Hello Greg, Alan,
In the 3.4 kernel, we have observed that the RNDIS function driver clears
the string id reference to 0 during unbind. And, during bind, we see that
the rndis_init() itself is initiated only when the RNDIS string id reference
is 0. Is this ok?
The issue behind this query is:
As rndis is requesting a new string id during every bind, during dynamic
function switches (e.g. enable tethering/disable tethering) RNDIS gets a
new string id for its descriptor from the composite framework. When connecting
to a Windows host, due to the string id changes, the device manager treats
this as a new instance of the device and requests for driver installation
during every connect.
When I did the below changes, the above issue is not seen anymore. Can
you provide
your views on this?
--Change--
diff --git a/drivers/usb/gadget/f_rndis.c b/drivers/usb/gadget/f_rndis.c
index 6cfbb5b..3c12c42 100644
--- a/drivers/usb/gadget/f_rndis.c
+++ b/drivers/usb/gadget/f_rndis.c
@@ -820,7 +820,7 @@ rndis_unbind(struct usb_configuration *c, struct
usb_function *f)
rndis_deregister(rndis->config);
rndis_exit();
- rndis_string_defs[0].id = 0;
if (gadget_is_superspeed(c->cdev->gadget))
usb_free_descriptors(f->ss_descriptors);
@@ -869,14 +869,13 @@ rndis_bind_config_vendor(struct
usb_configuration *c, u8 ethaddr[ETH_ALEN],
if (!can_support_rndis(c) || !ethaddr)
return -EINVAL;
+ /* ... and setup RNDIS itself */
+ status = rndis_init();
+ if (status < 0)
+ return status;
+
/* maybe allocate device-global string IDs */
if (rndis_string_defs[0].id == 0) {
-
- /* ... and setup RNDIS itself */
- status = rndis_init();
- if (status < 0)
- return status;
-
/* control interface label */
status = usb_string_id(c->cdev);
if (status < 0)
--End--
Thanks,
Balakumar R
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
