[PATCH] D112306: [ScheduleDAGInstrs] Call adjustSchedDependency in more cases

Jay Foad via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Oct 22 04:31:07 PDT 2021


foad created this revision.
Herald added subscribers: hiraditya, MatzeB.
foad requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

This removes a condition and the corresponding FIXME comment, because
the Hexagon assertion it refers to has apparently been fixed, probably
by D76134 <https://reviews.llvm.org/D76134>.

NFCI. This just gives targets the opportunity to adjust latencies that
were set to 0 by the generic code because they involve "implicit pseudo"
operands.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D112306

Files:
  llvm/lib/CodeGen/ScheduleDAGInstrs.cpp


Index: llvm/lib/CodeGen/ScheduleDAGInstrs.cpp
===================================================================
--- llvm/lib/CodeGen/ScheduleDAGInstrs.cpp
+++ llvm/lib/CodeGen/ScheduleDAGInstrs.cpp
@@ -271,15 +271,10 @@
       if (!ImplicitPseudoDef && !ImplicitPseudoUse) {
         Dep.setLatency(SchedModel.computeOperandLatency(SU->getInstr(), OperIdx,
                                                         RegUse, UseOp));
-        ST.adjustSchedDependency(SU, OperIdx, UseSU, UseOp, Dep);
       } else {
         Dep.setLatency(0);
-        // FIXME: We could always let target to adjustSchedDependency(), and
-        // remove this condition, but that currently asserts in Hexagon BE.
-        if (SU->getInstr()->isBundle() || (RegUse && RegUse->isBundle()))
-          ST.adjustSchedDependency(SU, OperIdx, UseSU, UseOp, Dep);
       }
-
+      ST.adjustSchedDependency(SU, OperIdx, UseSU, UseOp, Dep);
       UseSU->addPred(Dep);
     }
   }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D112306.381510.patch
Type: text/x-patch
Size: 959 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20211022/a494a0fe/attachment.bin>


More information about the llvm-commits mailing list