No, I mean, I'm going to reboot single user RO, run the btrfs checks, catch the logs, and dump a copy of the partition, so follow ups don't require RO reboots. On Sun, Dec 29, 2019 at 3:25 PM Chris Murphy <lists@xxxxxxxxxxxxxxxxx> wrote: > > On Sun, Dec 29, 2019 at 4:19 PM Patrick Erley <pat-lkml@xxxxxxxxx> wrote: > > > > Nah, this 'corruption' hasn't affected anything other than not being > > able to go to a newer kernel. The system doesn't do anything > > important. I'll grab the logs, and clone an image of the drive so I > > don't need to go single user RO mode to get dumps for devs when follow > > up requests come in. > > A clone of any rw file system is likewise unreliable, for the same > reasons that a forced btrfs check on a rw file system is unreliable. > So unreliable it's a waste of time for you and the developer, so I > wouldn't even bother with it. > > > -- > Chris Murphy
