Ok, I understand. And is there a way to disable them during filesystem creation (e.g. some option in mkfs.btrfs)? Thanks, Dejan -----Original Message----- From: Qu Wenruo <quwenruo.btrfs@xxxxxxx> Sent: utorak, 16. jun 2020. 15:45 To: Rebraca Dejan (BSOT/PJ-ES1-Bg) <Dejan.Rebraca@xxxxxxxxxxxx>; linux-btrfs@xxxxxxxxxxxxxxx Subject: Re: Having troubles to disable inline extents On 2020/6/16 下午9:41, Rebraca Dejan (BSOT/PJ-ES1-Bg) wrote: > Hi, > > We are trying to add support for BTRFS in our project, so we started to examine this filesystem. > For the moment, we don't want inline extents for our tests, but we have difficulties to turn them off. I'm using 'max_inline=0' mount option to disable them, but I still see them for small files (< 50 Bytes) using FS_IOC_FIEMAP ioctl. Kernel log when executing mount: max_inline only affects new writes. So existing inlined extent won't be affected. You need to defrag such small files to convert them back to regular extents. Thanks, Qu > [11051.642976] BTRFS info (device loop0): max_inline at 0 > [11051.642978] BTRFS info (device loop0): disk space caching is > enabled [11051.642979] BTRFS info (device loop0): has skinny extents > > Environment: > - 4.15.0-96-generic #97~16.04.1-Ubuntu SMP Wed Apr 1 03:03:31 UTC 2020 > x86_64 x86_64 x86_64 GNU/Linux > - btrfs-progs v4.4 > > I would really appreciate your support on this. > Tnx. > > Best regards, > > Dejan Rebraca > BSOT/PJ-ES1-Bg >
