Re: [TuxOnIce-devel] Latest updates.

Re: [TuxOnIce-devel] Latest updates.

[Nathan Scott]

----- "Martin Steigerwald" <Martin@lichtvoll.de> wrote:

> Please keep CC to TuxOnIce devel list.
> 
> 
> Hi!
> 
> Nigel, who develops the alternative TuxOnIce hibernation / snapshot 
> infrastructure, implemented checking last mount time of filesystem as
> a 
> safety feature for Ext 2, 3 and 4.
> 
> He would like this also for XFS. Does XFS record the last mount time 
> somewhere?

Nope.

> If not, could this be added?

Anythings possible ... its a relatively trivial change, and there's room
in the superblock.  Could probably even be done without a superblock version
bump - if you were prepared to live with the possibility of finding zero in
the ondisk location & dealing with it appropriately (i.e. last mount time
unknown).

Guess you'd have to convince the XFS folks doing most of the work these
days, but really it'd be quite trivial to implement.  xfs_mount_log_sb()
would be the place to start looking...

cheers.

-- 
Nathan

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [TuxOnIce-devel] Latest updates.

[Dave Chinner]

On Wed, Jan 13, 2010 at 02:09:14PM +1100, Nathan Scott wrote:
> 
> ----- "Martin Steigerwald" <Martin@lichtvoll.de> wrote:
> 
> > Please keep CC to TuxOnIce devel list.
> > 
> > 
> > Hi!
> > 
> > Nigel, who develops the alternative TuxOnIce hibernation / snapshot 
> > infrastructure, implemented checking last mount time of filesystem as
> > a 
> > safety feature for Ext 2, 3 and 4.
> > 
> > He would like this also for XFS. Does XFS record the last mount time 
> > somewhere?
> 
> Nope.
> 
> > If not, could this be added?
> 
> Anythings possible ... its a relatively trivial change, and there's room
> in the superblock.  Could probably even be done without a superblock version
> bump - if you were prepared to live with the possibility of finding zero in
> the ondisk location & dealing with it appropriately (i.e. last mount time
> unknown).
> 
> Guess you'd have to convince the XFS folks doing most of the work these
> days, but really it'd be quite trivial to implement.  xfs_mount_log_sb()
> would be the place to start looking...

Agreed that it is trivial to implement but there are still some
definite traps - like the fact the sb change transaction may be logged
immediately but the physical superblock may not get written for some
time after the mount.

Really it comes down to the interface used to read the mount time
fromt eh superblock at suspend and resume - can anyone tell me what
that is or point me at the code?

Personally I don't see why this needs to be in the superblock - an
extended attribute on the root directory of the root filesystem that is
written:

	a) by a startup script at boot time;
	b) by the suspend code;
	c) copied into the suspend image; and
	d) read and verified during resume

seems sufficient to me to provide what you want. This has the
benefit of being completely filesystem independent and doesn't
require on-disk format changes for every filesystem that ppl want
TuxOnIce to supprt.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [TuxOnIce-devel] Latest updates.

[Nathan Scott]

[Looks like tuxonice is a subscriber-only list...]

----- "Dave Chinner" <david@fromorbit.com> wrote:

> On Wed, Jan 13, 2010 at 02:09:14PM +1100, Nathan Scott wrote:
> > > 
> Agreed that it is trivial to implement but there are still some
> definite traps - like the fact the sb change transaction may be
> logged
> immediately but the physical superblock may not get written for some
> time after the mount.

*nod* ... it will be written when unmounted though...

> 
> Really it comes down to the interface used to read the mount time
> fromt eh superblock at suspend and resume - can anyone tell me what
> that is or point me at the code?

(not me!)

> Personally I don't see why this needs to be in the superblock - an
> extended attribute on the root directory of the root filesystem that
> is written:

Yeah, that was my first thought too, but I suspect they want to be
able to query this when the filesystem is unmounted; in which case
the interface is probably "read a fs-specific magic spot ondisk".
For this not-mounted query, the EA route would be incredibly painful.

cheers.

-- 
Nathan

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [TuxOnIce-devel] Latest updates.

[Dave Chinner]

On Wed, Jan 13, 2010 at 05:05:18PM +1100, Nathan Scott wrote:
> [Looks like tuxonice is a subscriber-only list...]
> 
> ----- "Dave Chinner" <david@fromorbit.com> wrote:
> 
> > On Wed, Jan 13, 2010 at 02:09:14PM +1100, Nathan Scott wrote:
> > > > 
> > Agreed that it is trivial to implement but there are still some
> > definite traps - like the fact the sb change transaction may be
> > logged
> > immediately but the physical superblock may not get written for some
> > time after the mount.
> 
> *nod* ... it will be written when unmounted though...

suspend in the kernel doesn't unmount filesystems. I have no idea
what tuxonice does these days, but last time I heard it left them
mounted but frozen over the suspend/resume cycle.

