Does ceph_write_iter() actually need to drop the inode I/O lock in order to handle EOLDSNAPC? I'm wondering if I can deal with it in the netfs request retry code - but that means dealing with it whilst the I/O lock is held. David
Does ceph_write_iter() actually need to drop the inode I/O lock in order to handle EOLDSNAPC? I'm wondering if I can deal with it in the netfs request retry code - but that means dealing with it whilst the I/O lock is held. David
![]() |