You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
if (pdata->hid_usage == 0)
{
// On Linux, libhidapi does not seem to return the HID usage from the report descriptor.
// We try to infer the board from the part information, until somebody fixes libhidapi.
// To use this workaround, the -F option is required.
avrdude_message(MSG_INFO, "%s: WARNING: Cannot detect board type (HID usage is 0)\n", progname);
This seems to happen to the libusb backend but not hidraw backend.
When I changed the device to the Logitech USB receiver:
Manufacturer String: Logitech
Product String: USB Receiver
Serial Number String: (1033)
Device Found
type: 046d c52b
path: 3-3:1.0
serial_number: (null)
Manufacturer: Logitech
Product: USB Receiver
Release: 1203
Interface: 0
Usage (page): 0x6 (0x1)
Indexed String 1: Logitech
Unable to send a feature report.
Unable to get a feature report: hid_error is not implemented yet
Unable to write(): hid_error is not implemented yet
Unable to write()/2: hid_error is not implemented yet
waiting...
waiting...
waiting...
waiting...
waiting...
waiting...
waiting...
waiting...
waiting...
waiting...
read() timeout
Activity
mcuee commentedon Aug 14, 2022
You can see from the following results that HID usage reporting is not good for the libusb backend.
JoergAtGithub commentedon Aug 14, 2022
@mcuee Could you try, what usbhid-dump returns for this device?
Youw commentedon Aug 14, 2022
This is a known and documented issue:
hidapi/libusb/hid.c
Line 729 in c3ae3d0
As I mentioned here - now there is a workaround (ask for USAGE after the device is open).
mcuee commentedon Aug 15, 2022
@Youw
In this case, should I close this issue as this is a known limitation and there is a workaround now?
For avrdude, I am suggesting to use the hidraw backend.
avrdudes/avrdude#1065 (comment)
Youw commentedon Aug 15, 2022
Lets close it after #438 is resolved - that's another related improvement.
mcuee commentedon Aug 16, 2022
Yes I can confirm that #444 is good for #438 which can serve as the workaround for this issue (to ask for USAGE after the device is open).
When I change the test device to Teensy 2.0.
When I changed the device to the Logitech USB receiver:
Youw commentedon Aug 16, 2022
A workaround is possible, based on current master.
mcuee commentedon Aug 16, 2022
Oh yes. Thanks for pointing that out. #432 makes the workaround possible, not #444.
avrdude probably will only use the workaround after 0.13 is released.