> From: "Jean-Francois De Rudder" <[EMAIL PROTECTED]>
> Date: Tue, 1 Jan 2002 20:34:24 -0000
> Subject: [Linux-usb-users] RE: Kernel panic from usb-ohci
> 
> The driver used is stock standard OHCI from the 2.4.26mdk kernel.

There's no "2.4.26" yet, 2.4.17 is current.  Try that, it may have some
of the bugs fixed in that DSL driver (which all but certainly holds the
bug, not the OHCI driver).


> I forgot to mention that I have succesfully installed the same config on
> my ACER laptop, which also uses an OHCI USB port. The diff seems to be
> with the add-on USB card. Whether it is not good enough or whether OHCI
> can't handle those add-on card (or just this one) properly (I doubt), or
> possibly my old motherboard/BIOS/PCI hardware too old (P1 133) to handle
> all this?

I could easily imagine the slower CPU changed the timing in such a way
that the bug in that DSL driver showed up, where previously it was for
some reason masked.

I've seen that phenomenon before... the fact that the bug is _detected_ in
the OHCI code has nothing to do with where the actual bug is located.

The usual way that message is seen is this:  device driver corrupts the
device refcount.  Then some request completes, causing OHCI to drop
the refcount it associates with the request.  At that point, OHCI notices
that the refcount is bogus, since it's _never_ allowed to go to zero in
an interrupt context, and reports it.

So you see, the issue is finding who's corrupting that device refcount.
As a rule, many device drivers don't keep that count correctly ... so the
first place to look is the device driver, not the controller driver.

- Dave


> 
> > -----Original Message-----
> > From: David Brownell [mailto:[EMAIL PROTECTED]] 
> > Sent: 01 January 2002 20:23
> > To: Jean-Francois De Rudder
> > Subject: Re: Kernel panic from usb-ohci
> > 
> > 
> > Since the issue is with the driver for that D-Link card,
> > you should take this up with the maintainer of that driver.
> > You didn't mention what driver that is.
> > 
> > You did verify the problem remains with current kernel, yes?
> > 
> > - Dave
> > 
> > 



_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users

Reply via email to