Bug 2472

Summary:	pegasus.c Tx timed out
Product:	Drivers	Reporter:	Sander Rijken (sr)
Component:	USB	Assignee:	Greg Kroah-Hartman (greg)
Status:	REJECTED INSUFFICIENT_DATA
Severity:	normal	CC:	bunk
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	2.6.5	Subsystem:
Regression:	---	Bisected commit-id:
Bug Depends on:
Bug Blocks:	5089

Description Sander Rijken 2004-04-08 04:37:41 UTC

Distribution: Gentoo Linux

Hardware Environment: 
Pentium 233 MMX
Usb controller:
00:0b.0 USB Controller: NEC Corporation USB (rev 02)


Software Environment: 
USB driver: ohci-hcd
USB NIC driver: pegasus
both are compiled in the kernel, but I've tried it with those as modules as 
well.

Problem Description: 
USB network adapter works perfectly on Windows, I've tried another USB pegasus 
device, and got the same errors

This is in my logs:

It starts with intr status -84

NIC is still up at that time. Then after a certain period of time NIC seems to 
hang, and this fills up my log:
NETDEV WATCHDOG: eth0: transmit timed out
drivers/usb/net/pegasus.c: eth0: Tx timed out.

It's impossible to bring the NIC down at that point, or remove the module


Steps to reproduce:
I'm not sure if _you_ can reproduce it. I'd suggest an ohci-hcd type usb 
controller, and a pegasus networkcard, and wait till things start to go wrong.

I've read that I'm not the only one experiencing the intr status -84 messages, 
but with them things keep on working.

Comment 1 Michel DEPEIGE 2004-11-02 09:26:29 UTC

I have exactly the same thing with all my 2.6 kernels.

My Hardware :
AMD K6-233
USB host : 0000:00:01.2 USB Controller: Intel Corp. 82371SB PIIX3 USB
[Natoma/Triton II] (rev 01)
(USB 1.0)

Software :
USB driver: uhci-hcd, pegasus, both modules
NIC hang after a few minutes (sometimes hours) OR heavy data rate ( > 100 Kb/s
). I will add that the NIC hang faster (only a few seconds) when the system is
under high load (2.00 and more). 

Now we know it isn't ohci specific. Here is the dmesg :Nov  2 13:51:33 BlueBox
kernel: drivers/usb/net/pegasus.c: eth2: Tx timed out.
Nov  2 13:51:43 BlueBox kernel: NETDEV WATCHDOG: eth2: transmit timed out
Nov  2 13:51:43 BlueBox kernel: drivers/usb/net/pegasus.c: eth2: Tx timed out.
Nov  2 13:51:53 BlueBox kernel: NETDEV WATCHDOG: eth2: transmit timed out
Nov  2 13:51:53 BlueBox kernel: drivers/usb/net/pegasus.c: eth2: Tx timed out.

(eth0 and eth1 are 3Com PCI NICs)

The only thing to do is to unplug/replug the NIC.

Also, ifconfig eth2 is strange :
eth2      Lien encap:Ethernet  HWaddr 00:05:1B:00:4C:B5
          adr inet6: fe80::205:1bff:fe00:4cb5/64 Scope:Lien
          UP BROADCAST RUNNING MULTICAST  MTU:1536  Metric:1
          RX packets:49804 errors:0 dropped:0 overruns:0 frame:0
          TX packets:57726 errors:10355 dropped:0 overruns:0 carrier:0
          collisions:0 lg file transmission:1000
          RX bytes:7542743 (7.1 MiB)  TX bytes:19227472 (18.3 MiB)

(only after a few seconds).

I tried on another system :
AMD K7 Barthon 2500+
Host : 0000:00:02.0 USB Controller: nVidia Corporation nForce2 USB Controller
(rev a4)

Works fine but lots of intr status -84 messages in kern.log (more than one per
second)

Comment 2 Adrian Bunk 2005-08-07 18:24:00 UTC

Is this problem still present in kernel 2.6.13-rc6?

Comment 3 Adrian Bunk 2006-02-10 09:26:13 UTC

I'm assuming this issue is already fixed.
Please reopen this bug if it's still present in recent 2.6 kernels.