Network Builds

Tue Mar 9 16:56:37 UTC 2004

I have been working on an RTEMS port to a custom hardware platform which I now have 
the basics up and running. I can create and run tasks, use semaphores and use serial IO. 
The board has CS8900 lan chip on and I am trying to compile in the network support for 
rtems and build one of the netdemo applications. 

I have configured my rtems tree using :

  /opt/rtems/rtems-4.6.0/configure --target=sh-rtems \
    --enable-posix --enable-networking --disable-cxx \
    --enable-rtemsbsp=fu0018 \
    --prefix=/opt/rtems/builds/fu0018_net

It all appears to build correctly. When I build netdemo however I get implicit declaration 
warnings  for function calls such as send, connect, getsockopt. I have <sys/socket.h>
included, I have also checked that this file is present in the built rtems tree. Am I missing
something required to get this to build? Can anyone also point me in the direction of some
more information on using the CS8900 driver located in libchip/ .

            Any help will be very appreciated.

                        James Yates

Return-Path: <vvv at oktetlabs.ru>
Mailing-List: contact rtems-users-help at rtems.com; run by ezmlm
Delivered-To: mailing list rtems-users at rtems.com
Received: (qmail 11462 invoked from network); 9 Mar 2004 17:02:32 -0000
Received: from unknown (HELO smtp.rol.ru) (194.67.21.9)
  by 0 with SMTP; 9 Mar 2004 17:02:32 -0000
Received: from ts1-a139.Spb.dial.rol.ru ([195.190.98.139]:63367 "EHLO
	oktetlabs.ru" smtp-auth: <none> TLS-CIPHER: <none> TLS-PEER-CN1: <none>)
	by gnome05.net.rol.ru with ESMTP id S17611700AbUCIRCa (ORCPT
	<rfc822;rtems-users at rtems.com>); Tue, 9 Mar 2004 20:02:30 +0300
Message-ID: <404DF8A3.9010500 at oktetlabs.ru>
Date:	Tue, 09 Mar 2004 20:02:27 +0300
From:	"Victor V. Vengerov" <vvv at oktetlabs.ru>
Organization: OKTET Labs.
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20030225
X-Accept-Language: en, ru
MIME-Version: 1.0
To:	Adriano Verardo <a.verardo at tecmav.fastwebnet.it>
CC:	rtems-users <rtems-users at rtems.com>
Subject: Re: DOSFS problems
References: <404CD3DC.15836.13FFDC at localhost> <404CF2C0.17667.8C1A6A at localhost> <404DE205.8010908 at tecmav.fastwebnet.it> <404DE41D.9030400 at oktetlabs.ru> <404DF21A.9010306 at tecmav.fastwebnet.it>
In-Reply-To: <404DF21A.9010306 at tecmav.fastwebnet.it>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Adriano,

As I remember, partition table is located at the end of first block.
Fragment you show looks like Intel boot loader code.

It seems, partition table starts from 0x1C0 offset (I've just dumped
my hard disk to check). And 55 AA signature should be at 0x1FE
offset.

Victor

Adriano Verardo wrote:

> Hi, Victor.
>
> Victor V. Vengerov wrote:
>
>> Adriano,
>>
>> As fine as I understand you need proper MBR on your HDD to run
>> dosfs test.
>
>
> Yes, QUANTUM ELS 127A, CHS‘9/16/17, DOS 6.22,
> Install from scratch, 16 MB partition.
> MBR:
> 00 - FA 33 C0 8E  D0 BC 00 7C 8B F4 50 07 50 1F FB FC
> 16 - BF 00 06  B9 00 01  F2 A5 EA  1D 06 00  00  BE  BE 07
> 32 - B3 04  80 3C  80 74  0E 80 3C 00  75 1C 83 C6 10 FE
> 48 -  CB 75 EF CD 18 8B 14 8B 4C 02 8B EE 83 C6 10 FE
> 64 -  CB 74 1A 80 3C 00 74 F4 BE 8B 06 AC 3C 00 74 0B
> ....
> Unexpected 07 50 in MBR[11,12]. Should be 00 02 according with the disk
> and DOSFS needs.
>
>>
>>
>> As I remember, you may re-initialize MBR on your HDD invoking
>>
>> fdisk /mbr
>> (or fdisk C: /mbr ?)
>
>
> Done, no changes.
>
>>
>> Then, run fdisk C: and ensure that your partition table is correct.
>> If problem persists, try to reduce partition size to, let say, 256Mb
>> using fdisk. 
>
>
> Tried 16 M ... unbelievable behaviour.
> Thank you, pleased to hear from you.
> Adriano
>
>>
>>
>> Regards,
>> Victor
>>
>> Adriano Verardo wrote:
>>
>>> Thomas Doerfler wrote:
>>>
>>>> Could you possible give the values that are present in your MBR and 
>>>> partition table? It may be that the RTEMS DOSFS code does not 
>>>> implement ALL possible partition types and options.
>>>>
>>> Hi Thomas, thank you again.
>>>
>>> The result of all tests is: the simplest possible DOS install always 
>>> gives
>>> working system disks with a MBR that's never accepted by DOSFS.
>>> For all tested DOS/fdisk/HW/FAT type/ etc etc. combinations:
>>> - HD MBR zeroed before installing DOS
>>> - no disk managers, installed DOS boots from C: and works
>>> - probably DOSFS fails before accessing the partition table:
>>>  * MBR[11,12] of the HD is always wrong (huge, ie: 07 50 instead of 
>>> 00 20
>>>     and changes using different fdisk/format tools)
>>>  * the final 55 AA signature on the HD MBR is sometimes absent (00 00)
>>> - the initial OEM FIELD on the HD MBR is never an ASCII string ?????
>>> - the partition tab is @ offset 446 ... standard ... DOS works
>>> - the MBR of the install floppy seems always to be "regular"... why 
>>> it is not
>>>   reproduced (with the obvious changes) on the HD ???
>>>
>>> Forcing rigth values in the DOSFS source don't work.
>>> Clearly there are other inconsistent values in the MBR.
>>>
>>> If this info are not precise enough I'll send some MBR dumps.
>>> It's not immediate to transfer them to my PC.
>>>
>>> Is there any other rtems-user who had a similar problem ?
>>> Using very standard but quite old HW can be a problem ?
>>> Is there any particular DOS install procedure/trick/option to
>>> force the layout of  the MBR of the HD ?
>>>
>>> Adriano
>>>
>>>
>>>
>>>
>>>
>>
>>
>
>

-- 
Victor V. Vengerov
OKTET Labs.
Phone: mobile +7 812 9389372, home +7 812 4281653 