Consistently use LLVM lifetime markers during codegen #72178
Conversation
|
Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @eddyb (or someone else) soon. If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes. Please see the contribution instructions for more information. |
rust-highfive
added
the
S-waiting-on-review
|
This looks fine to me, but I'm wondering if there would be value in adding the testcase from #72154 as an asan test? |
tmiasko
force-pushed
the
inliner-lifetimes
branch
from
88d1015
to
3e8f12d
Compare
|
I added a regression test and made equivalent changes to the flow using a new LLVM pass manager (although, as far as I can see it is not affected by the original issue, since inliner used there does not merge allocas). |
Ensure that inliner inserts lifetime markers if they have been emitted during codegen. Otherwise if allocas from inlined functions are merged together, lifetime markers from one function might invalidate load & stores performed by the other one.
tmiasko
force-pushed
the
inliner-lifetimes
branch
from
3e8f12d
to
bbb63d4
Compare
|
@bors r+ |
|
📌 Commit bbb63d4 has been approved by |
bors
added
S-waiting-on-bors
|
☀️ Test successful - checks-azure |
Ensure that inliner inserts lifetime markers if they have been emitted during
codegen. Otherwise if allocas from inlined functions are merged together,
lifetime markers from one function might invalidate load & stores performed
by the other one.
Fixes #72154.