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] scsi: ufs: Fix proper API to send HPB pre-request
To:     Christoph Hellwig <hch@infradead.org>,
        James Bottomley <jejb@linux.ibm.com>
Cc:     daejun7.park@samsung.com, ALIM AKHTAR <alim.akhtar@samsung.com>,
        "avri.altman@wdc.com" <avri.altman@wdc.com>,
        "martin.petersen@oracle.com" <martin.petersen@oracle.com>,
        "huobean@gmail.com" <huobean@gmail.com>,
        Keoseong Park <keosung.park@samsung.com>,
        "linux-scsi@vger.kernel.org" <linux-scsi@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
References: <CGME20211027223619epcms2p60bbc74c9ba9757c58709a99acd0892ff@epcms2p6>
 <20211027223619epcms2p60bbc74c9ba9757c58709a99acd0892ff@epcms2p6>
 <0f9229c3c4c7859524411a47db96a3b53ac89c90.camel@linux.ibm.com>
 <YXrBTHmu/fiAaZH5@infradead.org>
From:   Bart Van Assche <bvanassche@acm.org>
Message-ID: <54b45df9-9339-c69d-73b5-9c293449b849@acm.org>
Date:   Thu, 28 Oct 2021 10:07:52 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.14.0
MIME-Version: 1.0
In-Reply-To: <YXrBTHmu/fiAaZH5@infradead.org>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 10/28/21 8:27 AM, Christoph Hellwig wrote:
> On Thu, Oct 28, 2021 at 10:28:01AM -0400, James Bottomley wrote:
>> If the block people are happy with this, then I'm OK with it, but it
>> doesn't look like you've solved the fanout deadlock problem because
>> this new mechanism is still going to allocate a new tag.
> 
> Yes, same problem as before.

Hi Christoph,

I spent some time looking around for other examples of allocating and
inserting a request from inside block layer callbacks. I only found one
such example, namely in the NVMe core. nvme_timeout() calls
nvme_alloc_request() and blk_execute_rq_nowait(). The difference between
what the UFS HPB code is doing and what nvme_timeout() does doesn't seem
that big to me. For clarity, I don't like the UFS HPB protocol nor how
support for that protocol has been implemented. However, I don't see how
the UFS HPB implementation would complicate maintenance of the block
layer core. Am I perhaps missing something?

Thanks,

Bart.