On Mon, Dec 03, 2012 at 07:27:05PM +0100, Michal Nazarewicz wrote:
> > On Fri, Nov 30, 2012 at 07:40:31PM +0100, Michal Nazarewicz wrote:
> >> IMO that's actually mute argument since it can be done regardless. You
> >> can modprobe the module anyway.
>
> On Fri, Nov 30 2012, Felipe Balbi wrote:
>
On 12/03/2012 07:23 PM, Michal Nazarewicz wrote:
On 11/30/2012 07:39 PM, Michal Nazarewicz wrote:
Yeah, but what Felipe just made me realise is that if that's the case,
than it makes no sense to have the functions directory outside of the
gadget. Instead we should have something like:
On Mon,
> On Fri, Nov 30, 2012 at 07:40:31PM +0100, Michal Nazarewicz wrote:
>> IMO that's actually mute argument since it can be done regardless. You
>> can modprobe the module anyway.
On Fri, Nov 30 2012, Felipe Balbi wrote:
> fair enough, I just want to avoid doing any memory allocations, or HW
> init
> On 11/30/2012 07:39 PM, Michal Nazarewicz wrote:
>> Yeah, but what Felipe just made me realise is that if that's the case,
>> than it makes no sense to have the functions directory outside of the
>> gadget. Instead we should have something like:
On Mon, Dec 03 2012, Sebastian Andrzej Siewior wr
On Mon, 3 Dec 2012, Sebastian Andrzej Siewior wrote:
> I think I can make iteven more complicated: Everything on the gadget
> side uses language code 0x409 for en-us. The host-side on Linux takes
> the first available language code and falls back to 0x409 if it (the
> first reported) didn't work.
On Mon, 3 Dec 2012, Sebastian Andrzej Siewior wrote:
> On 11/30/2012 11:07 PM, Felipe Balbi wrote:
> > On Fri, Nov 30, 2012 at 08:09:38PM +0100, Sebastian Andrzej Siewior wrote:
> >> On 11/30/2012 05:43 PM, Alan Stern wrote:
> >>> The iProduct,... names are already defined in the USB spec, and the
On 12/03/2012 03:45 PM, Sebastian Andrzej Siewior wrote:
then how about we drop the hungarian notation altogether ? Instead of
sManufacturer we follow the host stack and just call it manufacturer.
Likewise for all other fields ??
that would be manufacturer, product, serial. Fine by me. Anyone?
On 11/30/2012 11:07 PM, Felipe Balbi wrote:
On Fri, Nov 30, 2012 at 08:09:38PM +0100, Sebastian Andrzej Siewior wrote:
On 11/30/2012 05:43 PM, Alan Stern wrote:
The iProduct,... names are already defined in the USB spec, and their
meanings are different from the way you are using them here. Si
On 11/30/2012 07:39 PM, Michal Nazarewicz wrote:
On Fri, Nov 30 2012, Sebastian Andrzej Siewior wrote:
That won't work. You can only symlink an instance to one UDC. It is an
instance after all. If you want two ACM ports you need two ACM folder
in your functions folder:
acm.ttyS0
acm.ttyS1
Ye
On 11/30/2012 05:04 PM, Felipe Balbi wrote:
Hi,
Hi,
# echo acm.2,obex.3,msc> \
/sys/kernel/config/usb_gadget/dwc3.0/gadget0/config0/functions
# echo ffs,sourcesink> \
/sys/kernel/config/usb_gadget/dwc3.0/gadget0/config1/functions
what is wrong with having the functions withi
Hi,
On Fri, Nov 30, 2012 at 07:40:31PM +0100, Michal Nazarewicz wrote:
> On Fri, Nov 30 2012, Felipe Balbi wrote:
> > why would you load a function before it's actually needed ? We could
> > end up with users who will just load the function and never actually
> > need them which will consume memor
Hi,
On Fri, Nov 30, 2012 at 08:09:38PM +0100, Sebastian Andrzej Siewior wrote:
> On 11/30/2012 05:43 PM, Alan Stern wrote:
> >By the way, can you name these string attributes
> >
> > sSerialNumber, sProduct, and sManufacturer (etc.)
> >
> >instead of
> >
> > iSerialNumber, iProduct, and iM
On 11/30/2012 05:43 PM, Alan Stern wrote:
By the way, can you name these string attributes
sSerialNumber, sProduct, and sManufacturer (etc.)
instead of
iSerialNumber, iProduct, and iManufacturer (etc.)?
Or is it too late to change?
Hehe :) Why should it be too late?
