On Wed 25-02-15 09:52:58, Ted Tso wrote: > On Tue, Feb 24, 2015 at 10:58:33AM -0800, Linus Torvalds wrote: > > On Tue, Feb 24, 2015 at 10:31 AM, Jan Kara <jack@xxxxxxx> wrote: > > > > > > any reaction here? > > > > Ted? Al? Should we just revert or at least disable the 'lazytime' option? > > Sorry, I've been recovering from conference travel. Lazytime is not > enabled by default, and except for ext4 it can't be enabled without > changes to the mount command. I'll take a look at it in the next > couple of days, and in the worst case we can disable it in > fs/ext4/super.c relatively simple. Yeah, that sounds reasonable. I've been thinking how to fix those time ordering issues and sadly it isn't trivial. We'll likely need a timestamp in the inode (although a coarse one is enough) remembering when inode was last written (i_dirtied_when needs to remember when dirty *data* were created so we cannot use it as I originally thought). And we'll need to sort inodes back into the list of inodes with dirty timestamps. It could be done in O(length of list + number of inodes added) if we are careful but it will be non-trivial. Alternatively we could decide that once per X hours we just walk the whole list of inodes with unwritten timestamps and write all inodes there. That's easy to implement but the downside is that we may generate peaks in inode writeback instead of slowly trickling inodes with dirty timestamps to disk. Honza -- Jan Kara <jack@xxxxxxx> SUSE Labs, CR -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
