[Bug 1206287] Kernel BUG fs/ext4/inode.c:3627 and kernel BUG at fs/ext4/super.c:493!
https://bugzilla.suse.com/show_bug.cgi?id=1206287 https://bugzilla.suse.com/show_bug.cgi?id=1206287#c8 --- Comment #8 from Jan Kara <jack@suse.com> --- For example I run Leap 15.3 on my system where I use KVM heavily. For openQA and OBS systems I believe the host OS is either some version of SLE or Leap but I'm not sure. Certainly quite some developers use KVM on their Tumbleweed systems as well. I'm not sure with the problem with VMCI does not trigger for Linux guests but only for Windows guests. Maybe Linux guest isn't using some feature Windows is using? Maybe they are pinning pages for shorter time so chances memory reclaim on host reclaims fs private metadata on the pages is lower? Or maybe the memory footprint of Linux guests is lower so host does not try to reclaim pages so aggressively? Many options here... -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@suse.com