[PATCH] D99790: [CGCall] Annotate `this` argument with alignment
Brooks Moses via Phabricator via cfe-commits
cfe-commits at lists.llvm.org
Thu Apr 8 18:18:01 PDT 2021
brooksmoses added a comment.
In D99790#2677919 <https://reviews.llvm.org/D99790#2677919>, @lebedev.ri wrote:
> In D99790#2677917 <https://reviews.llvm.org/D99790#2677917>, @brooksmoses wrote:
>
>> As a heads up, I'm seeing segfaults on internal code as a result of this change, as well as errors in Eigen's unalignedassert.cpp test (specifically, this line asserts: https://github.com/madlib/eigen/blob/master/test/unalignedassert.cpp#L151).
>
> Would be good to have a small standalone reproducer.
> Not really sure how we can end up with a misaligned `this`, but it sounds like UB.
Indeed, it's looking like all of the various segfaults are resulting from undefined behavior, just like the Eigen assert was (per @jyknight's comment). One of the segfaults is in the OpenJDK runtime -- albeit our internal copy, so it's possible it might not be in the external versions -- so that's fun. Luckily it shows up in the bootstrapping part of the build, rather than lying in wait to bite people after it's deployed.
In any case, thanks for the quick reply, and I'll figure out a small reproducer if we find something that isn't UB.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D99790/new/
https://reviews.llvm.org/D99790
More information about the cfe-commits
mailing list