From: Boaz Harrosh <bharrosh@panasas.com>
To: Kent Overstreet <koverstreet@google.com>
Cc: axboe@kernel.dk, dm-devel@redhat.com, neilb@suse.de,
linux-kernel@vger.kernel.org, linux-bcache@vger.kernel.org,
mpatocka@redhat.com, vgoyal@redhat.com, yehuda@hq.newdream.net,
tj@kernel.org, sage@newdream.net, agk@redhat.com,
drbd-dev@lists.linbit.com
Subject: Re: [Drbd-dev] [PATCH v4 05/12] block: Kill bi_destructor
Date: Wed, 25 Jul 2012 14:39:57 +0300 [thread overview]
Message-ID: <500FDB0D.4070605@panasas.com> (raw)
In-Reply-To: <1343160689-12378-6-git-send-email-koverstreet@google.com>
On 07/24/2012 11:11 PM, Kent Overstreet wrote:
> Now that we've got generic code for freeing bios allocated from bio
> pools, this isn't needed anymore.
>
> This also changes the semantics of bio_free() a bit - it now also frees
> bios allocated by bio_kmalloc(). It's also no longer exported, as
> without bi_destructor there should be no need for it to be called
> anywhere else.
>
> Signed-off-by: Kent Overstreet <koverstreet@google.com>
<snip>
> diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
> index be65582..9338d0602 100644
> --- a/drivers/target/target_core_iblock.c
> +++ b/drivers/target/target_core_iblock.c
> @@ -467,15 +467,7 @@ iblock_get_bio(struct se_cmd *cmd, sector_t lba, u32 sg_num)
> }
>
> bio->bi_bdev = ib_dev->ibd_bd;
> -<<<<<<< HEAD
> bio->bi_private = cmd;
> - bio->bi_destructor = iblock_bio_destructor;
> -||||||| merged common ancestors
> - bio->bi_private = task;
> - bio->bi_destructor = iblock_bio_destructor;
> -=======
> - bio->bi_private = task;
> ->>>>>>> block: Generalized bio pool freeing
> bio->bi_end_io = &iblock_bio_done;
> bio->bi_sector = lba;
> return bio;
Merge conflict allmodconfig compilation please?
> diff --git a/fs/bio.c b/fs/bio.c
> index 252e253..a301071 100644
> --- a/fs/bio.c
> +++ b/fs/bio.c
> @@ -56,6 +56,9 @@ static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = {
> */
> struct bio_set *fs_bio_set;
>
> +/* Only used as a sentinal value */
> +static struct bio_set bio_kmalloc_pool;
> +
Id rather if you use a define like:
#define BIO_KMALLOC_POOL ((void *)~0)
So any code access actually crashes in the bug case where
this leaks out. (And there is no actual unused storage allocated)
BTW I like this reuse of the bi_pool member as a flag as well.
Thanks
Boaz
> /*
> * Our slab pool management
> */
> @@ -232,10 +235,21 @@ fallback:
> return bvl;
> }
>
> -void bio_free(struct bio *bio, struct bio_set *bs)
> +void bio_free(struct bio *bio)
> {
> + struct bio_set *bs = bio->bi_pool;
> void *p;
>
> + BUG_ON(!bs);
> +
> + if (bs == &bio_kmalloc_pool) {
> + /* Bio was allocated by bio_kmalloc() */
> + if (bio_integrity(bio))
> + bio_integrity_free(bio, fs_bio_set);
> + kfree(bio);
> + return;
> + }
> +
> if (bio_has_allocated_vec(bio))
> bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio));
>
> @@ -347,13 +361,6 @@ struct bio *bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs)
> }
> EXPORT_SYMBOL(bio_alloc);
>
> -static void bio_kmalloc_destructor(struct bio *bio)
> -{
> - if (bio_integrity(bio))
> - bio_integrity_free(bio, fs_bio_set);
> - kfree(bio);
> -}
> -
> /**
> * bio_kmalloc - allocate a bio for I/O using kmalloc()
> * @gfp_mask: the GFP_ mask given to the slab allocator
> @@ -380,7 +387,7 @@ struct bio *bio_kmalloc(gfp_t gfp_mask, unsigned int nr_iovecs)
> bio->bi_flags |= BIO_POOL_NONE << BIO_POOL_OFFSET;
> bio->bi_max_vecs = nr_iovecs;
> bio->bi_io_vec = bio->bi_inline_vecs;
> - bio->bi_destructor = bio_kmalloc_destructor;
> + bio->bi_pool = &bio_kmalloc_pool;
>
> return bio;
> }
> @@ -418,12 +425,7 @@ void bio_put(struct bio *bio)
> */
> if (atomic_dec_and_test(&bio->bi_cnt)) {
> bio_disassociate_task(bio);
> - bio->bi_next = NULL;
> -
