Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 14 Oct 2021 23:30:03 -0700
From:   Christoph Hellwig <hch@infradead.org>
To:     Shiyang Ruan <ruansy.fnst@fujitsu.com>
Cc:     linux-kernel@vger.kernel.org, linux-xfs@vger.kernel.org,
        nvdimm@lists.linux.dev, linux-mm@kvack.org,
        linux-fsdevel@vger.kernel.org, djwong@kernel.org,
        dan.j.williams@intel.com, david@fromorbit.com, hch@infradead.org,
        jane.chu@oracle.com
Subject: Re: [PATCH v7 1/8] dax: Use rwsem for dax_{read,write}_lock()
Message-ID: <YWkf63g44Puobxrp@infradead.org>
References: <20210924130959.2695749-1-ruansy.fnst@fujitsu.com>
 <20210924130959.2695749-2-ruansy.fnst@fujitsu.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210924130959.2695749-2-ruansy.fnst@fujitsu.com>
Precedence: bulk

On Fri, Sep 24, 2021 at 09:09:52PM +0800, Shiyang Ruan wrote:
> In order to introduce dax holder registration, we need a write lock for
> dax.  Because of the rarity of notification failures and the infrequency
> of registration events, it would be better to be a global lock rather
> than per-device.  So, change the current lock to rwsem and introduce a
> write lock for registration.

I don't think taking the rw_semaphore everywhere will scale, as
basically any DAX based I/O will take it (in read mode).

So at a minimum we'd need a per-device percpu_rw_semaphore if we want
to go there.