cvs commit: src/sys/dev/pccbb pccbb.c

[Sam Leffler]

M. Warner Losh wrote:
> In message: <20060412212843.GH1258 at haakonia.hitnet.RWTH-Aachen.DE>
>             Christian Brueffer <brueffer at FreeBSD.org> writes:
> : On Thu, Mar 30, 2006 at 04:25:45AM +0000, Warner Losh wrote:
> : > imp         2006-03-30 04:25:45 UTC
> : > 
> : >   FreeBSD src repository
> : > 
> : >   Modified files:
> : >     sys/dev/pccbb        pccbb.c 
> : >   Log:
> : >   On some laptops, under very high loads, the socket event register read
> : >   in the ISR doesn't read the actual socket event register, but instead
> : >   reads garbage (usually 0xffffffff, but other times other things).
> : >   This totally violates the PCI spec, but happens rarely enough that a
> : >   workaround is in order.  This adds one test when we have a real
> : >   interrupt to service (which is very rare), and doesn't affect the
> : >   usualy 'nothing to see here' case at all.
> : >   
> : >   Problem reported by many, but sam@ gave me this workaround after
> : >   diagnosing the problem.
> : >   
> : 
> : This breaks pccard/cardbus support on my Thinkpad T41p (confirmed by
> : backing this change out).  With this change, I just get an "Interrupt
> : storm detected on irq 5; throttling" message.
> : 
> : What kind of debug info can I provide?
> 
> A Thinkpad T41p. :-)
> 
> Failing that, can you set hw.cbb.debug=1 and provide me with the
> output.

We should just back this out until we can resolve the issue.  For other 
folks: the change was originally done for an hp nx6125 laptop where 
running a wireless card in monitor mode would cause spontaneous 
interface detach's under load.  It sounds like the root cause is elsewhere.

	Sam