lkml.org 
[lkml]   [2006]   [Oct]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] Freeze bdevs when freezing processes.
On Tue, Oct 24, 2006 at 05:29:59PM +0200, Rafael J. Wysocki wrote:
> On Tuesday, 24 October 2006 16:44, David Chinner wrote:
> > On Mon, Oct 23, 2006 at 12:36:53PM +0200, Rafael J. Wysocki wrote:
> > > On Monday, 23 October 2006 06:12, Nigel Cunningham wrote:
> > > > XFS can continue to submit I/O from a timer routine, even after
> > > > freezeable kernel and userspace threads are frozen. This doesn't seem to
> > > > be an issue for current swsusp code,
> > >
> > > So it doesn't look like we need the patch _now_.
> > >
> > > > but is definitely an issue for Suspend2, where the pages being written could
> > > > be overwritten by Suspend2's atomic copy.
> > >
> > > And IMO that's a good reason why we shouldn't use RCU pages for storing the
> > > image. XFS is one known example that breaks things if we do so and
> > > there may be more such things that we don't know of. The fact that they
> > > haven't appeared in testing so far doesn't mean they don't exist and
> > > moreover some things like that may appear in the future.
> >
> > Could you please tell us which XFS bits are broken so we can get
> > them fixed? The XFS daemons should all be checking if they are
> > supposed to freeze (i.e. they call try_to_freeze() after they wake
> > up due to timer expiry) so I thought they were doing the right
> > thing.
> >
> > However, I have to say that I agree with freezing the filesystems
> > before suspend - at least XFS will be in a consistent state that can
> > be recovered from without corruption if your machine fails to
> > resume....
>
> Do you mean calling sys_sync() after the userspace has been frozen
> may not be sufficient?

In most cases it probably is, but sys_sync() doesn't provide any
guarantees that the filesystem is not being used or written to after
it completes. Given that every so often I hear about an XFS filesystem
that was corrupted by suspend, I don't think this is sufficient...

Cheers,

Dave.
--
Dave Chinner
Principal Engineer
SGI Australian Software Group
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2006-10-24 18:39    [W:0.096 / U:10.572 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site