linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: NeilBrown <neilb@suse.com>
To: Jinpu Wang <jinpu.wang@profitbricks.com>
Cc: linux-raid@vger.kernel.org, Shaohua Li <shli@fb.com>,
	Nate Dailey <nate.dailey@stratus.com>
Subject: Re: [BUG] MD/RAID1 hung forever on freeze_array
Date: Thu, 15 Dec 2016 14:20:00 +1100	[thread overview]
Message-ID: <87fulpyj33.fsf@notabene.neil.brown.name> (raw)
In-Reply-To: <CAMGffEnvn4YQ5ADFXrOm2vOeEwApo38V84wB7VF3D1eZihF-Zw@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 2300 bytes --]


> Hi Neil,
>
> I found a old mail thread below
> http://www.spinics.net/lists/raid/msg52792.html
>
> Likely Alex is trying to fix same bug, right?
> in one reply you suggested to modify the call in make_request
>
> @@ -1207,7 +1207,8 @@ read_again:
>                                 sectors_handled;
>                         goto read_again;
>                 } else
> -                       generic_make_request(read_bio);
> +                       reschedule_retry(r1_bio);
>                 return;
>         }
>
>
> I append above change, it looks fix the bug, I've run same tests over
> one hour,  no hung task anymore.
>
> Do you think this is right fix? Do we still need the change you
> suggested with punt_bios_to_rescuer?

I don't really like that fix.  I suspect it would probably hurt
performance.

I'd prefer to fix generic_make_request() to process queued requests in a
more sensible order.
Can you please try the following (with all other patches removed)?
Thanks,
NeilBrown

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c0740dc0..3436b6fc3ef8 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -2036,10 +2036,31 @@ blk_qc_t generic_make_request(struct bio *bio)
 		struct request_queue *q = bdev_get_queue(bio->bi_bdev);
 
 		if (likely(blk_queue_enter(q, false) == 0)) {
+			struct bio_list hold;
+			struct bio_list lower, same;
+
+			/* Create a fresh bio_list for all subordinate requests */
+			bio_list_merge(&hold, &bio_list_on_stack);
+			bio_list_init(&bio_list_on_stack);
 			ret = q->make_request_fn(q, bio);
 
 			blk_queue_exit(q);
 
+			/* sort new bios into those for a lower level
+			 * and those for the same level
+			 */
+			bio_list_init(&lower);
+			bio_list_init(&same);
+			while ((bio = bio_list_pop(&bio_list_on_stack)) != NULL)
+				if (q == bdev_get_queue(bio->bi_bdev))
+					bio_list_add(&same, bio);
+				else
+					bio_list_add(&lower, bio);
+			/* now assemble so we handle the lowest level first */
+			bio_list_merge(&bio_list_on_stack, &lower);
+			bio_list_merge(&bio_list_on_stack, &same);
+			bio_list_merge(&bio_list_on_stack, &hold);
+
 			bio = bio_list_pop(current->bio_list);
 		} else {
 			struct bio *bio_next = bio_list_pop(current->bio_list);

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

  reply	other threads:[~2016-12-15  3:20 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-25 13:30 [BUG] MD/RAID1 hung forever on freeze_array Jinpu Wang
2016-11-25 13:59 ` Jinpu Wang
2016-11-28  4:47   ` Coly Li
2016-11-28  8:24     ` Jinpu Wang
2016-11-28  8:54       ` Coly Li
2016-11-28  9:02         ` Jinpu Wang
2016-11-28  9:10           ` Coly Li
2016-11-29 11:15             ` Jinpu Wang
2016-12-07 14:17               ` Jinpu Wang
2016-12-08  3:17                 ` NeilBrown
2016-12-08  9:50                   ` Jinpu Wang
2016-12-09  6:01                     ` NeilBrown
2016-12-09 15:28                       ` Jinpu Wang
2016-12-09 15:36                       ` Jinpu Wang
2016-12-12  0:59 ` NeilBrown
2016-12-12 13:10   ` Jinpu Wang
2016-12-12 21:53     ` NeilBrown
2016-12-13 15:08       ` Jinpu Wang
2016-12-13 22:18         ` NeilBrown
2016-12-14 10:22           ` Jinpu Wang
2016-12-14 12:13             ` Jinpu Wang
2016-12-14 14:49               ` Jinpu Wang
2016-12-15  3:20                 ` NeilBrown [this message]
2016-12-15  9:24                   ` Jinpu Wang
     [not found]                   ` <CAMGffEkufeaDytaHxtLR02iiQifZDhcwkLdzMj3X8_yaitSoFQ@mail.gmail.com>
2016-12-19 14:56                     ` Jinpu Wang
2016-12-19 22:45                     ` NeilBrown
2016-12-20 10:34                       ` Jinpu Wang
2016-12-20 21:23                         ` NeilBrown
2016-12-21 12:48                           ` Jinpu Wang
2016-12-21 23:51                             ` NeilBrown
2016-12-22  8:35                               ` Jinpu Wang

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=87fulpyj33.fsf@notabene.neil.brown.name \
    --to=neilb@suse.com \
    --cc=jinpu.wang@profitbricks.com \
    --cc=linux-raid@vger.kernel.org \
    --cc=nate.dailey@stratus.com \
    --cc=shli@fb.com \
    /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).