-
Notifications
You must be signed in to change notification settings - Fork 1.8k
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
panic: VERIFY0(spa_open(dsname, &spa, FTAG)) failed (0 == 2) #16988
Comments
It seems FreeBSD for some reason includes two copies of ZTS, and As I can see, the test tries to destroy the pool during receive. And setting received properties can't handle the error if the pool was destroyed meanwhile. It seems the assertion where it has tripped should be replaces with proper error handling, since obviously it is not an impossible error. |
Do you mean two copies of the test suite? That's because we originally ported it from Solaris to FreeBSD when I worked for Spectra Logic. See FreeBSD commit 2fae26bd8b752cfae083962a152f4b1ee54ada17. But later the ZoL people independently ported it, apparently ignorant of our earlier work. So now there are two test suites, and they are difficult to reconcile. This particular test case is one that I wrote.
+1 |
Yes.
It would be good to port it and whatever else might be in FreeBSD only to OpenZFS upstream. That way it would regularly be executed by OpenZFS CI on both FreeBSD and Linux, and the issue would be spotted long ago. |
Agreed. But it's hard, especially because ZoL made two important design decisions differently than I did:
I put a lot of hours into that test suite at Spectra Logic. And frankly, the fact that ZoL completely overlooked all of my effort without a second thought is very dispiriting to me. That's why I'm not investing any more of my volunteer hours into the test suite. |
System information
Describe the problem you're observing
I can produce this panic semi-reliably by using the ZFS test suite (on 3 out of 4 attempts). I am running FreeBSD 15.0-CURRENT with KMSAN enabled, plus a change that allows me to use the original raidz math routines rather than the SIMD ones (though this panic doesn't look raidz related).
Describe how to reproduce the problem
Run cli_root/zpool_destroy/zpool_destroy_test:zpool_destroy_004_pos in FreeBSD's ZFS test suite.
Include any warning/errors/backtraces from the system logs
The text was updated successfully, but these errors were encountered: