Messages in this thread | | | Date | Sat, 14 Aug 2010 09:27:57 +0800 | Subject | Re: [PATCH] USB: Gadget: g_multi: added INF file for gadget with multiple configurations | From | Xiaofan Chen <> |
| |
On Sat, Aug 14, 2010 at 4:35 AM, Michal Nazarewicz <mina86@mina86.com> wrote: > Xiaofan Chen <xiaofanc@gmail.com> writes: > >> On Fri, Aug 13, 2010 at 5:31 PM, Dries Van Puymbroeck >> <dries.van.puymbroeck@gmail.com> wrote: >>> When a mutlifunction gadget has more than 1 configuration, Windows does not >>> use its generic driver for multifunctional USB devices. >> >> Sorry but what do you mean by the above? >> As per the previous discussions, it seems that a registry entry hack >> (for Vista and Windows 7) is needed. >> http://msdn.microsoft.com/en-us/library/ff538059%28VS.85%29.aspx >> >>> This commit adds an INF file that will load the required Windows driver >>> as well as a short description in gadget_multi.txt >> >> It seems to me your inf file forces the USB Generic >> Parent Driver (for USB Composite Device) to load. But for which >> configuration? I do not have any device to test myself. But >> I tend to guess it will still be the first configuration (either >> RNDIS or CDC ECM). > > 2-configuration g_multi has configuration with RNDIS as the first one so > that's pretty perfect for us.
I guess so. But I am wondering if the inf file really needed or not in this case.
Basically I am just wondering if the following statement is true or not. >>> When a mutlifunction gadget has more than 1 configuration, Windows does not >>> use its generic driver for multifunction USB devices.
I am thinking the inf file is not necessary. "When the system detects that a USB device is composite, it loads the USB common class generic parent driver (usbccgp.sys)." http://msdn.microsoft.com/en-us/library/ff538059%28VS.85%29.aspx
And since 1st configuration RNDIS is what desired, nothing needs to be done.
Obviously you have the device to test, I would be interested to know your test results under XP, Vista and Win7.
-- Xiaofan
| |