Bug 63306

I have observed the same problems with the following hardware:

Supermicro dual Pentium III server motherboard with dual onboard
NIC.

/sbin/lspci -n

00:00.0 Class 0600: 1166:0009 (rev 05)
00:00.1 Class 0600: 1166:0009 (rev 05)
00:03.0 Class 0300: 1002:4752 (rev 27)
00:04.0 Class 0200: 8086:1229 (rev 08)
00:06.0 Class 0200: 8086:1229 (rev 08)
00:0f.0 Class 0601: 1166:0200 (rev 4f)
00:0f.1 Class 0101: 1166:0211
00:0f.2 Class 0c03: 1166:0220 (rev 04)
01:01.0 Class 0100: 1119:01d6
01:03.0 Class 0100: 9005:008f (rev 02)


/sbin/lspci -v 

...

00:04.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08)
        Subsystem: Intel Corporation EtherExpress PRO/100+ Server Adapter
(PILA8470B)
        Flags: bus master, medium devsel, latency 64, IRQ 22
        Memory at feafc000 (32-bit, non-prefetchable) [size=4K]
        I/O ports at d000 [size=64]
        Memory at fe800000 (32-bit, non-prefetchable) [size=1M]
        Capabilities: <available only to root>

00:06.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08)
        Subsystem: Intel Corporation EtherExpress PRO/100+ Server Adapter
(PILA8470B)
        Flags: bus master, medium devsel, latency 64, IRQ 31
        Memory at feafd000 (32-bit, non-prefetchable) [size=4K]
        I/O ports at d400 [size=64]
        Memory at fe900000 (32-bit, non-prefetchable) [size=1M]
        Capabilities: <available only to root>

...

The symptoms are extremely bad performance, or complete loss of connectivity
accompanied by "eepro100: wait_for_cmd_done timeout!" errors in the logs.
This error has been reproduced using both the eepro100 and e100 drivers,
in kernels 2.4.9-31enterprise (7.2 updates), and 2.4.18-0.13smp (skipjack beta
2). I have also tested with a version of the driver published by Donald Backer,
with his revision number 1.20, but this drivers causes a kernel panic under load.

The system in question is not yet in production use, so I am willing to test
possible fixes.

I can confirm that the bug also exists in kernel 2.4.7-10 enterprise.

I have also experienced this issue.  I am using a Gateway E-1400 with an Intel 
i82557 chipset.  I am booting on a 2.4.7-10BOOT kernel, and am booting via 
PXE.  The problem is intermittent, but when it does occur, the machine has no 
network connectivity.

I'm seeing this too, on an IBM NetVista with the intel 810e chipset.  It's
preventing me from netloading the system and forcing me to burn a CD or shuffle
drives around (both major pains).

I managed to load the system from some burned CD's, and I'm still getting the
error.  Since I can now get a debug kernel on the machine (eventually), what
should I try?  I'd love to know if this is a hardware error or not.

Have you disabled sleep mode on the card firmware?
ie with ftp://ftp.scyld.com/pub/diag/eepro100-diag.c

I solved this problem on another kernel disabling sleep mode. This is a common
issue.

Note that Red Hat ships eepro100-diag in kernel-utils RPM also.

*** Bug 62799 has been marked as a duplicate of this bug. ***

So the quesiton is, shouldn't network install kernels have this disabled by default?

It is not obvious from the comments that this is definitely the problem.

Does disabling sleep mode via eepro100-diag fix the problem, for all bug reporters?

No response after multiple months.  I believe this problem to be fixed in the
current Red Hat kernel.  Please try with 8.1 beta or the latest errata kernel
for 7.1/7.2/8.0 and file a new bug if the behavior persists.

Sorry, but playing around here with latest RHL kernel 2.4.18-18.7.x still show this problem after some minutes (let say around 10). NIC is connected to a 10 MBit single speed hub, mii-tool proper tell "no autonegotianion, 10baseT-HD, link ok".

Does this problem occur with e100 driver?

We are deprecating eepro100 in favor of the latest e100.