Who have TCP/IP problem ?

Tekno Soft Snc (MC3641@mclink.it)
Tue, 7 Jan 97 20:00:24 CET

Messages sorted by: [ date ][ thread ][ subject ][ author ]
Next message: Simon Kenyon: "net alias eth0:1 transmits"
Previous message: Paul Christenson: "Re: DHCP again"

Hi all,

I've notice this using dmesg on 2.1.20:

TCPv4 bad checksum from 0b010184:0401 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0b010184:0401 to 01010184:008b, len=1480/1480/1500
[...deleted...]
TCPv4 bad checksum from 0b010184:0401 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0b010184:0401 to 01010184:008b, len=1480/1480/1500
[...deleted...]
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1480/1480/1500
[...deleted...]
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1480/1480/1500
TCPv4 bad checksum from 0a010184:0402 to 01010184:008b, len=1248/1248/1268

on the 2.1.13 (my previous kernel) isn't reported any messages.

The machines 0a010184 (132.1.1.10) and 0b010184 (132.1.1.11) are in local
net
with Samba on the server 01010184 (132.1.1.1). The client are Windoze 95
(ver 4.0.950a) and Dos with msclient 2.2c.

The problem are _only_ on Windoze 95 machines 0a010184/0b010184, the dos
client don't show this problem.

I think there are problem on the checksum to the Windoze 95 TCP/IP stack.

Any idea ?

Roberto Fichera - email MC3641@mclink.it

Xmeo ver.3.0 Alpha for Linux & Windows

Next message: Simon Kenyon: "net alias eth0:1 transmits"
Previous message: Paul Christenson: "Re: DHCP again"