NanoPI M4

[Dante4]

Can someone post result (if possible with logs) of

glmark2-es2

And

glmark2-es2 --off-screen 

? Please

[@lex]

2 hours ago, mio75 said:

Does anybody have an idea what could be wrong?

Not sure it is related to your Domoticz crash but if i recall correctly the "warning" message is related to Wifi and should be harmless.

There was a patch to fix that and i think the message is due to the missing password or wrong password to connect to wifi.

BUT if you were connected to wifi before the crash you could try running without wifi configured and see what happens.

Hope this helps.

[Ruslan Kutdusov]

Does anyone know how to wake up the board after suspend?

[Ruslan Kutdusov]

On 1/31/2019 at 3:48 PM, Dante4 said:

Can someone post result (if possible with logs) of

glmark2-es2

And

glmark2-es2 --off-screen 

? Please

 

glmark2-es2.log

glmark2-es2-offscreen.log

[Dante4]

3 hours ago, Ruslan Kutdusov said:

 

glmark2-es2.log

glmark2-es2-offscreen.log

Thank you!

[mio75]

On 1/31/2019 at 12:19 PM, @lex said:

Not sure it is related to your Domoticz crash but if i recall correctly the "warning" message is related to Wifi and should be harmless.

There was a patch to fix that and i think the message is due to the missing password or wrong password to connect to wifi.

BUT if you were connected to wifi before the crash you could try running without wifi configured and see what happens.

Hope this helps.

 

Thanks for writing back

I have tried now to disable wifi (I only use 1G eth), to see if that helps...

 

Does anybody know if the Friendlyarm dist. has same issue?

When is that patch included in Armbian build?

 

THANKS!

[@lex]

6 hours ago, mio75 said:

Does anybody know if the Friendlyarm dist. has same issue?

Yes, it does have the same issue.

 

This harmless annoying warning message just halted my NanoPi M4 in a matter of 24 hrs in idle mode. Flooded 5 GB of data and kernel was stalled on next day.

That's why i thought this info could help.

I would not waste time on this issue, make sure you disable wlan completely and watch for the message, it should not appears, clean up your SD card and run domoticz.

If this fix your crash then you should hunt for the fix.

[mio75]

On 2/3/2019 at 9:59 PM, @lex said:

Yes, it does have the same issue.

 

This harmless annoying warning message just halted my NanoPi M4 in a matter of 24 hrs in idle mode. Flooded 5 GB of data and kernel was stalled on next day.

That's why i thought this info could help.

I would not waste time on this issue, make sure you disable wlan completely and watch for the message, it should not appears, clean up your SD card and run domoticz.

If this fix your crash then you should hunt for the fix.

Hi @lex,

 

MANY thanks for the info - much appreciated!

I have been running with wlan disabled and powered-down for a few days now, and so far no warnings/errors *fingers crossed*

So even though it's only warnings and should be harmless it fills logs and halts other important tasks - we will have to see if it is indeed the problem causing Domoticz to crash...

 

I hope somebody commits a fix for this...

 

/M

[dragonlost]

Hello.

I would like to know if it is possible to access the 2x pcie missing on the nanopi-m4?
Are the 4 USB 3.0 ports independent or 2x2 or all on the same hub?
The usb type C is it OTG?

 

Do you have a release date for the hat sata?
It is very very interesting! (Currently I use the 2 socket 2.0 USB port).

Thanks for the answers !

[mikegleasonjr]

Hi, I have issues with external hard drives. First and foremost, I have a proper power supply and a proper cable. In fact, tried several of them. I have a 3a and a 4.2a charger and really, really good cables.

Now.

 

1 time out of 3 when I issue a reboot command the USB hard disk is not "seen" on reboot. The device /dev/sd* doesn't exist. Even if I unplug the drive and plug it back, nothing. I have to issue a poweroff command and unplug the power supply and plug it back for the hard drive to show up on reboot.

 

I have 3 ext4 hard drives, all show the same behaviour. (the 3 are not plugged at the same time).

 

So to repeat the bug: on reboot the hard disk is not seen by the os 1 out of 3 times. Unplugging and plugging it back doesn't help, I have to completely unplug the NanoPi and plug it back to see the HD again.

 

Once in a while also during normal operations, the hard drive will vanish, much like described by @datajerk on nov 9. I have the same issue as him. He also tried multiple power sources and cables, even powering his board through headers and everyone here said it was a power issue. I don't think it is, you just have to reboot a couple of times to see that the drive will get lost.

 

I didn't tried the FriendlyElec distro to see if it exibits the same behaviour. Pretty much happening on every version of Armbian I tried, even the latest kernel in nightly.

