[Fwd: problem involving wait_on_bh]

Jason Bratton (jbratton@usa.net)
Sat, 18 Dec 1999 22:54:26 -0600

I'm having a slight problem I was hoping someone on this list could help
me with. I'm not sure if this is a kernel problem or if it's a hardware
issue, but here is what's happening. During normal use of my computer, I
receive the following error:
Dec 18 22:23:36 whatever kernel: wait_on_bh, CPU 1:
Dec 18 22:23:36 whatever kernel: irq: 0 [0 0]
Dec 18 22:23:36 whatever kernel: bh: 1 [1 0]
Dec 18 22:27:58 whatever kernel: <[c010b2ad]> <[c017031e]> <[c0170399]>
<[c017fb35]> <[c015ccc7]> <[c013258b]> <[c0132a76]>

The c010b2ad references back to synchonize_bh. This exact same error
will occur under extremely busy ethernet usage (quake for example) and
will leave the system locked up with only sysrq allowing for any
response. The usage of any sysrq key will result in the above error
being echoed to the current window repeatedly. Also, running ifconfig
eth0 down will result in the system locking up for the same reason. The
system is a dual processor intel celeron (Mendocino) with both
processors being 466 mhz with 128k of cache. I'm at a real loss on why
this error is showing up so commonly when dealing with the ethernet
card. It's a 3com 905B NIC running off the 3c90x module released by
3com and built with SMP support. I'm running linux kernel 2.2.13 on a
Slackware 7.0 distribution. The error I pasted above can occur on
either CPU0 or CPU1. The version of ifconfig that can lock up my
machine with the eth0 down command is 1.39 from the net-tools version
1.52. I have tried changing everything I can think of in irq.c, but I
am unable to fix this problem. Any help would be greatly appreciated.
As I'm not on this list, please send any replies directly to me at
jbratton@usa.net. Thanks. -- Jason Bratton

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/