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: <20200811170611.GB1507044@carbon.DHCP.thefacebook.com>
Date:   Tue, 11 Aug 2020 10:06:11 -0700
From:   Roman Gushchin <guro@...com>
To:     Johannes Weiner <hannes@...xchg.org>
CC:     Andrew Morton <akpm@...ux-foundation.org>,
        Dennis Zhou <dennis@...nel.org>, Tejun Heo <tj@...nel.org>,
        Christoph Lameter <cl@...ux.com>,
        Michal Hocko <mhocko@...nel.org>,
        Shakeel Butt <shakeelb@...gle.com>, <linux-mm@...ck.org>,
        <kernel-team@...com>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 4/5] mm: memcg: charge memcg percpu memory to the
 parent cgroup

On Tue, Aug 11, 2020 at 11:27:37AM -0400, Johannes Weiner wrote:
> On Tue, Jun 23, 2020 at 11:45:14AM -0700, Roman Gushchin wrote:
> > Memory cgroups are using large chunks of percpu memory to store vmstat
> > data.  Yet this memory is not accounted at all, so in the case when there
> > are many (dying) cgroups, it's not exactly clear where all the memory is.
> > 
> > Because the size of memory cgroup internal structures can dramatically
> > exceed the size of object or page which is pinning it in the memory, it's
> > not a good idea to simple ignore it.  It actually breaks the isolation
> > between cgroups.
> > 
> > Let's account the consumed percpu memory to the parent cgroup.
> > 
> > Signed-off-by: Roman Gushchin <guro@...com>
> > Acked-by: Dennis Zhou <dennis@...nel.org>
> 
> Acked-by: Johannes Weiner <hannes@...xchg.org>

Thank you!

> 
> This makes sense, and the accounting is in line with how we track and
> distribute child creation quotas (cgroup.max.descendants and
> cgroup.max.depth) up the cgroup tree.
> 
> I have one minor comment that isn't a dealbreaker for me:
> 
> > @@ -5069,13 +5069,15 @@ static int alloc_mem_cgroup_per_node_info(struct mem_cgroup *memcg, int node)
> >  	if (!pn)
> >  		return 1;
> >  
> > -	pn->lruvec_stat_local = alloc_percpu(struct lruvec_stat);
> > +	pn->lruvec_stat_local = alloc_percpu_gfp(struct lruvec_stat,
> > +						 GFP_KERNEL_ACCOUNT);
> >  	if (!pn->lruvec_stat_local) {
> >  		kfree(pn);
> >  		return 1;
> >  	}
> >  
> > -	pn->lruvec_stat_cpu = alloc_percpu(struct lruvec_stat);
> > +	pn->lruvec_stat_cpu = alloc_percpu_gfp(struct lruvec_stat,
> > +					       GFP_KERNEL_ACCOUNT);
> >  	if (!pn->lruvec_stat_cpu) {
> >  		free_percpu(pn->lruvec_stat_local);
> >  		kfree(pn);
> > @@ -5149,11 +5151,13 @@ static struct mem_cgroup *mem_cgroup_alloc(void)
> >  		goto fail;
> >  	}
> >  
> > -	memcg->vmstats_local = alloc_percpu(struct memcg_vmstats_percpu);
> > +	memcg->vmstats_local = alloc_percpu_gfp(struct memcg_vmstats_percpu,
> > +						GFP_KERNEL_ACCOUNT);
> >  	if (!memcg->vmstats_local)
> >  		goto fail;
> >  
> > -	memcg->vmstats_percpu = alloc_percpu(struct memcg_vmstats_percpu);
> > +	memcg->vmstats_percpu = alloc_percpu_gfp(struct memcg_vmstats_percpu,
> > +						 GFP_KERNEL_ACCOUNT);
> >  	if (!memcg->vmstats_percpu)
> >  		goto fail;
> >  
> > @@ -5202,7 +5206,9 @@ mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css)
> >  	struct mem_cgroup *memcg;
> >  	long error = -ENOMEM;
> >  
> > +	memalloc_use_memcg(parent);
> >  	memcg = mem_cgroup_alloc();
> > +	memalloc_unuse_memcg();
> 
> The disconnect between 1) requesting accounting and 2) which cgroup to
> charge is making me uneasy. It makes mem_cgroup_alloc() a bit of a
> handgrenade, because accounting to the current task is almost
> guaranteed to be wrong if the use_memcg() annotation were to get lost
> in a refactor or not make it to a new caller of the function.
> 
> The saving grace is that mem_cgroup_alloc() is pretty unlikely to be
> used elsewhere. And pretending it's an independent interface would be
> overengineering. But how about the following in mem_cgroup_alloc() and
> alloc_mem_cgroup_per_node_info() to document that caller relationship:
> 
> 	/* We charge the parent cgroup, never the current task */
> 	WARN_ON_ONCE(!current->active_memcg);

I have nothing against.

Andrew, can you, please, squash the following diff into the patch?

Thanks!

--

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 130093bdf74b..e25f2db7e61c 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -5137,6 +5137,9 @@ static int alloc_mem_cgroup_per_node_info(struct mem_cgroup *memcg, int node)
        if (!pn)
                return 1;
 
+       /* We charge the parent cgroup, never the current task */
+       WARN_ON_ONCE(!current->active_memcg);
+
        pn->lruvec_stat_local = alloc_percpu_gfp(struct lruvec_stat,
                                                 GFP_KERNEL_ACCOUNT);
        if (!pn->lruvec_stat_local) {
@@ -5219,6 +5222,9 @@ static struct mem_cgroup *mem_cgroup_alloc(void)
                goto fail;
        }
 
+       /* We charge the parent cgroup, never the current task */
+       WARN_ON_ONCE(!current->active_memcg);
+
        memcg->vmstats_local = alloc_percpu_gfp(struct memcg_vmstats_percpu,
                                                GFP_KERNEL_ACCOUNT);
        if (!memcg->vmstats_local)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