Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH] nvme-tcp: Check if request has started before processing
 it
To:     Daniel Wagner <dwagner@suse.de>, Hannes Reinecke <hare@suse.de>
Cc:     Keith Busch <kbusch@kernel.org>, Jens Axboe <axboe@fb.com>,
        Christoph Hellwig <hch@lst.de>, linux-nvme@lists.infradead.org,
        linux-kernel@vger.kernel.org
References: <20210212181738.79274-1-dwagner@suse.de>
 <c3a682d3-58f7-f5cc-caaa-75c36ca464e2@grimberg.me>
 <20210212210929.GA3851@redsun51.ssa.fujisawa.hgst.com>
 <ddf87227-1ad3-b8be-23ba-460433f70a85@grimberg.me>
 <73e4914e-f867-c899-954d-4b61ae2b4c33@suse.de>
 <20210215104020.yyithlo2hkxqvguj@beryllium.lan>
From:   Sagi Grimberg <sagi@grimberg.me>
Message-ID: <a2064070-b511-ba6d-bd64-0b3abc208356@grimberg.me>
Date:   Mon, 15 Feb 2021 13:29:45 -0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.6.1
MIME-Version: 1.0
In-Reply-To: <20210215104020.yyithlo2hkxqvguj@beryllium.lan>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk


>>>>>> blk_mq_tag_to_rq() will always return a request if the command_id is
>>>>>> in the valid range. Check if the request has been started. If we
>>>>>> blindly process the request we might double complete a request which
>>>>>> can be fatal.
>>>>>
>>>>> How did you get to this one? did the controller send a completion for
>>>>> a completed/bogus request?
>>>>
>>>> If that is the case, then that must mean it's possible the driver could
>>>> have started the command id just before the bogus completion check. Data
>>>> iorruption, right?
> 
> 'during TCP LIF toggles and aggr relocates' testing the host
> crashes. TBH, I do not really know what is happening or what the test
> does. Still trying to figure out what's going on.

Well, I think we should probably figure out why that is happening first.

> I was just very
> surprised how much the code trusts the other side to behave correctly./

What does pci/rdma/fc does differently? What does scsi do here?

>>> Yes, which is why I don't think this check is very useful..
>>
>> I actually view that as a valid protection against spoofed frames.
>> Without it it's easy to crash the machine by injecting fake completions with
>> random command ids.
> 
> In this test scenario it's not even a spoofed frame; maybe just a
> confused controller.

Maybe... I am still not sure how this patch helps here though...