Re: [PATCH 2/2] [MTD] Call bdi_init() and bdi_register()

[Jens Axboe <jens.axboe@oracle.com>]

On Mon, Apr 26 2010, Jörn Engel wrote:
> On Mon, 26 April 2010 19:02:48 +0200, Jens Axboe wrote:
> > On Mon, Apr 26 2010, Jörn Engel wrote:
> > > Otherwise we hit a BUG_ON in bdi_queue_work().
> > > 
> > > Signed-off-by: Joern Engel <joern@logfs.org>
> > > ---
> > >  drivers/mtd/mtdcore.c |   17 ++++++++++++++++-
> > >  1 files changed, 16 insertions(+), 1 deletions(-)
> > > 
> > > diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
> > > index cb4858b..8dd3e46 100644
> > > --- a/drivers/mtd/mtdcore.c
> > > +++ b/drivers/mtd/mtdcore.c
> > > @@ -322,6 +322,12 @@ int add_mtd_device(struct mtd_info *mtd)
> > >  		if (!mtd_table[i]) {
> > >  			struct mtd_notifier *not;
> > >  
> > > +			err = bdi_register(mtd->backing_dev_info, NULL, "mtd%d",
> > > +					i);
> > > +			if (err) {
> > > +				/* We lose the errno information :( */
> > > +				break;
> > > +			}
> > 
> > This is not a good idea, even if it'll currently work.
> > 
> > > @@ -692,6 +698,15 @@ static int __init init_mtd(void)
> > >  	int ret;
> > >  	ret = class_register(&mtd_class);
> > >  
> > > +	ret = bdi_init(&mtd_bdi_unmappable);
> > > +	if (ret)
> > > +		return ret;
> > > +	ret = bdi_init(&mtd_bdi_ro_mappable);
> > > +	if (ret)
> > > +		return ret;
> > > +	ret = bdi_init(&mtd_bdi_rw_mappable);
> > > +	if (ret)
> > > +		return ret;
> > 
> > Do the bdi_register() here as well.
> 
> -ENOBRAIN
> 
> Initially I wanted to do just than.  Then I looked at the block layer
> and thought we could create one backing_dev_info per mtd as well.  Not
> necessarily a bad if I would actually _create_ them and not just reuse
> the same ones over and over again.

I cooked up that patch myself, here:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=0661b1ac5d48eb47c8a5948c0554fea25e0895ab

Care to give it a quick spin?

-- 
Jens Axboe