From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751767AbdG1E5T (ORCPT ); Fri, 28 Jul 2017 00:57:19 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:37068 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750785AbdG1E5S (ORCPT ); Fri, 28 Jul 2017 00:57:18 -0400 Date: Thu, 27 Jul 2017 21:57:12 -0700 From: Greg Kroah-Hartman To: Jacob von Chorus Cc: Insop Song , devel@driverdev.osuosl.org, linux-kernel@vger.kernel.org, dan.carpenter@oracle.com Subject: Re: [PATCH v3/resubmit 1/3] staging: gs_fpgaboot: add buffer overflow checks Message-ID: <20170728045712.GA32490@kroah.com> References: <20170727011359.2265-1-jacobvonchorus@cwphoto.ca> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170727011359.2265-1-jacobvonchorus@cwphoto.ca> User-Agent: Mutt/1.8.3 (2017-05-23) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jul 26, 2017 at 09:13:57PM -0400, Jacob von Chorus wrote: > Four fields in struct fpgaimage are char arrays of length MAX_STR (256). > The amount of data read into these buffers is controlled by a length > field in the bitstream file read from userspace. If a corrupt or > malicious firmware file was supplied, kernel data beyond these buffers > can be overwritten arbitrarily. > > This patch adds a check of the bitstream's length value to ensure it > fits within the bounds of the allocated buffers. An error condition is > returned from gs_read_bitstream if any of the reads fail. > > Signed-off-by: Jacob von Chorus > > v3: > - use >= to prevent an integer overflow in the comparison > - use get_unaligned_be functions to interpret length fields > - fix remainder of file to use valid error codes > > v2: > - char arrays converted to u8 arrays > - replace error return value with proper error code in > gs_read_bitstream > --- All of the v2: and such needs to go below the --- line, as Documentation/SubmittingPatches says to do. Please fix that up and resend the series. thanks, greg k-h