projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: f29256a) | patch
[X86] lowerToAddSubOrFMAddSub - lower 512-bit ADDSUB patterns to blend(fsub,fadd)
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Thu, 20 Jan 2022 15:15:46 +0000 (15:15 +0000)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Thu, 20 Jan 2022 15:16:05 +0000 (15:16 +0000)
commit866311e71c8f1f3724754d006eb50299424d5b1b
tree57f9f5f7b6afb982c88f48d8b8f28b30ef25f33atree | snapshot
parentf29256a64ac11cf59cea878c8d1ba9537db4f523commit | diff
[X86] lowerToAddSubOrFMAddSub - lower 512-bit ADDSUB patterns to blend(fsub,fadd)

AVX512 doesn't provide a ADDSUB instruction, but if we've built this from a build vector of scalar fsub/fadd elements we can still lower to blend(fsub,fadd)
llvm/lib/Target/X86/X86ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/X86/sse3-avx-addsub-2.ll diff | blob | history
Domain: System / Toolchain;