linux-kernel - Re: [RFC PATCH v4 05/13] workqueue, ktask: renice helper threads to prevent starvation

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: <20181120170320.2qvlwnpohzxow4bm@ca-dmjordan1.us.oracle.com>
Date:   Tue, 20 Nov 2018 09:03:21 -0800
From:   Daniel Jordan <daniel.m.jordan@...cle.com>
To:     Tejun Heo <tj@...nel.org>
Cc:     Daniel Jordan <daniel.m.jordan@...cle.com>, linux-mm@...ck.org,
        kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        aarcange@...hat.com, aaron.lu@...el.com, akpm@...ux-foundation.org,
        alex.williamson@...hat.com, bsd@...hat.com,
        darrick.wong@...cle.com, dave.hansen@...ux.intel.com,
        jgg@...lanox.com, jwadams@...gle.com, jiangshanlai@...il.com,
        mhocko@...nel.org, mike.kravetz@...cle.com,
        Pavel.Tatashin@...rosoft.com, prasad.singamsetty@...cle.com,
        rdunlap@...radead.org, steven.sistare@...cle.com,
        tim.c.chen@...el.com, vbabka@...e.cz
Subject: Re: [RFC PATCH v4 05/13] workqueue, ktask: renice helper threads to
 prevent starvation

On Tue, Nov 20, 2018 at 08:33:19AM -0800, Tejun Heo wrote:
> On Mon, Nov 19, 2018 at 08:45:54AM -0800, Daniel Jordan wrote:
> > So instead of flush_work_at_nice, how about this?:
> > 
> > void renice_work_sync(work_struct *work, long nice);
> 
> Wouldn't renice_or_cancel make more sense?

I guess you mean, for renicing if the work is started and canceling if it
hasn't?

Then yes, it would in this case, since if a ktask work hasn't start, there's no
point in running it--there are no more chunks for it to work on.

Was attempting to generalize for other cases when the work did need to be run,
but designing for the future can be dicey, and I'm fine either way.  So absent
other opinions, I'll do renice_or_cancel.