Received-SPF: pass (google.com: domain of linux-kernel+bounces-29272-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Message-ID: <5ff36535-7cd3-4e8f-a477-9d2a98d18dd9@acm.org>
Date: Wed, 17 Jan 2024 09:29:44 -0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [RFC PATCH 1/1] block: introduce activity based ioprio
Content-Language: en-US
To: "zhaoyang.huang" <zhaoyang.huang@unisoc.com>, Jens Axboe
 <axboe@kernel.dk>, linux-block@vger.kernel.org,
 linux-kernel@vger.kernel.org, Zhaoyang Huang <huangzhaoyang@gmail.com>,
 steve.kang@unisoc.com
References: <20240117092348.2873928-1-zhaoyang.huang@unisoc.com>
From: Bart Van Assche <bvanassche@acm.org>
In-Reply-To: <20240117092348.2873928-1-zhaoyang.huang@unisoc.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 1/17/24 01:23, zhaoyang.huang wrote:
> +static enum dd_prio dd_req_ioprio(struct request *rq)
> +{
> +	enum dd_prio prio;
> +	const u8 ioprio_class = dd_rq_ioclass(rq);
> +#ifdef CONFIG_ACTIVITY_BASED_IOPRIO
> +	struct bio *bio;
> +	struct bio_vec bv;
> +	struct bvec_iter iter;
> +	struct page *page;
> +	int gen = 0;
> +	int cnt = 0;
> +
> +	if (req_op(rq) == REQ_OP_READ) {
> +		__rq_for_each_bio(bio, rq) {
> +			bio_for_each_bvec(bv, bio, iter) {
> +				page = bv.bv_page;
> +				gen += PageWorkingset(page) ? 1 : 0;
> +				cnt++;
> +			}
> +		}
> +		prio = (gen >= cnt / 2) ? ioprio_class_to_prio[IOPRIO_CLASS_RT] :
> +			ioprio_class_to_prio[ioprio_class];
> +	} else
> +		prio = ioprio_class_to_prio[ioprio_class];
> +#else
> +	prio = ioprio_class_to_prio[ioprio_class];
> +#endif
> +	return prio;
> +}

I don't like it that code is introduced in the mq-deadline scheduler
that accesses page cache information. Isn't that a layering violation?
Additionally, this approach only works for buffered I/O and not for
direct I/O. Shouldn't the I/O submitter set the I/O priority instead of
deciding the I/O priority in the mq-deadline scheduler?

Bart.