projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 6dbf894) | patch
[X86] Add support for using Sched/Codesize information to `X86FixupInstTuning` Pass.
authorNoah Goldstein <goldstein.w.n@gmail.com>
Wed, 22 Feb 2023 16:59:17 +0000 (10:59 -0600)
committerNoah Goldstein <goldstein.w.n@gmail.com>
Wed, 8 Mar 2023 17:57:38 +0000 (11:57 -0600)
commit6b29a6f27d40cfc4225fe14d0f0a298fde611cad
tree0667aa566b82d40b328d13a4b169a2f69e1d3707tree | snapshot
parent6dbf894b806665b53c148844b80f0631a3543698commit | diff
[X86] Add support for using Sched/Codesize information to `X86FixupInstTuning` Pass.

Use this to handle new transform: `{v}unpck{l|h}pd` -> `{v}shufps`. We
need the sched information here as `{v}shufps` is 1 more byte of code
size, so we only want to make this transformation if `{v}shufps` is
actually faster.

Differential Revision: https://reviews.llvm.org/D144570
llvm/lib/Target/X86/X86FixupInstTuning.cpp diff | blob | history
llvm/test/CodeGen/X86/tuning-shuffle-unpckpd-avx512.ll diff | blob | history
llvm/test/CodeGen/X86/tuning-shuffle-unpckpd.ll diff | blob | history
Domain: System / Toolchain;