[all-commits] [llvm/llvm-project] eed35c: [AArch64] Avoid selecting XAR for reverse operatio...

[Ricardo Jesus via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: eed35c9e7fba1d0e4b905520350f431c5178ca5d
      https://github.com/llvm/llvm-project/commit/eed35c9e7fba1d0e4b905520350f431c5178ca5d
  Author: Ricardo Jesus <rjj at nvidia.com>
  Date:   2026-02-11 (Wed, 11 Feb 2026)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
    M llvm/test/CodeGen/AArch64/sve2-xar.ll
    M llvm/test/CodeGen/AArch64/xar.ll

  Log Message:
  -----------
  [AArch64] Avoid selecting XAR for reverse operations. (#178706)

Rotations that implement reverse operations, for example:
```c
uint64x2_t foo(uint64x2_t r) {
  return (r >> 32) | (r << 32);
}
```
Are currently lowered as XAR (when available):
```gas
foo:
  movi    v1.2d, #0000000000000000
  xar     v0.2d, v0.2d, v1.2d, #32
  ret
```
This is suboptimal as REV* instructions typically have higher throughput
than XAR and do not require the zero operand.

This patch combines half-rotations to Neon or SVE REVs so that they're
no longer selected as XAR.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications