linux-kernel - Re: [PATCH -next 2/2] blk-throttle: fix lower control under super low iops limit

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date: Fri, 14 Jun 2024 09:09:18 +0800
From: Yu Kuai <yukuai1@...weicloud.com>
To: Tejun Heo <tj@...nel.org>, Yu Kuai <yukuai1@...weicloud.com>
Cc: josef@...icpanda.com, axboe@...nel.dk, cgroups@...r.kernel.org,
 linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
 yi.zhang@...wei.com, yangerkun@...wei.com, "yukuai (C)" <yukuai3@...wei.com>
Subject: Re: [PATCH -next 2/2] blk-throttle: fix lower control under super low
 iops limit

Hi,

在 2024/06/14 6:32, Tejun Heo 写道:
> Hello,
> 
> On Mon, May 13, 2024 at 08:08:48PM +0800, Yu Kuai wrote:
> ...
>> However, if limit is quite low, the result can be 0, then
>> allowed IO in the slice is 0, this will cause missing dispatch and
>> control will be lower than limit.
>>
>> For example, set iops_limit to 5 with HD disk, and test will found that
>> iops will be 3.
> 
> Hmm... can't this be solved by starting the next slice with the right
> credit?

Of course, this looks like feasible. I'll look into this.
> 
>>   static unsigned int tg_throtl_slice(struct throtl_grp *tg, int rw)
>>   {
>> +	if (tg->throtl_slice[rw])
>> +		return tg->throtl_slice[rw];
>>   	return tg->td->throtl_slice;
> 
> Because this is a bit nasty. If we want to use difference throttling slices
> for different cgroups, we might as well do it universally.

I do thought about this, however, because different cgroups have
different slice start, I convinced myself this is fine to have different
slice as well.

Thanks,
Kuai

> 
> Thanks.
>