> - if (bio->bi_pool)
> - bio_free(bio, bio->bi_pool);
> - else
> - bio->bi_destructor(bio);
> + bio_free(bio);
> }
> }
> EXPORT_SYMBOL(bio_put);
> diff --git a/include/linux/bio.h b/include/linux/bio.h
> index ba60319..393c87e 100644
> --- a/include/linux/bio.h
> +++ b/include/linux/bio.h
> @@ -216,7 +216,7 @@ extern struct bio *bio_alloc(gfp_t, unsigned int);
> extern struct bio *bio_kmalloc(gfp_t, unsigned int);
> extern struct bio *bio_alloc_bioset(gfp_t, int, struct bio_set *);
> extern void bio_put(struct bio *);
> -extern void bio_free(struct bio *, struct bio_set *);
> +extern void bio_free(struct bio *);
>
> extern void bio_endio(struct bio *, int);
> struct request_queue;
> diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
> index 40411e2..fa45a12 100644
> --- a/include/linux/blk_types.h
> +++ b/include/linux/blk_types.h
> @@ -84,11 +84,8 @@ struct bio {
> struct bio_integrity_payload *bi_integrity; /* data integrity */
> #endif
>
> - /* If bi_pool is non NULL, bi_destructor is not called */
> struct bio_set *bi_pool;
>
> - bio_destructor_t *bi_destructor; /* destructor */
> -
> /*
> * We can inline a number of vecs at the end of the bio, to avoid
> * double allocations for a small number of bio_vecs. This member
next prev parent reply other threads:[~2012-07-26 9:35 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-07-24 20:11 [Drbd-dev] [PATCH v4 00/12] Block cleanups Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 01/12] block: Generalized bio pool freeing Kent Overstreet
2012-07-25 11:06 ` Boaz Harrosh
2012-07-25 23:38 ` Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 02/12] dm: Use bioset's front_pad for dm_rq_clone_bio_info Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 03/12] block: Add bio_reset() Kent Overstreet
2012-07-25 11:19 ` Boaz Harrosh
2012-07-25 22:56 ` Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 04/12] pktcdvd: Switch to bio_kmalloc() Kent Overstreet
2012-07-25 11:29 ` Boaz Harrosh
2012-07-25 23:01 ` Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 05/12] block: Kill bi_destructor Kent Overstreet
2012-07-25 11:39 ` Boaz Harrosh [this message]
2012-07-25 23:15 ` Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 06/12] block: Add an explicit bio flag for bios that own their bvec Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 07/12] block: Rename bio_split() -> bio_pair_split() Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 08/12] block: Introduce new bio_split() Kent Overstreet
2012-07-25 11:55 ` Boaz Harrosh
2012-07-25 23:26 ` Kent Overstreet
2012-07-27 0:50 ` [Drbd-dev] [PATCH] A possible deadlock with stacked devices (was: [PATCH v4 08/12] block: Introduce new bio_split()) Mikulas Patocka
2012-08-15 23:07 ` Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 09/12] block: Rework bio_pair_split() Kent Overstreet
2012-07-25 12:03 ` Boaz Harrosh
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 10/12] block: Add bio_clone_kmalloc() Kent Overstreet
2012-07-25 12:05 ` Boaz Harrosh
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 11/12] block: Add bio_clone_bioset() Kent Overstreet
2012-07-24 20:11 ` [Drbd-dev] [PATCH v4 12/12] block: Only clone bio vecs that are in use Kent Overstreet
2012-08-07 3:17 ` Muthu Kumar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=500FDB0D.4070605@panasas.com \
--to=bharrosh@panasas.com \
--cc=agk@redhat.com \
--cc=axboe@kernel.dk \
--cc=dm-devel@redhat.com \
--cc=drbd-dev@lists.linbit.com \
--cc=koverstreet@google.com \
--cc=linux-bcache@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mpatocka@redhat.com \
--cc=neilb@suse.de \
--cc=sage@newdream.net \
--cc=tj@kernel.org \
--cc=vgoyal@redhat.com \
--cc=yehuda@hq.newdream.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox