[PATCH] D60686: Asan use-after-scope: don't poison allocas if there were untraced lifetime intrinsics in the function (PR41481)

[Evgenii Stepanov via Phabricator via llvm-commits]

eugenis added inline comments.


================
Comment at: llvm/test/Instrumentation/AddressSanitizer/stack-poisoning-and-lifetime.ll:223
+  ; Since we cannot account for all lifetime intrinsics in this function, we
+  ; might have missed a lifetime.start one and therefore shouldn't poison the
+  ; allocas at function entry.
----------------
eugenis wrote:
> hans wrote:
> > eugenis wrote:
> > > did you mean "should poison"?
> > No, I mean shouldn't poison. Since we don't know the lifetimes we can't poison it. Or am I missing something?
> Well, we shouldn't poison them at lifetime intrinsic.
> We should poison them at function entry (and only there).
> 
Ah, I get what you mean. We should not poison alloca contents, only their redzones.


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D60686/new/

https://reviews.llvm.org/D60686