Here's some more info about the "unending timouts problem".
Reminder:
(1) Alcatel speedtouch USB modem plugged into VIA uhci hub
(2) URB sent to endpoint 0x81 (interrupt endpoint)
(3) when another device is plugged into the hub, the urb comes
back with error -EILSEQ. All future urbs to that endpoint of the
modem immediately return with -EILSEQ.
I stuck lots of printks in the code that deals with device connection
to see exactly where things go wrong for the modem when another
device is plugged in. The urb in (2) comes back exactly when the line
marked HERE below is executed (usb_hub_port_reset is called from
usb_hub_port_connect_change).
hub.c:
static int usb_hub_port_reset(struct usb_device *hub, int port,
struct usb_device *dev, unsigned int delay)
{
int i, status;
/* Reset the port */
for (i = 0; i < HUB_RESET_TRIES; i++) {
usb_set_port_feature(hub, port + 1, USB_PORT_FEAT_RESET); <== HERE
>From the syslog (between the line HERE and the next):
kernel: uhci-hcd.c: uhci_result_interrupt/bulk() failed with status 440000
kernel: [cac520c0] link (0ac52092) element (08eba1e0)
kernel: 0: [c8eba1e0] link (00000001) e0 IOC Stalled CRC/Timeo Length=7ff MaxLen=f
DT0 EndPt=1 Dev=2, PID
=69(IN) (buf=04102c80)
kernel: hcd.c: giveback urb ca736f80 status -84 len 0
It struck me that maybe the wrong device is being reset - the modem
rather than the newly plugged in device.
I will track this down further later - but now it's time to go to bed!
All the best,
Duncan.
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
