Ocfs2 devel - Jan 2015 - [PATCH 1/4] quota: Don't store flags for v2 quota format

On Thu 15-01-15 01:40:34, Christoph Hellwig wrote:
> On Wed, Jan 14, 2015 at 07:27:08PM +0100, Jan Kara wrote:
> > Currently, v2 quota format blindly stored flags from in-memory dqinfo
on
> > disk, although there are no flags supported. Since it is stupid to
store
> > flags which have no effect, just store 0 unconditionally and don't
> > bother loading it from disk.
> > 
> > Note that userspace could have stored some flags there via Q_SETINFO
> > quotactl and then later read them (although flags have no effect) but
> > I'm pretty sure noone does that (most definitely quota-tools
don't and
> > quota interface doesn't have too much other users).
> 
> What about future proofing?  Current kernels can store flags on disk,
> so the best is to reserve the currently (and possibly previously)
> assigned values, and mask them out when reading from disk.
  Hum, I'm not sure I follow you. Current kernels will store any 32-bit
number user sets in flags field. So if we wanted to be 100% safe, we'd have
to just ignore that field. Which isn't currently a problem since quota code
doesn't use the field for anything (it was added just for future
extensions). But since I'm pretty certain noone actually relies on values
of that field, I though we could just get away with forcibly zeroing the
field now and if there's a need to use the field in a few years, we could
start using it.

								Honza

-- 
Jan Kara <jack at suse.cz>
SUSE Labs, CR

On Thu, Jan 15, 2015 at 11:13:10AM +0100, Jan Kara
wrote:
>   Hum, I'm not sure I follow you. Current kernels will store any 32-bit
> number user sets in flags field. So if we wanted to be 100% safe, we'd
have
> to just ignore that field. Which isn't currently a problem since quota
code
> doesn't use the field for anything (it was added just for future
> extensions). But since I'm pretty certain noone actually relies on
values
> of that field, I though we could just get away with forcibly zeroing the
> field now and if there's a need to use the field in a few years, we
could
> start using it.
Oh, I misread the code and your description.  I thought we would just
store any potentially valid in-core flag on disk.

I guess for now the best case would be to stop storing anything, and
then just make an educated decision if/when we need a flags field.