Jump to content

FredrikA

Members
  • Posts

    21
  • Joined

  • Last visited

Everything posted by FredrikA

  1. After further experiments where I copied a successful setup from another computer, it seems like the bug is in putty. I.e. if I have a session, without X forward, and change it to X forward in the gui it does not actually change the x forward setting. And by the way I am using putty to connect and then xming for the x session.
  2. I have been away from the forum for a while because my armbian has worked as a charm for a long while. Now however I have a minor problem All my devices (nvidia-jetson nano, raspberry pi 4, raspberry pi 2W, raspberry pi 3) can do x-windows to xming (running windows on my main computer). However none of my armbian can (running kernel Linux orangepi4 5.10.63-rockchip64, sunxi and sunxi64) start xeyes on my windows computer while raspberry and tegra can, so nothing is wrong with my setup. What has changed regarding X11 since kernel 4.x? Using "export DISPLAY=localhost:10.0" in all SBC:s ~/.bashrc On armbian I get "Error: Can't open display: localhost:10.0" On the others I get the xeyes app displaying as a x11 application.
  3. Linux orangepi4 5.10.63-rockchip64 #21.08.2 SMP PREEMPT Wed Sep 8 10:57:23 UTC 2021 aarch64 GNU/Linux Mine has renamed to 1 and 2 Filesystem 1K-blocks Used Available Use% Mounted on pi@orangepi4:~/ext-src/opencv/build$ df | grep mmc /dev/mmcblk1p1 60651184 22786748 37212048 38% / /dev/mmcblk2p1 14961544 12154596 2027224 86% /mnt/mmc Previously the were 0 and 1 using kernel 4.4.213-rk3399
  4. Board orangepi4b uname -a Linux orangepi4 5.10.63-rockchip64 #21.08.2 SMP PREEMPT Wed Sep 8 10:57:23 UTC 2021 aarch64 GNU/Linux Got syntax error on armbianmonitor -u: armbianmonitor -u "System diagnosis information will now be uploaded to syntax error at or near "" Please post the URL in the forum where you've been asked for" set -x armbianmonitor -u Network/firewall problem detected. Not able to upload debug info. Please fix this or use "-U" instead and upload whole output manually attached file err is output from armbianmonitor -U Anyway I had to reflash after a microSD broke. Previously I used 4.4.213-rk3399 (don't recall the armbian version) and then built in wifi worked. Now I used the latest image 5.10.63-rockchip64, and I can't get the built in wifi to work. As I think I recall right, i added the module rtl8189fs with "insmod 8189fs" and the kernel log said [ 4584.914559] RTL871X: module init start [ 4584.914579] RTL871X: rtl8189fs v4.3.24.8_22657.20170607 [ 4584.914783] RTL871X: module init ret=0 but I can't find the device with nmtui or lshw. What Am I doing wrong? err
  5. For my board to originally boot, I needed the USB TTL serial. Please recall the speed is high, and a lot of letters are missing during boot, Don't let that worry you. just press enter a couple of times to get it to sync once dmesg boot gets scilent. After this once, It was possible to boot from sd card.
  6. FredrikA

    Orange pi 4

    After a couple of loggings from different starts mine always ends with [ 113.070097] rockchip-dmc dmc: Get wrong frequency, Request 856000000, Current 800000000 and on the 8:th try. Then it is happy. What kernel do you run? Justin Wills has solved it for another board here:
  7. FredrikA

    Orange pi 4

    Got some headless measurements of current. OK-DO (original Raspberry PI PSU supplier 5V, 3A). "Cabriolet" since my 3d printer got some problems, cannot build cabinet right now. Headless. Only an USB memory stick attached (and the SD-card). Idle 0.61A 7z -b on repeat 1.79A (did go up to 1.82A but I was too late for picture)
  8. FredrikA

    Orange pi 4

    So I am not sure I can answer, but I will give it a try I got a Orange 4B I do not have the error line from the X-server [ 116.359] (EE) open /dev/fb0: No such file or directory instead on mine I've got in the file system pi@orangepi4:/var/log$ ls -l /dev/fb0 crw-rw---- 1 root video 29, 0 May 4 12:17 /dev/fb0 also dmesg|grep rockchip-drm [ 2.448079] rockchip-drm display-subsystem: defer getting devfreq [ 2.448650] rockchip-drm display-subsystem: bound ff900000.vop (ops 0xffffff8008c63878) [ 2.448885] rockchip-drm display-subsystem: bound ff8f0000.vop (ops 0xffffff8008c63878) [ 2.449080] rockchip-drm display-subsystem: failed to bind ff940000.hdmi (ops 0xffffff8008c58940): -517 [ 2.449330] rockchip-drm display-subsystem: master bind failed: -517 [ 3.301493] rockchip-drm display-subsystem: defer getting devfreq [ 3.301831] rockchip-drm display-subsystem: bound ff900000.vop (ops 0xffffff8008c63878) [ 3.301952] rockchip-drm display-subsystem: bound ff8f0000.vop (ops 0xffffff8008c63878) [ 3.305640] rockchip-drm display-subsystem: bound ff940000.hdmi (ops 0xffffff8008c58940) [ 3.306271] rockchip-drm display-subsystem: failed to parse loader memory [ 3.306347] rockchip-drm display-subsystem: No connectors reported connected with modes [ 3.373987] rockchip-drm display-subsystem: fb0: frame buffer device As far as I can see, you are lacking the three last lines from dmesg. I am currently running headless due to other projects (nvidia jetson nano) uname -a Linux orangepi4 4.4.192-rk3399 #6 SMP Tue Oct 8 18:32:46 CEST 2019 aarch64 GNU/Linux Maybe the resolution negotiation fails? I got so much problem with my 4k Samsung 55" TV, so I bought a dirt cheap 21" 1920x1080 display.
  9. FredrikA

