[PATCH] D41981: [SROA] fix assetion failure

[Hiroshi Inoue via Phabricator via llvm-commits]

inouehrs added a comment.

In https://reviews.llvm.org/D41981#975406, @efriedma wrote:

> It looks like this doesn't properly fix PR35657; it fixes the crash, yes, but it looks like the optimizer is introducing undefined behavior.


I updated the test case to avoid undef in the generated code due to loads from uninitialized alloca.
Do you expect other behavior (such as not applying SROA for out-of-bounds accesses)?


https://reviews.llvm.org/D41981