Hi Lukas, On 7/27/22 10:25, Lukas Czerner wrote:
I don't think this is the right fix. We're in ext4_writepages, so at this point I don't think an inode should have any actual inline data in it. If it does it's a bug and the question is how did this get here? The inode is likely corrupted and it should have been noticed earliler and it should never get here.
Yes, that was just an attempt fix something that I'm not quite familiar with. Jan sent already a patch for that fixes it: https://lore.kernel.org/all/20220727155753.13969-1-jack@xxxxxxx/ -- Thanks, Tadeusz