    Orange pi 4

    I am not sure. I get these messages in dmesg only (and /var/log/armbian-hardware-monitor.log) dmesg|grep rockchip-dmc [ 2.687760] rockchip-dmc dmc: unable to get devfreq-event device : dfi [ 3.541515] rockchip-dmc dmc: Looking up center-supply from device tree [ 3.541712] rockchip-dmc dmc: Failed to get leakage [ 3.541726] rockchip-dmc dmc: Looking up center-supply from device tree [ 3.541755] rockchip-dmc dmc: Failed to get pvtm [ 3.541882] rockchip-dmc dmc: avs=0 [ 3.541901] rockchip-dmc dmc: failed to get vop pn to msch rl [ 3.542023] rockchip-dmc dmc: l=-2147483648 h=2147483647 hyst=0 l_limit=0 h_limit=0 [ 3.542042] rockchip-dmc dmc: could not find power_model node [ 3.542084] rockchip-dmc dmc: Get wrong frequency, Request 416000000, Current 800000000 [ 3.609043] rockchip-dmc dmc: Get wrong frequency, Request 416000000, Current 800000000 [ 7.293079] rockchip-dmc dmc: Get wrong frequency, Request 416000000, Current 800000000 .... [ 101.345126] rockchip-dmc dmc: Get wrong frequency, Request 416000000, Current 800000000 [ 101.397080] rockchip-dmc dmc: Get wrong frequency, Request 416000000, Current 800000000 [ 101.398339] rockchip-dmc dmc: Get wrong frequency, Request 856000000, Current 800000000 then it stops, so something happens for me.
  10. FredrikA

    Orange pi 4

    You can measure the power consumption and see if you are at risk. I can try to see if I can find my clamp ampere meter, I got it around somewhere, maybe I can make some measurement. If you are at risk, the processor cores draws a lot more when at high frequency, maybe you could implement a more restrictive throttling.
  11. FredrikA

    Orange pi 4

    I have no stability problem, since I got a cooling fan. I am using an OKDO usb-c power supply (the company that makes the PSU for raspberry pi). It has real 3A supply. I had it halt twice before Once because overheating Once because memory was exhausted when I tried to build tensorflow.
  12. http://ix.io/2glp I have updated to latest armbian And the built in wifi was still unstable After a while the link goes down and I get the following in dmesg 2767.152762] RTL8723BS: linked_status_chk(wlan0) disconnect or roaming It is annoying but no catastrophe. I run this bash script once every hour in crontab to restart wifi0 #!/bin/bash list=`ifconfig|grep 192|awk '{print $2}'` for l in $list; do #echo $list count=$(($count + 1)) done; #echo $count # if we got less than 2 interfaces, restart networking if [ $count -lt 2 ] ; then echo restarting network and network-manager sudo systemctl stop network-manager sudo systemctl stop networking sudo systemctl start networking sudo systemctl start network-manager fi
  13. FredrikA

    FredrikA

  14. FredrikA

    Orange pi 4

    I will answer my own question. After adding voltage debug to statistics I have concluded that regulator 10 is for the four ARM A53 and regulator 23 is for the two ARM A72. if anyone interested files here https://github.com/Freddan-67/rpimonitor-armbian.git
  15. Thanks @Werner, I actually believed I got upgrades automatically. I have updated now by re-flashing the SD card and will supervise if wifi is stable. Seems Ok after a couple of hours, but previously it could take a day or two before the instability showed. (Now armbianmonitor -u at http://ix.io/2g61 )
  16. My board works pretty well upto 1 000 00 Hz without fan and up to 1 200 000 Hz with (noctua) fan. However I am still using an usb wifi card, to be able to connect with ssh. Anyone having wifi with the built in wifi card work properly? Mine just shuts off. With an external wifi usb card, it works like a charm. Since my 3-d printers are in the living room I i have the orange pi:s in the living room. They are primary used for octoprint. I would like to have the built in wifi on orangepi prime to work.
  17. FredrikA

