[all-commits] [llvm/llvm-project] bf684a: [RISCV] Don't emit vxrm writes for vnclip(u).wi wi...

[Craig Topper via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: bf684a97f37b12ccf2c98b007b8222c54c7480f5
      https://github.com/llvm/llvm-project/commit/bf684a97f37b12ccf2c98b007b8222c54c7480f5
  Author: Craig Topper <craig.topper at sifive.com>
  Date:   2024-01-02 (Tue, 02 Jan 2024)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVInsertWriteVXRM.cpp
    M llvm/test/CodeGen/RISCV/rvv/fpclamptosat_vec.ll
    M llvm/test/CodeGen/RISCV/rvv/trunc-sat-clip.ll

  Log Message:
  -----------
  [RISCV] Don't emit vxrm writes for vnclip(u).wi with shift of 0. (#76578)

If there's no shift being performed, the rounding mode doesn't matter.

We could do the same for vssra and vssrl, but they are no-ops with a
shift of 0 so would be better off being removed earlier.