If Ubuntu is updating the kernel, or boot scripts then you will most likely see a failure. This image is meant to keep the default BSP kernel and boot scripts in the image that is originally flashed.
Hello team,
AS i wanted RT kernel v5.4 and its available with LSDK20.12 i build that source with DDR changes as suggested earlier and again i started to get same issue.
https://github.com/SolidRun/arm-trusted-firmware/commit/8c756b166b2132290af53d9a9e9a5f307828b84e.patch
any other patch specific to LSDK 20.12
It would be great if the default image is ready so that i can download and verify parallel
You should look at your arm-trusted-firmware repository and make sure that the patch was applied. I don’t believe it will work by default with older versions of the source repo. Again an image with the patch applied won’t be available any time soon. We have a list of validated memory modules which will need to be re-validated with any changes.
Hello team,
As the DDR changes was not working with LSDK20.12 + LSDK20.12-v.5.4-RT , I switched to the default LSDK 21.08 with kernel 5.10.35 present in the runme.sh with patch.
So - the issue is kernel boot stops in emmc boot of the kernel
“can’t access tty: job control turned off”
How can i get rid of this so that i can proceed for proper boot i can proceed ,even with lan connected i cant assign the IP.
It would great we can have debug session to sort out or steps. so my basic kernel boots fine and get login prompt.
regards,
ranga
please pastebin your full console log and link to it here. It looks to me like linux is using the wrong device for the root filesystem. Have you followed the instructions and switched the boot jumpers to boot from the eMMC, and also removed the installer SDHC card?
1 Like
Hello team,
I removed the SD card.
changed the switch to OFF,ON,ON,OFF, No change
observed same - log attached
log file
let me know which bin is required -Lx2160acex7_2000_700_3200_8_5_2-a393e2e.img file .
kindly guide the same to proceed.
regards,
ranga
your log file is blocked.
it looks like your original image is corrupt. Can you check the contents of /bin and see if there are executables inside it? Most likely when you installed your new software it has removed or replaced part of the init process.
and /sbin/init is a link to systemd?
/sbin/init -> /lib/systemd/systemd
does /lib/systemd/system exist? It looks to me like when you updated your image Ubuntu removed all the init systems.
How to recover the same any steps .
yes the /lib/systemd/system is present
As i understand its not the issue with build code image (lx2160acex7_2000_700_3200_8_5_2-a393e2e.img ). when i updated on board it was corrupted.
regards,
ranga
That is really a question for Ubuntu. We do not control their software updates or repositories. Apt does have the ability to show you the changes it is going to make before it makes any changes to the system.
Hello team,
I have updated the lx2160acex7_2000_700_3200_8_5_2-a393e2e.img in SD card.
then once the board is booted and break in u boot.
=>load mmc 0:1 0xa4000000 ubuntu-core.img
=>mmc dev 1
=>mmc write 0xa4000000 0 0xd2000
=>boot
once booted i am seeing this. So i am blocked i also checked in other board also same issue is observed.
let me know how i can proceed.
regards,
ranga
It seems as if your self built image did not complete correctly. Could you check the logs of the build process and see if there are any errors present? I have just tested on our prebuilt images and they boot from eMMC without any issues.
Hello team ,
Is there any way - so that i take the prebuild image from solid run forum lx2160acex7_2000_700_3200_8_5_2-a393e2e.img and push my built aft uboot with fix on this img with patch and flash on SD and boot
if any possibility let me know so that i will get correct ubuntu package.
regards,
ranga
I had a few corrupted builds, and found out that they were caused by disk-full errors.
-I also found that if I used an APT-proxy (apt-cacher), then something went wrong, so try disabling that if you have it enabled.
I recommend that you check your free space both in …
/var/lib/docker
… and in your preferred build location.
On my system, I have 32GB free space on the build-volume and 17GB space free on the docker volume.
(I think the docker image only needs a few GB, but I know that 17GB seem to work).
Also, I’d recommend removing the ‘images’ directory before you build; it shouldn’t be necessary to remove other directories.
Personally, I’m using u-boot and initially wrote the root-image to eMMC, but I did succeed in writing it to SATA 0 partition 10 and boot it there (requires some modification and messing with PARTUUID).
I think you could likely attach a USB-drive (stick or harddisk), so you could make modifications on another Linux-box directly while testing (I didn’t try this myself yet).
BTW: I am using a Kingston 64GB SDXC card in the MicroSD slot and can boot from a second partition written to that card, but I don’t feel any speed-difference between SDHC and SDXC (yet).