    Orange pi 4

    @IanHI am connected wireless (only), due to lacking ethernet in the living room. I still get some crashes on ethernet in dmesg. I am running: uname -a Linux orangepi4 4.4.213-rk3399 #15 SMP Thu Mar 19 21:33:44 CET 2020 aarch64 GNU/Linux (Armbian_20.02.5_Orangepi4_buster_legacy_4.4.213) I wonder if anyone knows how to get the Gyrfalcon Lightspeeur 2801S running in armbian? Also want to make the RPI monitor look nice, including core voltages, does anyone knows which voltage regulators go where? I found four regulators with suspected cpu voltages, but have no clue on how they are connected /sys/devices/platform/vdd-log/regulator/regulator.8/microvolts /sys/devices/platform/ff3c0000.i2c/i2c-0/0-001b/regulator/regulator.9/microvolts /sys/devices/platform/ff3c0000.i2c/i2c-0/0-001b/regulator/regulator.10/microvolts /sys/devices/platform/ff3c0000.i2c/i2c-0/0-0041/regulator/regulator.24/microvolts
  18. If you can start it you can stop it. when you start it keep a reference to the PID, then follow this guide https://www.thegeekstuff.com/2011/02/send-signal-to-process/
  19. This was tricky enough on raspberry pi, an unsupported screen on a supported platform, where I finally got it working. I haven't even tried to get it working on this half-supported platform. What I had to do on raspberry was add one line to /boot/config.txt dtoverlay=ads7846,cs=0,penirq=25,penirq_pull=2,speed=10000,keep_vref_on=0,swap_xy=1,pmax=255,xohms=150,xmin=5837,xmax=35819,ymin=59198,ymax=27978 please note that the [xy][min|max] are calibrated values for my screen. And I needed this in a new file in /usr/share/X11/x.org.conf called 98-calibration.conf Section "InputClass" Identifier "calibration" MatchProduct "ADS7846 Touchscreen" Option "MinX" "5837" Option "MaxX" "35819" Option "MinY" "27978" Option "MaxY" "59198" Option "SwapXY" "1" Option "SwapAxes" "1" Option "InvertX" "0" Option "InvertY" "1" Option "TransformationMatrix" "1 0 0 0 -1 1 0 0 1" EndSection I think I added these max and min x and y values in one place more than I needed, but when it finally worked,..... The dtoverlay ads7846 is not compiled for armbian as far as I know. I don't even have the source code. good luck.
  20. Thank you Igor much appreciated. Yes I got /dev/media0 (for the USB camera). However I got three small problems left, probably due to this image in naming is for Orange Pi PC, not Orange Pi PC+ 1) how do I enable wifi? It was mostly working on the previous kernel (minor hickups a couple of times per day) 2) how do I enable the GC2035 camera? It was working on the previous kernel. Now the module is missing. I haven't compiled kernel since 2.4 but I guess I will have to dive into bitbake/yocto. 3) how do I get my eMMC back? It was working out of the box on the old kernel. (And I got the old kernel on the eMMC) old kernel gave: $grep mmcblk1 dmesg.old-kernel mmcblk1: mmc2:0001 8WPD3R 7.28 GiB [ 1.315152] mmcblk1boot0: mmc2:0001 8WPD3R partition 1 4.00 MiB [ 1.315384] mmcblk1boot1: mmc2:0001 8WPD3R partition 2 4.00 MiB [ 1.316269] mmcblk1: p1 [ 1.317898] mmcblk1boot1: unknown partition table [ 1.319265] mmcblk1boot0: unknown partition table [ 1.319493] mmcblk mmc2:0001: Card claimed for testing. [ 7.509553] EXT4-fs (mmcblk1p1): mounted filesystem with ordered data mode. Opts: (null) [ 10.316476] EXT4-fs (mmcblk1p1): re-mounted. Opts: commit=600,errors=remount-ro but noting now. $dmesg|grep mmcblk1 $
  21. Hi I managed to get my usb camera taking pictures alright on Orange PI PC+ (VERSION=5.37.180112), $uname -a Linux orangepipcplus 3.4.113-sun8i #4 SMP PREEMPT Wed Nov 22 13:45:28 CET 2017 armv7l armv7l armv7l GNU/Linux And I even got the GC2035 working sort of with: $sunxi-pio -m "PG11<1><0><1><1>" $modprobe gc2035 $modprobe vfe_v4l2 But I have no /dev/media0 or /dev/media1. So I cannot re-focus the web camera (that supports it; Logitec C920). I can re-focus my usb camera on my Orange PI Prime (VERSION=5.37.180120) using $uvcdynctrl -v -d /dev/video0 --set='Focus (absolute)' $COUNT. Where i also have the /dev/media0 (or 1 if I plug in 2 cams) So why does it not work with Orange Pi PC+? I know it is a different kernel & chipset. Orange PI Prime says $uname -a Linux orangepiprime 4.14.14-sunxi64 #61 SMP Sat Jan 20 03:18:42 CET 2018 aarch64 aarch64 aarch64 GNU/Linux (did not get a /boot/dtb tree for 4.14.15 or 4.14.16 so I hang around with the old one) (Actually I find Orange PI Prime more mature than Orange PI PC+, with less number of crashes especially after tuning down ram speed.) Both are primary used for controlling 3d printers with octoprint, and taking snapshots with fswebcam.
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines