From: Jens Axboe <jens.axboe@oracle.com>
To: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: tglx@linutronix.de, James.Bottomley@HansenPartnership.com,
jengelh@medozas.de, bharrosh@panasas.com,
linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-ide@vger.kernel.org
Subject: Re: [BUG] 2.6.29-rc6-2450cf in scsi_lib.c (was: Large amount of scsi-sgpool)objects
Date: Thu, 5 Mar 2009 11:14:36 +0100 [thread overview]
Message-ID: <20090305101436.GV11787@kernel.dk> (raw)
In-Reply-To: <20090305100901.GU11787@kernel.dk>
On Thu, Mar 05 2009, Jens Axboe wrote:
> On Thu, Mar 05 2009, FUJITA Tomonori wrote:
> > Oops, somehow I forgot to CC Jens...
> >
> > On Thu, 5 Mar 2009 17:39:17 +0900
> > FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> wrote:
> >
> > > On Thu, 5 Mar 2009 17:36:13 +0900
> > > FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> wrote:
> > >
> > > > CC'ed Jens,
> > > >
> > > > On Wed, 04 Mar 2009 22:56:29 +0000
> > > > James Bottomley <James.Bottomley@HansenPartnership.com> wrote:
> > > >
> > > > > On Wed, 2009-03-04 at 22:45 +0100, Thomas Gleixner wrote:
> > > > > > On Wed, 4 Mar 2009, Thomas Gleixner wrote:
> > > > > >
> > > > > > Instrumented the code and the result of the failing request is
> > > > > > below. Looks like the function which sets up the request gets
> > > > > > nr_phys_segments wrong by one.
> > > > > >
> > > > > > If you need further trace data feel free to ask.
> > > > >
> > > > > OK, the mapping all checks out correctly ... there must be something
> > > > > wrong with the way we count before mapping.
> > > >
> > > > Yeah, looks we miscalculate nr_phys_segments in the merging path.
> > > >
> > > > blk_recount_segments() needs to set bi_seg_front_size and
> > > > bi_seg_back_size for ll_merge_requests_fn()?
> > > >
> > > > =
> > > > diff --git a/block/blk-merge.c b/block/blk-merge.c
> > > > index a104593..efb65b6 100644
> > > > --- a/block/blk-merge.c
> > > > +++ b/block/blk-merge.c
> > > > @@ -111,12 +111,19 @@ void blk_recalc_rq_segments(struct request *rq)
> > > >
> > > > void blk_recount_segments(struct request_queue *q, struct bio *bio)
> > > > {
> > > > + unsigned int seg_size;
> > > > struct bio *nxt = bio->bi_next;
> > > >
> > > > bio->bi_next = NULL;
> > > > - bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, NULL);
> > > > + bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, &seg_size);
> > > > bio->bi_next = nxt;
> > > > bio->bi_flags |= (1 << BIO_SEG_VALID);
> > > > +
> > > > + if (bio->bi_phys_segments == 1 && seg_size > bio->bi_seg_front_size)
> > > > + bio->bi_seg_front_size = seg_size;
> > > > + if (bio->bi_phys_segments > bio->bi_seg_back_size)
> > > > + bio->bi_seg_back_size = seg_size;
> > > > +
> > > > }
> > > > EXPORT_SYMBOL(blk_recount_segments);
> > >
> > > Duh, here's the proper patch.
> > >
> > > diff --git a/block/blk-merge.c b/block/blk-merge.c
> > > index a104593..06e0db4 100644
> > > --- a/block/blk-merge.c
> > > +++ b/block/blk-merge.c
> > > @@ -111,12 +111,19 @@ void blk_recalc_rq_segments(struct request *rq)
> > >
> > > void blk_recount_segments(struct request_queue *q, struct bio *bio)
> > > {
> > > + unsigned int seg_size;
> > > struct bio *nxt = bio->bi_next;
> > >
> > > bio->bi_next = NULL;
> > > - bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, NULL);
> > > + bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, &seg_size);
> > > bio->bi_next = nxt;
> > > bio->bi_flags |= (1 << BIO_SEG_VALID);
> > > +
> > > + if (bio->bi_phys_segments == 1 && seg_size > bio->bi_seg_front_size)
> > > + bio->bi_seg_front_size = seg_size;
> > > + if (seg_size > bio->bi_seg_back_size)
> > > + bio->bi_seg_back_size = seg_size;
> > > +
> > > }
> > > EXPORT_SYMBOL(blk_recount_segments);
>
> Good catch, I merged it with a slight change of layout and clearing
> seg_size initially, to avoid gcc silly errors.
While merging that, I think we can do better than this. Essentially we
just need to have __blk_recalc_rq_segments() track the back bio as well,
then we don't have to pass in a pointer for segment sizes.
