[clang] [llvm] [AMDGPU] Emit a waitcnt instruction after each memory instruction (PR #79236)

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Mon Jan 29 00:57:59 PST 2024


================
@@ -2561,6 +2567,70 @@ bool SIMemoryLegalizer::expandAtomicCmpxchgOrRmw(const SIMemOpInfo &MOI,
   return Changed;
 }
 
+bool SIMemoryLegalizer::GFX9InsertWaitcntForPreciseMem(MachineFunction &MF) {
+  const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
+  const SIInstrInfo *TII = ST.getInstrInfo();
+  IsaVersion IV = getIsaVersion(ST.getCPU());
+
+  bool Changed = false;
+
+  for (auto &MBB : MF) {
+    for (auto MI = MBB.begin(); MI != MBB.end();) {
+      MachineInstr &Inst = *MI;
+      ++MI;
+      if (Inst.mayLoadOrStore() == false)
----------------
arsenm wrote:

Don't use == false 

https://github.com/llvm/llvm-project/pull/79236


More information about the llvm-commits mailing list