Messages in this thread | | | Date | Wed, 14 Jan 2009 15:37:56 +0100 | From | Jan Kara <> | Subject | Re: ext2 + -osync: not as easy as it seems |
| |
On Wed 14-01-09 09:12:04, Theodore Tso wrote: > On Wed, Jan 14, 2009 at 03:05:32PM +0100, Jan Kara wrote: > > On Wed 14-01-09 08:21:46, Theodore Tso wrote: > > > > > > If we optimize out the journal commit when there are no blocks > > > attached to the transaction, we could change the patch to only force a > > > flush if inode->i_state did not have I_DIRTY before the call to > > > sync_inode(). Does that sound sane? > > Yes. And also add a flush in case of fdatasync(). > > Um, we have that already; the sync_inode() followed by > blkdev_issue_flush() is the path taken by fdatasync(), I do believe. Maybe ext4-patch-queue changes that area but in Linus's tree I see:
if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) goto out;
So if we just overwrite some data, we send them to disk via fdatawrite() and then we quickly bail out from ext4_sync_file() without doing blkdev_issue_flush().
> > Well, I thought that a barrier, as an abstraction, only guarantees that > > any IO which happened before the barrier hits the iron before any IO which > > has been submitted after a barrier. This is actually enough for a > > journalling to work correctly but it's not enough for fsync() guarantees. > > But I might be wrong... > > Ah, yes, that's what you're getting at. True, but for better or for > worse, we have no other interface other than blkdev_issue_flush(). > This will guarantee that the data has made it to the disk controller, > but it won't necessarily guarantee that it will have made it onto the > disk platter, as I understand things; but I don't think we have any > other interfaces available to us at this point. As Jens wrote, it seems barrier guarantees more than I thought so we are correct.
Honza -- Jan Kara <jack@suse.cz> SUSE Labs, CR
| |