From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754394Ab0EIWFP (ORCPT ); Sun, 9 May 2010 18:05:15 -0400 Received: from mail-yw0-f198.google.com ([209.85.211.198]:47037 "EHLO mail-yw0-f198.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753817Ab0EIWEv convert rfc822-to-8bit (ORCPT ); Sun, 9 May 2010 18:04:51 -0400 MIME-Version: 1.0 In-Reply-To: <1273011539-13994-1-git-send-email-agust@denx.de> References: <1273011539-13994-1-git-send-email-agust@denx.de> From: Grant Likely Date: Mon, 10 May 2010 00:04:30 +0200 X-Google-Sender-Auth: 3f208bfce532a988 Message-ID: Subject: Re: [PATCH] serial: mpc52xx_uart: fix null pointer dereference To: Anatolij Gustschin Cc: linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org, leonardo.lists@gmail.com, gregkh@suse.de, Andrew Morton Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, May 5, 2010 at 12:18 AM, Anatolij Gustschin wrote: > Commit 6acc6833510db8f72b5ef343296d97480555fda9 > introduced NULL pointer dereference and kernel crash > on ppc32 machines while booting. Fix this bug now. > > Reported-by: Leonardo Chiquitto > Signed-off-by: Anatolij Gustschin Thanks for being so quick to pick up on this Anatolij. I've been traveling and not able to respond to stuff. I see that Greg has already picked this up (thanks Greg!), but FWIW: Acked-by: Grant Likely Actually, now that I look, this fix needs to go into Linus' tree right away (not just linux-next) since the offending patch went in during the 2.6.34 merge window. Greg, I can either pick this up and push it out to Linus tomorrow, or let you do it. Whichever you prefer. Let me know so I don't cause conflicts in your tree. On another note, this patch isn't actually the right fix, but that's because this driver is such a horribly bad example of handling private data that doing the right thing requires some re-architecting. The driver should not need any of that extra code in the module_init hook. For now this is the best thing to do, but I'll dig into it tomorrow and see if I can churn out a patch to fix it properly for the 2.6.35 merge window. Cheers and thanks, g. > --- >  drivers/serial/mpc52xx_uart.c |    2 +- >  1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/drivers/serial/mpc52xx_uart.c b/drivers/serial/mpc52xx_uart.c > index a176ab4..02469c3 100644 > --- a/drivers/serial/mpc52xx_uart.c > +++ b/drivers/serial/mpc52xx_uart.c > @@ -1467,7 +1467,7 @@ mpc52xx_uart_init(void) >        /* >         * Map the PSC FIFO Controller and init if on MPC512x. >         */ > -       if (psc_ops->fifoc_init) { > +       if (psc_ops && psc_ops->fifoc_init) { >                ret = psc_ops->fifoc_init(); >                if (ret) >                        return ret; > -- > 1.7.0.4 > > -- Grant Likely, B.Sc., P.Eng. Secret Lab Technologies Ltd.