Date:   Mon, 30 Jan 2023 22:12:43 -0500
From:   "Theodore Ts'o" <tytso@mit.edu>
To:     zhanchengbin <zhanchengbin1@huawei.com>
Cc:     linux-ext4@vger.kernel.org, linfeilong <linfeilong@huawei.com>,
        liuzhiqiang26@huawei.com
Subject: Re: [PATCH] unix_io.c: fix deadlock problem in unix_write_blk64
Message-ID: <Y9iHKxQ8jszr+/g4@mit.edu>
References: <310fb77f-dfed-1196-c4ee-30d5138ee5a2@huawei.com>
 <Y9DJnh9P+FijguS2@mit.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Y9DJnh9P+FijguS2@mit.edu>
Precedence: bulk

On Wed, Jan 25, 2023 at 01:18:06AM -0500, Theodore Ts'o wrote:
> 
> Fortunately, we're safe on the read side, because we currently very
> carefully do not call raw_read_blk() while holding the CACHE_MUTEX.
> Instead, we write the data from the user-supplied buffer, and *then*
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> take the cache mutex, and then save the data from the user-supplied
> buffer into the cache.

Correction: the last sentence of this paragraph should begin:

"Instead, we READ the data INTO the user-supplied buffer, ..."

I have a series of patches that should address the deadlock issue,
that I'll be sending out shortly.

						- Ted