> > Really it comes down to the interface used to read the mount time
> > fromt eh superblock at suspend and resume - can anyone tell me what
> > that is or point me at the code?
> 
> (not me!)
> 
> > Personally I don't see why this needs to be in the superblock - an
> > extended attribute on the root directory of the root filesystem that
> > is written:
> 
> Yeah, that was my first thought too, but I suspect they want to be
> able to query this when the filesystem is unmounted; in which case
> the interface is probably "read a fs-specific magic spot ondisk".
> For this not-mounted query, the EA route would be incredibly painful.

I know, but it's the "read from the magic spot on disk" approach
that I'm most concerned about. It's so easy to implement, but oh so
easy get wrong because block device access is not coherent with the
filesystem. Not to mention that if we change the format of the
"magic spot" then the application breaks, whereas that will not
happen if access is via extended attributes.

Essentially, I don't want to see another boot related, filesystem
interacting application having random problems with XFS and then
doing stupid stuff to try to work around problems caused by
fundamentally broken assumptions made by the application. IMO,
learning from the mistakes the world's most screwed up bootlader
(grub) made is a good idea....

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [TuxOnIce-devel] Latest updates.

[Dave Chinner]

On Wed, Jan 13, 2010 at 05:57:40PM +1100, Dave Chinner wrote:
> On Wed, Jan 13, 2010 at 05:05:18PM +1100, Nathan Scott wrote:
> > [Looks like tuxonice is a subscriber-only list...]
> > 
> > ----- "Dave Chinner" <david@fromorbit.com> wrote:
> > 
> > > On Wed, Jan 13, 2010 at 02:09:14PM +1100, Nathan Scott wrote:
> > > > > 
> > > Agreed that it is trivial to implement but there are still some
> > > definite traps - like the fact the sb change transaction may be
> > > logged
> > > immediately but the physical superblock may not get written for some
> > > time after the mount.
> > 
> > *nod* ... it will be written when unmounted though...
> 
> suspend in the kernel doesn't unmount filesystems. I have no idea
> what tuxonice does these days, but last time I heard it left them
> mounted but frozen over the suspend/resume cycle.

I found the code - it doesn't unmount filesystems. The git tree is
here:

http://git.kernel.org/?p=linux/kernel/git/nigelc/tuxonice-head.git;a=summary

It appears that it walks the list of superblocks to grab the block
device off each mounted superblock to do it's work, so I don't see
any fundamental problem with using an attribute hanging off
sb->s_root to hold the boot time...

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [TuxOnIce-devel] Latest updates.

[Nigel Cunningham]

Hi.

Dave Chinner wrote:
> On Wed, Jan 13, 2010 at 05:57:40PM +1100, Dave Chinner wrote:
>> On Wed, Jan 13, 2010 at 05:05:18PM +1100, Nathan Scott wrote:
>>> [Looks like tuxonice is a subscriber-only list...]
>>>
>>> ----- "Dave Chinner" <david@fromorbit.com> wrote:
>>>
>>>> On Wed, Jan 13, 2010 at 02:09:14PM +1100, Nathan Scott wrote:
>>>> Agreed that it is trivial to implement but there are still some
>>>> definite traps - like the fact the sb change transaction may be
>>>> logged
>>>> immediately but the physical superblock may not get written for some
>>>> time after the mount.
>>> *nod* ... it will be written when unmounted though...
>> suspend in the kernel doesn't unmount filesystems. I have no idea
>> what tuxonice does these days, but last time I heard it left them
>> mounted but frozen over the suspend/resume cycle.
> 
> I found the code - it doesn't unmount filesystems. The git tree is
> here:
> 
> http://git.kernel.org/?p=linux/kernel/git/nigelc/tuxonice-head.git;a=summary
> 
> It appears that it walks the list of superblocks to grab the block
> device off each mounted superblock to do it's work, so I don't see
> any fundamental problem with using an attribute hanging off
> sb->s_root to hold the boot time...

Yeah. The idea is to have very simple code that looks at a fixed number
of bytes at a known offset into the partition. The value of the bytes is
stored in the image header, and compared prior to starting to resume. If
the data is different, we assume the filesystem has been mounted since
the image was written and it's unsafe to resume. It doesn't _need_ to be
a time - just something that's guaranteed to have changed if the
filesystem gets mounted afresh.

By the way, sorry for the slow replies so far and for slow ones in the
coming 24 hours. I'm on holidays and my father-in-law's funeral is tomorrow.

Nigel

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs