-
-
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
Occasional segfaults when running with @threads #44019
Comments
Did you try |
I had not. On trying, it appears to fail an assertion and crash: Output of `julia --bug-report=rr-local segfaulting_file.jl`
Interesting to note that this is a different assertion than the one that trips when I try to run |
Note sure how precise valgrind is in this case, but it consistently points to this location, looks like (at least) a null pointer.
|
I've gotten a working version of A packed tarball of the trace can be found at https://www.dropbox.com/s/yv3zuvy7ojod7nb/gh-julia-44019-rr-1.tar.xz?dl=0 Console output when running `rr`
|
v1.7.1 is a very old release to be relying on threading. You will need to try again with a newer copy of julia (preferably nightly) and possibly try updating |
Can confirm that this appears to have been solved by the upgrade to 1.7.2. |
I spoke too soon. The segfault takes longer to occur than it used to, but still occurs, both on 1.7.2 and nightly. Should I open a new issue? |
Executive summary: I occasionally get segfaults when running an
@threads
loop in Julia 1.7.1. It is a Heisenbug, and I'm still trying to getrr
working on my system.I've been trying to pare down the example code, but this is definitely a heisenbug and it often takes me several tries to confirm that the program still segfaults with a set of changes, so progress has been slow. I also have not been able to get an rr trace since
rr
refuses to run at all on my system at the moment, though I'll work towards getting that working in my spare time.Running on an Arch system (confirmed that this still happens after a reboot w/ no updates), julia installed from the
julia-bin
package which downloads and unpacks the official binaries.I see segfaults with 32 threads, they're much rarer with just 16 but can still occur after a very long time.
Current Code
I will continue to pare this down to get as minimal an example as possible.Output of `versioninfo()`
Stack trace from the one time in many that the program managed to print one before dying
The line that originates in my code (line 65 at the top) is just the
@threads
loop.Let me know if any other information would be helpful!
The text was updated successfully, but these errors were encountered: