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Outdated suggestions cannot be applied.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.
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Ensure
ptr::read
gets all the same LLVMload
metadata that dereferencing does #109035Ensure
ptr::read
gets all the same LLVMload
metadata that dereferencing does #109035Changes from 4 commits
b2c717f
0b96fee
1f70bb8
87696fd
e7c6ad8
dfc3377
File filter
Filter by extension
Conversations
Jump to
There are no files selected for viewing
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This confused me a bit — at first I though that
read_via_copy
only works for pointers to locals; While it seems like actually it can only be called with a pointer which is itself a local (i.e.read_via_copy(x)
✅,read_via_copy(s.f)
❎).Maybe the docs can be clarified a bit.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This seems like a strange (and very syntactic) restriction? Isn't there a high risk that some other program transformation might, for instance, turn
into
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That can be fixed via introducing a temporary if it becomes a problem (and Jakob mentioned that on... Zulip?), but it seems the current tendency in the MIR is to aggressively desugar everything and introduce temporaries everywhere, then roll them back up in opt passes. If the implementation works and produces less MIR than the alternative, I think there's a merit in not introducing One More Temporary for the crab to claw through.
Maybe we should note why this "bug" was not "fixed", though, so that if anyone comes by and it needs to be fixed, they can immediately change it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, this is about as big an optimization footgun as can exist. What saves us is that this runs before optimizations, so they don't have to care
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Still seems rather fragile, and needs at least a comment explaining the situation.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
To show my work for these numbers,
With this PR:
Previously:
That's of course not a semantic problem, as both work fine, but I think it makes a nice demonstration of how this ends up being practically useful, especially in conjunction with MIR inlining.