 

Thanks for your help!

 

(I have the 4GB board)

[dragonlost]

I have exactly the same problem with a usb 3.0 stick (64 go sandisk) and a usb 3.0 SSD. Unable to retrieve the port without disconnecting the power supply. By cons I have a USB keyboard that has no problem.

[mikegleasonjr]

9 hours ago, dragonlost said:

I have exactly the same problem with a usb 3.0 stick (64 go sandisk) and a usb 3.0 SSD. Unable to retrieve the port without disconnecting the power supply. By cons I have a USB keyboard that has no problem.

 

The good news is the official FriendlyElec distro doesn't have this problem so it is not hardware related. The bad news (for me) is the FriendlyElec distro doesn't have the overlay module in the kernel so I have trouble setting up the latest Docker version. The kernel is also not great and I have problems with apt sometimes.

 

I like Armbian but I bought the M4 for a NAS solution so it doesn't work

[pavelectric]

Good afternoon, dear forum members. I have long had a question about the operation of the PCI-e bus on the M4 card. Who tried to use it?

I have an adapter with a PCI-e x1 to PCI-e x16, one that uses a USB 3.0 cable.

Theoretically, you need to take 6 pins from the M4 board - PCIE_RX0_P + PCIE_RX0_N + PCIE_REF_CLKP + PCIE_REF_CLKN + PCIE_TX0P + PCIE_TX0N - is this correct?

Will there be support for dannyo tires in Armbian? After all, the hat "M4 sata hat" obviously on the way.

I also want to connect for example a network card with a PCIe bus or a modem on the same bus.

[pavelectric]

Meanwhile ... a 4x SATA HAT for NanoPi M4 appeared on the market. Hats & Docks

Already ordered, I will try. Support is planned by the Armbian?

[mikegleasonjr]

@datajerk @dragonlost

 

Good news. I wrote previously that I had problems with USB drives connected. The issues seems to have been resolved. I just installed a fresh install of Armbian 5.75 (4.4.174).

 

No matter how much I rebooted and worked with the unit, the usb hard drive always was present.

 

OTOH it seems the bluetooth service can't start. I don't mind since I am not using it, but here an armbianmonitor -u output: http://ix.io/1BA5

 

Here's the systemctl status nanopi-bluetooth.service output:

 

● nanopi-bluetooth.service - Bluetooth nanopi
   Loaded: loaded (/lib/systemd/system/nanopi-bluetooth.service; enabled; vendor preset: enabled)
   Active: failed (Result: exit-code) since Wed 2019-02-20 16:24:13 +01; 4min 36s ago
  Process: 812 ExecStart=/usr/bin/hciattach /dev/ttyS0 bcm43xx 115200 > /dev/null (code=exited, status=203/EXEC)
  Process: 800 ExecStartPre=/bin/echo 1 > /sys/class/rfkill/rfkill0/state (code=exited, status=0/SUCCESS)
      CPU: 10ms

Feb 20 16:24:13 m4a systemd[1]: Starting Bluetooth nanopi...
Feb 20 16:24:13 m4a echo[800]: 1 > /sys/class/rfkill/rfkill0/state
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Control process exited, code=exited status=203
Feb 20 16:24:13 m4a systemd[1]: Failed to start Bluetooth nanopi.
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Unit entered failed state.
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Failed with result 'exit-code'.

 

Let me know if you need anything else.

[dragonlost]

On 2/14/2019 at 8:23 PM, pavelectric said:

Good afternoon, dear forum members. I have long had a question about the operation of the PCI-e bus on the M4 card. Who tried to use it?

I have an adapter with a PCI-e x1 to PCI-e x16, one that uses a USB 3.0 cable.

Theoretically, you need to take 6 pins from the M4 board - PCIE_RX0_P + PCIE_RX0_N + PCIE_REF_CLKP + PCIE_REF_CLKN + PCIE_TX0P + PCIE_TX0N - is this correct?

Will there be support for dannyo tires in Armbian? After all, the hat "M4 sata hat" obviously on the way.

I also want to connect for example a network card with a PCIe bus or a modem on the same bus.

 

I use only the part 2x usb 2.0 host with success

 

[Clu2t3r]

On 2/20/2019 at 4:31 PM, mikegleasonjr said:

@datajerk @dragonlost

 

Good news. I wrote previously that I had problems with USB drives connected. The issues seems to have been resolved. I just installed a fresh install of Armbian 5.75 (4.4.174).

 

No matter how much I rebooted and worked with the unit, the usb hard drive always was present.

