• Content Count

  • Joined

  • Last visited

Everything posted by jeanrhum

  1. Hi, I received this week my M1 obtained from the giveaway . I will try to report regularly the images I'll test, but there is 2 kinds of images: csc armbian and armbianTV images. Which one should be preferred?
  2. Hi Christos, Thanks for reporting this issue too. What is your kernel? I suppose that you tried to decompile from fs, but have you tried to decompile from the dtb in use from /boot/dtb ?
  3. Hi, I can confirm that latest available testing image based on groovy, downloaded directly from armbian pine64 page works well !
  4. I already use it on x86 like some others tools by the same author/team: fd, hyperfind and ripgrep. Their softwares are developped in rust and provide nice features and performances.
  5. ok, the only problem I face was when using dtc with the fs as input.
  6. Sorry, I don't know how to decompile by hand. With a small script, I tested the packaged dtc with all the dtbs from the boot folders of my edge-v (allwinner, amlogic and rockchip) and got no segfaults. With this board, I have much more warnings when decompiling from fs:
  7. Compiling from, I'm able to generate a dts file, but using the one installed from apt I got the following output on my pine64h-b: $ dtc -I fs /sys/firmware/devicetree/base -O dts -o custom.dts custom.dts: Warning (unit_address_vs_reg): /cpu-opp-table/opp@480000000: node has a unit name, but no reg property custom.dts: Warning (unit_address_vs_reg): /cpu-opp-table/opp@1704000000: node has a unit name, but no reg property custom.dts: Warning (unit_address_vs_reg): /cpu-opp-table/opp@720000000: node has a unit name, but no reg property custom.dts: Warning (unit_
  8. Your welcome! It's a pleasure to contribute to armbian. I also tried your proposition with the reading of the device-tree folder with this command: dtc -I fs -O dts /sys/firmware/devicetree/base -o custom.dts But on my edge-v (kernel 5.9), pine64h-b and opi1 with kernel 5.8, it generates a segmentation fault, whereas on my gt1 ultimate running armbianTV with kernel 4.20 it works.
  9. Testers needed. More especially, people with samsung, nexell or even marvell are expected. The main issue for now is the way the dtb is named and located in the boot folder. I tried to infer it using the comptible property of the device tree, but I don't own all supported hardware . I already tested sucessfully on several kernels (aml 4.20, rk and aw on 5.8, 5.9).
  10. First pull request for me: It is a basic implementation with my humble knowledge in bash. It may be improved by experts if possible. I tested on my armbian dev config and it seems to work. I hope to have follow the right way to make a PR since it is my first one.
  11. I don't see real questions in your post, but here is my personal experience. I run armbian on sbcs since 2016 and used it on one of them for home server. My current server is based on a s912 tv box with a 4.20 kernel installed maybe two years ago (I don't remember exactly ): Welcome to ARMBIAN 5.75 user-built Debian GNU/Linux 9 (stretch) 4.20.5-aml-s912 System load: 0.03 0.02 0.00 Up time: 135 days The system is amazingly stable and the only improvement I'm looking at, is maybe upgrading to a board/sbc having usb3 or sata(but since my local network is only 100MB, n
  12. As I saw in these data: [ 2.285937] sun6i-mipi-dsi 1ca0000.dsi: supply vcc-dsi not found, using dummy regulator Maybe it comes from there, because after searching about this error, I found these 2 links: - - And looking at your dts file with panel, I saw that the dldo1 node (phandle 0x04) is labelled as "vcc-hdmi" and it is referenced both in the dsi node (vcc-dsi-supply) an
  13. I think this could be an interesting soc, but in how many months (or years)? This soc is not referenced in sunxi status matrix and nothing is planned in kernel 5.11, so I fear that support is not even started from their side. Maybe it's close enough to h6, but I didn't see any information about somebody able to start linux with it and it has a different gpu.
  14. I think you refer to JMCC work, but it only works with kernel 4.4 if I'm right. The problem is to be able to compile rkmpp with mainline and then ffmeg or gstreamer.
  15. I gave some bucks for this powerful server. Paypal was perfect for me. Long live to armbian! I hope the target will be attained soon.
  16. One of the coming issue is also that Balbes stops supporting amlogic socs for his next builds, so don't expect new robust builds. If you have tv boxes, then it depends of the soc(s) you have. On my side, I was able to use the rockpi4 armbian image with a rk3399 board (edge-v) and changing after in armbianEnv.txt to the right dtb. I have a working up-to-date environment that I installed in emmc but I may have problems if the next kernel/uboot upgrades breaks things. I also tried to do that with another box I have with a s912 soc, but there is no official armbian build with this s
  17. Looking at the documentation of jellyfin, only the rpi3 or 4 seems to be the supported boards based on an arm architecture. Other than that, I think that JMCC have a script to install such stuff on most rk boards based on 4.4 kernel if I remember well, but I'm not sure transcoding is working: Myy is also working on porting vpu rk lib to mainline for rk3399; I'm not up to date about cedrus features for transcoding concerning Allwinner socs. I'm quite sure that my previous server sbc based on H3 was able to transcode using plex several years ago. My current server is based
  18. I think you misunderstand balbes. I think he was talking about u-boot for instance or any other low level software that can make things difficult when it does not fit the hardware. You can use the new images if you want to test, but there will be no support from him and no warranty that it'll work on your hardware. It is a bit sad, but when you look at the tv box market, amlogic boxes quality are worst and worst (many variants of the same reference with cheap chips and mainline support is mainly done slowly by Baylibre - only for the aml soc and not for all additional chips that may never
  19. You shouldn't you android recovery. Have you installed a custom firmware or any alternative system that may have modified the original u-boot? If so, just flash an original firmware and try again. If it doesn't boot directly try the toothpick method.
  20. Is it normal on your armbian storage, that the arm32 images are named Armbian_20.09_Arm-64_* ? Exactly like the arm64 ones.
  21. You have an error in your extlinux.conf: you should comment the rk3399 lines and uncomment 1 line of amlogic starting with fdt and the last one starting with append. Here is the one I use for s912 boxes (you may change the dtb if the generic q200 is not relevant): And you may also change the extlinux.conf-menu file:
  22. Even if Balbes makes a great work, I agree with Igor about updates and the use of a supported board. I'm not a good example since I use a S912 box with a Balbes image for my home server. But I can say that I stay on an "old" 4.19 or 4.20 since recent ones have several issues like GbE and usb even if some patches or hand modifications can be used to solve them. Rk3399 is powerful but heat a lot, so a good cooling is preferable if you plan to have heavy loads. Odroid c4 with amlogic S905x3 could be a nice alternative, since it is a supported board, you can add emmc and the
  23. I don't know if somebody make it work, but @@lex talked about that in another topic:
  24. Excelvan T2 (s912) using q201 dtb started from a slow usb key: The system seems usable. It is slow but I think it comes from my usb key. EDIT: same test with an H96pro+ using vim2 dtb using a uhs1 sd card : The system is faster than my previous test and desktop is reactive. On both config video playback was smooth.