[all-commits] [llvm/llvm-project] 8e091b: [RISCV] Prefer vmsle.vi vX, vY, -1 over vslt.vx vX...

[Craig Topper via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 8e091b1220e09c0c1000ed676c1c92a09e871129
      https://github.com/llvm/llvm-project/commit/8e091b1220e09c0c1000ed676c1c92a09e871129
  Author: Craig Topper <craig.topper at sifive.com>
  Date:   2023-04-24 (Mon, 24 Apr 2023)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVInstrInfoVSDPatterns.td
    M llvm/lib/Target/RISCV/RISCVInstrInfoVVLPatterns.td
    M llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-setcc.ll
    M llvm/test/CodeGen/RISCV/rvv/setcc-integer.ll
    M llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll

  Log Message:
  -----------
  [RISCV] Prefer vmsle.vi vX, vY, -1 over vslt.vx vX, vY, x0.

If a target hasn't optimized scalar to vector transfers for x0,
using an immediate may be more efficient.