[dm-devel] Re: [PATCH v9] dm snapshot: allow live exception store handover between tables

[Mike Snitzer]

On Fri, Nov 20 2009 at  1:00pm -0500,
Mikulas Patocka <mpatocka at redhat.com> wrote:

> 
> 
> On Thu, 19 Nov 2009, Mike Snitzer wrote:
> 
> > Permit in-use snapshot exception data to be 'handed over' from one
> > snapshot instance to another.  This is a pre-requisite for patches
> > that allow the changes made in a snapshot device to be merged back into
> > its origin device and also allows device resizing.
> > 
> > The basic call sequence is:
> > 
> >   dmsetup load new_snapshot (referencing the existing in-use cow device)
> >      - the ctr code detects that the cow is already in use and allows the
> >        two snapshot target instances to be linked together
> >   dmsetup suspend original_snapshot
> >   dmsetup resume new_snapshot
> >      - the new_snapshot becomes live, and if anything now tries to access
> >        the original one it will receive -EIO
> >   dmsetup remove original_snapshot
> > 
> > (There can only be two snapshot targets referencing the same cow device
> > simultaneously.)
> > 
> > Signed-off-by: Mike Snitzer <snitzer at redhat.com>
> 
> Signed-off-by: Mikulas Patocka <mpatocka at redhat.com>
> Acked-by: Mikulas Patocka <mpatocka at redhat.com>
> 
> > +	down_read(&_origins_lock);
> > +	(void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest);
> > +	if (snap_src && snap_dest) {
> > +		down_write_nested(&snap_src->lock, SINGLE_DEPTH_NESTING);
> > +		down_write(&snap_dest->lock);
> 
> I'd suggest to use the "_nested" version for the nested lock, it is a 
> convention.
> 
> down_write(&snap_src->lock);
> down_write_nested(&snap_dest->lock, SINGLE_DEPTH_NESTING);

OK, I've tested this to work fine.  Thanks for spotting this.

Alasdair, are you OK with editing v9 of this patch to include Mikulas'
change?  And also add his 'Signed-off-by'?

Or should I send v10?

Mike