 

OTOH it seems the bluetooth service can't start. I don't mind since I am not using it, but here an armbianmonitor -u output: http://ix.io/1BA5

 

Here's the systemctl status nanopi-bluetooth.service output:

 

● nanopi-bluetooth.service - Bluetooth nanopi
   Loaded: loaded (/lib/systemd/system/nanopi-bluetooth.service; enabled; vendor preset: enabled)
   Active: failed (Result: exit-code) since Wed 2019-02-20 16:24:13 +01; 4min 36s ago
  Process: 812 ExecStart=/usr/bin/hciattach /dev/ttyS0 bcm43xx 115200 > /dev/null (code=exited, status=203/EXEC)
  Process: 800 ExecStartPre=/bin/echo 1 > /sys/class/rfkill/rfkill0/state (code=exited, status=0/SUCCESS)
      CPU: 10ms

Feb 20 16:24:13 m4a systemd[1]: Starting Bluetooth nanopi...
Feb 20 16:24:13 m4a echo[800]: 1 > /sys/class/rfkill/rfkill0/state
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Control process exited, code=exited status=203
Feb 20 16:24:13 m4a systemd[1]: Failed to start Bluetooth nanopi.
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Unit entered failed state.
Feb 20 16:24:13 m4a systemd[1]: nanopi-bluetooth.service: Failed with result 'exit-code'.

 

Let me know if you need anything else.

 

 

To activate Bluetooth on the Nanopi M4 you just need to execute the following commands (as root):

 

rfkill unblock all
echo "0" > /sys/class/rfkill/rfkill0/state
echo "1" > /sys/class/rfkill/rfkill0/state
hciattach /dev/ttyS0 bcm43xx 1500000 flow bdaddr   42:29:B1:55:01:01

 

Edited March 3, 2019 by Clu2t3r

[Da Alchemist]

I have just received my SATA HAT, it is recognized and working on Armbian out of the Box  (4.4.176-rk3399).  

[mitag]

14 minutes ago, Da Alchemist said:

I have just received my SATA HAT, it is recognized and working on Armbian out of the Box  (4.4.176-rk3399).  

Me too, mine arrived yesterday, but the 48W switching power supply arrived today. Only the installation was a bit fiddly. I had copied Armbian_5.65_Nanopim4_Ubuntu_bionic_default_4.4.162_desktop to an SD at the end of October and transferred it from there to an eMMC using armbian-config. In December I had packed a mSATA SSD into a USB3 case, connected it to the NanoPi and very conveniently, with armbian-config, integrated the USB SSD so that it boots from the eMMC.

 

So far so good, I was satisfied with this solution. When I then installed the SATA HAT, the NanoPi didn't want to boot from a current Armbian Ubuntu, nor from an Armbian Stretch, nor from an older version, as long as the eMMC was installed at the same time. Then I installed an SD version of Friendly Core Bionic and repartitioned the eMMC. Then I could boot normally with Armbian_5.75_Nanopim4_Debian_stretch_default_4.4.174 from SD and set up the eMMC first, reboot and without SD, set up the SATA SSD in a second pass. Fiddling with a SMD tweezer to insert or remove the eMMC under the SATA HAT was a bit annoying. But with the RockPi 4 it's even narrower under the cooler.

[dogshome]

Hi all. sata hat also on way.

 

Regarding the two microphone inputs: how is the trrs jack connection enabled? I think the default is the solder pad connection. I've looked in alsa mixer and that has various inputs for the codec but nothing appearing in sound. 

 

I'm using friendly OS right now, but I guess it is the same or very similar like the audio output in most versions of Linux.

[dogshome]

PS Like to use the mic for spectrum analysis. I have two tweaked TAS5630 class D amps, one in PBTL and several watts into a 90L sub, plus BTL and also several into my 2 way 10" mids and somewhat ample horn tweeters. 48V DC 20A supply.... Mic through the 'phone says flat down to 20Hz. I have a subsonic filter below that. So sound from this board all OK to my ears!

 

PPS The speakers happily take the TPA3116 amps into clipping on 24VDC. A bit more headroom with 48.

 

Live music in my shed :-)

[Da Alchemist]

 @dogshome  For real speaker measurements, get yourself a calibrated Microphone like the MiniDSP UMIK and use RoomEQWizard as Software. Too fool yourself just go ahead taking measurements with your mobile.

[dogshome]

I am playing with a Dayton calibrated mic. Hence wanting mic input on the M4 to compare with it plugged into 'phone i.e via trrs jack.

 

The calibrated mic plugged into my Wileyfox swift plus on Android seems to be reasonably truthful. The works iPhone gets all confused with bass and shows response when there is nothing there.

