netdev - Re: [PATCH bpf-next v5 06/34] bpf: prepare for memcg-based memory accounting for bpf maps

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: <20201113194023.GD2955309@carbon.dhcp.thefacebook.com>
Date:   Fri, 13 Nov 2020 11:40:23 -0800
From:   Roman Gushchin <guro@...com>
To:     Song Liu <songliubraving@...com>
CC:     bpf <bpf@...r.kernel.org>, Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        Andrii Nakryiko <andrii@...nel.org>,
        Shakeel Butt <shakeelb@...gle.com>,
        "linux-mm@...ck.org" <linux-mm@...ck.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Kernel Team <Kernel-team@...com>
Subject: Re: [PATCH bpf-next v5 06/34] bpf: prepare for memcg-based memory
 accounting for bpf maps

On Fri, Nov 13, 2020 at 09:46:49AM -0800, Song Liu wrote:
> 
> 
> > On Nov 12, 2020, at 2:15 PM, Roman Gushchin <guro@...com> wrote:
> 
> [...]
> 
> > 
> > +#ifdef CONFIG_MEMCG_KMEM
> > +static __always_inline int __bpf_map_update_elem(struct bpf_map *map, void *key,
> > +						 void *value, u64 flags)
> > +{
> > +	struct mem_cgroup *old_memcg;
> > +	bool in_interrupt;
> > +	int ret;
> > +
> > +	/*
> > +	 * If update from an interrupt context results in a memory allocation,
> > +	 * the memory cgroup to charge can't be determined from the context
> > +	 * of the current task. Instead, we charge the memory cgroup, which
> > +	 * contained a process created the map.
> > +	 */
> > +	in_interrupt = in_interrupt();
> > +	if (in_interrupt)
> > +		old_memcg = set_active_memcg(map->memcg);
> 
> set_active_memcg() checks in_interrupt() again. Maybe we can introduce another
> helper to avoid checking it twice? Something like
> 
> static inline struct mem_cgroup *
> set_active_memcg_int(struct mem_cgroup *memcg)
> {
>         struct mem_cgroup *old;
> 
>         old = this_cpu_read(int_active_memcg);
>         this_cpu_write(int_active_memcg, memcg);
>         return old;
> }

Yeah, it's a good idea!

in_interrupt() check is very cheap (like checking some bits in a per-cpu variable),
so I don't think there will be any measurable difference. So I suggest to implement
it later as an enhancement on top (maybe in the next merge window), to avoid an another
delay. Otherwise I'll need to send a patch to mm@, wait for reviews and an inclusion
into the mm tree, etc). Does it work for you?

Thanks!