Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [RESEND PATCH] blk-mq: fix hang caused by freeze/unfreeze
 sequence
To:     Bart Van Assche <bvanassche@acm.org>, linux-block@vger.kernel.org
Cc:     shirley.ma@oracle.com, martin.petersen@oracle.com,
        Roman Pen <roman.penyaev@profitbricks.com>,
        Akinobu Mita <akinobu.mita@gmail.com>,
        Tejun Heo <tj@kernel.org>, Jens Axboe <axboe@kernel.dk>,
        Christoph Hellwig <hch@lst.de>, linux-kernel@vger.kernel.org
References: <20190409090828.16282-1-bob.liu@oracle.com>
 <0763cb5a-5598-69e3-e5ac-765989aab5b1@acm.org>
From:   Bob Liu <bob.liu@oracle.com>
Message-ID: <9ca4314d-fe45-9b24-8ba0-44d80f790502@oracle.com>
Date:   Sun, 14 Apr 2019 21:09:43 +0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.5.1
MIME-Version: 1.0
In-Reply-To: <0763cb5a-5598-69e3-e5ac-765989aab5b1@acm.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 4/13/19 11:42 AM, Bart Van Assche wrote:
> On 4/9/19 2:08 AM, Bob Liu wrote:
>>  void blk_freeze_queue_start(struct request_queue *q)
>>  {
>> -	int freeze_depth;
>> -
>> -	freeze_depth = atomic_inc_return(&q->mq_freeze_depth);
>> -	if (freeze_depth == 1) {
>> +	mutex_lock(&q->mq_freeze_lock);
>> +	if (++q->mq_freeze_depth == 1) {
>>  		percpu_ref_kill(&q->q_usage_counter);
>> +		mutex_unlock(&q->mq_freeze_lock);
>>  		if (queue_is_mq(q))
>>  			blk_mq_run_hw_queues(q, false);
>> +	} else {
>> +		mutex_unlock(&q->mq_freeze_lock);
>>  	}
>>  }
> Have you considered to move the mutex_unlock() call to the end of the function
> such that there is only one mutex_unlock() call instead of two? In case you> would be worried about holding the mutex around the code that runs the queue,
> how about changing the blk_mq_run_hw_queues() call such that the queues are
> run async?
>
 
Yes, but this bug is difficult to reproduce. I'd prefer keep the change minimum.