[AMDGPU] Enable -combiner-topological-sorting by default
Open
backend:AMDGPU
missed-optimization
- Dominant language
- LLVM
- Stars
- 40.5k
- Forks
- 18.7k
- PR merge metrics
- PR metrics pending
Description
- Locally enable `-combiner-topological-sorting` by default.
- Run the lit codegen tests and any other workloads of interest.
- Triage and fix the (many, many) code quality regressions one by one.
- Stop when we can enable it by default without any significant regression.
Contributor guide
Assessment
This issue has not been assessed yet.