From: Xiao Ni <xni@redhat.com>
To: Yu Kuai <yukuai1@huaweicloud.com>,
mpatocka@redhat.com, heinzm@redhat.com,
blazej.kucman@linux.intel.com, agk@redhat.com,
snitzer@kernel.org, dm-devel@lists.linux.dev, song@kernel.org,
yukuai3@huawei.com, jbrassow@f14.redhat.com, neilb@suse.de,
shli@fb.com, akpm@osdl.org
Cc: linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org,
yi.zhang@huawei.com, yangerkun@huawei.com
Subject: Re: [PATCH v5 05/14] md: don't suspend the array for interrupted reshape
Date: Thu, 29 Feb 2024 10:10:54 +0800 [thread overview]
Message-ID: <60d75867-8fb7-4c67-96f7-3e5ba65bdbd9@redhat.com> (raw)
In-Reply-To: <20240201092559.910982-6-yukuai1@huaweicloud.com>
在 2024/2/1 下午5:25, Yu Kuai 写道:
> From: Yu Kuai <yukuai3@huawei.com>
>
> md_start_sync() will suspend the array if there are spares that can be
> added or removed from conf, however, if reshape is still in progress,
Hi Kuai
Why md_start_sync can run when reshape is still in progress?
md_check_recovery should return without queue sync_work, right?
> this won't happen at all or data will be corrupted(remove_and_add_spares
> won't be called from md_choose_sync_action for reshape), hence there is
> no need to suspend the array if reshape is not done yet.
>
> Meanwhile, there is a potential deadlock for raid456:
>
> 1) reshape is interrupted;
>
> 2) set one of the disk WantReplacement, and add a new disk to the array,
> however, recovery won't start until the reshape is finished;
>
> 3) then issue an IO across reshpae position, this IO will wait for
> reshape to make progress;
>
> 4) continue to reshape, then md_start_sync() found there is a spare disk
> that can be added to conf, mddev_suspend() is called;
I c. The answer for my above question is reshape is interrupted and then
it continues the reshape, right?
Best Regards
Xiao
>
> Step 4 and step 3 is waiting for each other, deadlock triggered. Noted
> this problem is found by code review, and it's not reporduced yet.
>
> Fix this porblem by don't suspend the array for interrupted reshape,
> this is safe because conf won't be changed until reshape is done.
>
> Fixes: bc08041b32ab ("md: suspend array in md_start_sync() if array need reconfiguration")
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
> ---
> drivers/md/md.c | 13 +++++++++----
> 1 file changed, 9 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/md/md.c b/drivers/md/md.c
> index 6c5d0a372927..85fde05c37dd 100644
> --- a/drivers/md/md.c
> +++ b/drivers/md/md.c
> @@ -9374,12 +9374,17 @@ static void md_start_sync(struct work_struct *ws)
> bool suspend = false;
> char *name;
>
> - if (md_spares_need_change(mddev))
> + /*
> + * If reshape is still in progress, spares won't be added or removed
> + * from conf until reshape is done.
> + */
> + if (mddev->reshape_position == MaxSector &&
> + md_spares_need_change(mddev)) {
> suspend = true;
> + mddev_suspend(mddev, false);
> + }
>
> - suspend ? mddev_suspend_and_lock_nointr(mddev) :
> - mddev_lock_nointr(mddev);
> -
> + mddev_lock_nointr(mddev);
> if (!md_is_rdwr(mddev)) {
> /*
> * On a read-only array we can:
next prev parent reply other threads:[~2024-02-29 2:11 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-01 9:25 [PATCH v5 00/14] dm-raid/md/raid: fix v6.7 regressions Yu Kuai
2024-02-01 9:25 ` [PATCH v5 01/14] md: don't ignore suspended array in md_check_recovery() Yu Kuai
