[PATCH] D126173: [VPlan] Use Exiting-block instead of Exit-block terminology (NFC).

Ayal Zaks via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Sun May 22 14:59:45 PDT 2022


Ayal accepted this revision.
Ayal added a comment.
This revision is now accepted and ready to land.

Thanks for following up! Minor nits to help keep consistency.



================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:179
+  assert(Parent->getExiting() == this &&
          "Block w/o successors not the exit of its parent.");
   return Parent->getEnclosingBlockWithSuccessors();
----------------
not the exiting?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:301
     } else {
       // PredVPBB is the exit block of a loop region. Connect its successor
       // outside the region.
----------------
the exiting block?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:308
+             LoopRegion->getExitingBasicBlock() == PredVPBB &&
              "PredVPBB must be the exit block of its parent region");
       assert(this == LoopRegion->getSingleSuccessor() &&
----------------
ditto


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:333
     // ExitBB can be re-used for the exit block of the Plan.
     NewBB = State->CFG.ExitBB;
     State->CFG.PrevBB = NewBB;
----------------
ExitBB should remain as such, right?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:348
     // C. when the current VPBB is an entry of a region replica - where PrevVPBB
     //    is the exit of this region from a previous instance, or the
     //    predecessor of this region.
----------------
should the above read "the exiting VPBB of this region"?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.cpp:1190
   // Due to "dot" we print an edge between two regions as an edge between the
   // exit basic block and the entry basic of the respective regions.
+  const VPBlockBase *Tail = From->getExitingBasicBlock();
----------------
// exiting basic block ?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.h:495
 
-  /// \return the VPBasicBlock that is the exit of this VPBlockBase,
+  /// \return the VPBasicBlock that is the exiting this VPBlockBase,
   /// recursively, if the latter is a VPRegionBlock. Otherwise, if this
----------------
... the exiting VPBasicBlock of this VPBlockBase?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlan.h:2136
     assert(Entry->getPredecessors().empty() && "Entry block has predecessors.");
-    assert(Exit->getSuccessors().empty() && "Exit block has successors.");
+    assert(Exiting->getSuccessors().empty() && "Exit block has successors.");
     Entry->setParent(this);
----------------
"Exiting block ...


================
Comment at: llvm/lib/Transforms/Vectorize/VPlanHCFGBuilder.cpp:332
   TopRegion->setEntry(PreheaderVPBB);
-  TopRegion->setExit(LoopExitVPBB);
+  TopRegion->setExiting(LoopExitVPBB);
   return TopRegion;
----------------
LoopExitingVPBB?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlanPredicator.cpp:133
                                                   VPRegionBlock *Region) {
   // Blocks that dominate region exit inherit the predicate from the region.
   // Return after setting the predicate.
----------------
region exit(ing) ? In terms of domination they are close...


================
Comment at: llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp:97
   const VPBlockBase *Entry = Region->getEntry();
-  const VPBlockBase *Exit = Region->getExit();
+  const VPBlockBase *Exit = Region->getExiting();
 
----------------
Exit >> Exiting?


================
Comment at: llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp:173
 
-  const VPBasicBlock *Exit = dyn_cast<VPBasicBlock>(TopRegion->getExit());
+  const VPBasicBlock *Exit = dyn_cast<VPBasicBlock>(TopRegion->getExiting());
   if (!Exit) {
----------------
ditto


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D126173/new/

https://reviews.llvm.org/D126173



More information about the llvm-commits mailing list