From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265098AbUIML65 (ORCPT ); Mon, 13 Sep 2004 07:58:57 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266069AbUIML65 (ORCPT ); Mon, 13 Sep 2004 07:58:57 -0400 Received: from clock-tower.bc.nu ([81.2.110.250]:65464 "EHLO localhost.localdomain") by vger.kernel.org with ESMTP id S265098AbUIML6z (ORCPT ); Mon, 13 Sep 2004 07:58:55 -0400 Subject: Re: [PATCH][2.4.28-pre3] USB drivers gcc-3.4 fixes From: Alan Cox To: Pete Zaitcev Cc: Mikael Pettersson , drivers@neukum.org, marcelo.tosatti@cyclades.com, sailer@ife.ee.ethz.ch, Linux Kernel Mailing List In-Reply-To: <20040912173018.44b9902f@lembas.zaitcev.lan> References: <200409121129.i8CBT5Bo015222@harpo.it.uu.se> <20040912173018.44b9902f@lembas.zaitcev.lan> Content-Type: text/plain Content-Transfer-Encoding: 7bit Message-Id: <1095072845.14359.9.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6 (1.4.6-2) Date: Mon, 13 Sep 2004 11:54:08 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Llu, 2004-09-13 at 01:30, Pete Zaitcev wrote: > > size -= pgrem; > > - (char *)buffer += pgrem; > > + buffer += pgrem; > > I'm pretty sure it's done that way on purpose. There were compilers which > did not allow any arithmetics on void*, and it had to be cast to char*. > So perhaps it's correct for 2.6, which requires gcc 3 anyway, but I have > my doubts about applicability of this to 2.4. Should be ok for the 2.* gcc we support. (BTW is anyone using anything less than 2.95 nowdays because we've still got 2.7.* workarounds in asm-i386/semaphore.h !) Otherwise in "valid" C it's buffer = ((char *)buffer) + pgrem;