RE: [microblaze-uclinux] 91c111 tcp/ip

["Steve Spano" <steve@xxxxxxxxxxxxx>]

Hi There Matthias,

When my firm develops new network appliances, the last thing I want to
see happen is it connected to a live-network and/or internet
explorer/etc. I would recommend this, so you can get a very simple
picture of what is happening

1) Follow this link
http://www.tamos.com/products/commview/
download the program CommView (it has an evaluation version)

2) Completely isolate the PCB by directly connecting it to a PC so that
no Ethernet traffic exists

3) Add some debug code so you can see every raw byte of data received by
the 91c111 and read-in by the Microblaze. (I think you already did
this?)

4) Use the CommView and select tools-->packetgenerator. This will allow
you to manually create a TCP/UDP/ICMP packet and transmit it from the PC
to your board. The only bytes you can't manually adjust are the preamble
(which your board will never see) and the 32-bit CRC at the end of the
packet. Other than than, you can send a TCP/UDP/ICMP or Matthias packet
format - basically anything you want. 

Inspect the results of #4 to see what bytes were read, bit errors, count
errors, etc. This will tell you if you are reading the data properly/etc
and let you know where your problem lies.

We did some work using the Xilinx webserver examples/etc and they are a
little crude. Are you using LwIP for the Ethernet stack? We have used
LwIP (with our own MAC core) and the Xilinx XMK operating system for
some simple network products.

Let me know how commview helps.

Steve Spano
President, Finger Lakes Engineering
(V) 607-277-1614 x223
(F) 800-835-7164
(C) 607-342-1150
steve@xxxxxxxxxxxxx
www.fl-eng.com

-----Original Message-----
From: owner-microblaze-uclinux@xxxxxxxxxxxxxx
[mailto:owner-microblaze-uclinux@xxxxxxxxxxxxxx] On Behalf Of Matthias
Heise
Sent: Tuesday, February 14, 2006 4:18 AM
To: microblaze-uclinux@xxxxxxxxxxxxxx
Subject: [microblaze-uclinux] 91c111 tcp/ip

good morning,

I can observe different behaviour between  Internet Explorer and Firefox

. When I try to connect with Internet Explorer the message "tcp: 
incorrect acknum" is thrown out and in ethereal I can see a http-packet 
with an incorrect tcp checksum, the system stays in a loop of obviously 
wrong transmissions and continues to throw the acknum error. With 
Firefox there is the above described RST loop and no http transmissions 
at all.
Why should the tcp checksum be calculated  incorrect when it works with 
the other TCP frames that are transmittet before ? By the way (to 
Georges question) I'm not working with linux but with a "bare" 
microblaze and a ping to the system is possible but there are two of 
four transmissions lost. Has anyone a deeper knowledge of all this and 
can help me ? Thank you for your time and patience.

 -- 
mit freundlichen Grüßen
best regards

Matthias Heise

ZEUTEC Opto-Elektronik GmbH
im Zentrum für Energie und Technik
Kieler Strasse 211
24768 Rendsburg
Germany

Tel.: +49-4331-136 799
Fax.: +49-4331-136 651
m.heise@xxxxxxxxx
http://www.zeutec.de 

___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive :
http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/





___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/