diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2024-11-06 18:27:43 +0000 |
|---|---|---|
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | 2024-11-06 18:35:03 +0000 |
| commit | 6ccbf1da6c9225fddaf6911e7bb49ee011e845a6 (patch) | |
| tree | 3da32db1cc70da38e0115bfd1edd9016f321b205 /llvm/lib/Target/AMDGPU/AMDGPUMachineFunction.cpp | |
| parent | 712c90e479f975f2e0c5ed4554dbf2f3a7a6d9d6 (diff) | |
| download | llvm-6ccbf1da6c9225fddaf6911e7bb49ee011e845a6.zip llvm-6ccbf1da6c9225fddaf6911e7bb49ee011e845a6.tar.gz llvm-6ccbf1da6c9225fddaf6911e7bb49ee011e845a6.tar.bz2 | |
[X86] combineSelect - use SelectableOp helper to match the zero operand as well as the target shuffle
For the "select(mask, extract_subvector(shuffle(x)), zero) --> extract_subvector(select(insert_subvector(mask), shuffle(x), zero))" fold, match the zero operand inside the SelectableOp helper.
Prep work for #113400 - we will be able to relax the zero operand requirement for some target shuffles.
Diffstat (limited to 'llvm/lib/Target/AMDGPU/AMDGPUMachineFunction.cpp')
0 files changed, 0 insertions, 0 deletions