All Activity

This stream auto-updates     

  1. Past hour
  2. You have to start here:
  3. do u have a serial-cable UART to read all system messages while rebooting // sent from mobile phone //
  4. Today
  5. I would like to make Orangepi zero to capture USB output destined for a printer in order to log the output from a system. 1 ) Mimic the connected USB printer (so that the machine thinks it still has the original printer attached) 2 ) Capture the output from the machine 3 ) Send the captured output to the printer.
  6. I used N6 dtb, rk3399-magicsee-n6-max.dtb, which works! Thanks so much! If you put up a tutorial, there will be more members to follow. Next question: where we can modify the rootfs info such as partition ext2, ext3, ext4 etc?
  7. Did you edit /boot/extlinux/extlinux.conf to FDT /dtb/rockchip/rk3399-rock-pi-4-2.dtb
  8. Yesterday
  9. Okay, the secondary card is a pretty good idea! Simplest solution first and all that. I was considering starting from scratch, I'd have to dig up all the information on the configuration etc from my memory but it should still be knocking around in there. Really I was just wanting to upgrade as when I was still on Jessie, eventually the repo's stopped responding as it had hit end-of-life and I was lazy with my upkeep of the little guy. Was kind of a pain to get from Jessie to Stretch, and I can't 100% say I did it correctly (though its been running well up until Sunday night when I was messing with it). TY for the advice!
  10. My system disk is a 120 GB SSD - i have one of those sata to usb cables that has usb 2.0 (power) and usb 3.0 (data) connectors so I guess the SSD has enough power. I tried and learned that the reason is probably the SD card itself. I tried with two SD cards. First is Sandisk Ultra 32 GB and second is Verbatim 8 GB class10. The process of preparing was the same - I've burned the last stretch image with etcher. SanDisk reboots and Verbatim doesen't. So I guess the main reason is the speed of the card (or maybe someting else)? It seems it is important even if the system is on fast SSD. If you have any additional comments I'll be glad to hear them. Thanks!
  11. I've uploaded Armbian_5.37_S9xxx_Ubuntu_xenial_3.14.29_server_20171226.img.xz to megashares (md5sum a867b08685f25304847acfed7cc88803) -!xqZgkQrB!g7HzYbusB0wifVxKbGwR-HPGETRvJMXyhu-u9SDy8pM If you want any others, let me know - I've got kernel 3.14.79 as well.
  12. I flashed the default img to sd card but it failed to boot up my N6 box with the dtb files you posted. it says that it cannot find ROOTFS. Can you put up a tutorial for howto? Thank for your help! I looked up armbianEnv.txt in which rootdev=UUID=e4f3cef1-6360-464a-a70d-21cfd96b763c rootfstype=BOOT_IMAGE=/boot/vmlinuz-5.0.0-32-generic rootdev=LABEL=ROOTFS rootfstype=ext4 but in the BOOT (dos) there is no file named /boot/vmlinuz-5.0.0-32-generic.
  13. I have a Jenkins server that builds kernels/images for several boards that I own, and, for the most part it works very well. The biggest problem that I currently have is the kernel configuration. I need to add a few patches, which generally works well, and change a couple of configuration options, which sometimes fails. The problem is that I go through a process of creating a modified kernel configuration by first building a stock configuration. I then go into the kernel source directory under cache, make my changes, do a "make oldconfig", grab the new config file and add it to my userpaches directory. That works for a while, but eventually the kernel configuration changes enough that I get non-bootable images and have to go through the process again, potentially for each board. Is there a better way of doing this? Is there a way to build the kernel using the default configuration with just a few options changed?
