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;
Sender: Tejun Heo <htejun@gmail.com>
Date:   Tue, 6 Sep 2022 07:13:21 -1000
From:   Tejun Heo <tj@kernel.org>
To:     Michal =?iso-8859-1?Q?Koutn=FD?= <mkoutny@suse.com>
Cc:     cgroups@vger.kernel.org, linux-kernel@vger.kernel.org,
        Zefan Li <lizefan.x@bytedance.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Dan Carpenter <dan.carpenter@oracle.com>
Subject: Re: [PATCH] cgroup: Reorganize css_set_lock and kernfs path
 processing
Message-ID: <Yxd/sUQ/NB3NlC6f@slm.duckdns.org>
References: <20220905170944.23071-1-mkoutny@suse.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20220905170944.23071-1-mkoutny@suse.com>
Precedence: bulk

Hello, Michal.

On Mon, Sep 05, 2022 at 07:09:44PM +0200, Michal Koutn? wrote:
> I considered adding get_cgroup() into current_cgns_cgroup_from_root to
> avoid reliance on the transitive pinning via css_set. 
> After reasoning about no asynchronous NS switch and v1 hiearchies kill_sb it
> didn't seem to bring that much benefit (it didn't compose well with
> BUG_ON(!cgrp) neither).

I still think this is too subtle and incidental. If we go this way, we'd
need to add comments explaining why this obviously confusing pattern (lock,
find obj, unlock, use obj) is being used which goes into how the object is
directly pinned through the css_set which happens to be pinned also
because... and so on. Even if the code looks a bit uglier, I'd much prefer
straight-forward pinning.

Thanks.

-- 
tejun