From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alex Williamson Subject: Re: VFIO compile failure on i386 allmodconfig in linux-next Date: Thu, 28 Jun 2012 16:25:47 -0600 Message-ID: <1340922347.3179.60.camel@ul30vt> References: <20120628175712.GA20250@windriver.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: Received: from mx1.redhat.com ([209.132.183.28]:12666 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754853Ab2F1WZs (ORCPT ); Thu, 28 Jun 2012 18:25:48 -0400 In-Reply-To: <20120628175712.GA20250@windriver.com> Sender: linux-next-owner@vger.kernel.org List-ID: To: Paul Gortmaker Cc: linux-next@vger.kernel.org On Thu, 2012-06-28 at 13:57 -0400, Paul Gortmaker wrote: > Hi Alex, > > Just a heads up that the i386 allmodconfig failed on VFIO, in case > you didn't see it. > > http://kisskb.ellerman.id.au/kisskb/buildresult/6603285/ > > arch/x86/include/asm/uaccess_32.h:211:26: error: call to 'copy_from_user_overflow' declared with attribute error: copy_from_user() buffer size is not provably correct > make[4]: *** [drivers/vfio/pci/vfio_pci_config.o] Error 1 Thank you for the report. This looks bogus to me, it's complaining about a copy_from_user into an __le32 because it can't determine the number of bytes to be copied. That number is a parameter to the function, but the function is static and only called with values of 1, 2 or 4. I've added a test that seems to make the compiler happy, so this will be fixed in tomorrow's tree. Thanks! Alex