diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2024-03-26 14:09:48 +0000 |
|---|---|---|
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | 2024-03-26 15:03:41 +0000 |
| commit | d18bee2313ab8274fad386b1c0227316bf4238b2 (patch) | |
| tree | 91536a67e75b73ec8b227b243d9861a591205172 /clang/lib/CodeGen/CodeGenFunction.cpp | |
| parent | e933c05cd2cb97974cfca9544ebbdb9af349d602 (diff) | |
| download | llvm-d18bee2313ab8274fad386b1c0227316bf4238b2.zip llvm-d18bee2313ab8274fad386b1c0227316bf4238b2.tar.gz llvm-d18bee2313ab8274fad386b1c0227316bf4238b2.tar.bz2 | |
[X86] combineConcatVectorOps - concatenate FADD/FSUB/FMUL ops if we don't increase the number of INSERT_SUBVECTOR nodes.
FADD/FSUB/FMUL are usually less port-bound than INSERT_SUBVECTOR, so only concatenate if it reduces the instruction count and doesn't introduce extra INSERT_SUBVECTOR nodes.
Diffstat (limited to 'clang/lib/CodeGen/CodeGenFunction.cpp')
0 files changed, 0 insertions, 0 deletions