[llvm] [AMDGPU] Do not put memory instructions in *ALU SchedGroups (PR #162560)

Jeffrey Byrnes via llvm-commits llvm-commits at lists.llvm.org
Fri Oct 10 09:18:07 PDT 2025


================
@@ -2394,15 +2394,18 @@ bool SchedGroup::canAddMI(const MachineInstr &MI) const {
   else if (((SGMask & SchedGroupMask::ALU) != SchedGroupMask::NONE) &&
            (TII->isVALU(MI) || TII->isMFMAorWMMA(MI) || TII->isSALU(MI) ||
             TII->isTRANS(MI)))
-    Result = true;
+    Result = !(MI.mayLoad() || MI.mayLoad());
----------------
jrbyrnes wrote:

Oops -- thanks

https://github.com/llvm/llvm-project/pull/162560


More information about the llvm-commits mailing list