Ignore NLL boring locals in polonius diagnostics #136299
Merged
+376
−152
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
rustbot
added
A-tidy
This comment was marked as resolved.
This comment was marked as resolved.
- describe how that data flows during borrowck - prepares for recording some liveness data for diagnostics, not just for the main analysis
this is used in diagnostics to focus on relevant live locals to match NLL diagnostics
also add explicit revisions for -Zpolonius=next
The blessed expectations were recently removed because they were only checked via the compare-mode. This switches to explicit revisions to ensure it doesn't happen again. - `assignment-to-differing-field` - `polonius-smoke-test` - `subset-relations`
- it still mentions AST borrowck - it tracks errors that shouldn't exist and are fixed by polonius
lqd
force-pushed
the
polonius-next-episode-9
branch
from
c19a34e to
a616fa5
Compare
18 tasks
jackh726
approved these changes
Feb 3, 2025
|
@bors r+ |
bors
added
S-waiting-on-bors
matthiaskrgr
added a commit
to matthiaskrgr/rust
that referenced
this pull request
Feb 3, 2025
…kh726 Ignore NLL boring locals in polonius diagnostics Another easy one `@jackh726` (the diff is inflated by blessed test expectations don't worry :) NLLs don't compute liveness for boring locals, and therefore cannot find them in causes explaining borrows. In polonius, we don't have this liveness optimization (we may be able to do something partially similar in the future, e.g. for function parameters and the like), so we do encounter these in diagnostics even though we don't want to. This PR: - restructures the polonius context into per-phase data, in spirit as you requested in an earlier review - stores the locals NLLs would consider boring into the errors/diagnostics data - ignores these if a boring local is found when trying to explain borrows This PR fixes around 80 cases of diagnostics differences between `-Zpolonius=next` and NLLs. I've also added explicit revisions to a few polonius tests (both for the in-tree implementation as well as the datalog implementation -- even if we'll eventually remove them). I didn't do this for all the "dead" expectations that were removed from rust-lang#136112 for that same reason, it's fine. I'll soon/eventually add explicit revisions where they're needed: there's only a handful of tests left to fix. r? `@jackh726`
This was referenced Feb 3, 2025
bors
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Feb 4, 2025
…iaskrgr Rollup of 8 pull requests Successful merges: - rust-lang#136289 (OnceCell & OnceLock docs: Using (un)initialized consistently) - rust-lang#136299 (Ignore NLL boring locals in polonius diagnostics) - rust-lang#136411 (Omit argument names from function pointers that do not have argument names) - rust-lang#136430 (Use the type-level constant value `ty::Value` where needed) - rust-lang#136476 (Remove generic `//@ ignore-{wasm,wasm32,emscripten}` in tests) - rust-lang#136484 (Notes on types/traits used for in-memory query caching) - rust-lang#136493 (platform-support: document CPU baseline for x86-32 targets) - rust-lang#136498 (Update books) r? `@ghost` `@rustbot` modify labels: rollup
rust-timer
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Feb 4, 2025
Rollup merge of rust-lang#136299 - lqd:polonius-next-episode-9, r=jackh726 Ignore NLL boring locals in polonius diagnostics Another easy one ``@jackh726`` (the diff is inflated by blessed test expectations don't worry :) NLLs don't compute liveness for boring locals, and therefore cannot find them in causes explaining borrows. In polonius, we don't have this liveness optimization (we may be able to do something partially similar in the future, e.g. for function parameters and the like), so we do encounter these in diagnostics even though we don't want to. This PR: - restructures the polonius context into per-phase data, in spirit as you requested in an earlier review - stores the locals NLLs would consider boring into the errors/diagnostics data - ignores these if a boring local is found when trying to explain borrows This PR fixes around 80 cases of diagnostics differences between `-Zpolonius=next` and NLLs. I've also added explicit revisions to a few polonius tests (both for the in-tree implementation as well as the datalog implementation -- even if we'll eventually remove them). I didn't do this for all the "dead" expectations that were removed from rust-lang#136112 for that same reason, it's fine. I'll soon/eventually add explicit revisions where they're needed: there's only a handful of tests left to fix. r? ``@jackh726``
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.
Another easy one @jackh726 (the diff is inflated by blessed test expectations don't worry :)
NLLs don't compute liveness for boring locals, and therefore cannot find them in causes explaining borrows. In polonius, we don't have this liveness optimization (we may be able to do something partially similar in the future, e.g. for function parameters and the like), so we do encounter these in diagnostics even though we don't want to. This PR:
This PR fixes around 80 cases of diagnostics differences between
-Zpolonius=nextand NLLs. I've also added explicit revisions to a few polonius tests (both for the in-tree implementation as well as the datalog implementation -- even if we'll eventually remove them). I didn't do this for all the "dead" expectations that were removed from #136112 for that same reason, it's fine. I'll soon/eventually add explicit revisions where they're needed: there's only a handful of tests left to fix.r? @jackh726