Skip pub structs with repr(c) and repr(transparent) in dead code analysis#127104
Skip pub structs with repr(c) and repr(transparent) in dead code analysis#127104mu001999 wants to merge 1 commit intorust-lang:masterfrom
Conversation
|
I think any changes to lints should go through T-lang. Sorry for taking long to review this. |
|
This came up on Zulip https://rust-lang.zulipchat.com/#narrow/stream/122651-general/topic/Status.20of.20fix.20for.20dead_code.20warnings.20on.20repr.28C.29.20structs. It seems rather unfortunate that large bindings crates are needing to If this proposed solution is decided not to be acceptable, unfortunately we should probably consider temporarily reverting or adjusting part of #125572. One suggestion is to emit a new lint within Unfortunately this behavior just branched, so: @rustbot label +beta-nominated |
|
This PR has been introduced to address a diagnostics regression. The original PR that introduced the behavior this change was meant to prevent regressing can no longer be cleanly reverted on beta. This is thus now the most minimal diff if we want to apply the change to beta, even if it means we introduce new code. |
|
☔ The latest upstream changes (presumably #128222) made this pull request unmergeable. Please resolve the merge conflicts. |
|
I think that all of this new dead code analysis landed recently should probably be reverted and reworked, since then we're not pressured to put up hacky fixes in favor of getting it correct and well documented the first time 👍 |
|
☔ The latest upstream changes (presumably #128404) made this pull request unmergeable. Please resolve the merge conflicts. |
|
Dropping beta-nominated, looks like the original PR was reverted on beta 1.81 (via #128606). |
Fixes #126169