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, 16 Aug 2022 12:42:18 -1000
From:   Tejun Heo <tj@kernel.org>
To:     Imran Khan <imran.f.khan@oracle.com>
Cc:     gregkh@linuxfoundation.org, viro@zeniv.linux.org.uk,
        linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org
Subject: Re: [RESEND PATCH 4/5] kernfs: Replace per-fs rwsem with hashed
 rwsems.
Message-ID: <YvwdShstDCK+uQ+R@slm.duckdns.org>
References: <20220810111017.2267160-1-imran.f.khan@oracle.com>
 <20220810111017.2267160-5-imran.f.khan@oracle.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220810111017.2267160-5-imran.f.khan@oracle.com>
Precedence: bulk

Hello,

I'm bandwidth constrained right now and can't really shepherd this patchset,
so I'm not gonna ack or nack the series. That said, here are my thoughts
after glancing through it:

* I find the returning-with-rwsem-held interface and usage odd. We return
  with locks held all the time, so that part in itself is fine but how it's
  used in the proposed patch is pretty alien.

* I don't understand why the topo_mutex is needed. What is its relationship
  with rename_lock?

* Can't the double/triple lock helpers loop over the sorted list instead of
  if'ing each case?

Thanks.

-- 
tejun