[PATCH] D126962: [SCCP] Don't mark edges feasible when resolving undefs

[Nikita Popov via Phabricator via llvm-commits]

nikic added inline comments.


================
Comment at: llvm/test/Transforms/FunctionSpecialization/bug52821-use-after-free.ll:37
   %phi2 = phi %mystruct* [ undef, %for.body2 ], [ null, %for.cond ]
-  br i1 undef, label %for.end, label %for.body2
+  br i1 false, label %for.end, label %for.body2
 
----------------
fhahn wrote:
> not sure if it would be better to use `i1` argument instead here and `llvm/test/Transforms/FunctionSpecialization/bug55000-read-uninitialized-value.ll`
I've picked the `i1 false` in these tests to retain the same behavior that that resolveUndefs would end up picking, so that the test expectation remains the same.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D126962/new/

https://reviews.llvm.org/D126962