Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230710183338.58531-1-ivan@cloudflare.com> <2023071039-negate-stalemate-6987@gregkh>
 <CABWYdi39+TJd1qV3nWs_eYc7XMC0RvxG22ihfq7rzuPaNvn1cQ@mail.gmail.com> <jy7ktvlb4tkg6pl2vll6u4gozfji7giddyseypj4w2d2ue4gvn@4tw7dmjy4hfv>
In-Reply-To: <jy7ktvlb4tkg6pl2vll6u4gozfji7giddyseypj4w2d2ue4gvn@4tw7dmjy4hfv>
From:   Ivan Babrou <ivan@cloudflare.com>
Date:   Tue, 25 Jul 2023 16:58:47 -0700
Message-ID: <CABWYdi1k1gNwkWT8TH7kPv=tA8qaZbjaaogYoMRnFPtBqvR_Uw@mail.gmail.com>
Subject: Re: [PATCH] kernfs: attach uuid for every kernfs and report it in fsid
To:     =?UTF-8?Q?Michal_Koutn=C3=BD?= <mkoutny@suse.com>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        linux-fsdevel@vger.kernel.org, kernel-team@cloudflare.com,
        linux-kernel@vger.kernel.org, cgroups@vger.kernel.org,
        Tejun Heo <tj@kernel.org>, Hugh Dickins <hughd@google.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Amir Goldstein <amir73il@gmail.com>,
        Christoph Hellwig <hch@lst.de>, Jan Kara <jack@suse.cz>,
        Zefan Li <lizefan.x@bytedance.com>,
        Johannes Weiner <hannes@cmpxchg.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Jul 25, 2023 at 7:07=E2=80=AFAM Michal Koutn=C3=BD <mkoutny@suse.co=
m> wrote:
>
> Hello.
>
> On Mon, Jul 10, 2023 at 02:21:10PM -0700, Ivan Babrou <ivan@cloudflare.co=
m> wrote:
> > I want to monitor cgroup changes, so that I can have an up to date map
> > of inode -> cgroup path, so that I can resolve the value returned from
> > bpf_get_current_cgroup_id() into something that a human can easily
> > grasp (think system.slice/nginx.service).
>
> Have you considered cgroup_path_from_kernfs_id()?

* It's not available from bpf from what I see (should I send a patch?)
* It turns short numeric keys into large string keys (you mention this belo=
w)
* It's a lot more work upfront for every event under a spinlock

> > Currently I do a full sweep to build a map, which doesn't work if a
> > cgroup is short lived, as it just disappears before I can resolve it.
> > Unfortunately, systemd recycles cgroups on restart, changing inode
> > number, so this is a very real issue.
>
> So, a historical map of cgroup id -> path is also useful for you, right?
> (IOW, cgroup_path_from_kernfs_id() is possible but it'd inflate log
> buffer size if full paths were stored instead of ids.)

For the most part the historical map would not be necessary if we had
cgroup paths (except for the points I mentioned above).

> (I think a similar map would be beneficial for SCM_CGROUP [1] idea too.)

Yes, it seems like it.

> > There's also this old wiki page from systemd:
> >
> > * https://freedesktop.org/wiki/Software/systemd/Optimizations
>
> The page also states:
>
> > Last edited Sat 18 May 2013 08:20:38 AM UTC
>
> Emptiness notifications via release_agent are so 2016 :-), unified
> hiearchy has more convenient API [2], this is FTR.

Sure, but these aren't arguments against having fanotify for cgroup filesys=
tem.