linux-kernel - Re: [PATCH 07/10] writeback: factor out wb_dirty_freerun to remove more repeated freerun code

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

Message-ID: <7fecbde8-d4b8-c733-847e-a760efb41571@huaweicloud.com>
Date: Mon, 6 May 2024 09:53:49 +0800
From: Kemeng Shi <shikemeng@...weicloud.com>
To: Tejun Heo <tj@...nel.org>
Cc: willy@...radead.org, akpm@...ux-foundation.org, jack@...e.cz,
 linux-fsdevel@...r.kernel.org, linux-mm@...ck.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH 07/10] writeback: factor out wb_dirty_freerun to remove
 more repeated freerun code


Hello,
on 5/2/2024 1:18 AM, Tejun Heo wrote:
> On Mon, Apr 29, 2024 at 11:47:35AM +0800, Kemeng Shi wrote:
> ...
>> +static void wb_dirty_freerun(struct dirty_throttle_control *dtc,
>> +			     bool strictlimit)
>> +{
> ...
>> +	/*
>> +	 * LOCAL_THROTTLE tasks must not be throttled when below the per-wb
>> +	 * freerun ceiling.
>> +	 */
>> +	if (!(current->flags & PF_LOCAL_THROTTLE))
>> +		return;
> 
> Shouldn't this set free_run to true?
Originally, we will go freerun if PF_LOCAL_THROTTLE is set and number of dirty
pages is under freerun ceil. So if PF_LOCAL_THROTTLE is *not* set, freerun
should be false. Maybe I miss something and please correct me, Thanks.
> 
> Also, wouldn't it be better if these functions return bool instead of
> recording the result in dtc->freerun?
As I try to factor out balance_wb_limits to calculate freerun, dirty_exceeded
and position ratio of wb, so wb_dirty_freerun and wb_dirty_exceeded will be
called indirectly and balance_dirty_pages has to retrieve freerun and
dirty_exceeded from somewhere like dtc where position ratio is retrieved.
Would like to know any better idea.
Thanks.
> 
> Thanks.
>