  14. @kevinplatt would it be enough to freeze the dtb-package (for not get rid of the usb & audio fix) because armbian-config does set much more packages on hold like root, kernel, config, firmware and u-boot? root@npi-a64( apt-mark showhold armbian-firmware linux-buster-root-dev-nanopia64 linux-dtb-dev-sunxi64 linux-image-dev-sunxi64 linux-u-boot-nanopia64-dev Today i compiled a image with kernel 5.3.7 and did by mistake/accident a apt update & apt upgrade, so the USB & audio fix was missing with the newer dtb[5.98.191022] Debian Buster with Armbian Linux 5.3.7-sunxi64 package bsp-kernel[5.98.191022] u-boot[5.98] dtb[5.98.191022] firmware[5.98] config[5.98] I copied the saved/fixed .dtb in /boot/dtb/allwinner and the fix does work again after reboot I f I only freeze the dtb - can I also have/use the other packages in the newer version? root@npi-a64( apt-mark hold linux-dtb-dev-sunxi64 linux-dtb-dev-sunxi64 set on hold. root@npi-a64( apt-mark showhold linux-dtb-dev-sunxi64 Or which package (additional to the dtb) could break the fix?
  15. I never use HDMI, so the above outputs are from UART log. The strange thing is that with "quiet" argument, I get "[ 37.019170] reboot: Restarting system", and without "quiet", it stay frozen ...
  16. I correctly understood that nothing is displayed on the HDMI screen and the system hangs on this message for a long time (from the last UART log) ?
  17. U-Boot itself seems to work, but starting kernel make a reboot loop : I've then decided to remove the "quiet" argument to get more details, but strangely, this time it didn't do reboot loop, but stuck trying to mount sdcard : I think the problem is that I'm using plain class 10 SDCard ... EDIT: I've compared with my running Armbian image, and SDCard is also shown as "ultra high speed DDR50 SDHC", but it also showing "[ 3.126895] mmcblk1: p1" which here with your image it stall even before showing partitions ...
  18. This works for us. When you get to the boot loader ... generic arm64 kernel config you will not get you far.
  19. Hello everybody, Does anyone has Ubuntu server image with Kernel 3.14?
  20. I'm trying to build a linux from source on a RK3399 nanopi-m4. I did the below steps: ### Download and compile ARM-TF ### $ git clone $ make CROSS_COMPILE=aarch64-linux-gnu- PLAT=rk3399 bl31 ### Download and complie U-Boot ### $ cd .. $ git clone $ cd u-boot/ $ git checkout v2019.10 $ make CROSS_COMPILE=arm-linux-gnueabi- nanopi-m4-rk3399_defconfig $ cp ../arm-trusted-firmware/build/rk3399/release/bl31/bl31.elf . $ make ARCH=arm CROSS_COMPILE=aarch64-linux-gnu- $ make ARCH=arm CROSS_COMPILE=aarch64-linux-gnu- u-boot.itb $ ./tools/mkimage -n rk3399 -T rksd -d ./tpl/u-boot-tpl-dtb.bin idbloader.img Image Type: Rockchip RK33 (SD/MMC) boot image Data Size: 47104 bytes $ cat spl/u-boot-spl-dtb.bin >> idbloader.img ### Build kernel ### $ git clone $ cd linux $ git checkout v5.1 **I've used the arm64 defconfig *** $ make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- defconfig $ make CROSS_COMPILE=aarch64-linux-gnu- ARCH=arm64 ### Create boot.img ### $ cd .. $ mkdir boot $ cp linux/arch/arm64/boot/dts/rockchip/rk3399-nanopi-m4.dtb boot/ $ cp arch/arm64/boot/Image boot/ $ cp linux/arch/arm64/boot/Image boot/ $ mkdir boot/extlinux $ nano boot/extlinux/extlinux.conf label NanoPi M4 5.1 kernel /boot/Image devicetree /boot/rk3399-nanopi-m4.dtb append earlycon=uart8250,mmio32,0xff1a0000 root=/dev/mmcblk1p1 rootwait $ dd if=/dev/zero of=boot.img bs=1M count=32 $ sudo mkfs.fat boot.img $ mkdir tmp $ sudo mount boot.img tmp/ $ cp -r boot/* tmp/ $ umount tmp *** copy the images to a folder *** $ mkdir images $ cp boot.img images/ $ cp u-boot/idbloader.img images/ $ cp u-boot/u-boot.itb images/ ### Create sd card image ### $ cd images/ *** I'm using a VM so it was easier to use a loop device to create the image *** $ dd if=/dev/zero of=sdcard.img bs=1024 count=0 seek=400000 || exit 1 $ sudo losetup -f /dev/loop18 $ sudo losetup /dev/loop18 sdcard.img $ sudo dd if=idbloader.img of=/dev/loop18 seek=64 $ sudo dd if=u-boot.itb of=/dev/loop18 seek=16384 $ sudo dd if=boot.img of=/dev/loop18 seek=32768 $ sync Flush the image(sdcard.img) with etcher to a card and place the card in the board. Plug the UART cable and power on the board. I receive the below message: U-Boot TPL board init sdram_init: data training failed rk3399_dmc_init DRAM init failed -5 Does anyone know what the issue is?
  21. "Armbian is community driven open source project" You are asking this yourself. A few of Armbian contributors does not deal with 4 years old bugs found in 3rd party wireless driver we don't support at all. Its not the first and not the last one that is in the code. This one is too small to bother with. You will receive this fix after its landed upstream and after next regular release. You can use automated daily builds, where this will land automatically but those are untested builds. Decide what is worse.
  22. On imx6q SoC there is: v4l2-ctl --device=/dev/video0 --all v4l2-ctl --device=/dev/video1 --all v4l2-ctl --device=/dev/video10 --all /dev/video10 requires at least kernel 5.4.0-rc1 to work out of the box. gst-inspect-1.0 exposes: So hardware accelerated video pipelines can be composed in all flavors.
  23. Those are automated builds and there only an automated way of information extraction is an option. And it will still be limited since it's hard to add relevant upstream (u-boot/kernel) information automatically. Even we proceed with versions we have on beta builds it will still be yet another small to medium project. And we are stuck again. 2 x small kids, 2 x full time work, moving to different location earlier this year ...
  24. As far as comments are somehow meaningful it's perfectly fine. Consider this example: $ sudo apt list --upgradable Listing... Done armbian-config/bionic,bionic 5.98.191022 all [upgradable from: 5.98.191016] It would be nice to know what has changed in armbian-config package version 5.98.191022 compared to 5.98.191016, if it's worth to upgrade or just dangerous to me (like kernel/uboot). Commit comments are sufficient, I can just go to github repo to investigate it further. I really know that any hands are welcome, I've been tracking project for more than year and still have no time to participate (kids, kids, kids, work, building a house) :/ But maybe...some day...:P
  25. Automatic? Yes. It is possible to make some automated releasing, but this will just merge commit comments into one file. Without human touch this will not be much bigger step from what it is now. One can see git log and understand ... but also adding upstream features/bug changes? On a moving development target? We never did that very good .... and since we struggle with maintaining without this I don't see to put more effort into doing that. Manual? I can't do it, can't hire and can only ask if you would like to help. If you can't, let's forget about or hope that someone else saw this and consider helping our efforts. On the other project we use Jira for keeping development under control (which we would like to implement exactly with the purpose to be able to extract such information without too much efforts), but at the end, person responsible for releasing, copy and pastes only important information from Jira to the release document. Major, Minor + patch which are not getting any release text. We have to fix versioning on the way ... and hopefully find someone who will be responsible for keeping information fresh at one place. Remember that two out of three most active people are inactive & third had to cut down involvement into the project to half.
  26. Sorry to necro an old thread, but I have managed to get the all the usb ports working on my MXQ Pro 4k s905w . I noticed there was no clock crystal soldered on near the usb hub chip, I happened to have a 12mhz crystal lying around and tried it. Not sure how they were making it work in android, but I have the android dts file if someone smarter than me wants to go through it. Now I'm just looking for an older image so I can have wifi as well. The older links in this thread seem dead.
  27. Anyone know where I can find one of these old images?
  1. Load more activity