>Hmm? I thought that NMI oopser is only applicable to 2.3 kernels..
I am maintaining it as a separate patch (included into IKD) for 2.2.x.
ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/ikd/v2.2/
The 2.2.x NMI oopser is more rudimental (it doesn't allow the timer irq to
scale in SMP) but it's simpler and it's equally useful :)
>But I see stack trace even without NMI oopser (from __global_cli)
That's the stack trace of the innocent CPU so it's not interesting.
Basically if you don't see `NMI watchdog deadlock etc..etc..` before the
Oops, and you instead see the `wait_on_irq` message, then press SYSRQ+P
lots of times and try to gather lots of EIP addresses.
>Or do you want to get trace from second CPU too?
Exactly, the buggy code is going to be the one running on the CPU that
doesn't show the stack trace.
>It is in BH handler. And I am suspect that it is in BH handler from serial...
We'll see :)
Andrea
-
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/