From: NeilBrown <neilb@suse.com>
To: Artur Paszkiewicz <artur.paszkiewicz@intel.com>,
Shaohua Li <shli@kernel.org>
Cc: Linux Raid <linux-raid@vger.kernel.org>
Subject: Re: [PATCH] md: create new workqueue for object destruction
Date: Thu, 19 Oct 2017 09:36:27 +1100 [thread overview]
Message-ID: <87376ghyms.fsf@notabene.neil.brown.name> (raw)
In-Reply-To: <6454f28e-4728-a10d-f3c3-b68afedec8d9@intel.com>
[-- Attachment #1: Type: text/plain, Size: 4473 bytes --]
On Wed, Oct 18 2017, Artur Paszkiewicz wrote:
> On 10/18/2017 09:29 AM, NeilBrown wrote:
>> On Tue, Oct 17 2017, Shaohua Li wrote:
>>
>>> On Tue, Oct 17, 2017 at 04:04:52PM +1100, Neil Brown wrote:
>>>>
>>>> lockdep currently complains about a potential deadlock
>>>> with sysfs access taking reconfig_mutex, and that
>>>> waiting for a work queue to complete.
>>>>
>>>> The cause is inappropriate overloading of work-items
>>>> on work-queues.
>>>>
>>>> We currently have two work-queues: md_wq and md_misc_wq.
>>>> They service 5 different tasks:
>>>>
>>>> mddev->flush_work md_wq
>>>> mddev->event_work (for dm-raid) md_misc_wq
>>>> mddev->del_work (mddev_delayed_delete) md_misc_wq
>>>> mddev->del_work (md_start_sync) md_misc_wq
>>>> rdev->del_work md_misc_wq
>>>>
>>>> We need to call flush_workqueue() for md_start_sync and ->event_work
>>>> while holding reconfig_mutex, but mustn't hold it when
>>>> flushing mddev_delayed_delete or rdev->del_work.
>>>>
>>>> md_wq is a bit special as it has WQ_MEM_RECLAIM so it is
>>>> best to leave that alone.
>>>>
>>>> So create a new workqueue, md_del_wq, and a new work_struct,
>>>> mddev->sync_work, so we can keep two classes of work separate.
>>>>
>>>> md_del_wq and ->del_work are used only for destroying rdev
>>>> and mddev.
>>>> md_misc_wq is used for event_work and sync_work.
>>>>
>>>> Also document the purpose of each flush_workqueue() call.
>>>>
>>>> This removes the lockdep warning.
>>>
>>> I had the exactly same patch queued internally,
>>
>> Cool :-)
>>
>>> but the mdadm test suite still
>>> shows lockdep warnning. I haven't time to check further.
>>>
>>
>> The only other lockdep I've seen later was some ext4 thing, though I
>> haven't tried the full test suite. I might have a look tomorrow.
>
> I'm also seeing a lockdep warning with or without this patch,
> reproducible with:
>
Thanks!
Looks like using one workqueue for mddev->del_work and rdev->del_work
causes problems.
Can you try with this addition please?
Thanks,
NeilBrown
diff --git a/drivers/md/md.c b/drivers/md/md.c
index d1dfc9879368..b3192943de7d 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -92,6 +92,7 @@ EXPORT_SYMBOL(md_cluster_mod);
static DECLARE_WAIT_QUEUE_HEAD(resync_wait);
static struct workqueue_struct *md_wq;
static struct workqueue_struct *md_del_wq;
+static struct workqueue_struct *rdev_del_wq;
static struct workqueue_struct *md_misc_wq;
static int remove_and_add_spares(struct mddev *mddev,
@@ -2265,7 +2266,7 @@ static void unbind_rdev_from_array(struct md_rdev *rdev)
synchronize_rcu();
INIT_WORK(&rdev->del_work, md_delayed_delete);
kobject_get(&rdev->kobj);
- queue_work(md_del_wq, &rdev->del_work);
+ queue_work(rdev_del_wq, &rdev->del_work);
}
/*
@@ -4307,7 +4308,7 @@ new_dev_store(struct mddev *mddev, const char *buf, size_t len)
return -EOVERFLOW;
/* Ensure old devices are fully deleted (rdev->del_work) */
- flush_workqueue(md_del_wq);
+ flush_workqueue(rdev_del_wq);
err = mddev_lock(mddev);
if (err)
@@ -7140,7 +7141,7 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
if (cmd == ADD_NEW_DISK)
/* need to ensure md_delayed_delete() has completed (rdev->del_work) */
- flush_workqueue(md_del_wq);
+ flush_workqueue(rdev_del_wq);
if (cmd == HOT_REMOVE_DISK)
/* need to ensure recovery thread has run */
@@ -9033,8 +9034,9 @@ static int __init md_init(void)
md_wq = alloc_workqueue("md", WQ_MEM_RECLAIM, 0);
md_del_wq = alloc_workqueue("md_del", 0, 0);
+ rdev_del_wq = alloc_workqueue("md_rdev_del", 0, 0);
md_misc_wq = alloc_workqueue("md_misc", 0, 0);
- if (!md_wq || !md_del_wq || !md_misc_wq)
+ if (!md_wq || !md_del_wq || !rdev_del_wq || !md_misc_wq)
goto err;
if ((ret = register_blkdev(MD_MAJOR, "md")) < 0)
@@ -9062,6 +9064,8 @@ static int __init md_init(void)
destroy_workqueue(md_wq);
if (md_del_wq)
destroy_workqueue(md_del_wq);
+ if (rdev_del_wq)
+ destroy_workqueue(rdev_del_wq);
if (md_misc_wq)
destroy_workqueue(md_misc_wq);
return ret;
@@ -9319,6 +9323,7 @@ static __exit void md_exit(void)
}
destroy_workqueue(md_misc_wq);
destroy_workqueue(md_del_wq);
+ destroy_workqueue(rdev_del_wq);
destroy_workqueue(md_wq);
}
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]
next prev parent reply other threads:[~2017-10-18 22:36 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-10-17 5:04 [PATCH] md: create new workqueue for object destruction NeilBrown
2017-10-18 6:21 ` Shaohua Li
2017-10-18 7:29 ` NeilBrown
2017-10-18 11:21 ` Artur Paszkiewicz
2017-10-18 22:36 ` NeilBrown [this message]
2017-10-19 8:27 ` Artur Paszkiewicz
2017-10-19 22:28 ` NeilBrown
2017-10-20 14:00 ` Artur Paszkiewicz
2017-10-22 23:31 ` NeilBrown
2017-10-27 10:44 ` Artur Paszkiewicz
2017-10-29 22:18 ` NeilBrown
2017-10-30 13:02 ` Artur Paszkiewicz
2017-11-01 3:57 ` NeilBrown
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=87376ghyms.fsf@notabene.neil.brown.name \
--to=neilb@suse.com \
--cc=artur.paszkiewicz@intel.com \
--cc=linux-raid@vger.kernel.org \
--cc=shli@kernel.org \
/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).