[Bug 276985] crash in LinuxKPI/drm
- In reply to: bugzilla-noreply_a_freebsd.org: "[Bug 276985] Crash in scheduler __curthread"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sat, 07 Sep 2024 21:07:09 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276985 --- Comment #62 from Tomasz "CeDeROM" CEDRO <tomek@cedro.info> --- after short consideration i will stick to 13.3 on my production. i really like to play but i need to work on that particular machine. there is no guarantee a hardware upgrade will remove all of the issues not introducing new ones and i have still some ports here unavailable on modern motherboards. i will keep the additional drive for testing and switch os when there is something to test. my main concern is serious filesystem damage during all those kernel panics: * i did use separate disk and UFS install not to touch production data on ZFS drives. also to test impact on filesystem. * each kernel panic caused UFS corruption. * even though UFS SU+J was not detected on next boot it finally got to point where init asked me to fsck by hand. in that case i had to run `fsck -fy /` several times until no problem was found (!!). then i could not run xorg or reinstall previously compiled drm ports because filesystem was corrupted anyway! * frequent power on/off also impacts hdd lifetime and condition and i have 6 hdd zraid. * it is not a problem for system packages that i can reinstall with pkg but a serious problem for config files or (NN TB of) data files that cannot be that easily restored. * i remember last time i upgraded to 14.0 and had frequent kernel panics from drm and then rolled back to 13.2 just to make sure no damage is done i noticed zfs corruption on big files reported by zpool status. those errors could not be fixed (scrub / resilver) and i had to delete them. luckily i did not upgrade zpool because there would be no way back. -- You are receiving this mail because: You are the assignee for the bug.