From: "yukuai (C)" <yukuai3@huawei.com>
To: Bart Van Assche <bvanassche@acm.org>, <axboe@kernel.dk>,
<andriy.shevchenko@linux.intel.com>, <john.garry@huawei.com>,
<ming.lei@redhat.com>
Cc: <linux-block@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<yi.zhang@huawei.com>
Subject: Re: [PATCH -next RFC v2 2/8] blk-mq: call 'bt_wait_ptr()' later in blk_mq_get_tag()
Date: Sat, 9 Apr 2022 10:09:52 +0800 [thread overview]
Message-ID: <fbe5565e-4dbe-bb71-e3c4-c33eb470a8d2@huawei.com> (raw)
In-Reply-To: <f6bcc53d-1419-7190-fd9a-8c5fa7178fe1@acm.org>
在 2022/04/08 22:20, Bart Van Assche 写道:
> On 4/8/22 00:39, Yu Kuai wrote:
>> bt_wait_ptr() will increase 'wait_index', however, if blk_mq_get_tag()
>> get a tag successfully after bt_wait_ptr() is called and before
>> sbitmap_prepare_to_wait() is called, then the 'ws' is skipped. This
>> behavior might cause 8 waitqueues to be unbalanced.
>>
>> Move bt_wait_ptr() later should reduce the problem when the disk is
>> under high io preesure.
>>
>> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
>> ---
>> block/blk-mq-tag.c | 4 +---
>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c
>> index 68ac23d0b640..228a0001694f 100644
>> --- a/block/blk-mq-tag.c
>> +++ b/block/blk-mq-tag.c
>> @@ -155,7 +155,6 @@ unsigned int blk_mq_get_tag(struct
>> blk_mq_alloc_data *data)
>> if (data->flags & BLK_MQ_REQ_NOWAIT)
>> return BLK_MQ_NO_TAG;
>> - ws = bt_wait_ptr(bt, data->hctx);
>> do {
>> struct sbitmap_queue *bt_prev;
>> @@ -174,6 +173,7 @@ unsigned int blk_mq_get_tag(struct
>> blk_mq_alloc_data *data)
>> if (tag != BLK_MQ_NO_TAG)
>> break;
>> + ws = bt_wait_ptr(bt, data->hctx);
>> sbitmap_prepare_to_wait(bt, ws, &wait, TASK_UNINTERRUPTIBLE);
>> tag = __blk_mq_get_tag(data, bt);
>> @@ -201,8 +201,6 @@ unsigned int blk_mq_get_tag(struct
>> blk_mq_alloc_data *data)
>> */
>> if (bt != bt_prev)
>> sbitmap_queue_wake_up(bt_prev);
>> -
>> - ws = bt_wait_ptr(bt, data->hctx);
>> } while (1);
>
> Is it necessary to call bt_wait_ptr() during every loop iteration or
> only if bt != bt_prev? Would calling bt_wait_ptr() only if bt != bt_prev
> help to reduce unfairness further?
Hi,
You are right, that sounds reasonable.
Thanks,
Kuai
>
> Thanks,
>
> Bart.
> .
>
next prev parent reply other threads:[~2022-04-09 2:09 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-04-08 7:39 [PATCH -next RFC v2 0/8] improve tag allocation under heavy load Yu Kuai
2022-04-08 7:39 ` [PATCH -next RFC v2 1/8] sbitmap: record the number of waiters for each waitqueue Yu Kuai
2022-04-08 7:39 ` [PATCH -next RFC v2 2/8] blk-mq: call 'bt_wait_ptr()' later in blk_mq_get_tag() Yu Kuai
2022-04-08 14:20 ` Bart Van Assche
2022-04-09 2:09 ` yukuai (C) [this message]
2022-04-08 7:39 ` [PATCH -next RFC v2 3/8] sbitmap: make sure waitqueues are balanced Yu Kuai
2022-04-15 6:31 ` Li, Ming
2022-04-15 7:07 ` yukuai (C)
2022-04-08 7:39 ` [PATCH -next RFC v2 4/8] blk-mq: don't preempt tag under heavy load Yu Kuai
2022-04-08 14:24 ` Bart Van Assche
2022-04-09 2:38 ` yukuai (C)
2022-04-08 7:39 ` [PATCH -next RFC v2 5/8] sbitmap: force tag preemption if free tags are sufficient Yu Kuai
2022-04-08 7:39 ` [PATCH -next RFC v2 6/8] blk-mq: force tag preemption for split bios Yu Kuai
2022-04-08 7:39 ` [PATCH -next RFC v2 7/8] blk-mq: record how many tags are needed for splited bio Yu Kuai
2022-04-08 7:39 ` [PATCH -next RFC v2 8/8] sbitmap: wake up the number of threads based on required tags Yu Kuai
2022-04-08 14:31 ` Bart Van Assche
2022-04-09 2:19 ` yukuai (C)
2022-04-08 21:13 ` Bart Van Assche
2022-04-09 2:17 ` yukuai (C)
2022-04-09 4:16 ` Bart Van Assche
2022-04-09 7:01 ` yukuai (C)
2022-04-12 3:20 ` Bart Van Assche
2022-04-08 19:10 ` [PATCH -next RFC v2 0/8] improve tag allocation under heavy load Jens Axboe
2022-04-09 2:26 ` yukuai (C)
2022-04-09 2:28 ` Jens Axboe
2022-04-09 2:34 ` yukuai (C)
2022-04-09 7:14 ` yukuai (C)
2022-04-09 21:31 ` Bart Van Assche
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=fbe5565e-4dbe-bb71-e3c4-c33eb470a8d2@huawei.com \
--to=yukuai3@huawei.com \
--cc=andriy.shevchenko@linux.intel.com \
--cc=axboe@kernel.dk \
--cc=bvanassche@acm.org \
--cc=john.garry@huawei.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ming.lei@redhat.com \
--cc=yi.zhang@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;
as well as URLs for NNTP newsgroup(s).