Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751722AbdGaTTZ (ORCPT ); Mon, 31 Jul 2017 15:19:25 -0400 Received: from mail-oi0-f68.google.com ([209.85.218.68]:37585 "EHLO mail-oi0-f68.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750971AbdGaTTX (ORCPT ); Mon, 31 Jul 2017 15:19:23 -0400 MIME-Version: 1.0 In-Reply-To: <1501517853.2466.12.camel@wdc.com> References: <20170731065016.2947796-1-arnd@arndb.de> <1501515117.2466.9.camel@wdc.com> <1501517853.2466.12.camel@wdc.com> From: Arnd Bergmann Date: Mon, 31 Jul 2017 21:19:22 +0200 X-Google-Sender-Auth: dS5ozP9N_xJg9d57P83buaSIWgA Message-ID: Subject: Re: [PATCH] infiniband: avoid overflow warning To: Bart Van Assche Cc: "linux-kernel@vger.kernel.org" , "keescook@chromium.org" , "linux-rdma@vger.kernel.org" , "parav@mellanox.com" , "monis@mellanox.com" , "Michal.Kalderon@cavium.com" , "sean.hefty@intel.com" , "danielmicay@gmail.com" , "Ariel.Elior@cavium.com" , "hal.rosenstock@gmail.com" , "davem@davemloft.net" , "dledford@redhat.com" , "noaos@mellanox.com" Content-Type: text/plain; charset="UTF-8" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1433 Lines: 29 On Mon, Jul 31, 2017 at 6:17 PM, Bart Van Assche wrote: > On Mon, 2017-07-31 at 18:04 +0200, Arnd Bergmann wrote: >> On Mon, Jul 31, 2017 at 5:32 PM, Bart Van Assche wrote: >> > So inetaddr_event() assigns AF_INET so .sin_family and gcc warns about code >> > that is only executed if .sin_family == AF_INET6? Since this warning is the >> > result of incorrect interprocedural analysis by gcc, shouldn't this be >> > reported as a bug to the gcc authors? >> >> I think the interprocedural analysis here is just a little worse than it could >> be, but it's not actually correct. It's not gcc that prints the warning (if >> it did, then I'd agree it would be a gcc bug) but the warning is triggered >> intentionally by the fortified version of memcpy in include/linux/string.h. >> >> The problem as I understand it is that gcc cannot guarantee that it >> tracks the value of addr->sa_family at least as far as the size of the >> stack object, and it has no strict reason to do so, so the inlined >> rdma_ip2gid() will still contain both cases. > > Hello Arnd, > > Had you already considered to uninline the rdma_ip2gid() function? Not really, that would prevent the normal optimization from happening, so that would be worse than uninlining addr_event() as I tried. It would of course get rid of the warning, so if you think that's a better solution, I won't complain. Arnd