Johan Verrept wrote:
> anyway, the usb-uhci log is at the end, it ends when my system
> hardlocks.
> can this be related ?
how about this:
the hub thread noticed the device is gone and calls disconnect.
I usb_unlink_urb() all the urbs. They get their completion handler
called which a certain status (which one?). After the unlink, the urb &
context memory is freed.
I suppose handler doesn't know the status and reschedules the urb. The
urb context pointer still points to a context that doesn't exist
anymore. This should give an oops though, not a hardlock...
Also, the interrupt problem thingie is still runnning...
maybe there is a race issue there somewhere...
aaaah! too complex, not enough knowledge of the internals...
/mode positive_attitude=on
It'll probably go away when i handle all errors correctly ;)
/mode positive_attitude=off
There is still a problem lurking somewhere though...:(
J.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
