projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 552e27c) | patch
[X86] combineX86ShufflesRecursively - peek through free extract_subvector nodes
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Sun, 5 Feb 2023 20:16:32 +0000 (20:16 +0000)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Sun, 5 Feb 2023 20:16:32 +0000 (20:16 +0000)
commitf9bf64798c1aee3d9f1ce754b2d625026e29d197
treee4466bad5dedf65e1d3a0d52460c2f19a5ced358tree | snapshot
parent552e27c5217cc41c857e47d47b1d9d1253c6cefdcommit | diff
[X86] combineX86ShufflesRecursively - peek through free extract_subvector nodes

If we're shuffling an op that is extracted from the lowest subvector of a larger vector then we should try to peek through as much as possible without exceeding the root size

Another step towards removing the widenSubVector call from combineX86ShufflesRecursively (Issue #45319)
llvm/lib/Target/X86/X86ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll diff | blob | history
Domain: System / Toolchain;