On Sat, 25 Jun 2005 [EMAIL PROTECTED] wrote:
> Selon Alan Stern <[EMAIL PROTECTED]>:
>
> > On Wed, 22 Jun 2005 [EMAIL PROTECTED] wrote:
> >
> > > > The patch for 2.6.12 is below. All you have to do is apply it and then
> > > > rebuild the uhci-hcd driver. If you use a text console (not an X
> > > > window)
> > > > when plugging in your device, the messages should show up on the screen.
> > > > Assuming the interrupt routine is getting called, that is.
> > >
> > > No message appears on plug :-( even with highest log level (alt sysrq 9).
> >
> > It was a low probability anyway... This is a very perplexing problem, no
> > doubt about it.
> >
> > Maybe it has something to do with the UHCI controller's suspend mode.
> > The 2.6 uhci-hcd driver suspends the controller if no devices are plugged
> > in and so does the 2.4 JE uhci driver, whereas the usb-uhci driver
> > doesn't. Then when you plug in a device, the controller is supposed to
> > automatically wake up from the suspended state. Some sort of hardware bug
> > might cause the controller to hang your system instead of waking up
> > correctly. That would explain why it works okay if the device is already
> > plugged in when the driver is loaded.
>
> ok. I don't know anything about writing drivers or USB hardware, but why not
> let
> the controler woken up ? if yes, can an option of the driver be implemented to
> control this ?
Below is a patch that will prevent the uhci-hcd driver from suspending the
controller. Since I don't have a copy of 2.6.12 with me here, I'm not
certain the patch will apply okay. If it doesn't I'll send you a
corrected version on Monday.
> > This is why I suggested you try using a different controller. It would
> > not have the hardware bug of the controller on the motherboard, so it
> > should work okay.
>
> mmm... I cannot afford buying another card, but I can borrow one to make a
> test
> and tell you what happens.
>
> Is there a way I can use to determine where in the kernel or in the driver the
> crash takes place ? is kdbg usable to do this ?
I don't think so. This is just a guess, but it looks like you're getting
a severe hardware crash. It's not caused by a bug in the kernel or in the
driver; it's caused by a bug in the USB controller. No changes to the
software can fix the problem, and no software will run at all following
that kind of hardware crash.
Alan Stern
--- a/drivers/usb/host/uhci-hub.c Tue May 10 09:29:14 2005
+++ b/drivers/usb/host/uhci-hub.c Sat Jun 25 19:53:46 2005
@@ -46,7 +46,7 @@
test_bit(port, &uhci->port_c_suspend))
return 1;
}
- return 0;
+ return 1;
}
static inline int get_hub_status_data(struct uhci_hcd *uhci, char *buf)
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users