The iP
On Fri, Nov 30 2012, Felipe Balbi wrote:
> why would you load a function before it's actually needed ? We could
> end up with users who will just load the function and never actually
> need them which will consume memory for no reason.
IMO that's actually mute argument since it can be done regardl
On Fri, Nov 30 2012, Sebastian Andrzej Siewior wrote:
> That won't work. You can only symlink an instance to one UDC. It is an
> instance after all. If you want two ACM ports you need two ACM folder
> in your functions folder:
>
> acm.ttyS0
> acm.ttyS1
Yeah, but what Felipe just made me realise i
On Fri, 30 Nov 2012, Sebastian Andrzej Siewior wrote:
> /sys/kernel/config/usb_gadget/gadgets/# mkdir oha
> /sys/kernel/config/usb_gadget/gadgets/oha# find
> .
> ./iSerialNumber
> ./iProduct
> ./iManufacturer
> ./bcdUSB
> ./idProduct
> ./idVendor
> ./bMaxPacketSize0
> ./bDeviceProtocol
> ./bDevice
Hi,
On Fri, Nov 30, 2012 at 04:24:43PM +0100, Sebastian Andrzej Siewior wrote:
> On 11/30/2012 03:46 PM, Felipe Balbi wrote:
> >Hi,
>
> Hi,
>
> >On Thu, Nov 29, 2012 at 05:43:21PM +0100, Sebastian Andrzej Siewior wrote:
> >>|# modprobe dummy_hcd num=2
> >>
> >>|# find /sys/kernel/config/ -ls
> >
On 11/30/2012 04:11 PM, Felipe Balbi wrote:
This loads function without even creating a gadget. This would allow
the same function to be used in several different gadgets.
why would you load a function before it's actually needed ? We could end
up with users who will just load the function an
On 11/30/2012 03:46 PM, Felipe Balbi wrote:
Hi,
Hi,
On Thu, Nov 29, 2012 at 05:43:21PM +0100, Sebastian Andrzej Siewior wrote:
|# modprobe dummy_hcd num=2
|# find /sys/kernel/config/ -ls
| 5570 drwxr-xr-x 3 root root0 Nov 29 17:26
/sys/kernel/config/
| 5580 d
Hi,
On Fri, Nov 30, 2012 at 04:07:22PM +0100, Michal Nazarewicz wrote:
> On Fri, Nov 30 2012, Felipe Balbi wrote:
> > Hi,
> >
> > On Thu, Nov 29, 2012 at 05:43:21PM +0100, Sebastian Andrzej Siewior wrote:
> >> |# modprobe dummy_hcd num=2
> >>
> >> |# find /sys/kernel/config/ -ls
> >>
> >> | 55
On Fri, Nov 30 2012, Felipe Balbi wrote:
> Hi,
>
> On Thu, Nov 29, 2012 at 05:43:21PM +0100, Sebastian Andrzej Siewior wrote:
>> |# modprobe dummy_hcd num=2
>>
>> |# find /sys/kernel/config/ -ls
>>
>> | 5570 drwxr-xr-x 3 root root0 Nov 29 17:26
>> /sys/kernel/config/
>> |
Hi,
On Thu, Nov 29, 2012 at 05:43:21PM +0100, Sebastian Andrzej Siewior wrote:
> |# modprobe dummy_hcd num=2
>
> |# find /sys/kernel/config/ -ls
>
> | 5570 drwxr-xr-x 3 root root0 Nov 29 17:26
> /sys/kernel/config/
> | 5580 drwxr-xr-x 5 root root0
On 11/29/2012 06:26 PM, Michal Nazarewicz wrote:
+#define MAX_NAME_LEN 40
+static struct config_group *function_make(
+ struct config_group *group,
+ const char *name)
+{
+ struct usb_function *f;
+ char buf[MAX_NAME_LEN];
+ char *func_name;
+
On Thu, Nov 29 2012, Sebastian Andrzej Siewior wrote:
> |# modprobe dummy_hcd num=2
>
> |# find /sys/kernel/config/ -ls
>
> | 5570 drwxr-xr-x 3 root root0 Nov 29 17:26
> /sys/kernel/config/
> | 5580 drwxr-xr-x 5 root root0 Nov 29 17:26
> /sys/kerne
|# modprobe dummy_hcd num=2
|# find /sys/kernel/config/ -ls
| 5570 drwxr-xr-x 3 root root0 Nov 29 17:26
/sys/kernel/config/
| 5580 drwxr-xr-x 5 root root0 Nov 29 17:26
/sys/kernel/config/usb_gadget
| 5610 drwxr-xr-x 4 root root
25 matches
Mail list logo
