From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755387Ab0BHKbj (ORCPT ); Mon, 8 Feb 2010 05:31:39 -0500 Received: from earthlight.etchedpixels.co.uk ([81.2.110.250]:48802 "EHLO bob.linux.org.uk" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753274Ab0BHKbh (ORCPT ); Mon, 8 Feb 2010 05:31:37 -0500 From: Alan Cox Subject: [PATCH 2/3] nozomi: Fix mutex handling To: greg@kroah.com, linux-kernel@vger.kernel.org, p.hardwick@option.com Date: Mon, 08 Feb 2010 10:07:04 +0000 Message-ID: <20100208100650.2794.82836.stgit@localhost.localdomain> In-Reply-To: <20100208100642.2794.73225.stgit@localhost.localdomain> References: <20100208100642.2794.73225.stgit@localhost.localdomain> User-Agent: StGIT/0.14.3 MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org The original author didn't realise the kernel lock was a drop while sleep lock so did clever (and wrong) things to work around the non need to avoid deadlocks. Remove the cleverness and the comment (as we don't hold the BKL now anyway in those paths) Signed-off-by: Alan Cox --- drivers/char/nozomi.c | 34 ++++++++++++---------------------- 1 files changed, 12 insertions(+), 22 deletions(-) diff --git a/drivers/char/nozomi.c b/drivers/char/nozomi.c index cd405bc..fac9157 100644 --- a/drivers/char/nozomi.c +++ b/drivers/char/nozomi.c @@ -1693,15 +1693,7 @@ static int ntty_write(struct tty_struct *tty, const unsigned char *buffer, if (!dc || !port) return -ENODEV; - if (unlikely(!mutex_trylock(&port->tty_sem))) { - /* - * must test lock as tty layer wraps calls - * to this function with BKL - */ - dev_err(&dc->pdev->dev, "Would have deadlocked - " - "return EAGAIN\n"); - return -EAGAIN; - } + mutex_lock(&port->tty_sem); if (unlikely(!port->port.count)) { DBG1(" "); @@ -1741,25 +1733,23 @@ exit: * This method is called by the upper tty layer. * #according to sources N_TTY.c it expects a value >= 0 and * does not check for negative values. + * + * If the port is unplugged report lots of room and let the bits + * dribble away so we don't block anything. */ static int ntty_write_room(struct tty_struct *tty) { struct port *port = tty->driver_data; - int room = 0; + int room = 4096; const struct nozomi *dc = get_dc_by_tty(tty); - if (!dc || !port) - return 0; - if (!mutex_trylock(&port->tty_sem)) - return 0; - - if (!port->port.count) - goto exit; - - room = port->fifo_ul.size - kfifo_len(&port->fifo_ul); - -exit: - mutex_unlock(&port->tty_sem); + if (dc) { + mutex_lock(&port->tty_sem); + if (port->port.count) + room = port->fifo_ul.size - + kfifo_len(&port->fifo_ul); + mutex_unlock(&port->tty_sem); + } return room; }