Received-SPF: pass (google.com: domain of linux-kernel+bounces-188332-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Precedence: bulk
MIME-Version: 1.0
References: <20240522074829.1750204-1-yunlong.xing@unisoc.com>
 <5166bc31-1fd9-4f7f-bc51-f1f50d9d5483@acm.org> <68cfbc08-6d39-4bc6-854d-5df0c94dbfd4@kernel.dk>
 <f6d3e1f2-e004-49bb-b6c1-969915ccab37@acm.org> <CA+3AYtS=5=_4cQK3=ASvgqQWWCohOsDuVwqiuDgErAnBJ17bBw@mail.gmail.com>
 <ab21593c-d32e-40b4-9238-60acdd402fd1@kernel.dk> <CA+3AYtTbkG_8KWNWJ8rZ-z=v-V+A9CqKCUUsXLPJyHZgL-FjwQ@mail.gmail.com>
 <fac0eb31-55f4-43fe-9e85-6363031aa5ce@kernel.dk>
In-Reply-To: <fac0eb31-55f4-43fe-9e85-6363031aa5ce@kernel.dk>
From: yunlong xing <yunlongxing23@gmail.com>
Date: Fri, 24 May 2024 13:26:16 +0800
Message-ID: <CA+3AYtR-6MxZg9_Mwqe0-w0jf15dr7=fOsMw-Y3Btbzo=jPd9A@mail.gmail.com>
Subject: Re: [PATCH] loop: inherit the ioprio in loop woker thread
To: Jens Axboe <axboe@kernel.dk>
Cc: Bart Van Assche <bvanassche@acm.org>, Yunlong Xing <yunlong.xing@unisoc.com>, niuzhiguo84@gmail.com, 
	Hao_hao.Wang@unisoc.com, linux-block@vger.kernel.org, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Jens Axboe <axboe@kernel.dk> =E4=BA=8E2024=E5=B9=B45=E6=9C=8823=E6=97=A5=E5=
=91=A8=E5=9B=9B 22:58=E5=86=99=E9=81=93=EF=BC=9A
>
> On 5/23/24 8:52 AM, yunlong xing wrote:
> > Jens Axboe <axboe@kernel.dk> ?2024?5?23??? 21:04???
> >>
> >> On 5/23/24 12:04 AM, yunlong xing wrote:
> >>> Bart Van Assche <bvanassche@acm.org> ?2024?5?23??? 02:12???
> >>>>
> >>>> On 5/22/24 10:57, Jens Axboe wrote:
> >>>>> On 5/22/24 11:38 AM, Bart Van Assche wrote:
> >>>>>> On 5/22/24 00:48, Yunlong Xing wrote:
> >>>>>>> @@ -1913,6 +1921,10 @@ static void loop_handle_cmd(struct loop_cm=
d *cmd)
> >>>>>>>            set_active_memcg(old_memcg);
> >>>>>>>            css_put(cmd_memcg_css);
> >>>>>>>        }
> >>>>>>> +
> >>>>>>> +    if (ori_ioprio !=3D cmd_ioprio)
> >>>>>>> +        set_task_ioprio(current, ori_ioprio);
> >>>>>>> +
> >>>>>>>     failed:
> >>>>>>>        /* complete non-aio request */
> >>>>>>>        if (!use_aio || ret) {
> >>>>>>
> >>>>>> Does adding this call in the hot path have a measurable performanc=
e impact?
> >>>>>
> >>>>> It's loop, I would not be concerned with overhead. But it does look=
 pretty
> >>>>> bogus to modify the task ioprio from here.
> >>>>
> >>>> Hi Jens,
> >>>>
> >>>> Maybe Yunlong uses that call to pass the I/O priority to the I/O sub=
mitter?
> >>>>
> >>>> I think that it is easy to pass the I/O priority to the kiocb submit=
ted by
> >>>> lo_rw_aio() without calling set_task_ioprio().
> >>>>
> >>>> lo_read_simple() and lo_write_simple() however call vfs_iter_read() =
/
> >>>> vfs_iter_write(). This results in a call of do_iter_readv_writev() a=
nd
> >>>> init_sync_kiocb(). The latter function calls get_current_ioprio(). T=
his is
> >>>> probably why the set_task_ioprio() call has been added?
> >>>
> >>> Yeah that's why I call set_task_ioprio.  I want to the loop kwoker
> >>> task?submit I/O to the real disk device?can pass the iopriority of th=
e
> >>> loop device request? both lo_rw_aio() and
> >>> lo_read_simple()/lo_write_simple().
> >>
> >> And that's a totally backwards and suboptimal way to do it. The task
> >> priority is only used as a last resort lower down, if the IO itself
> >> hasn't been appropriately marked.
> >>
> >> Like I said, it's back to the drawing board on this patch, there's no
> >> way it's acceptable in its current form.
> >>
> >> --
> >> Jens Axboe
> >>
> > Thanks for your advice. So, you can't accept pass the ioprio by
> > set_task_ioprio?
>
> Not sure how many times I'd have to state that, no.
Of course, I understand what you mean. I would like to ask if you only
disagree with this part. Sorry for missing a word "just".

Back to the patch,  I couldn't find a better way to pass the ioprio in the
lo_read/write_simple(). Do you have some suggestions or ideas=EF=BC=9F
>
> > If only the method of lo_rw_aio() counld you accept? I don't want to
> > submit this part of the modifications separately. I just want to know,
> > this is ok to you or not?
>
> Inheriting the kiocb ioprio from the request is the right approach, so
> yeah that part is fine.
>
> --
> Jens Axboe
>