Description of Problem: possibly related to bug # 27440 kernel-smp-2.4.3-7 build qa0523.0 The eepro100 NICS seem to be on the fritz w/ the Intel and the Compaq. The Compaq started acting up about the middle of last week, so I swapped out the NIC with a tulip card. Unfortunately I didn't get any info from lspci before it was yanked out, but I'm sure I can get it if needed. I didn't think much of it until I started noticing the same behavior on the Intel on Friday afternoon. Net installs start choking at random points during package installation. It doesn't hard lock the system, but package installation progress comes to a complete halt. The Intel currently has two adapters and I've verified the behavior with both the onboard and PCI adapters. The system at my desk doesn't have any network related problems during or after installation, and the NICs seem to be almost identical to the second device installed in the Intel in the lab. (I can perform net installs w/ both adapters in the system at my desk). How Reproducible: 100% Additional Information: Here is the info from "lspci -v" for both systems: Intel (lab) -- 4 CPUs / 4 Gigs RAM: 00:05.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08) Subsystem: Intel Corporation 82559 Fast Ethernet LAN on Motherboard Flags: bus master, medium devsel, latency 32, IRQ 55 Memory at 00000000f3fe0000 (32-bit, non-prefetchable) [size=4K] I/O ports at 6f00 [size=64] Memory at 00000000f3e00000 (32-bit, non-prefetchable) [size=1M] Capabilities: [dc] Power Management version 2 04:01.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08) Subsystem: Intel Corporation EtherExpress PRO/100+ Management Adapter Flags: bus master, medium devsel, latency 32, IRQ 63 Memory at 00000000e3ff0000 (32-bit, non-prefetchable) [size=4K] I/O ports at af00 [size=64] Memory at 00000000e3e00000 (32-bit, non-prefetchable) [size=1M] Expansion ROM at 00000000e3d00000 [disabled] [size=1M] Capabilities: [dc] Power Management version 2 Big Sur (at my desk) -- 1 CPU / 1 Gig RAM: 00:05.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08) Subsystem: Intel Corporation EtherExpress PRO/100+ Management Adapter Flags: bus master, medium devsel, latency 32, IRQ 49 Memory at 00000000fdeef000 (32-bit, non-prefetchable) [size=4K] I/O ports at 1000 [size=64] Memory at 00000000fdf00000 (32-bit, non-prefetchable) [size=1M] Capabilities: [dc] Power Management version 2 01:02.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08) Subsystem: Intel Corporation EtherExpress PRO/100+ Management Adapter Flags: bus master, medium devsel, latency 32, IRQ 68 Memory at 00000000fbdff000 (32-bit, non-prefetchable) [size=4K] I/O ports at 2000 [size=64] Memory at 00000000fbf00000 (32-bit, non-prefetchable) [size=1M] Expansion ROM at <unassigned> [disabled] [size=1M] Capabilities: [dc] Power Management version 2
Any net installs or just NFS?
FTP, HTTP, and NFS
Intel apparently released a 1.6.6 driver of their driver with the hardware workaround.... I can't locate the driver on their site though; everything has moved and most of their links are stale
I found it... ftp://download.intel.com/df-support/2749/eng/e100-1.6.6.tar.gz The parent page is: http://appsr.intel.com/scripts-df/filter_results.asp?strOSs=39&strTypes=PLU% 2CBIO%2CDRV%2CSPH%2CUTL&ProductID=417&OSFullName=Linux*&submit=Go%21
Yes. The relevant change in that driver is that it frobs a byte in the eeprom. (!)
The current plan is to not ship this, as shipping a driver that frobs the eeprom and forces a reboot is not acceptable; this is a fix that should be done in firmware, or by a userspace program.
closing
actually, we changed our minds, and we're shipping e100-1.6.6 anyway.