From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754825AbaHLCDL (ORCPT ); Mon, 11 Aug 2014 22:03:11 -0400 Received: from linuxhacker.ru ([217.76.32.60]:34387 "EHLO fiona.linuxhacker.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753121AbaHLCDJ convert rfc822-to-8bit (ORCPT ); Mon, 11 Aug 2014 22:03:09 -0400 Subject: Re: [PATCH] drivers: staging: lustre: fix sparse warnings / delete unused function Mime-Version: 1.0 (Apple Message framework v1283) Content-Type: text/plain; charset=windows-1252 From: Oleg Drokin In-Reply-To: <4F91D9A1-B87D-498F-8089-7008F6BD4F30@intel.com> Date: Mon, 11 Aug 2014 22:02:54 -0400 Cc: Greg KH , Benedict Boerger , , , Content-Transfer-Encoding: 8BIT Message-Id: <1A7F74DC-3922-4626-BDDF-E8E6C90C221C@linuxhacker.ru> References: <1407515182-24287-1-git-send-email-benedict.boerger@cs.tu-dortmund.de> <20140810043443.GA12200@kroah.com> <4F91D9A1-B87D-498F-8089-7008F6BD4F30@intel.com> To: Oleg Drokin X-Mailer: Apple Mail (2.1283) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Aug 11, 2014, at 9:58 PM, Oleg Drokin wrote: > > On Aug 10, 2014, at 12:34 AM, Greg KH wrote: >> Very odd, with this patch applied, I now get a build warning: >> >> drivers/staging/lustre/lnet/lnet/config.c: In function ‘lnet_parse_ip2nets’: >> drivers/staging/lustre/lnet/lnet/config.c:1193:2: warning: ‘ipaddrs’ may be used uninitialized in this function [-Wmaybe-uninitialized] >> lnet_ipaddr_free_enumeration(ipaddrs, nip); >> ^ > > What's even more odd is that the very previous line also uses ipaddrs, yet it does not trigger the warning. > In any case I believe it's a false positive, since the only way it could remain unitialized is if > libcfs_ipif_enumerate returned negative value (stored in nif in this function) and we check for it being more than zero > before getting to that part of code. > > I'll submit a patch assigning ipaddrs to NULL as a workaround shortly. Err, looking at the wrong function there, it's lnet_ipaddr_enumerate that returns negative if ipaddrs was not changed. Anyway still false positive.