Thanks for using Compiler Explorer
Sponsors
Jakt
C++
Ada
Analysis
Android Java
Android Kotlin
Assembly
C
C3
Carbon
C++ (Circle)
CIRCT
Clean
CMake
CMakeScript
COBOL
C++ for OpenCL
MLIR
Cppx
Cppx-Blue
Cppx-Gold
Cpp2-cppfront
Crystal
C#
CUDA C++
D
Dart
Elixir
Erlang
Fortran
F#
Go
Haskell
HLSL
Hook
Hylo
ispc
Java
Julia
Kotlin
LLVM IR
LLVM MIR
Modula-2
Nim
Objective-C
Objective-C++
OCaml
OpenCL C
Pascal
Pony
Python
Racket
Ruby
Rust
Snowball
Scala
Solidity
Spice
Swift
LLVM TableGen
Toit
TypeScript Native
V
Vala
Visual Basic
Zig
Javascript
GIMPLE
llvm source #1
Output
Compile to binary object
Link to binary
Execute the code
Intel asm syntax
Demangle identifiers
Verbose demangling
Filters
Unused labels
Library functions
Directives
Comments
Horizontal whitespace
Debug intrinsics
Compiler
clang (assertions trunk)
clang (trunk)
clang 10.0.0
clang 10.0.1
clang 11.0.0
clang 11.0.1
clang 12.0.0
clang 12.0.1
clang 13.0.0
clang 14.0.0
clang 15.0.0
clang 16.0.0
clang 17.0.1
clang 18.1.0
clang 4.0.1
clang 5.0.0
clang 6.0.0
clang 7.0.0
clang 8.0.0
clang 9.0.0
llc (assertions trunk)
llc (trunk)
llc 10.0.0
llc 10.0.1
llc 11.0.0
llc 11.0.1
llc 12.0.0
llc 12.0.1
llc 13.0.0
llc 14.0.0
llc 15.0.0
llc 16.0.0
llc 17.0.1
llc 18.1.0
llc 3.2
llc 3.3
llc 3.9.1
llc 4.0.0
llc 4.0.1
llc 5.0.0
llc 6.0.0
llc 7.0.0
llc 8.0.0
llc 9.0.0
opt (assertions trunk)
opt (trunk)
opt 10.0.0
opt 10.0.1
opt 11.0.0
opt 11.0.1
opt 12.0.0
opt 12.0.1
opt 13.0.0
opt 14.0.0
opt 15.0.0
opt 16.0.0
opt 17.0.1
opt 18.1.0
opt 3.2
opt 3.3
opt 3.9.1
opt 4.0.0
opt 4.0.1
opt 5.0.0
opt 6.0.0
opt 7.0.0
opt 8.0.0
opt 9.0.0
Options
Source code
define void @load_i64_stride3_vf16(<48 x i64>* noalias %in.vec, <16 x i64>* noalias %out.vec0, <16 x i64>* noalias %out.vec1, <16 x i64>* noalias %out.vec2) nounwind { %wide.vec = load <48 x i64>, <48 x i64>* %in.vec, align 32 %strided.vec0 = shufflevector <48 x i64> %wide.vec, <48 x i64> poison, <16 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21, i32 24, i32 27, i32 30, i32 33, i32 36, i32 39, i32 42, i32 45> %strided.vec1 = shufflevector <48 x i64> %wide.vec, <48 x i64> poison, <16 x i32> <i32 1, i32 4, i32 7, i32 10, i32 13, i32 16, i32 19, i32 22, i32 25, i32 28, i32 31, i32 34, i32 37, i32 40, i32 43, i32 46> %strided.vec2 = shufflevector <48 x i64> %wide.vec, <48 x i64> poison, <16 x i32> <i32 2, i32 5, i32 8, i32 11, i32 14, i32 17, i32 20, i32 23, i32 26, i32 29, i32 32, i32 35, i32 38, i32 41, i32 44, i32 47> ; call void @f() store <16 x i64> %strided.vec0, <16 x i64>* %out.vec0, align 32 store <16 x i64> %strided.vec1, <16 x i64>* %out.vec1, align 32 store <16 x i64> %strided.vec2, <16 x i64>* %out.vec2, align 32 ret void } declare void @f()
