> > This was when plugging in another (low speed) trackball, and
> > the OOPS was clearly caused by accessing poisoned pci_pool
> > memory.  Seems like something in_irq() was trying to walk a
> > pointer that had been poisoned -- actually to offset it, it was an
> > access to an address like like a7a7a7b6 which oopesed.
> 
> But doesn't that mean that there's a real bug still somewhere?

Right, accessing poisoned memory is not l33t!


> I'm using 2.5.7, so I should be ok, if it has your pci_pool patch,
> right?

What that patch will do is make the UHCI drivers poison as
religiously as the OHCI and EHCI drivers, when the kernel
is CONFIG_DEBUG_SLAB enabled.  I now suspect it's been
a while since they got that treatment for their hardware level
interactions, unfortunately.

Which I suspect is having the good/bad feature of turning up
some bugs as oopses that were previously (mostly) latent.
I'd call that "OK", sure ... :)

Though there could easily be other gremlins at play.

- Dave



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

Reply via email to