2024-02-16 6:58 ` Xiao Ni
2024-02-18 1:14 ` Yu Kuai
2024-02-18 1:33 ` Xiao Ni
2024-02-18 1:46 ` Yu Kuai
2024-02-18 2:27 ` Xiao Ni
2024-02-18 2:34 ` Yu Kuai
2024-02-18 3:15 ` Xiao Ni
2024-02-18 3:24 ` Yu Kuai
2024-02-18 5:07 ` Xiao Ni
2024-02-18 6:22 ` Yu Kuai
2024-02-18 8:07 ` Xiao Ni
2024-02-18 8:47 ` Yu Kuai
2024-02-19 7:10 ` Xiao Ni
2024-02-19 8:19 ` Yu Kuai
2024-02-01 9:25 ` [PATCH v5 02/14] md: don't ignore read-only " Yu Kuai
2024-02-01 9:25 ` [PATCH v5 03/14] md: make sure md_do_sync() will set MD_RECOVERY_DONE Yu Kuai
2024-02-18 5:56 ` Xiao Ni
2024-02-18 6:51 ` Yu Kuai
2024-02-18 8:41 ` Xiao Ni
2024-02-18 8:59 ` Yu Kuai
2024-02-01 9:25 ` [PATCH v5 04/14] md: don't register sync_thread for reshape directly Yu Kuai
2024-02-28 12:07 ` Xiao Ni
2024-02-28 12:44 ` Yu Kuai
2024-02-28 12:57 ` Xiao Ni
2024-02-01 9:25 ` [PATCH v5 05/14] md: don't suspend the array for interrupted reshape Yu Kuai
2024-02-29 2:10 ` Xiao Ni [this message]
2024-02-29 2:14 ` Yu Kuai
2024-02-01 9:25 ` [PATCH v5 06/14] md: fix missing release of 'active_io' for flush Yu Kuai
2024-02-08 7:47 ` Song Liu
2024-02-01 9:25 ` [PATCH v5 07/14] md: export helpers to stop sync_thread Yu Kuai
2024-02-15 22:27 ` Song Liu
2024-02-18 2:35 ` Yu Kuai
2024-02-01 9:25 ` [PATCH v5 08/14] md: export helper md_is_rdwr() Yu Kuai
2024-02-01 9:25 ` [PATCH v5 09/14] dm-raid: really frozen sync_thread during suspend Yu Kuai
2024-02-18 4:53 ` Xiao Ni
2024-02-18 6:34 ` Yu Kuai
2024-02-19 7:27 ` Xiao Ni
2024-02-19 7:53 ` Yu Kuai
2024-02-19 8:45 ` Xiao Ni
2024-02-01 9:25 ` [PATCH v5 10/14] md/dm-raid: don't call md_reap_sync_thread() directly Yu Kuai
2024-02-01 9:25 ` [PATCH v5 11/14] dm-raid: add a new helper prepare_suspend() in md_personality Yu Kuai
2024-02-01 9:25 ` [PATCH v5 12/14] md/raid456: fix a deadlock for dm-raid456 while io concurrent with reshape Yu Kuai
2024-02-01 9:25 ` [PATCH v5 13/14] dm-raid: fix lockdep waring in "pers->hot_add_disk" Yu Kuai
2024-02-01 9:25 ` [PATCH v5 14/14] dm-raid: remove mddev_suspend/resume() Yu Kuai
2024-02-03 3:19 ` [PATCH v5 00/14] dm-raid/md/raid: fix v6.7 regressions Benjamin Marzinski
2024-02-04 1:35 ` Yu Kuai
2024-02-04 7:00 ` Yu Kuai
2024-02-05 19:35 ` Benjamin Marzinski
2024-02-06 1:36 ` Yu Kuai
2024-02-06 3:57 ` Benjamin Marzinski
2024-02-06 7:03 ` Yu Kuai
2024-02-08 8:04 ` Song Liu
2024-02-08 23:17 ` Benjamin Marzinski
2024-02-09 22:37 ` Song Liu
2024-02-26 7:58 ` Su Yue
2024-02-12 22:30 ` Song Liu
2024-02-15 22:24 ` Song Liu
2024-02-16 5:46 ` Benjamin Marzinski
2024-02-18 1:24 ` Yu Kuai
2024-02-19 16:05 ` Benjamin Marzinski
2024-02-20 3:09 ` Yu Kuai
2024-02-22 9:00 ` Yu Kuai
2024-02-28 1:19 ` Benjamin Marzinski
2024-02-28 1:35 ` Yu Kuai
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=60d75867-8fb7-4c67-96f7-3e5ba65bdbd9@redhat.com \
--to=xni@redhat.com \
--cc=agk@redhat.com \
--cc=akpm@osdl.org \
--cc=blazej.kucman@linux.intel.com \
--cc=dm-devel@lists.linux.dev \
--cc=heinzm@redhat.com \
--cc=jbrassow@f14.redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-raid@vger.kernel.org \
--cc=mpatocka@redhat.com \
--cc=neilb@suse.de \
--cc=shli@fb.com \
--cc=snitzer@kernel.org \
--cc=song@kernel.org \
--cc=yangerkun@huawei.com \
--cc=yi.zhang@huawei.com \
--cc=yukuai1@huaweicloud.com \
--cc=yukuai3@huawei.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