inline asm!: ban OpReturn/OpReturnValue (they're always UB).
#1006
Conversation
This file contains 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
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Outdated suggestions cannot be applied.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.
This PR fixes #1002 by banning the misuse of inline
asm!(see the issue for more details).We used to get away with the UB, but now MIR inlining is wreaking havoc (e.g. inlining
asm!("ret")will return from the caller it was inlined into - one of the many ways in which the UB can manifest).Most of this PR is actually making progress towards:
MaybeUninit#981However, I'm not considering #981 as being fixed just yet, because the original usecase involves arrays, which have their own complications, and I focused on only opaque handles (as
AccelerationStructureconstructors need it).I've also not changed any unaffected use of
asm!, leaving the migration toMaybeUninitfor an issue:asm!should useMaybeUninit<T>instead oflet mut result = T::default();. #1007Each commit should be reviewed separately (or maybe I should've split this into two PRs?).