coverage: Infer instances_used from pgo_func_name_var_map
#144530
Merged
Conversation
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
In obscure circumstances, we would sometimes emit a covfun record for a
function with no physical coverage counters, causing `llvm-cov` to fail with
the cryptic error message:
malformed instrumentation profile data: function name is empty
We can eliminate this mismatch by removing `instances_used` entirely, and
instead inferring its contents from the keys of `pgo_func_name_var_map`.
This makes it impossible for a "used" function to lack a PGO name entry.
Zalathar
added
the
A-code-coverage
|
r? @SparrowLii rustbot has assigned @SparrowLii. Use |
rustbot
added
A-LLVM
|
r? lqd This change looks good to me. It'd be better to have a test but we don't necessarily need it in this PR, as getting an MCVE from #141577 is not easy. Reducing the issue seems doable to me, but the current repro involving fuzzing is still quite involved, and it seems it'd be painful to produce a minimization in this current state. We should have such a test before relanding #144298 though, so the issue doesn't reappear a 3rd time. With that, @bors r+ |
bors
added
S-waiting-on-bors
bors
added a commit
that referenced
this pull request
Jul 28, 2025
Rollup of 7 pull requests Successful merges: - #144072 (update `Atomic*::from_ptr` and `Atomic*::as_ptr` docs) - #144151 (`tests/ui/issues/`: The Issues Strike Back [1/N]) - #144300 (Clippy fixes for miropt-test-tools) - #144399 (Add a ratchet for moving all standard library tests to separate packages) - #144472 (str: Mark unstable `round_char_boundary` feature functions as const) - #144503 (Various refactors to the codegen coordinator code (part 3)) - #144530 (coverage: Infer `instances_used` from `pgo_func_name_var_map`) r? `@ghost` `@rustbot` modify labels: rollup
rust-timer
added a commit
that referenced
this pull request
Jul 28, 2025
Rollup merge of #144530 - Zalathar:instances-used, r=lqd coverage: Infer `instances_used` from `pgo_func_name_var_map` In obscure circumstances involving macro-expanded spans, we would sometimes emit a covfun record for a function with no physical coverage counters, and therefore no corresponding entry in the “PGO names” section of the binary. The absence of that name entry causes `llvm-cov` to fail with the cryptic error message: ```text malformed instrumentation profile data: function name is empty ``` We can eliminate this mismatch by removing `instances_used` entirely, and instead inferring its contents from the keys of `pgo_func_name_var_map`. This makes it impossible for a "used" function to lack a PGO name entry. --- This is an attempt to eliminate the cause of #141577 when re-landing changes like #144298 in the future. I haven't been able to reproduce the underlying issue in an in-tree test, because the only known repro involves a non-trivial derive proc-macro that relies on `syn` and `proc-macro2`. But I have manually verified in a separate branch that this change would have prevented the reoccurrence of #141577 (comment).
Zalathar
added a commit
to Zalathar/rust
that referenced
this pull request
Jul 29, 2025
coverage: Regression test for "function name is empty" bug Regression test for rust-lang#141577, which was triggered by rust-lang#144298. The bug was triggered by a particular usage of the `?` try operator in a proc-macro expansion. Thanks to lqd for the minimization at rust-lang#144571 (comment). --- I have manually verified that reverting the relevant follow-up fixes (rust-lang#144480 and rust-lang#144530) causes this test to reproduce the bug: ```sh git revert -m1 8aa3d41 c462895 ``` --- r? compiler
Zalathar
added a commit
to Zalathar/rust
that referenced
this pull request
Jul 29, 2025
coverage: Regression test for "function name is empty" bug Regression test for rust-lang#141577, which was triggered by rust-lang#144298. The bug was triggered by a particular usage of the `?` try operator in a proc-macro expansion. Thanks to lqd for the minimization at rust-lang#144571 (comment). --- I have manually verified that reverting the relevant follow-up fixes (rust-lang#144480 and rust-lang#144530) causes this test to reproduce the bug: ```sh git revert -m1 8aa3d41 c462895 ``` --- r? compiler
Zalathar
added a commit
to Zalathar/rust
that referenced
this pull request
Jul 29, 2025
coverage: Regression test for "function name is empty" bug Regression test for rust-lang#141577, which was triggered by rust-lang#144298. The bug was triggered by a particular usage of the `?` try operator in a proc-macro expansion. Thanks to lqd for the minimization at rust-lang#144571 (comment). --- I have manually verified that reverting the relevant follow-up fixes (rust-lang#144480 and rust-lang#144530) causes this test to reproduce the bug: ```sh git revert -m1 8aa3d41 c462895 ``` --- r? compiler
rust-timer
added a commit
that referenced
this pull request
Jul 29, 2025
Rollup merge of #144616 - Zalathar:try-in-macro, r=jieyouxu coverage: Regression test for "function name is empty" bug Regression test for #141577, which was triggered by #144298. The bug was triggered by a particular usage of the `?` try operator in a proc-macro expansion. Thanks to lqd for the minimization at #144571 (comment). --- I have manually verified that reverting the relevant follow-up fixes (#144480 and #144530) causes this test to reproduce the bug: ```sh git revert -m1 8aa3d41 c462895 ``` --- r? compiler
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.
In obscure circumstances involving macro-expanded spans, we would sometimes emit a covfun record for a function with no physical coverage counters, and therefore no corresponding entry in the “PGO names” section of the binary. The absence of that name entry causes
llvm-covto fail with the cryptic error message:We can eliminate this mismatch by removing
instances_usedentirely, and instead inferring its contents from the keys ofpgo_func_name_var_map.This makes it impossible for a "used" function to lack a PGO name entry.
This is an attempt to eliminate the cause of #141577 when re-landing changes like #144298 in the future.
I haven't been able to reproduce the underlying issue in an in-tree test, because the only known repro involves a non-trivial derive proc-macro that relies on
synandproc-macro2. But I have manually verified in a separate branch that this change would have prevented the reoccurrence of #141577 (comment).