Hi,
On Mon, Apr 10, 2000 at 01:55:46PM +0200, Jamie Lokier wrote:
> I suppose simply making sure the device request queue cannot go empty as
> long as there are dirty blocks would be equivalent. Now that we have
> device request queues :-)
Partially --- though Linus is vehemently against adding any extra
semantics to the existing buffer cache.
>
> > It even defers completion of the close() of the last
> > writer until the sync has completed, so that if you copy anything to the
> > device, you can be sure that the media is synced up once the copy
> > completes.
>
> Now this is a nice feature.
It's even better: because it knows that the media has no more writers,
once the write completes supermount can remount the fs as readonly. So,
if you have an ext2 filesystem, once your copy to the disk has completed,
the disk can be safely ejected and you won't get any warnings about
needing fsck the next time you mount it.
--Stephen
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
