64-LABEL: test: +; X64: # %bb.0: # %entry +; X64-NEXT: pushq %rbp +; X64-NEXT: .cfi_def_cfa_offset 16 +; X64-NEXT: .cfi_offset %rbp, -16 +; X64-NEXT: movq %rsp, %rbp +; X64-NEXT: .cfi_def_cfa_register %rbp +; X64-NEXT: pushq %r14 +; X64-NEXT: pushq %rbx +; X64-NEXT: andq $-64, %rsp +; X64-NEXT: subq $64, %rsp +; X64-NEXT: .cfi_offset %rbx, -32 +; X64-NEXT: .cfi_offset %r14, -24 +; X64-NEXT: movq %rdi, %r14 +; X64-NEXT: vmovaps 16(%rbp), %zmm8 +; X64-NEXT: vp2intersectd %zmm1, %zmm0, %k0 +; X64-NEXT: kmovw %k0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill +; X64-NEXT: kmovw %k1, {{[0-9]+}}(%rsp) +; X64-NEXT: vp2intersectd %zmm3, %zmm2, %k0 +; X64-NEXT: kmovw %k0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill +; X64-NEXT: kmovw %k1, {{[0-9]+}}(%rsp) +; X64-NEXT: vp2intersectd %zmm5, %zmm4, %k0 +; X64-NEXT: kmovw %k0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill +; X64-NEXT: kmovw %k1, {{[0-9]+}}(%rsp) +; X64-NEXT: vp2intersectd %zmm7, %zmm6, %k0 +; X64-NEXT: kmovw %k0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill +; X64-NEXT: kmovw %k1, {{[0-9]+}}(%rsp) +; X64-NEXT: vp2intersectd 80(%rbp), %zmm8, %k0 +; X64-NEXT: kmovw %k0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill +; X64-NEXT: kmovw %k1, {{[0-9]+}}(%rsp) +; X64-NEXT: vzeroupper +; X64-NEXT: callq dummy +; X64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k0 # 4-byte Folded Reload +; X64-NEXT: kmovw {{[0-9]+}}(%rsp), %k1 +; X64-NEXT: kmovw %k0, %eax +; X64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k0 # 4-byte Folded Reload +; X64-NEXT: kmovw {{[0-9]+}}(%rsp), %k1 +; X64-NEXT: kmovw %k0, %ecx +; X64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k0 # 4-byte Folded Reload +; X64-NEXT: kmovw {{[0-9]+}}(%rsp), %k1 +; X64-NEXT: kmovw %k0, %edx +; X64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k0 # 4-byte Folded Reload +; X64-NEXT: kmovw {{[0-9]+}}(%rsp), %k1 +; X64-NEXT: kmovw %k0, %esi +; X64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k0 # 4-byte Folded Reload +; X64-NEXT: kmovw {{[0-9]+}}(%rsp), %k1 +; X64-NEXT: kmovw %k0, %edi +; X64-NEXT: kmovw %k1, %ebx +entry: + %0 = call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %a0, <16 x i32> %b0) + %1 = call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %a1, <16 x i32> %b1) + %2 = call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %a2, <16 x i32> %b2) + %3 = call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %a3, <16 x i32> %b3) + %4 = call { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32> %a4, <16 x i32> %b4) + + %5 = extractvalue { <16 x i1>, <16 x i1> } %0, 0 + %6 = extractvalue { <16 x i1>, <16 x i1> } %1, 0 + %7 = extractvalue { <16 x i1>, <16 x i1> } %2, 0 + %8 = extractvalue { <16 x i1>, <16 x i1> } %3, 0 + %9 = extractvalue { <16 x i1>, <16 x i1> } %4, 0 + %10 = extractvalue { <16 x i1>, <16 x i1> } %0, 1 + %11 = extractvalue { <16 x i1>, <16 x i1> } %1, 1 + + call void @dummy() + + %12 = bitcast <16 x i1> %5 to i16 + %13 = bitcast <16 x i1> %6 to i16 + %14 = bitcast <16 x i1> %7 to i16 + %15 = bitcast <16 x i1> %8 to i16 + %16 = bitcast <16 x i1> %9 to i16 + %17 = bitcast <16 x i1> %10 to i16 + %18 = bitcast <16 x i1> %11 to i16 + + %19 = add i16 %12, %13 + %20 = add i16 %14, %15 + %21 = add i16 %16, %17 + %22 = add i16 %19, %21 + %23 = add i16 %22, %20 + + store i16 %23, i16* %m0, align 16 + ret void +} + +declare { <16 x i1>, <16 x i1> } @llvm.x86.avx512.vp2intersect.d.512(<16 x i32>, <16 x i32>) +declare void @dummy() |