 

I'd like to have it working on the M4, but the system does not see the jack input. I think it's configured for the second mic (solder pads) input ? 

Edited March 17, 2019 by dogshome
more testing

[mitag]

Because everything ran so smoothly and thanks to the SATA HAT there is enough space for experiments, I installed the desktop under Armbian Stretch in armbian-config. Then I installed the Arduino IDE 1.8.9 for Linux64, cif-utils to get to my sketches and tightvncserver to work comfortably from my PC. Everything was chic until I tried to set up the IDE for my STM32 and SAMD boards.

Spoiler

This will be a fumbling.

 

Dance with me the BOSSAnova.

 

[martinayotte]

16 minutes ago, mitag said:

until I tried to set up the IDE for my STM32 and SAMD boards.

That is pretty normal since one of the bundled package is bringing cross-compiler which is compile for Intel.

It can probably be workarounded : Install them on a PC, then copying over network the arduino-1.8.9/hardware/<stm32> onto Nano, then try to recompile gcc cross-compilers for the platform.

[mitag]

I copied the whole Arduino15 folder. 50 thousand files, 6GB, even compressed still 3GB. Shortly before the end of the upload, vnc got out of step, only had to log me off and on with the client, then I could see that the copy process was finished cleanly. Unfortunately the Arduino IDE doesn't show the installed boards or libraries. This is getting too off-topic. But the Armbian did well with this action.

Spoiler

 

 

 

[martinayotte]

36 minutes ago, mitag said:

I copied the whole Arduino15 folder.

This maybe not enough, since some files are in arduino-1.8.9/hardware ...

[pavelectric]

Good day. I had a question - why not Ubuntu Core? I don't like the Xfce environment.

Of course, it can be removed, until it is not so convenient if it were a clean Core.

I don't like Debian Core.

 

[SriLK]

Hello

I am a new user here! weeee :-)

 

I have some questions, 1st one: Have anybody discover what are the points to short in order to enter MASKROM mode? I have enter it booting the M4 without sdcard or emmc, just to test that I can connect to the M4 via rkflashkit through the USB-C bus... and it does. So, I would need to to the same but with the emmc connected to the M4 and short the EMMC_CLK to GND (I guess). Have anybody discovered these points in the M4 PCB and willing to show them at hi-res image? :-)

 

This would help a lot, since I could backup directly the rootfs partitions with rkflashkit, mount it in my linux box, modify it and flash it back to the emmc again with rkflashkit. Very straight forward! 

 

But, If these MASKROM points are too complicated for trying even with 2 needles (connected with wire) then I will explain further my problem. Since maybe there is another solution to my problem.

 

Some time ago I have flashed to the M4 the image rk3399-eflasher-friendlydesktop-bionic-4.4-arm64-20181219 to my emmc (16GB) and well... M4 is still with Ubuntu Core OS

I have 1 external HD connected to the nanopi M4 and added to /etc/fstab (by UUID) for automount on boot (sadly I didn't add the param nofail). HD now died. Boot process gets stuck. No matter how much I wait, I get permanent blank screen, no sshd, nor any other way to interact with the OS (tried pushing every key in the keyboard with any change).

 

Ok, 1st thing I backed emmc to an image (dd)

Then I started to try to access the rootfs partition, in order to modify the /etc/fstab ... without success!

The emmc should have this structure:

FIRMWARE_VER: 6.0.1
MACHINE_MODEL: RK3399
MACHINE_ID: 007
MANUFACTURER: RK3399
MAGIC: 0x5041524B
ATAG: 0x00200800
MACHINE: 3399
CHECK_MASK: 0x80
PWR_HLD: 0,0,A,0,1
#KERNEL_IMG: 0x00280000
#FDT_NAME: rk-kernel.dtb
#RECOVER_KEY: 1,1,0,20,0
#in section; per section 512(0x200) bytes
CMDLINE: root=/dev/mmcblk1p7 rw rootfstype=ext4 mtdparts=rk29xxnand:

0x00002000@0x00002000(uboot),
0x00002000@0x00004000(trust),
0x00002000@0x00006000(misc),
0x00006000@0x00008000(resource),
0x00010000@0x0000e000(kernel),
0x00010000@0x0001e000(boot),
-@0x00030000(rootfs) consoleblank=0

So, I tried mounting from backup image:

mount -t ext4 m4-image /media/img/ -o loop,offset=100663296  (offset for rootfs should be 0x30000 * 512)

and fails with "wrong fs type"

then I list the image's structure with: fdisk -l m4-image

Device         Boot   Start      End  Sectors  Size Id Type
m4-image1           3342336 15234373 11892038  5.7G 83 Linux
m4-image2            196608  3342335  3145728  1.5G 83 Linux

So... I try m4-image1 partition instead:

mount -t ext4 m4-image /media/img/ -o loop,offset=1811939328  (offset for rootfs should be 3342336 * 512 + 0x30000 * 512, in case rootfs is inside m4-image1 partition?)

same error as before, "wrong fs type"

 

This make me think that I would need some kind of tool for extracting the data from m4-image. Here is my 2nd question: Does that tool exists, can I download from somewhere? ..or the reason for these mount failures is that I am calculating wrongly the offsets? if so, what the correct offset for rootfs in m4-image is?

