Alexey Guskov

Looks like the source of the conflict is here: &uart0 { ... pinctrl-0 = <&uart0_xfer>; // RK_PC0 is rx and RK_PC1 is tx ... }; &uart1 { ... bluetooth { ... device-wake-gpios = <&gpio2 RK_PC1 GPIO_ACTIVE_HIGH>; host-wake-gpios = <&gpio2 RK_PC0 GPIO_ACTIVE_HIGH>; shutdown-gpios = <&gpio2 RK_PB7 GPIO_ACTIVE_HIGH>; pinctrl-names = "default"; pinctrl-0 = <&bt_host_wake_l &bt_wake_l &bt_enable_h>; // RK_PB4 and RK_PC1 ... }; }; So, indeed, bluetooth device is conflicting with uart0

Okay, silly me, of course a device wont work if there's no driver for it

Nope # find /proc/device-tree/serial-number find: ‘/proc/device-tree/serial-number’: No such file or directory Don't see a 'serial-number' device in .dts files for rk356x, are you sure it works on Station M2 for you?

M2

I'll start with questions then: 1) i got the armbian 5.17 kernel, patched the dts file, built a new .dtd, tried that on my device 2) Now i can see OTP device: # find /sys/firmware/devicetree/base/otp\@fe38c000/ /sys/firmware/devicetree/base/otp@fe38c000/ /sys/firmware/devicetree/base/otp@fe38c000/#address-cells /sys/firmware/devicetree/base/otp@fe38c000/cpu-leakage@1a /sys/firmware/devicetree/base/otp@fe38c000/cpu-leakage@1a/reg .... and cpuinfo device: # find /sys/firmware/devicetree/base/cpuinfo /sys/firmware/devicetree/base/cpuinfo /sys/firmware/devicetree/base/cpuinfo/nvmem-cells /sys/firmware/devicetree/base/cpuinfo/compatible /sys/firmware/devicetree/base/cpuinfo/nvmem-cell-names /sys/firmware/devicetree/base/cpuinfo/name But if I understand correctly, devices with NVMEM cells should be exposed to userspace as /sys/bus/nvmem/devices/somethingsomething/nvmem, but i got nothing there: # find /sys/bus/nvmem/ /sys/bus/nvmem/ /sys/bus/nvmem/uevent /sys/bus/nvmem/drivers_autoprobe /sys/bus/nvmem/devices /sys/bus/nvmem/drivers_probe /sys/bus/nvmem/drivers So, am I doing something wrong or this is expected behaviour? Do I need a loaded and active device driver for nvmem cells to be available via /sys/bus/nvmem or correct device tree definition should be enough?

Bluetooth device does not appear on my armbian (kernel is 5.17.5-media, tried on edge-media and current-media as well). # hciconfig version Can't get device info: No such device # bluetoothctl Agent registered [bluetooth]# devices No default controller available HCI socket does not seem to be created. Got this in dmesg: [ 11.990128] Bluetooth: HCI device and connection manager initialized [ 11.990145] Bluetooth: HCI socket layer initialized [ 11.990164] Bluetooth: L2CAP socket layer initialized [ 11.990188] Bluetooth: SCO socket layer initialized [ 12.081220] Bluetooth: HCI UART driver ver 2.3 [ 12.081245] Bluetooth: HCI UART protocol H4 registered [ 12.081250] Bluetooth: HCI UART protocol BCSP registered [ 12.081352] Bluetooth: HCI UART protocol LL registered [ 12.081358] Bluetooth: HCI UART protocol ATH3K registered [ 12.081397] Bluetooth: HCI UART protocol Three-wire (H5) registered [ 12.081760] Bluetooth: HCI UART protocol Intel registered [ 12.082139] Bluetooth: HCI UART protocol Broadcom registered [ 12.082205] Bluetooth: HCI UART protocol QCA registered [ 12.082211] Bluetooth: HCI UART protocol AG6XX registered [ 12.082335] Bluetooth: HCI UART protocol Marvell registered [ 12.082530] rockchip-pinctrl pinctrl: pin gpio0-17 already requested by fdd50000.serial; cannot claim for serial0-0 [ 12.082544] rockchip-pinctrl pinctrl: pin-17 (serial0-0) status -22 [ 12.082550] rockchip-pinctrl pinctrl: could not request pin 17 (gpio0-17) from group bt-enable-h on device rockchip-pinctrl [ 12.082557] hci_uart_bcm serial0-0: Error applying setting, reverse things back [ 12.297655] cfg80211: failed to load regulatory.db [ 12.308441] brcmfmac mmc2:0001:1: Direct firmware load for brcm/brcmfmac43455-sdio.firefly,rk3566-roc-pc.txt failed with error -2 [ 12.308466] brcmfmac mmc2:0001:1: Falling back to sysfs fallback for: brcm/brcmfmac43455-sdio.firefly,rk3566-roc-pc.txt [ 12.549006] brcmfmac: brcmf_fw_alloc_request: using brcm/brcmfmac43455-sdio for chip BCM4345/6 [ 12.589256] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4345/6 wl0: Feb 27 2018 03:15:32 version 7.45.154 (r684107 CY) FWID 01-4fbe0b04 Looks like wifi firmware loaded okay (not sure if the chip uses same blob for bluetooth and wifi, probably not), but there's some conflict with fdd50000.serial (which is uart0, another strange thing). To me it looks like a bug in .dts, but could not find it so far. Tried removing uart0 from .dtd. HCI device appears after that, but HCI socket is down and not responding. (Maybe I need to trigger gpio pin to enable bluetooth, but why doesn't it happen automatically?) Any ideas?

I know there's no support, I checked the sources. But the driver seems to be simple enough and .dts changes are trivial. I can port the code from firefly kernel and create PR, but will probably need some help.

sorry, a typo. Of course i meant /proc/cpuinfo

Hey guys, looks like there's no support for board serial number in latest images for Station M2. Serial number on rk356x is coming from otp memory, so: 1) there are no bindings for `otp` and `cpuinfo` devices in .dts, like here 2) there are no kernel driver to extract data from otp memory, like here I'm ready to contribute, but not sure what's the future plan for those.