eepro100.c 1.09t 29/9/99 still has problems

Keith K Chau keithc@unitech-com.com
Wed Nov 17 05:27:20 1999


Hi all,

I think the "Transmit timeout error" problem is very very well known.
Someone suggested that 1.09t dated 29/9/99 is very stable and urged others
to give it a try.  While 1.09t does seem to solve the infamous timeout
problem, it introduces another even more nasty one.

I have tried it on a few machines with different configurations, yet
yielding the *SAME* result.  All machines crashed at the end with a "Aiee,
killing interrupt handler."

The machines that I tried include:
1. ASUS P3B with 1GB of RAM, Adaptec 78xx Ultra2 controller with 2 9.1GB
U2SCSI drives, Intel Pro/100 FE adapter, RH 5.2, Kernel 2.2.10, eepro100.c
1.09t
2. Dell PE2300 with 1GB of RAM, PERC/2 SC U2 controller, 6 x 9.1GB U2SCSI
drives in RAID5, OEM Intel Pro/100 adapter, RH 6.1, Kernel 2.2.10,
eepro100.c 1.09t

I have reproduced the error codes below for your kind reference.  This is
from the Dell server.

[<80......		# a lot of lines of this kind
[<80......
[<801c8f58>][<801ca48e>][......
Code: 89 02 85 c0 74 03 89 50 05 b8 01 00 00 00 eb 03 90 31 c0 c7
Aiee, killing interrupt handler.
Unable to handle kernel paging request at virtual address 50890374.

Rgds,
Keith