Details of a "desktop" (4.4.13-ntc-mlc) flashed C.H.I.P


As requested by @marc - but others are welcome to ask for additional details and I’ll post as I can.

(I reserve the right to flash it again multiple times while testing things, so I may not be able to respond immediately.)

Kernel Config

Generated via:

 cat /boot/config-4.4.13-ntc-mlc | curl -F 'clbin=<-'

This is the config in use for the kernel (kernel source is available on github - along with other things like u-boot:

Big thanks to @computermouth and the rest for this - much, much easier than trying to hunt through google for various bits of Chinese documentation for my other sunxi / allwinner / mali projects (: )

List of pwm directory:

Sadly, there is nothing deeper than /sys/class/pwm .

I’ll update this if this needs to be enabled via some loaded module or similar.


root@chip:/sys/class# lsmod
Module                  Size  Used by
r8152                  33150  0
bnep                    9842  2
hci_uart               47655  1
btbcm                   5507  1 hci_uart
btintel                 5911  1 hci_uart
cmac                    2555  0
ecb                     1807  0
bluetooth             397077  23 bnep,btbcm,hci_uart,btintel
8723bs               1244907  0
cfg80211              217810  1 8723bs
sun4i_codec            12377  3
snd_soc_core          116548  1 sun4i_codec
snd_pcm_dmaengine       2943  1 snd_soc_core
snd_pcm                68852  2 snd_soc_core,snd_pcm_dmaengine
snd_timer              17776  1 snd_pcm
snd                    47207  3 snd_soc_core,snd_timer,snd_pcm
soundcore               4991  1 snd
ac97_bus                1300  1 snd_soc_core
mali                  183862  0

Further requests

Hit me up here in this thread if there’s something additional you’d like me to document.

3D Acceleration for C.H.I.P., Improved Web Flasher, and Alpha C.H.I.P.s are back!

Could you check if NetworkManager randomizes the MAC address when scanning? I’m trying to do my own debootstrapping, and the wifi doesn’t work unless I disable this feature.

Does it warn about failing to set the MAC address and proceed to scan anyway?


out of the box, it is not recording any errors or any sort; and on subsequent reboots it appears to be assigning the MAC address of 38:a2:8c:5e:86:36 to my wlan0.

(If I recall correctly, that can be set by uboot, so I’ll try to grab a full boot from uboot onward via serial and post a link to that. It may help with diagnosing anything set via uboot, and also show boot time / kprintf logging.


I captured that info, and will post links in a second, but I do note that within the uboot variables the ethaddr is 02:d7:08:02:df:4f . Having done that research, I am thinking that is NOT related to the wlan0 mac address at all, but instead is the mac address of the device connected to the ARM itself, used for pxe booting. (As for how that is attached to anything, I’m not sure it is.)


Capture: full boot output over UART

Capture: uboot / help / printenv

You can pull the raw output of down by leaving off the ?hl - curl pulls that just fine, a browser is likely to want you to save it to disk.


Could you check if systemctl stop getty@ttyGS0 works? From SSH, not from the USB cable. I’m trying to do my own debootstrapping (with Debian unstable), and attempting to do this makes systemctl stop working. I think it causes a deadlock in the kernel.