From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jonathan Corbet <corbet@lwn.net>
Subject: Re: [RFC PATCH for-next 3/4] epoll: struct epoll support
Date: Mon, 24 Feb 2014 11:59:07 -0700
Message-ID: <20140224115907.07092911@lwn.net>
References: <1393206162-18151-1-git-send-email-n1ght.4nd.d4y@gmail.com>
	<1393206162-18151-4-git-send-email-n1ght.4nd.d4y@gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 8bit
Cc: viro@zeniv.linux.org.uk, linux-fsdevel@vger.kernel.org,
	linux-kernel@vger.kernel.org
To: Nathaniel Yazdani <n1ght.4nd.d4y@gmail.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <1393206162-18151-4-git-send-email-n1ght.4nd.d4y@gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: linux-fsdevel.vger.kernel.org

So I was just looking things over quickly, and something jumped out at
me.  In ep_control():

> +	} else if (!(*io) && epi) {
> +		/* delete this eventpoll entry */
> +		if (is_file_epoll(target)) {
> +			tep = target->private_data;
> +			mutex_lock_nested(&tep->mtx, 1);
> +		}
> +		if (is_file_epoll(target))
> +			mutex_lock_nested(&tep->mtx, 1);

How could that possibly work?  I can't imagine tep->mtx is going to
react well to being locked a second time...

jon