 

I have also tried entering the M4 with the original rk3399-eflasher-friendlydesktop-bionic-4.4-arm64-20181219 image flashed the sdcard (and booting from sdcard. so, emmc as storage).

I connect to sshd (user:root/pass:fa) I can access the device. But, what I see there is:

root@NanoPi-M4:/dev# ls -al | grep mmc
brw-rw----  1 root disk    179, 128 Apr 14 08:54 mmcblk0
brw-rw----  1 root disk    179, 129 Apr 14 08:54 mmcblk0p1             <--- mounted at /mnt/sdcard
brw-rw----  1 root disk    179, 130 Apr 14 08:54 mmcblk0p2             <--- mounted at /
brw-rw----  1 root disk    179,   0 Apr 14 08:54 mmcblk1
brw-rw----  1 root disk    179,  32 Apr 14 08:54 mmcblk1boot0
brw-rw----  1 root disk    179,  64 Apr 14 08:54 mmcblk1boot1
brw-rw----  1 root disk    179,   1 Apr 14 08:54 mmcblk1p1             <--- this seems to be same as m4-image1 backed up with dd
brw-rw----  1 root disk    179,   2 Apr 14 08:54 mmcblk1p2             <--- this seems to be same as m4-image2 backed up with dd
brw-rw----  1 root disk    179,  96 Apr 14 08:54 mmcblk1rpmb

I can not see the other mmcblk devices, So, I am missing the devices that actually I would know how to mount. These are the devices when Ubuntu Core OS (flashed with image rk3399-eflasher-friendlydesktop-bionic-4.4-arm64-20181219) is booted:

root@NanoPi-M4:/dev# ls -al | grep mmc
brw-rw----  1 root disk    179,   0 Jan 28  2018 mmcblk1
brw-rw----  1 root disk    179,  32 Jan 28  2018 mmcblk1boot0
brw-rw----  1 root disk    179,  64 Jan 28  2018 mmcblk1boot1
brw-rw----  1 root disk    179,   1 Jan 28  2018 mmcblk1p1             <--- uboot (Not the same as when booted with eflasher sdcard image)
brw-rw----  1 root disk    179,   2 Jan 28  2018 mmcblk1p2             <--- trust (Not the same as when booted with eflasher sdcard image)
brw-rw----  1 root disk    179,   3 Jan 28  2018 mmcblk1p3             <--- misc
brw-rw----  1 root disk    179,   4 Jan 28  2018 mmcblk1p4             <--- resource
brw-rw----  1 root disk    179,   5 Jan 28  2018 mmcblk1p5             <--- kernel
brw-rw----  1 root disk    179,   6 Jan 28  2018 mmcblk1p6             <--- boot
brw-rw----  1 root disk    179,   7 Jan 28  2018 mmcblk1p7             <--- rootfs *** This one I'd like to mount
brw-rw----  1 root disk    179,  96 Jan 28  2018 mmcblk1rpmb

So, my 3rd question, Is there a way to populate mmcblk1p7 from eflasher so I can mount it, so I can modify the /etc/fstab?

Any reply to any of the 3 questions would help greatly.

For more explanations the 3 questions could be be answered, that would be awesome. I am very interested in the info about the MASKROM mode points :-)

 

Thanks!

Edited April 14, 2019 by SriLK
minor typos and more light

[SriLK]

Btw, here http://wiki.friendlyarm.com/wiki/images/7/7a/NanoPi-M4-4GB-1807-Schematic.pdf I could see the MASKROM points... I guess the one to short is the one at page 7 of 25 under the symbol: EMMC_CLKO, right on R194 33R or R0201.

 

Before in this thread it has been hinted that it would be eMMC_RST (with point at C349) shown at page 25 of 25, but it looks more likely to me (after reading about the MASKROM mode) that EMMC_CLKO is the one to put to GND while boot (shorting for 3 or 4 seconds just after connecting the USB-C)

 

No matter which ones are the correct ones, my question is: where they would be in the board? :-)