Hi,
On Wed, Oct 08, 2014 at 01:32:31PM +0200, Andrzej Pietrasiewicz wrote:
> Some not-so-well-behaving USB hosts with a popular proprietary operating
> system sometimes issue per-device requests even though they mean requests
> for a particular function, e.g. for ptp (picture transfer protocol).
which request would that be ?
> This patch adds optional "for_device" function in usb_configuration so that
> it can be used as the last fallback when handling unknown requests. If it
> is not set then the composite_setup() behaviour remains unchanged.
>
> Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@xxxxxxxxxxx>
> ---
> drivers/usb/gadget/composite.c | 16 ++++++++++------
> include/linux/usb/composite.h | 1 +
> 2 files changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
> index 6178353..32b0104 100644
> --- a/drivers/usb/gadget/composite.c
> +++ b/drivers/usb/gadget/composite.c
> @@ -1785,12 +1785,16 @@ unknown:
> }
>
> /* try the only function in the current config */
> - if (!list_is_singular(&c->functions))
> - goto done;
> - f = list_first_entry(&c->functions, struct usb_function,
> - list);
> - if (f->setup)
> - value = f->setup(f, ctrl);
> + if (list_is_singular(&c->functions)) {
> + f = list_first_entry(&c->functions,
> + struct usb_function, list);
> + if (f->setup)
> + value = f->setup(f, ctrl);
> + } else {
> + f = c->for_device;
> + if (f && f->setup)
> + value = f->setup(f, ctrl);
> + }
instead, why don't you just list_for_each_entry() here ?
--
balbi
Attachment:
signature.asc
Description: Digital signature
