linux-kernel - Re: [PATCH 4/4] writeback, memcg: Implement foreign dirty flushing

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: <20190807183434.GN136335@devbig004.ftw2.facebook.com>
Date:   Wed, 7 Aug 2019 11:34:34 -0700
From:   Tejun Heo <tj@...nel.org>
To:     Andrew Morton <akpm@...ux-foundation.org>
Cc:     axboe@...nel.dk, jack@...e.cz, hannes@...xchg.org,
        mhocko@...nel.org, vdavydov.dev@...il.com, cgroups@...r.kernel.org,
        linux-mm@...ck.org, linux-block@...r.kernel.org,
        linux-kernel@...r.kernel.org, kernel-team@...com, guro@...com
Subject: Re: [PATCH 4/4] writeback, memcg: Implement foreign dirty flushing

Hello,

On Tue, Aug 06, 2019 at 04:03:06PM -0700, Andrew Morton wrote:
> > +	if (i < MEMCG_CGWB_FRN_CNT) {
> > +		unsigned long update_intv =
> > +			min_t(unsigned long, HZ,
> > +			      msecs_to_jiffies(dirty_expire_interval * 10) / 8);
> 
> An explanation of what's going on here would be helpful.
> 
> Why "* 1.25" and not, umm "* 1.24"?

Just because /8 is cheaper.  It's likely that a fairly wide range of
numbers are okay for the above.  I'll add some comment to explain that
and why the specific constants are picked.

> > +void mem_cgroup_flush_foreign(struct bdi_writeback *wb)
> > +{
> > +	struct mem_cgroup *memcg = mem_cgroup_from_css(wb->memcg_css);
> > +	unsigned long intv = msecs_to_jiffies(dirty_expire_interval * 10);
> 
> Ditto.

This is just dirty expiration.  If the dirty data has expired,
writeback must already be in progress by its bdi_wb, so there's no
reason to scheduler foreign writeback.  Will add a comment.

Thanks.

-- 
tejun