On Fri, May 18, 2012 at 09:21:42AM -0700, Tejun Heo wrote:
> Hello,
>
> > @@ -417,8 +409,11 @@ void bio_put(struct bio *bio)
> >
> > if (bio->bi_pool)
> > bio_free(bio, bio->bi_pool);
> > - else
> > - bio->bi_destructor(bio);
> > + else {
> > + if (bio_integrity(bio))
> > + bio_integrity_free(bio, fs_bio_set);
> > + kfree(bio);
>
> if {
> } else {
> }
>
> And wouldn't it be better to make bio_free() handle kfreeing too?
That'd kind of change the semantics of bio_free() - but I suppose it'd
make sense.
I'm not terribly happy with how the bio integrity data belongs to
fs_bio_set for kmalloced bios. Maybe I'll change that at some point.
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
