When optimizing vmv.s.x/vmv.v.x's of scalar loads, if VL is known to be
1 then we don't need to perform a stride of x0, and can just do a
regular load.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Unit Tests
Unit Tests
Time | Test | |
---|---|---|
990 ms | x64 debian > LLVM.CodeGen/RISCV::fpclamptosat_vec.ll |
Could VM_S_X_VL nodes be lowered to non-strided loads with VL=1? Since only the first lane needs to be written to.