[llvm] 558f3ea - [AMDGPU] Remove #if 0 code for indexed resources in SIInsertWaitcnts (#92905)

via llvm-commits llvm-commits at lists.llvm.org
Tue May 21 05:51:47 PDT 2024


Author: Jay Foad
Date: 2024-05-21T13:51:42+01:00
New Revision: 558f3ea4aef718f953d3f38a45363d8a91988292

URL: https://github.com/llvm/llvm-project/commit/558f3ea4aef718f953d3f38a45363d8a91988292
DIFF: https://github.com/llvm/llvm-project/commit/558f3ea4aef718f953d3f38a45363d8a91988292.diff

LOG: [AMDGPU] Remove #if 0 code for indexed resources in SIInsertWaitcnts (#92905)

I do not understand what optimization this was supposed to implement.
It has never been enabled. I suspect it no longer applies to GCN/RDNA
architectures.

Added: 
    

Modified: 
    llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
index 0a50973e939b0..230443313d72a 100644
--- a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
@@ -2244,17 +2244,6 @@ bool SIInsertWaitcnts::insertWaitcntInBlock(MachineFunction &MF,
 
     updateEventWaitcntAfter(Inst, &ScoreBrackets);
 
-#if 0 // TODO: implement resource type check controlled by options with ub = LB.
-    // If this instruction generates a S_SETVSKIP because it is an
-    // indexed resource, and we are on Tahiti, then it will also force
-    // an S_WAITCNT vmcnt(0)
-    if (RequireCheckResourceType(Inst, context)) {
-      // Force the score to as if an S_WAITCNT vmcnt(0) is emitted.
-      ScoreBrackets->setScoreLB(LOAD_CNT,
-      ScoreBrackets->getScoreUB(LOAD_CNT));
-    }
-#endif
-
     if (ST->isPreciseMemoryEnabled() && Inst.mayLoadOrStore()) {
       AMDGPU::Waitcnt Wait = WCG->getAllZeroWaitcnt(
           Inst.mayStore() && !SIInstrInfo::isAtomicRet(Inst));


        


More information about the llvm-commits mailing list