From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 26155 invoked by alias); 23 Jul 2007 19:23:57 -0000 Received: (qmail 26146 invoked by uid 22791); 23 Jul 2007 19:23:56 -0000 X-Spam-Check-By: sourceware.org Received: from unknown.mitsi.com (HELO mail.mitsi.com) (206.190.248.2) by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 23 Jul 2007 19:23:48 +0000 Received: from HELIUM ([10.6.0.32]) by mitsi.com with MailEnable ESMTP; Mon, 23 Jul 2007 14:27:12 -0500 Message-ID: <006701c7cd5e$fa663290$2000060a@lan.mitsi.com> From: "Chris Balthrop" To: Date: Mon, 23 Jul 2007 19:23:00 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Outlook Express 6.00.2800.1409 X-IsSubscribed: yes Mailing-List: contact ecos-discuss-help@ecos.sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: ecos-discuss-owner@ecos.sourceware.org Subject: [ECOS] Redboot on load part2 X-SW-Source: 2007-07/txt/msg00179.txt.bz2 Hello All, The "Redboot on load" discussion has prompted me to ask an opinion from our experts about a proposed solution to booting a new system that we are now writing the software for. I apologize for the long post and if this is not the correct forum for this question. Background: We have been using eCos/Redboot for about 2+ years, and on the previous 3 systems that we have built. These were all Xilinx PowerPC based systems with Flash and SDram. The boot sequence was controlled by a CPLD that loaded the Xilinx FPGA with a bitfile from the flash memory and then a small boot monitor program was run from within the FPGA block ram memory. This program loaded Redboot from the flash into SDram and jumped to it. This allowed us to use Redboot, with the serial ports, to debug our application. For the deliverable systems, we enabled a Redboot FIS script in the FLASH that called back into our boot monitor that then loaded one of the several application images from FLASH into SDram and ran it. New boot sequence: In the new system there is no CPLD or FLASH. The CPLD/FLASH has been replaced by a small ARM processor and an SDCard. The SDCard is formatted FAT32 so that it can be read or written on a PC. The ARM processor reads the SDCard and loads the Xilinx bitfile. The boot monitor then reads the SDCard and loads the Redboot image and an application into SDram. The proposed solution for the deliverable units was to use one of the older systems to initialize the Redboot FIS file system with a boot script and copy this to a file image that could be loaded by the boot monitor into the SDram. Then a Redboot would be built that thought that there was flash at the location that the FIS image was loaded. If the Redboot found the Sdram "FIS" it would then run the script from the Sdram to run our application otherwise it would stop in Redboot for debugging. Question: Is this a viable solution, or is there a simpler one? Can Redboot create an FIS in the SDram that we could save to the SDCard and have the boot monitor reload into Sdram so that Redboot could run it if it was there? We would like to keep the Redboot debug standalone and/or run script functionality in the final system. I thank you for your comments. Good Huntin' cb Chris Balthrop Sr. Engineer Micro Technology Services Inc. chris@mitsi.com -- Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss