On 03-May-01 Warner Losh wrote:
> In message <[EMAIL PROTECTED]> Kevin Day writes:
>: The PCI target itself isn't doing anything like that, but it's possible that
>: the PCI-PCI bridge we're going through might be. In any case, getting the
>: NMI isn't really all that bad, it's stopping the chipset from getting hung
>: on a infinite retry. My only concern is the NMI handler while in the kernel
>: may be too aggressive in causing a panic.
>
> Yes. The NMI handler is a little too agressive about panicing. Also,
> current has problems where sometimes it will panic when the nmi
> happens with GIANT held.
Correction, with a spin lock held. It may try to acquire Giant (not sure _why_
it acquires Giant) and then it pukes. Getting a traceback would be most
helpful. :) Also, the output of 'show locks' from ddb could help, too.
--
John Baldwin <[EMAIL PROTECTED]> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.baldwin.cx/~john/pgpkey.asc
"Power Users Use the Power to Serve!" - http://www.FreeBSD.org/
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
