From mboxrd@z Thu Jan  1 00:00:00 1970
From: Bart Van Assche <bart.vanassche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
Subject: Re: [PATCH v2 0/3] new ib_drain_qp() API
Date: Thu, 11 Feb 2016 07:15:06 -0800
Message-ID: <56BCA57A.4000500@sandisk.com>
References: <cover.1454969695.git.swise@chelsio.com>
 <010901d16375$1a023210$4e069630$@opengridcomputing.com>
 <011601d1637b$8c01a3e0$a404eba0$@opengridcomputing.com>
 <56BA540B.4040405@sandisk.com>
 <011901d1637d$b5286400$1f792c00$@opengridcomputing.com>
 <012801d16384$f68884f0$e3998ed0$@opengridcomputing.com>
 <56BB11F0.9090203@dev.mellanox.co.il> <56BB4479.8090009@sandisk.com>
 <56BC6686.8030301@dev.mellanox.co.il>
Mime-Version: 1.0
Content-Type: text/plain; charset="utf-8"; format=flowed
Content-Transfer-Encoding: 7bit
Return-path: <linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
In-Reply-To: <56BC6686.8030301-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
To: Sagi Grimberg <sagig-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>, Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>, "linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" <linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Cc: 'Sagi Grimberg' <sagig-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>, 'Christoph Hellwig' <hch-jcswGhMUV9g@public.gmane.org>, 'Chuck Lever' <chuck.lever-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
List-Id: linux-rdma@vger.kernel.org

On 02/11/16 02:46, Sagi Grimberg wrote:
>>> I'd suggest to look at the CQ context and act accordingly, something
>>> like:
>>>
>>>      if (cq->poll_ctx == IB_POLL_DIRECT) {
>>>          while (!wait_for_completion_timeout(&sdrain.done,
>>>                  mescs_to_jiffies(100))
>>>              ib_process_cq_direct(cq, 1024)
>>>      } else {
>>>          wait_for_completion(&sdrain.done);
>>>      }
>>>
>>> Thoughts?
>>
>> Hello Sagi,
>>
>> The above code will only work as expected if the caller won't call
>> ib_process_cq_direct() for the same queue from another context.
>
> Why? ib_poll_cq is safe from multiple contexts. Is it because the
> completion structures are on stack? because if so we can easily move
> them to the cq...

Hello Sagi,

The srp_send_done() function accesses the ch->free_tx list without 
locking. This is safe because all existing 
ib_process_cq_direct(ch->send_cq, ...) calls occur while holding 
ch->lock. The approach suggested above breaks that assumption.

Bart.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html