llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT--> @llvm/pr-subscribers-llvm-transforms Author: Matt Arsenault (arsenm) <details> <summary>Changes</summary> Theoretically this does a worse job with globals but this is not covered by existing tests --- Full diff: https://github.com/llvm/llvm-project/pull/135019.diff 1 Files Affected: - (modified) llvm/lib/Transforms/IPO/IROutliner.cpp (+2-1) ``````````diff diff --git a/llvm/lib/Transforms/IPO/IROutliner.cpp b/llvm/lib/Transforms/IPO/IROutliner.cpp index 8c5946fb07c65..eec869d57a6db 100644 --- a/llvm/lib/Transforms/IPO/IROutliner.cpp +++ b/llvm/lib/Transforms/IPO/IROutliner.cpp @@ -1126,7 +1126,8 @@ static void analyzeExitPHIsForOutputUses( // outside of the single PHINode we should not skip over it. for (unsigned Idx : IncomingVals) { Value *V = PN.getIncomingValue(Idx); - if (outputHasNonPHI(V, Idx, PN, PotentialExitsFromRegion, RegionBlocks)) { + if (!isa<Constant>(V) && + outputHasNonPHI(V, Idx, PN, PotentialExitsFromRegion, RegionBlocks)) { OutputsWithNonPhiUses.insert(V); OutputsReplacedByPHINode.erase(V); continue; `````````` </details> https://github.com/llvm/llvm-project/pull/135019 _______________________________________________ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
