linker: Never use whole-archive linking unless explicitly requested #85144
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
|
r? @jackh726 (rust-highfive has picked a reviewer for you, use r? to override) |
rust-highfive
added
the
S-waiting-on-review
This comment has been minimized.
This comment has been minimized.
|
Interesting, I haven't observed any errors on a full |
petrochenkov
added
S-waiting-on-author
JohnCSimon
added
S-waiting-on-author
This comment has been minimized.
This comment has been minimized.
|
FWIW LLD shows even more missing symbols: |
|
@petrochenkov I'm not a good reviewer for this. Any idea who might be better? |
|
r? @bjorn3 I guess I obviously made some wrong assumptions about dylibs and I need to figure out what exactly I got wrong, but I wasn't able to get to this PR since May for prioritization reasons, so this PR is pretty far from needing a review right now. |
JohnCSimon
added
S-waiting-on-author
|
The job Click to see the possible cause of the failure (guessed by this bot) |
JohnCSimon
added
S-inactive
JohnCSimon
removed
the
S-inactive
|
☔ The latest upstream changes (presumably #88227) made this pull request unmergeable. Please resolve the merge conflicts. |
|
I'm going to close this for now. This PR contains two orthogonal changes:
|
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Outdated suggestions cannot be applied.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.
Reminder: "whole-archive" linking for a static library means including all the object files from the library into the output binary instead only those that are needed ("needed" is determined by the current list of unresolved symbols).
We have two automatic uses of whole-archive right now.
The use of whole-archive in this case is pretty random and inconsistent, and also entirely unnecessary in regular cases.
Moreover, when whole-archive was properly implemented for MSVC targets in linker: MSVC supports linking static libraries as a whole archive #72785 it caused regressions (see links to other issues from that PR).
The comment tells that it's done because "
dylibcan be reused as an intermediate artifact", perhaps it was necessary in the past, but now it shouldn't be necessary because everything usable from a dylib should be already kept alive by export lists (fn export_symbols).I wanted to land this sooner in the release cycle in case I'm missing something and there are regressions.
However, it may also make sense to land this together with stabilizing the
+whole-archivelinking modifier (#81490) to make sure there's a fix available on stable in cases wherewhole-archivewas actually required in one these specific situations (which should be a rare case).