From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 7860 invoked by alias); 3 Sep 2008 13:22:27 -0000 Received: (qmail 7852 invoked by uid 22791); 3 Sep 2008 13:22:26 -0000 X-Spam-Check-By: sourceware.org Received: from fg-out-1718.google.com (HELO fg-out-1718.google.com) (72.14.220.155) by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 03 Sep 2008 13:21:31 +0000 Received: by fg-out-1718.google.com with SMTP id 16so151417fgg.44 for ; Wed, 03 Sep 2008 06:21:28 -0700 (PDT) Received: by 10.86.81.9 with SMTP id e9mr4133507fgb.63.1220448087931; Wed, 03 Sep 2008 06:21:27 -0700 (PDT) Received: from localhost ( [212.41.76.200]) by mx.google.com with ESMTPS id 4sm9785425fgg.4.2008.09.03.06.21.24 (version=SSLv3 cipher=RC4-MD5); Wed, 03 Sep 2008 06:21:25 -0700 (PDT) Date: Wed, 03 Sep 2008 13:22:00 -0000 To: "Andrew Lunn" , "Robert Brusa" Reply-To: Bob.Brusa@gmail.com Cc: ecos-discuss@ecos.sourceware.org Content-Type: text/plain; format=flowed; delsp=yes; charset=utf-8 MIME-Version: 1.0 References: <48B8137D.7040803@mlbassoc.com> <20080829162608.GV31987@lunn.ch> <20080901093643.GT25863@lunn.ch> <20080902084120.GA22673@lunn.ch> Content-Transfer-Encoding: 7bit Message-ID: In-Reply-To: <20080902084120.GA22673@lunn.ch> User-Agent: Opera Mail/9.52 (Linux) From: Robert Brusa 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: Re: [ECOS] Fwd: AT91SAM7X-EK with TC/IP - unresolved conflicts X-SW-Source: 2008-09/txt/msg00037.txt.bz2 On Tue, 02 Sep 2008 10:41:20 +0200, Andrew Lunn wrote: > On Tue, Sep 02, 2008 at 10:11:11AM +0200, Robert Brusa wrote: >> On Mon, 01 Sep 2008 11:36:43 +0200, Andrew Lunn wrote: >> >>> >>> I never said that. I just said the FreeBSD stack will not work. The >>> lwip stack has been shown to work on the AT91SAM7X. Go google lwip. >>> >> running through ecosconfig new at91sam7xek lwip_eth >> etc. seems to work. But when I try to build my previously working >> projects, flash support is no longer available and the cor. calls >> produce >> errors (e. g. a call to flash_program) > > So you probably need to add the flash package. > I did - and more - see below >> On the other hand, when attempting to configure the same thing using >> configtool, there are 5 conflicts the tool is not able to resolve. From >> a >> first look at the ecc-file, it seems to be a problem that somewhere a >> restriction for the AT91SAM7S is required. Any adice how to overcome >> this >> hurdle? > > No, not without you giving us the details... > Meanwhile I have tried a lot, but was not able to produce a working eCOS with TCP/IP. Finally I succeeded in producing a version free of conflicts. I have put the files I am referencing in the next lines on the web at www.brusag.ch/xyz script used to generate eCOS: ecosconfig.sh copy of command line to invoke the script (with redirection of output) terminal.txt the output o the script: config.log the ecc-file: ecos.ecc a test program: tracker.c As one can see in terminal.txt and config.log, there is a problem when generating the tests. I decided to ignore this and make a test with a small user program tracker.c. The program does not use TCP/IP and (under this configuration) it behaves strangly. I then found out that when running thread_0 the stackpointer points to nirwana! Any help to generate a working configuration for my Eval-Kit that support TCP/IP is highly appreciated. Thanks. Robert >> I am talking about RFC 1831 (which is the successor of 1057). I have >> built a small example using Suns rpcgen-tool and it worked fine using >> two >> linux-PCs linked by ethernet. Now the next step would be to port the >> serverapp to my AT91SAM7X target. The 2nd step then would be to "expand" >> this little server app to handle the routines my app requires - by hand >> or using e. g. rpcgen-produced templates. Is this a realistic scenario? >> Robert > > I've never used these tools, so it is hard to say. What libraries do > you need? How dependent are these on having true BSD sockets? The lwip > stack has something like BSD sockets, but it is not true BSD > sockets. You might run into problems when porting the RPC libraries > etc. > > Andrew Well, given the present status, RPC is only a dream... -- Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss