-
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
strange CI error running gc code #53026
Comments
More apparent examples of GC issues with this test
https://buildkite.com/julialang/julia-master/builds/33000#018d64bc-df85-41f7-bf7d-cfa160941109 |
I started doing a bisect on this, but cycle time is fairly slow since this takes a while to reproduce in the GC tests. I plan to post an update here when I'm done bisecting. |
If helpful, there is also an rr trace of this in #52757 |
Unable to reproduce it after leaving it running for a few hundred times in a loop... Will look at the rr trace. |
This aims to slightly simplify the synchronization by making `n_threads_marking` the sole memory location of relevance for it, it also removes the fast path, because being protected by the lock is quite important so that the observed gc state arrays are valid. Fixes: #53350 Fixes: #52757 Maybe fixes: #53026 Co-authored-by: Jameson Nash <vtjnash@gmail.com> (cherry picked from commit a96726b)
This aims to slightly simplify the synchronization by making `n_threads_marking` the sole memory location of relevance for it, it also removes the fast path, because being protected by the lock is quite important so that the observed gc state arrays are valid. Fixes: JuliaLang#53350 Fixes: JuliaLang#52757 Maybe fixes: JuliaLang#53026 Co-authored-by: Jameson Nash <vtjnash@gmail.com>
This aims to slightly simplify the synchronization by making `n_threads_marking` the sole memory location of relevance for it, it also removes the fast path, because being protected by the lock is quite important so that the observed gc state arrays are valid. Fixes: JuliaLang#53350 Fixes: JuliaLang#52757 Maybe fixes: JuliaLang#53026 Co-authored-by: Jameson Nash <vtjnash@gmail.com>
This seems it clearly shouldn't happen, as the Matrix is thread-local and the size is not changed, so there is not much way the length of the Memory inside should be corrupted, but yet somehow, we see this failure:
https://buildkite.com/julialang/julia-master/builds/32516#018d2921-a653-440d-aefb-55eb83d5595c
The text was updated successfully, but these errors were encountered: