Prevent race condition in genpy via named mutex #2118
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
Chronial
force-pushed
the
feature/genpy_mutex
branch
from
418fe20
to
40ab8de
Compare
Chronial
force-pushed
the
feature/genpy_mutex
branch
from
40ab8de
to
efb3e43
Compare
|
@Chronial your PR currently has conflicts |
|
👍 resolved |
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Outdated suggestions cannot be applied.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.
As described in its comment, #1601 did not fully fix the problems of #1600. It massively reduced the problem and the duration of the race condition, but did not remove it.
Opening many instances of the same com server in parallel can still cause exceptions in finish_writer. This is caused by two things:
os.renamebetween ouros.unlinkand ouros.rename, causing our rename to fail. This could be fixed by simply replacing theos.renamewithos.replace, but then:os.replace, causing the replace to fail.The only proper solution for this issue I could think of is to use a named mutex to force mutual exclusion between all the operations that might interfere with each other. This solution is implemented in this PR.