[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Bug in linux-2_4_26.diff in combination with a Deboard LX?



Hello Ferdy,

Thanks for the information, I'll investigate it and include a solution
in the next release of the SDK..

Best Regards,
Anders

> -----Original Message-----
> From: owner-dev-etrax@xxxxxxx.com 
> [mailto:owner-dev-etrax@xxxxxxx.com] On Behalf Of Ferdy Hanssen
> Sent: den 2 juli 2004 23:43
> To: dev-etrax
> Subject: Bug in linux-2_4_26.diff in combination with a Deboard LX?
> 
> 
> Dear all,
> 
> I believe I have found a bug in the linux-2_4_26.diff file, 
> dated June 3rd, 2004, as it can currently be found on the 
> developer.axis.com website in the download/linux section and 
> distributed as part of the 
> download/devboard_lx/R2_2_0/devboard_lx-R2_2_0.tar.gz SDK 
> release 2.2.0 (June 3rd, 2004 edition) for the devboard LX.  
> I applied the fix to the 'flashit' script (jffs => jffs2). 
> 
> Using stock Linux 2.4.26 with the patch, as supplied by Axis, 
> the resulting kernel will not boot using the 'ktest' script 
> provided in the SDK.  At least no output appears on serial 
> port 2 (or any other port).
> 
> After some testing I found that the kernel will boot when a 
> slight change is applied.  I replaced the file 
> arch/cris/kernel/debugport.c with the same file as provided 
> by the stock kernel and removed support for the
> rs_debug_write_function() in arch/cris/drivers/serial.c.  The 
> attached patch will do this (apply this patch to the kernel 
> after applying the Axis patch).
> 
> I would like to know whether others can verify this, 
> specifically the poster of message 
> http://mhonarc.axis.se/dev-etrax/msg04459.html (Eric Sorton), 
> as his post seems to indicate the same problem.
> 
> My solution is probably too coarse, but I did not study the 
> code extensively to find a more elegant solution.
> 
> For reference, I don't know if it matters, my board is an 
> Axis Devboard LX, and the numbers provided on the board say: 
> PCB AXIS DEVBOARD E5, 19784, R2, and the Etrax 100LX chip is 
> a V2 variant, as indicated by the number 18816 on the chip.
> 
> Kind regards,
> 
> -- 
> Ferdy Hanssen, DIstributed and Embedded Systems group; 
> University of Twente, Dept. of Computer Science;
> e-mail: hanssen@xxxxxxx.nl; tel: +31 53 489 3744;
> fax: +31 53 489 4590; PO Box 217, 7500AE Enschede, NL
>