analysis source #2
Output
Compile to binary object
Link to binary
Execute the code
Intel asm syntax
Demangle identifiers
Verbose demangling
Filters
Unused labels
Library functions
Directives
Comments
Horizontal whitespace
Debug intrinsics
Compiler
OSACA (0.5.2)
llvm-mca (assertions trunk)
llvm-mca (trunk)
Options
Source code
.text .intel_syntax noprefix .file "example.ll" .globl load_i64_stride3_vf16 # -- Begin function load_i64_stride3_vf16 .p2align 4, 0x90 .type load_i64_stride3_vf16,@function load_i64_stride3_vf16: # @load_i64_stride3_vf16 # %bb.0: vmovdqa ymm3, ymmword ptr [rdi + 224] vmovdqa ymm6, ymmword ptr [rdi + 192] vmovdqa ymm5, ymmword ptr [rdi + 320] vmovdqa ymm7, ymmword ptr [rdi + 288] vmovdqa ymm8, ymmword ptr [rdi + 32] vmovdqa ymm9, ymmword ptr [rdi] vmovdqa ymm10, ymmword ptr [rdi + 128] vmovdqa ymm11, ymmword ptr [rdi + 96] vinserti128 ymm0, ymm0, xmmword ptr [rdi + 160], 1 vpermq ymm1, ymm11, 236 # ymm1 = ymm11[0,3,2,3] vpblendd ymm1, ymm1, ymm10, 48 # ymm1 = ymm1[0,1,2,3],ymm10[4,5],ymm1[6,7] vpblendd ymm13, ymm1, ymm0, 192 # ymm13 = ymm1[0,1,2,3,4,5],ymm0[6,7] vinserti128 ymm1, ymm0, xmmword ptr [rdi + 64], 1 vpermq ymm2, ymm9, 236 # ymm2 = ymm9[0,3,2,3] vpblendd ymm2, ymm2, ymm8, 48 # ymm2 = ymm2[0,1,2,3],ymm8[4,5],ymm2[6,7] vpblendd ymm14, ymm2, ymm1, 192 # ymm14 = ymm2[0,1,2,3,4,5],ymm1[6,7] vinserti128 ymm2, ymm0, xmmword ptr [rdi + 352], 1 vpermq ymm4, ymm7, 236 # ymm4 = ymm7[0,3,2,3] vpblendd ymm4, ymm4, ymm5, 48 # ymm4 = ymm4[0,1,2,3],ymm5[4,5],ymm4[6,7] vpblendd ymm15, ymm4, ymm2, 192 # ymm15 = ymm4[0,1,2,3,4,5],ymm2[6,7] vinserti128 ymm4, ymm0, xmmword ptr [rdi + 256], 1 vpermq ymm12, ymm6, 236 # ymm12 = ymm6[0,3,2,3] vpblendd ymm12, ymm12, ymm3, 48 # ymm12 = ymm12[0,1,2,3],ymm3[4,5],ymm12[6,7] vpblendd ymm4, ymm12, ymm4, 192 # ymm4 = ymm12[0,1,2,3,4,5],ymm4[6,7] vpblendd ymm6, ymm6, ymm3, 240 # ymm6 = ymm6[0,1,2,3],ymm3[4,5,6,7] vpalignr ymm3, ymm3, ymm6, 8 # ymm3 = ymm6[8,9,10,11,12,13,14,15],ymm3[0,1,2,3,4,5,6,7],ymm6[24,25,26,27,28,29,30,31],ymm3[16,17,18,19,20,21,22,23] vpbroadcastq ymm6, qword ptr [rdi + 272] vpblendd ymm3, ymm3, ymm6, 192 # ymm3 = ymm3[0,1,2,3,4,5],ymm6[6,7] vpblendd ymm6, ymm11, ymm10, 240 # ymm6 = ymm11[0,1,2,3],ymm10[4,5,6,7] vpalignr ymm6, ymm10, ymm6, 8 # ymm6 = ymm6[8,9,10,11,12,13,14,15],ymm10[0,1,2,3,4,5,6,7],ymm6[24,25,26,27,28,29,30,31],ymm10[16,17,18,19,20,21,22,23] vpbroadcastq ymm10, qword ptr [rdi + 176] vpblendd ymm6, ymm6, ymm10, 192 # ymm6 = ymm6[0,1,2,3,4,5],ymm10[6,7] vpblendd ymm9, ymm9, ymm8, 240 # ymm9 = ymm9[0,1,2,3],ymm8[4,5,6,7] vpalignr ymm8, ymm8, ymm9, 8 # ymm8 = ymm9[8,9,10,11,12,13,14,15],ymm8[0,1,2,3,4,5,6,7],ymm9[24,25,26,27,28,29,30,31],ymm8[16,17,18,19,20,21,22,23] vpbroadcastq ymm9, qword ptr [rdi + 80] vpblendd ymm8, ymm8, ymm9, 192 # ymm8 = ymm8[0,1,2,3,4,5],ymm9[6,7] vpblendd ymm7, ymm7, ymm5, 240 # ymm7 = ymm7[0,1,2,3],ymm5[4,5,6,7] vpalignr ymm5, ymm5, ymm7, 8 # ymm5 = ymm7[8,9,10,11,12,13,14,15],ymm5[0,1,2,3,4,5,6,7],ymm7[24,25,26,27,28,29,30,31],ymm5[16,17,18,19,20,21,22,23] vpbroadcastq ymm7, qword ptr [rdi + 368] vpblendd ymm5, ymm5, ymm7, 192 # ymm5 = ymm5[0,1,2,3,4,5],ymm7[6,7] vmovaps xmm7, xmmword ptr [rdi + 112] vblendps xmm7, xmm7, xmmword ptr [rdi + 128], 12 # xmm7 = xmm7[0,1],mem[2,3] vpermpd ymm9, ymmword ptr [rdi + 160], 196 # ymm9 = mem[0,1,0,3] vblendps ymm7, ymm7, ymm9, 240 # ymm7 = ymm7[0,1,2,3],ymm9[4,5,6,7] vmovaps xmm0, xmmword ptr [rdi + 16] vblendps xmm0, xmm0, xmmword ptr [rdi + 32], 12 # xmm0 = xmm0[0,1],mem[2,3] vpermpd ymm9, ymmword ptr [rdi + 64], 196 # ymm9 = mem[0,1,0,3] vblendps ymm0, ymm0, ymm9, 240 # ymm0 = ymm0[0,1,2,3],ymm9[4,5,6,7] vmovaps xmm1, xmmword ptr [rdi + 304] vblendps xmm1, xmm1, xmmword ptr [rdi + 320], 12 # xmm1 = xmm1[0,1],mem[2,3] vpermpd ymm9, ymmword ptr [rdi + 352], 196 # ymm9 = mem[0,1,0,3] vblendps ymm1, ymm1, ymm9, 240 # ymm1 = ymm1[0,1,2,3],ymm9[4,5,6,7] vmovaps xmm2, xmmword ptr [rdi + 208] vblendps xmm2, xmm2, xmmword ptr [rdi + 224], 12 # xmm2 = xmm2[0,1],mem[2,3] vpermpd ymm9, ymmword ptr [rdi + 256], 196 # ymm9 = mem[0,1,0,3] vblendps ymm2, ymm2, ymm9, 240 # ymm2 = ymm2[0,1,2,3],ymm9[4,5,6,7] vmovdqa ymmword ptr [rsi + 64], ymm4 vmovdqa ymmword ptr [rsi + 96], ymm15 vmovdqa ymmword ptr [rsi], ymm14 vmovdqa ymmword ptr [rsi + 32], ymm13 vmovdqa ymmword ptr [rdx + 96], ymm5 vmovdqa ymmword ptr [rdx], ymm8 vmovdqa ymmword ptr [rdx + 32], ymm6 vmovdqa ymmword ptr [rdx + 64], ymm3 vmovaps ymmword ptr [rcx + 64], ymm2 vmovaps ymmword ptr [rcx + 96], ymm1 vmovaps ymmword ptr [rcx], ymm0 vmovaps ymmword ptr [rcx + 32], ymm7 vzeroupper ret .Lfunc_end0: .size load_i64_stride3_vf16, .Lfunc_end0-load_i64_stride3_vf16 # -- End function .section ".note.GNU-stack","",@progbits
Become a Patron
Sponsor on GitHub
Donate via PayPal
Source on GitHub
Mailing list
Installed libraries
Wiki
Report an issue
How it works
Contact the author
CE on Mastodon
About the author
Statistics
Changelog
Version tree