From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 4851 invoked by alias); 28 Oct 2004 17:27:31 -0000 Mailing-List: contact ecos-devel-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Post: List-Help: , Sender: ecos-devel-owner@sources.redhat.com Received: (qmail 4837 invoked from network); 28 Oct 2004 17:27:28 -0000 Received: from unknown (HELO mail08a.verio.de) (213.198.55.73) by sourceware.org with SMTP; 28 Oct 2004 17:27:28 -0000 Received: from www.cesys.com (213.198.46.19) by mail08a.verio.de (RS ver 1.0.95vs) with SMTP id 3-0924434447; Thu, 28 Oct 2004 19:27:27 +0200 (CEST) Date: Thu, 28 Oct 2004 17:27:00 -0000 To: ecos-devel@sources.redhat.com, "v2p@mind.be" Subject: Problem with Virtex II Pro PowerPC 405 in 'hal_cpu_int_merge' From: Stefan Rosinger Content-Type: text/plain; format=flowed; charset=iso-8859-15 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-ID: User-Agent: Opera7.21/Win32 M2 build 3218 X-Loop-Detect:1 X-SW-Source: 2004-10/txt/msg00090.txt.bz2 Hi all! I have the same problem described here http://ecos.sourceware.org/ml/ecos-discuss/2004-01/msg00080.html I am working with an adaption of the Mind port of eCos for the Virtex II Pro on an own developed hardware. I am building a ROM startup configuration and don't use Redboot. I debug with GDB over a JTAG connection. I read the mail above and also disabled the MMU and Data/Instruction caches. That brought no effect. I also unchecked the 'use seperate interrupt stack'... The instruction where the stack crashes: .macro hal_cpu_int_merge sr mfmsr r0 rlwimi r0,\sr,0,16,16 sync mtmsr r0 sync .endm at the mtmsr r0. the content of r0 at runtime is 0x9000 I have no idea why! Who knows shomething about this problem.? For more informations about hardware or something else, write me a mail! Thx for help Stefan