[llvm] 69d0746 - [Lint] Use poison instead of undef for self-referential values

[Nikita Popov via llvm-commits]

Author: Nikita Popov
Date: 2024-06-24T16:32:03+02:00
New Revision: 69d07465c5cb0194d15dfe75b058957fd6ab16c8

URL: https://github.com/llvm/llvm-project/commit/69d07465c5cb0194d15dfe75b058957fd6ab16c8
DIFF: https://github.com/llvm/llvm-project/commit/69d07465c5cb0194d15dfe75b058957fd6ab16c8.diff

LOG: [Lint] Use poison instead of undef for self-referential values

Added: 
    

Modified: 
    llvm/lib/Analysis/Lint.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Analysis/Lint.cpp b/llvm/lib/Analysis/Lint.cpp
index 1ab856ac8830a..e9ed0c22c8e37 100644
--- a/llvm/lib/Analysis/Lint.cpp
+++ b/llvm/lib/Analysis/Lint.cpp
@@ -647,7 +647,7 @@ Value *Lint::findValueImpl(Value *V, bool OffsetOk,
                            SmallPtrSetImpl<Value *> &Visited) const {
   // Detect self-referential values.
   if (!Visited.insert(V).second)
-    return UndefValue::get(V->getType());
+    return PoisonValue::get(V->getType());
 
   // TODO: Look through sext or zext cast, when the result is known to
   // be interpreted as signed or unsigned, respectively.