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
Updated version with issue: 2a95baf, also tried 96e00d70bdadcbce16c98bdfd75451205dbbd5ce .
On top of these baseline versions I have added a custom copy of the infinity_led.bash / infinity_led_alphabet.bash script to enable my custom layers present in the kll repo (also updated). After updating the controller repo I have realigned that custom script with any changes in the base script.
Another difference, although I don't think it matters, is that previously I was building with a compiler installed natively on my Gentoo install, while now I have built via the Archlinux docker container, built from the dockerfile provided in the controller repo. Gentoo no longer provides an arm-none-eabi compiler as an official package.
Resulting Issue:
I have been using my custom layout, locally built with the controller repo, for a long time. Recently I decided to upgrade the controller repo and build again. But with the updated firmware I notice two issues:
USB initialization is very slow. On the original firmware I could plug the keyboard and start typing within a second. With the updated firmware, it takes more than 10 seconds before I can type. In 'dmesg' output you see that some of the HID devices are very slow to come up.
With the original firmware: (dmesg -H, showing delta timestamps)
[ +0.250161] usb 1-4: new full-speed USB device number 34 using xhci_hcd
[ +0.128011] usb 1-4: New USB device found, idVendor=308f, idProduct=000f, bcdDevice= 3.91
[ +0.000009] usb 1-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ +0.000004] usb 1-4: Product: Keyboard - Infinity_60_LED:alphabet PixelMap USB
[ +0.000002] usb 1-4: Manufacturer: Input Club
[ +0.000003] usb 1-4: SerialNumber: FFFFFFFFFFFF0029000E001224404E45 - mk20dx128
[ +0.004977] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:308F:000F.0056/input/input126
[ +0.053157] hid-generic 0003:308F:000F.0056: input,hidraw4: USB HID v1.11 Keyboard [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input0
[ +0.001491] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.1/0003:308F:000F.0057/input/input127
[ +0.052526] hid-generic 0003:308F:000F.0057: input,hidraw5: USB HID v1.11 Keyboard [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input1
[ +0.001396] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB System Control as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.2/0003:308F:000F.0058/input/input128
[ +0.051863] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB Consumer Control as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.2/0003:308F:000F.0058/input/input129
[ +0.000467] hid-generic 0003:308F:000F.0058: input,hidraw6: USB HID v1.11 Device [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input2
[ +0.001011] cdc_acm 1-4:1.3: ttyACM0: USB ACM device
[ +0.002022] hid-generic 0003:308F:000F.0059: hiddev99,hidraw7: USB HID v1.11 Device [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input5
[ +0.002261] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.6/0003:308F:000F.005A/input/input130
[ +0.052666] hid-generic 0003:308F:000F.005A: input,hidraw8: USB HID v1.11 Mouse [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input6
with the updated firmware:
[ +0.249314] usb 1-4: new full-speed USB device number 32 using xhci_hcd
[ +0.129555] usb 1-4: New USB device found, idVendor=308f, idProduct=000f, bcdDevice= 4.f6
[ +0.000018] usb 1-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ +0.000008] usb 1-4: Product: Keyboard - Infinity_60_LED:alphabet PixelMap USB
[ +0.000006] usb 1-4: Manufacturer: Input Club
[ +0.000005] usb 1-4: SerialNumber: FFFFFFFFFFFF0029000E001224404E45 - mk20dx128
[ +0.005642] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:308F:000F.0051/input/input122
[ +0.052637] hid-generic 0003:308F:000F.0051: input,hidraw4: USB HID v1.11 Keyboard [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input0
[ +0.001780] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.1/0003:308F:000F.0052/input/input123
[ +0.052491] hid-generic 0003:308F:000F.0052: input,hidraw5: USB HID v1.11 Keyboard [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input1
[ +0.001819] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.2/0003:308F:000F.0053/input/input124
[ +0.052212] hid-generic 0003:308F:000F.0053: input,hidraw6: USB HID v1.11 Device [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input2
[ +5.166210] input: Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB as /devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.3/0003:308F:000F.0054/input/input125
[ +0.000225] hid-generic 0003:308F:000F.0054: input,hidraw7: USB HID v1.11 Mouse [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input3
[ +5.121160] hid-generic 0003:308F:000F.0055: hiddev99,hidraw8: USB HID v1.11 Device [Input Club Keyboard - Infinity_60_LED:alphabet PixelMap USB] on usb-0000:00:14.0-4/input4
Notice twice a 5 second delay.
The second problem I encounter is that the /dev/ttyACM0 device is no longer present with the updated firmware. In the dmesg output for the original firmware, you could see: cdc_acm 1-4:1.3: ttyACM0: USB ACM device and with the new one this line is not present.
I found some references to a new debug protocol HID-IO and the old mechanism removed from the repo, but it is not clear to me how to access the CLI using this new protocol. Should it still show a /dev/ttyACM0 device? Or does it work differently? I could not find references to this in the documentation or commits.
I have checked the kernel config options related to HID and although I'm not sure what could be additionally required, the enabled set seems correct.
For the moment I stepped back to the original firmware, but it would be great to sort these issues out.
The text was updated successfully, but these errors were encountered:
Device:
Infinity IC60-LED (Massdrop)
Firmware Origin:
controller.git repo, built locally.
OS:
Gentoo Linux, but controller built in Archlinux docker image
Version:
On top of these baseline versions I have added a custom copy of the infinity_led.bash / infinity_led_alphabet.bash script to enable my custom layers present in the kll repo (also updated). After updating the controller repo I have realigned that custom script with any changes in the base script.
Another difference, although I don't think it matters, is that previously I was building with a compiler installed natively on my Gentoo install, while now I have built via the Archlinux docker container, built from the dockerfile provided in the controller repo. Gentoo no longer provides an arm-none-eabi compiler as an official package.
Resulting Issue:
I have been using my custom layout, locally built with the controller repo, for a long time. Recently I decided to upgrade the controller repo and build again. But with the updated firmware I notice two issues:
With the original firmware: (dmesg -H, showing delta timestamps)
with the updated firmware:
Notice twice a 5 second delay.
cdc_acm 1-4:1.3: ttyACM0: USB ACM device
and with the new one this line is not present.I found some references to a new debug protocol HID-IO and the old mechanism removed from the repo, but it is not clear to me how to access the CLI using this new protocol. Should it still show a /dev/ttyACM0 device? Or does it work differently? I could not find references to this in the documentation or commits.
I have checked the kernel config options related to HID and although I'm not sure what could be additionally required, the enabled set seems correct.
For the moment I stepped back to the original firmware, but it would be great to sort these issues out.
The text was updated successfully, but these errors were encountered: