[llvm] [AMDGPU] Classify FLAT instructions as VMEM (PR #137148)
Robert Imschweiler via llvm-commits
llvm-commits at lists.llvm.org
Thu Apr 24 07:19:39 PDT 2025
================
@@ -1425,8 +1418,8 @@ static bool shouldRunLdsBranchVmemWARHazardFixup(const MachineFunction &MF,
for (auto &MBB : MF) {
for (auto &MI : MBB) {
HasLds |= SIInstrInfo::isDS(MI);
- HasVmem |=
- SIInstrInfo::isVMEM(MI) || SIInstrInfo::isSegmentSpecificFLAT(MI);
+ HasVmem |= (SIInstrInfo::isVMEM(MI) && !SIInstrInfo::isFLAT(MI)) ||
+ SIInstrInfo::isSegmentSpecificFLAT(MI);
----------------
ro-i wrote:
This is addressed by #137170 because otherwise this PR wouldn't be NFC anymore
https://github.com/llvm/llvm-project/pull/137148
More information about the llvm-commits
mailing list