Totally untested, comments welcome...
diff --git a/block/blk-merge.c b/block/blk-merge.c
index 6be3797..5a244f0 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -39,14 +39,13 @@ void blk_recalc_rq_sectors(struct request *rq, int nsect)
}
static unsigned int __blk_recalc_rq_segments(struct request_queue *q,
- struct bio *bio,
- unsigned int *seg_size_ptr)
+ struct bio *bio)
{
unsigned int phys_size;
struct bio_vec *bv, *bvprv = NULL;
int cluster, i, high, highprv = 1;
unsigned int seg_size, nr_phys_segs;
- struct bio *fbio;
+ struct bio *fbio, *bbio;
if (!bio)
return 0;
@@ -87,41 +86,28 @@ new_segment:
seg_size = bv->bv_len;
highprv = high;
}
+ bbio = bio;
}
- if (seg_size_ptr)
- *seg_size_ptr = seg_size;
+ if (nr_phys_segs == 1 && seg_size > fbio->bi_seg_front_size)
+ fbio->bi_seg_front_size = seg_size;
+ if (seg_size > bbio->bi_seg_back_size)
+ bbio->bi_seg_back_size = seg_size;
return nr_phys_segs;
}
void blk_recalc_rq_segments(struct request *rq)
{
- unsigned int seg_size = 0, phys_segs;
-
- phys_segs = __blk_recalc_rq_segments(rq->q, rq->bio, &seg_size);
-
- if (phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size)
- rq->bio->bi_seg_front_size = seg_size;
- if (seg_size > rq->biotail->bi_seg_back_size)
- rq->biotail->bi_seg_back_size = seg_size;
-
- rq->nr_phys_segments = phys_segs;
+ rq->nr_phys_segments = __blk_recalc_rq_segments(rq->q, rq->bio);
}
void blk_recount_segments(struct request_queue *q, struct bio *bio)
{
struct bio *nxt = bio->bi_next;
- unsigned int seg_size = 0;
bio->bi_next = NULL;
- bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, NULL);
-
- if (bio->bi_phys_segments == 1 && seg_size > bio->bi_seg_front_size)
- bio->bi_seg_front_size = seg_size;
- if (seg_size > bio->bi_seg_back_size)
- bio->bi_seg_back_size = seg_size;
-
+ bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio);
bio->bi_next = nxt;
bio->bi_flags |= (1 << BIO_SEG_VALID);
}
--
Jens Axboe
next prev parent reply other threads:[~2009-03-05 10:14 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-03 1:28 Large amount of scsi-sgpool objects Jan Engelhardt
2009-03-03 9:31 ` Boaz Harrosh
2009-03-03 15:21 ` James Bottomley
2009-03-03 16:08 ` Jan Engelhardt
2009-03-03 16:24 ` Boaz Harrosh
2009-03-03 17:59 ` Alan Stern
2009-03-03 20:56 ` Ingo Molnar
2009-03-03 21:06 ` Alan Stern
2009-03-03 16:25 ` James Bottomley
2009-03-03 17:19 ` Thomas Gleixner
2009-03-03 22:07 ` [BUG] 2.6.29-rc6-2450cf in scsi_lib.c (was: Large amount of scsi-sgpool)objects Thomas Gleixner
2009-03-03 22:22 ` Jan Engelhardt
2009-03-03 23:07 ` Thomas Gleixner
2009-03-03 22:26 ` James Bottomley
2009-03-04 2:01 ` Thomas Gleixner
2009-03-04 18:55 ` James Bottomley
2009-03-04 21:45 ` Thomas Gleixner
2009-03-04 22:56 ` James Bottomley
2009-03-05 0:13 ` James Bottomley
2009-03-05 8:36 ` FUJITA Tomonori
2009-03-05 8:39 ` FUJITA Tomonori
2009-03-05 9:29 ` FUJITA Tomonori
2009-03-05 10:09 ` Jens Axboe
2009-03-05 10:14 ` Jens Axboe [this message]
2009-03-05 10:27 ` FUJITA Tomonori
2009-03-05 10:30 ` Jens Axboe
2009-03-05 10:41 ` FUJITA Tomonori
2009-03-05 11:10 ` Jens Axboe
2009-03-05 11:40 ` FUJITA Tomonori
2009-03-05 10:41 ` Ingo Molnar
2009-03-05 11:05 ` Jens Axboe
2009-03-05 11:07 ` Ingo Molnar
2009-03-05 12:09 ` Thomas Gleixner
2009-03-05 23:16 ` Thomas Gleixner
2009-03-05 19:32 ` Ingo Molnar
2009-03-05 10:15 ` Ingo Molnar
2009-03-03 19:22 ` Large amount of scsi-sgpool objects Ingo Molnar
2009-03-03 21:25 ` James Bottomley
2009-03-03 21:44 ` Ingo Molnar
2009-03-03 22:39 ` James Bottomley
2009-03-03 23:03 ` Ingo Molnar
2009-03-03 23:32 ` Stefan Richter
2009-03-03 23:48 ` Ingo Molnar
2009-03-04 6:39 ` Stefan Richter
2009-03-04 7:12 ` Mike Galbraith
2009-03-04 7:50 ` Stefan Richter
2009-03-04 8:00 ` Mike Galbraith
2009-03-04 9:06 ` Thomas Gleixner
2009-03-04 11:12 ` Ingo Molnar
2009-03-04 11:28 ` Stefan Richter
2009-03-04 11:47 ` Ingo Molnar
2009-03-04 12:02 ` Stefan Richter
2009-03-04 11:24 ` Ingo Molnar
2009-03-04 19:11 ` Andrew Morton
2009-03-04 20:09 ` Thomas Gleixner
2009-03-04 0:01 ` James Bottomley
2009-03-04 0:39 ` Ingo Molnar
2009-03-04 0:30 ` Thomas Gleixner
2009-03-04 0:47 ` Ingo Molnar
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=20090305101436.GV11787@kernel.dk \
--to=jens.axboe@oracle.com \
--cc=James.Bottomley@HansenPartnership.com \
--cc=bharrosh@panasas.com \
--cc=fujita.tomonori@lab.ntt.co.jp \
--cc=jengelh@medozas.de \
--cc=linux-ide@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=tglx@linutronix.de \
/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;
as well as URLs for NNTP newsgroup(s).