Subject: Re: [PATCH 5/9] block: add ability to flag write back caching on a
 device
To: Christoph Hellwig <hch@infradead.org>
References: <1459350477-16404-1-git-send-email-axboe@fb.com>
 <1459350477-16404-6-git-send-email-axboe@fb.com>
 <20160330154241.GA13131@infradead.org> <56FBF4D8.9000203@fb.com>
CC: <linux-kernel@vger.kernel.org>, <linux-fsdevel@vger.kernel.org>,
        <linux-block@vger.kernel.org>
From: Jens Axboe <axboe@fb.com>
Message-ID: <56FBFD96.9020501@fb.com>
Date: Wed, 30 Mar 2016 10:23:50 -0600
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101
 Thunderbird/38.6.0
MIME-Version: 1.0
In-Reply-To: <56FBF4D8.9000203@fb.com>
Content-Type: text/plain; charset="windows-1252"; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1132
Lines: 27

On 03/30/2016 09:46 AM, Jens Axboe wrote:
> On 03/30/2016 09:42 AM, Christoph Hellwig wrote:
>> On Wed, Mar 30, 2016 at 09:07:53AM -0600, Jens Axboe wrote:
>>> Add an internal helper and flag for setting whether a queue has
>>> write back caching, or write through (or none). Add a sysfs file
>>> to show this as well, and make it changeable from user space.
>>
>> As per previous discussion:  I don't really care about the way how we
>> tell the block layer we have a writeback cache, but a NAK to having
>> each driver call two interfaces to convey the same information.
>>
>> Please just look at q->flush_flag & REQ_FLUSH for now, and then
>> improvem the interface if you don't like it (I don't particularly like
>> it, but not to the point that I'm motivated enough to fix it :))
>
> That's fine, I don't mind making that change, just didn't do it for this
> version. I prefer if we change the cache flagging to be the primary way
> to signal it, it's more intuitive than REQ_FLUSH.
>
> It'll be in the next version.

Conversion series in here now:

http://git.kernel.dk/cgit/linux-block/log/?h=wb-buf-throttle

-- 
Jens Axboe