It uses VGPR_32.RegTypes which includes 16 bit types. As a
result DS_WRITE_B32 may be generated for "store i16" which
is a bug. The only reason we do not hit it now is relative
patterns complexity and sorting. Should DS_WRITE_B16 pattern
complexity become higher and the bug appears.
Details
Details
- Reviewers
arsenm - Commits
- rG03954a12aecb: [AMDGPU] Fix DS_WRITE_B32 patterns
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo