codegen: skip stores for entirely-uninit constant aggregate fields#157797
Open
glandium wants to merge 1 commit into
Open
codegen: skip stores for entirely-uninit constant aggregate fields#157797glandium wants to merge 1 commit into
glandium wants to merge 1 commit into
Conversation
rustbot
added
S-waiting-on-review
Collaborator
|
r? @jackh726 rustbot has assigned @jackh726. Use Why was this reviewer chosen?The reviewer was selected based on:
|
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
MIR GVN (since rust-lang#147827) propagates MaybeUninit::uninit() as `const <uninit>` in aggregate constructions. Without this fix, codegen would emit a memcpy from an `[N x i8] undef` global for each such field, which LLVM materializes as zero-initialization. This mirrors the existing `all_bytes_uninit` skip already present for `Rvalue::Use` (added in rust-lang#147827) into the `Rvalue::Aggregate` field loop.
Collaborator
|
Member
|
Uhh...not code I'm familiar with. Let's see if a reroll helps. @rustbot reroll |
Member
|
@rustbot reroll |
Member
|
@rustbot reroll |
oli-obk
requested changes
Jun 26, 2026
Comment on lines
+193
to
+201
| // Do not generate stores for entirely uninit constant fields. | ||
| // Leaving the field uninitialized is correct for MaybeUninit and avoids | ||
| // LLVM materializing the undef as zeros (causing unnecessary memset/stores). | ||
| if let mir::Operand::Constant(const_op) = operand { | ||
| let val = self.eval_mir_constant(const_op); | ||
| if val.all_bytes_uninit(self.cx.tcx()) { | ||
| continue; | ||
| } | ||
| } |
Contributor
There was a problem hiding this comment.
Hmm... seems like sth we should just do generally in codegen_operand (could add a new OperandValue::Uninit that can be handled at the call sites that are interested in it). In this case the store_with_annotation should probably just not generate any code.
rustbot
added
S-waiting-on-author
Collaborator
|
Reminder, once the PR becomes ready for a review, use |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
MIR GVN (since #147827) propagates MaybeUninit::uninit() as
const <uninit>in aggregate constructions. Without this fix, codegen would emit a memcpy from an[N x i8] undefglobal for each such field, which LLVM materializes as zero-initialization.This mirrors the existing
all_bytes_uninitskip already present forRvalue::Use(added in #147827) into theRvalue::Aggregatefield loop.Fixes: #157743