On Fri, 1 Jul 2005, Matthias Urlichs wrote:
> > But during a normal rmmod the HC _is_ running. So how does this
> > end up solving your problem? Does ohci-hcd somehow learn that the
> > controller is _not_ running?
> >
> During a normal rmmod the HC is running; however, PCMCIA is special in
> t
Hi,
Alan Stern:
> I get it. And that call to finish_unlinks is made from within the
> ohci_endpoint_disable routine. It looks like the "clean up the list
> without looking at the hardware state" happens only when the HC isn't
> running.
Right.
> But during a normal rmmod the HC _is_ running.
On Thu, 30 Jun 2005, Matthias Urlichs wrote:
> > usb_disconnect does not disable the chip -- that happens after it returns.
> > The only way usb_disconnect interacts with the controller is through the
> > hcd_endpoint_disable routine. That in turn does a normal unlink (which
> > you say will no
Hi,
Alan Stern:
> There's one other thing I forgot to ask about... Do you know if the URB
> that is refusing to unlink was sent to the root hub, or was it sent to an
> external device? The unlink operation differs between those two cases.
>
It's queued to the device (USB_DT_DEVICE).
--
Matt
On Thu, 30 Jun 2005, Alan Stern wrote:
> On Thu, 30 Jun 2005, Matthias Urlichs wrote:
>
> > > If the unlink is broken then there's pretty much nothing you can do. The
> > > HCD drivers won't exit while there are still outstanding URBs. It
> > > doesn't
> > > matter what's locked or unlocked.
Hi,
Alan Stern:
> usb_disconnect does not disable the chip -- that happens after it returns.
> The only way usb_disconnect interacts with the controller is through the
> hcd_endpoint_disable routine. That in turn does a normal unlink (which
> you say will not work) and then calls the HCD's endp
On Thu, 30 Jun 2005, Matthias Urlichs wrote:
> > If the unlink is broken then there's pretty much nothing you can do. The
> > HCD drivers won't exit while there are still outstanding URBs. It doesn't
> > matter what's locked or unlocked.
> >
> Umm, not exactly. ;-) If the hub is not locked,
Hi,
Alan Stern:
> Did you really mean that the transfer _did_ time out, but chip issues
> prevent the _unlink_ from working?
>
Exactly. The URBs sit on the ed_rm queue, and they stay there. :-/
> If the unlink is broken then there's pretty much nothing you can do. The
> HCD drivers won't exit
On Thu, 30 Jun 2005, Matthias Urlichs wrote:
> Hi, Alan Stern schrub am Wed, 29 Jun 2005 10:46:39 -0400:
>
> > I'm afraid there's no way around it. Timeouts _have_ to work; pretty much
> > every device driver depends on them crucially. The best way to fix this
> > is to take care of those bro
Hi, Alan Stern schrub am Wed, 29 Jun 2005 10:46:39 -0400:
> I'm afraid there's no way around it. Timeouts _have_ to work; pretty much
> every device driver depends on them crucially. The best way to fix this
> is to take care of those broken OPTi chip issues you mentioned.
>
And the best way
10 matches
Mail list logo
