Hi,

> I tried using kernel 2.5.15 and I still get the same oops message. 

I suspected you would, but that's good to confirm (given how
much changed since 2.5.8).


>  There's no high speed device connected to the host controller.  You 
> said end_unlink_async function should never be called when there is no 
> device connected.  Is there a way to find out which function is calling 
> end_unlink_async function? 

"grep" reports only one call site, so ... :)


>     This only happens during power up.  If I 
> plug the cardbus in after the system has finished the boot up process, I 
> don't get any oops message.

Which makes me suspect the real issue is the other option I
pointed out:  something clobbering the EHCI data structure
holding the "reclaim_ready" flag.  There's a lot going on during
system powerup, I could believe that something's misbehaving
there.  (Even maybe the EHCI driver, though I suspect not.)

You could put debug code in the end_unlink_async() code to
print the "reclaim_ready" flag (its legal values are 0 and 1),
to see if that looks bogus.  That flag is initted to zero, so the
first question is how it gets any other value in your scenario.
You can put other debug statements in the places that the
EHCI driver does so.

- Dave




_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: [EMAIL PROTECTED]
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to