Displaying 4 results from an estimated 4 matches for "libusb_get_config_descriptor".
2023 Nov 05
1
Passing hid_rep_index to libusb_get_config_descriptor is wrong?
...ivers/libusb1.c b/drivers/libusb1.c
index 17f4b8f74..f49cc78aa 100644
--- a/drivers/libusb1.c
+++ b/drivers/libusb1.c
@@ -420,7 +420,7 @@ static int nut_libusb_open(libusb_device_handle **udevp,
upsdebugx(2, "Reading first configuration descriptor");
ret = libusb_get_config_descriptor(device,
- (uint8_t)usb_subdriver.hid_rep_index,
+ 0, //(uint8_t)usb_subdriver.hid_rep_index,
&conf_desc);
/*ret = libusb_get_active_config_descriptor(device,
&conf_desc);*/
When running from commit ab55bc0...
2023 Nov 05
1
Passing hid_rep_index to libusb_get_config_descriptor is wrong?
...; index 17f4b8f74..f49cc78aa 100644
> --- a/drivers/libusb1.c
> +++ b/drivers/libusb1.c
> @@ -420,7 +420,7 @@ static int nut_libusb_open(libusb_device_handle
> **udevp,
>
> upsdebugx(2, "Reading first configuration descriptor");
> ret = libusb_get_config_descriptor(device,
> - (uint8_t)usb_subdriver.hid_rep_index,
> + 0, //(uint8_t)usb_subdriver.hid_rep_index,
> &conf_desc);
> /*ret = libusb_get_active_config_descriptor(device,
> &conf_desc);*/
>
>...
2023 Nov 05
1
Passing hid_rep_index to libusb_get_config_descriptor is wrong?
...r
things like a USB device operating in high-speed or low-speed mode, or a
debug vs normal mode, or maybe high vs low power mode. Given this,
hid_rep_index (which libusb1.h defines as "number of the interface we use
in a composite USB device") is absolutely not the right thing to pass to
libusb_get_config_descriptor(), because the number of configurations are
not related to the index of the interface being used. Passing 0 would be
better than the current code, because 0 is what happens with not-composite
devices (most UPS devices) and I bet even among the composite ones, there's
only one config descriptor...
2023 Nov 05
1
Passing hid_rep_index to libusb_get_config_descriptor is wrong?
...device operating in high-speed or low-speed mode, or a
> debug vs normal mode, or maybe high vs low power mode. Given this,
> hid_rep_index (which libusb1.h defines as "number of the interface we use
> in a composite USB device") is absolutely not the right thing to pass to
> libusb_get_config_descriptor(), because the number of configurations are
> not related to the index of the interface being used. Passing 0 would be
> better than the current code, because 0 is what happens with not-composite
> devices (most UPS devices) and I bet even among the composite ones, there's
> only on...