[PATCH] D105617: [PowerPC] Fix addi-load bias scheduling heuristics

[Jinsong Ji via Phabricator via llvm-commits]

jsji added inline comments.


================
Comment at: llvm/lib/Target/PowerPC/PPCMachineScheduler.cpp:27
+         Cand.SU->getInstr()->getOpcode() == PPC::ADDI8 ||
+         Cand.SU->getInstr()->getOpcode() == PPC::LI ||
+         Cand.SU->getInstr()->getOpcode() == PPC::LI8;
----------------
This heuristics was intended for the addi in loop induction variable calculation, not for all addi. 
So `LI/LI8` shouldn't be counted.

Do you have any specific example that we have LI/LI8 in loop and should schedule them earlier? 


================
Comment at: llvm/lib/Target/PowerPC/PPCMachineScheduler.cpp:173
 
-  if (isADDIInstr(TryCand) && !isADDIInstr(Cand)) {
-    TryCand.Reason = Stall;
-    return true;
+  if (tryGreater(isADDIInstr(TryCand) && Cand.SU->getInstr()->mayLoad(),
+                 isADDIInstr(Cand) && TryCand.SU->getInstr()->mayLoad(),
----------------
The heuristics was to try to schedule addi as early as possible, not necessary just before mayload instructions.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D105617/new/

https://reviews.llvm.org/D105617