[all-commits] [llvm/llvm-project] ecf8b0: [AMDGPU] Recompute EXEC liveness in SIWholeQuadMod...

[Igor Wodiany via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: ecf8b039e5e80a609a4446b29e83600062179cc6
      https://github.com/llvm/llvm-project/commit/ecf8b039e5e80a609a4446b29e83600062179cc6
  Author: Igor Wodiany <igor.wodiany at amd.com>
  Date:   2026-06-02 (Tue, 02 Jun 2026)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIWholeQuadMode.cpp

  Log Message:
  -----------
  [AMDGPU] Recompute EXEC liveness in SIWholeQuadMode::toExact (#200866)

This is required as stale liveness information can lead to an incorrect
optimization in `SIOptimizeExecMaskingPreRA`. For example, when `hi16`
is removed from EXEC, `optimizeElseBranch` produces an incorrect result
by removing `S_AND`.

This is caused by the following code:

```c++
  SlotIndex StartIdx = LIS->getInstructionIndex(SaveExecMI);
  SlotIndex EndIdx = LIS->getInstructionIndex(*AndExecMI);
  for (MCRegUnit Unit : TRI->regunits(ExecReg)) {
    LiveRange &RegUnit = LIS->getRegUnit(Unit);
    if (RegUnit.find(StartIdx) != std::prev(RegUnit.find(EndIdx)))
      return false;
  }
```

When `hi16` is available there are two `RegUnit`s, one for `hi16` and
one for `lo16`. In the case of `wqm.ll` test it produces two live
ranges:

```
0: [320r,320d:3)[368r,368d:2)[736r,736d:4)[832r,832d:1)[944r,944d:0) 0 at 944r 1 at 832r 2 at 368r 3 at 320r 4 at 736r
1: [12r,12d:1)[320r,320d:5)[368r,368d:4)[736r,736d:6)[744r,744d:0)[832r,832d:3)[944r,944d:2) 0 at 744r 1 at 12r 2 at 944r 3 at 832r 4 at 368r 5 at 320r 6 at 736r
```

When `hi16` is removed there is only one range:

```
0: [320r,320d:3)[368r,368d:2)[736r,736d:4)[832r,832d:1)[944r,944d:0) 0 at 944r 1 at 832r 2 at 368r 3 at 320r 4 at 736r
```

If only the first range is considered the loop will finish without
returning false and continue to remove `S_AND`. It is because EXEC
register for `S_AND` added by `SIWholeQuadMode::toExact` is not in the
range. This is incorrect.

When `hi16` is present the incorrect behavior is masked by the second
live range coming from `hi16`, causing the loop to return. This can be
most likely explained by the fact that `hi16` is only computed after the
new `S_AND` was added so contains more up to date information. Removing
EXEC from `LiveIntervals` forces the live range to be recomputed.



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