Fred St-Pierre

  • Content Count

    51
  • Joined

  • Last visited

About Fred St-Pierre

  • Rank
    Advanced Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Yeah, that's what I thought. I'll recompile everything again and reattempt loading the kernel from nvme
  2. Well correct me if I'm wrong, but if it doesn't initialize pcie, it won't boot off the nvme drive?
  3. That could explain why I couldn't boot the kernel a couple of weeks ago. I'll probably pull another attempt later this week, so I don't have to mess with initramfs off SPI.
  4. Power it with POE for now, before I was using a 60W PD power brick during staging
  5. Normal message on first boot. Just enter u-boot without a boot drive, type saveen. If you want, you can use printenv to see your environment variables and modify them to your liking (for example edit boot_targets) and then type saveen. CRC error should dissapear.
  6. I dirty modified the config code to include this... Booting from nvme doesn't seem like it's an issue, it's more the fact the initramfs on the amarula image is part of their SPI image, while the other needs to be added to the armbian image to boot properly. I don't have much time to play around with it lately, so I'll give it some time before I try it again. But the SPI image from Amarula boots the armbian latest compiled image without issues, so I'll keep this for now and play around with it further when I have time. Thanks!
  7. So further trying to get the board to boot, I tried to compile Linux-next, copied all files to the /boot folder, flashed spi.img as recommended and the board is dead in the water again. The u-boot from the amarula release seems to have initramfs built in to the image... So... More attempts required to get the board to boot with self-compiled u-boot and initramfs, I guess.
  8. Ok, so progress: flashed u-boot from this release: https://github.com/amarula/bsp-rockchip/releases/tag/v1.3-roc-rk3399-pc and board boots into Armbian with a freshly compiled Armbian build. HUZZAH! Boots off nvme too... Now I'm kind of afraid to see what will happen if I update the board, as I think this is u-boot 5.5... At least it's a step in the right direction.
  9. Yeah, just flashed SD card and tested, seems like my build does not work... It doesn't even fetch extlinux.conf to begin boot. More digging to figure it all out. Baby steps and progress. Learning how to compile slowly.
  10. This board is doing its best to drive me totally bonkers. Still can't boot anything... It sees NVME, it sees my USB, able to see all files are there, but it just doesn't fire up anything. More digging required, I guess. Used Armbian compiler for most, compiled initramfs with linux-next instructions, followed most recommendations here. At least I figured out that having the board plugged into my laptop to power it, if I type ums 0 nvme 0 from u-boot, I can access the nvme drive from my Ubuntu VM and fully manipulate it from there. Learning experience
  11. Yep, I had a previous bl31.elf built, that was not the issue. I had to enable CONFIG_DM_KEYBOARD=y in my defconfig... Keyboard now works. Thanks for the help! Next step, boot from nvme.
  12. Thanks for this, I just tried it yesterday and I actually get a post from SPI. It's a step in the right direction! I made a build environment in a Vmware player VM, which lets me manipulate the board as well. Compiling Armbian right now. My issue is that I still have no keyboard input. I only have Unifying wireless keyboards, both do not work. I checked the config file and CONFIG_USB_KEYBOARD is set. I'm at a loss why it would not work. In the meantime, I'll flash to a USB stick to test and then migrate that to the nvme drive...
  13. The last pre-compiled u-boot I tried worked from SPI, but I couldn't interact with it. It didn't accept any keyboard input.
  14. The project was closed off by the support team. They didn't complete sending of hardware (I am out 100$ of gear which they are ignoring me on) and the last update on their IGG is they completed upstreaming all code and have effectively ended the project. To be honest, I never got u-boot to work on this board other than their horrible pre-compiled images which are just used to flash to SD, never got the nvme to work to boot the device. So effectively, this thing is a huge, expensive paperweight for me and just gathering dust in a drawer. This board had so much potential, but Libre
  15. https://github.com/amarula/bsp-rockchip/releases Finally. PCIE support mainlined in u-boot, 4k support, USB fixes, etc etc etc. This board will now be usable the way I intended to use it. Time to play around with NVMe boot and SPI flash.