From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753063AbdENEOJ (ORCPT ); Sun, 14 May 2017 00:14:09 -0400 Received: from mail-pf0-f177.google.com ([209.85.192.177]:34057 "EHLO mail-pf0-f177.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752372AbdENEOH (ORCPT ); Sun, 14 May 2017 00:14:07 -0400 Date: Sat, 13 May 2017 21:14:03 -0700 From: Bjorn Andersson To: Henri Roosen Cc: linux-remoteproc@vger.kernel.org, Ohad Ben-Cohen , open list Subject: Re: [PATCH 1/1] remoteproc: fix elf_loader da_to_va translation and writing beyond segment Message-ID: <20170514041403.GD69278@Bjorns-MacBook-Pro-2.local> References: <1493813529-19184-1-git-send-email-henri.roosen@ginzinger.com> <20170511000511.GV15143@minitux> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.8.0 (2017-02-23) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu 11 May 09:12 PDT 2017, Henri Roosen wrote: > On 05/11/2017 02:05 AM, Bjorn Andersson wrote: > > On Wed 03 May 05:12 PDT 2017, Henri Roosen wrote: > > > > > Consider a system with 2 memory regions: > > > 0x1fff8000 - 0x1fffffff: iram > > > > So I presume there's a hole here. > > > > > 0x21000000 - 0x21007fff: dram > > > > > > The .elf file for this system contains the following Program Headers: > > > Program Headers: > > > Type Offset VirtAddr PhysAddr FileSiz MemSiz Flg Align > > > LOAD 0x000094 0x1fff8000 0x1fff8000 0x00240 0x00240 R 0x4 > > > LOAD 0x0002e0 0x1fff8240 0x1fff8240 0x03d1c 0x03d1c RWE 0x10 > > > LOAD 0x003ffc 0x21000000 0x1fffbf5c 0x001cc 0x048a0 RW 0x4 > > > > > > > Your ELF header states that there is a segment of 0x48a0 bytes starting > > at 0x1fffbf5c, but your iram ends after 0x40a3 bytes. I assume your > > MemSiz comes from some linker script, which would mean that your > > firmware expects to be able to use all 0x48a0 bytes, which should be > > invalid. > > I had a closer look at the linker script. The .data section uses the > "AT"-keyword to place the initialized .data right after the .text section > (0x1fffbf5c/PhysAddr). > > Some run-time startup-code is responsible for copying the initialized data > to its runtime address (0x21000000/VirtAddr). The run-time startup-code is > also responsible for zero-ing the .bss section. > > The size of the initialized data is 0x1cc (FileSiz). The size of the whole > 3rd segment at run-time is 0x048a0 (MemSiz), starting from 0x21000000 > (VirtAddr), which also includes the .bss .heap and .stack sections. > [1] specifies that p_memsz is the size of the memory segment and that the difference between p_filesz and p_memsz are defined to hold the value 0. So I believe that your segment list states that the physical range 0x1fffbf5c through 0x200007fc is valid and should be populated. [1] https://refspecs.linuxfoundation.org/elf/elf.pdf Regards, Bjorn