This is one of those "doesn't reproduce for me" sort of things,
so I can't really help with that.

You said this modem is a "USB 2.0" device ... fine, but is it also a
high speed device?  Your kernel messages have been edited to remove
important details like that ... that's a bad habit when you're asking
for any kind of help.

If it's a high speed device, then this looks to be the same thing
that a few other people have reported.  But agin, since it doesn't
reproduce for me, at this point there's nothing I can do to help.

I strongly suspect the attached patch will _not_ help you at all,
but give it a try anyway.

- Dave





Tweak some early enumeration code paths:

 - After a set_address() error, reset the port.  Failures in that particular
   protocol request leave the device's address indeterminate.

 - Let devices settle longer after successful SET_ADDRESS

EXPERIMENTAL ... 

Index: g26/drivers/usb/core/hub.c
===================================================================
--- g26.orig/drivers/usb/core/hub.c	2006-05-30 21:00:25.000000000 -0700
+++ g26/drivers/usb/core/hub.c	2006-06-02 18:11:33.000000000 -0700
@@ -2165,6 +2165,8 @@ static int hub_set_address(struct usb_de
 	if (retval == 0) {
 		usb_set_device_state(udev, USB_STATE_ADDRESS);
 		ep0_reinit(udev);
+		/* allow HCD and peripheral time to settle */
+		msleep(40);
 	}
 	return retval;
 }
@@ -2338,24 +2340,27 @@ hub_port_init (struct usb_hub *hub, stru
 #undef GET_DESCRIPTOR_BUFSIZE
 		}
 
+		/* after set_address faults it's not clear what the device's
+		 * address will be.  reset and be sure.
+		 *
+		 * REVISIT peripherals could trigger erroneous faults...
+		 */
 		for (j = 0; j < SET_ADDRESS_TRIES; ++j) {
 			retval = hub_set_address(udev);
 			if (retval >= 0)
 				break;
-			msleep(200);
+			(void) hub_port_reset(hub, port1, udev, delay);
 		}
 		if (retval < 0) {
 			dev_err(&udev->dev,
 				"device not accepting address %d, error %d\n",
 				udev->devnum, retval);
-			goto fail;
+			continue;
 		}
- 
+
 		/* cope with hardware quirkiness:
-		 *  - let SET_ADDRESS settle, some device hardware wants it
 		 *  - read ep0 maxpacket even for high and low speed,
-  		 */
-		msleep(10);
+		 */
 		if (USE_NEW_SCHEME(retry_counter))
 			break;
 
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to