[pine64] boot issue (cpu stuck)

[kometchtech]

Hello, all

When pine64+ the you try to start in Armbian 5.20, you can not start with an error such as the following.

Starting kernel ...

[mmc]: MMC Device 2 not found
[mmc]: mmc 2 not find, so not exit
INFO:    BL3-1: Next image address = 0x41080000
INFO:    BL3-1: Next image spsr = 0x3c9
Loading, please wait...
starting version 229
Begin: Loading essential drivers ... done.
Begin: Running /scripts/init-premount ... done.
Begin: Mounting root file system ... Begin: Running /scripts/local-top ... done.
Begin: Running /scripts/local-premount ... [   36.484607] BUG: soft lockup - CPU#0 stuck for 22s]
[   36.492947] Call trace:
[   36.495655] Call trace:
[   36.498351] Call trace:
[   36.501060] Call trace:
<snip>

Thanks,

[zador-blood-stained]

Legacy or vanilla kernel? Image from download page or self built?

[kometchtech]

Hello, @zador-blood-stained

This problem is, what you have downloaded the legacy image (xenial) from the download page, is what was dd to the SD card.

Thanks,

[martinayotte]

A bit strange that u-boot try to find mmc2. (On my PinA64 5.17 self-build, the SDCard is on mmc0)

[zador-blood-stained]

A bit strange that u-boot try to find mmc2. (On my PinA64 5.17 self-build, the SDCard is on mmc0)

Looks loke strangely implemented cleanup procedure: https://github.com/longsleep/u-boot-pine64/blob/55c9c8c8ac005b1c00ac948386c60c4a741ebaa9/drivers/mmc/mmc.c#L2852

Anyway I can reproduce this issue, so I'll try to check if it is caused by Armbian patches or DT changes.

[zador-blood-stained]

After updating pine64-default kernel configuration I can't reproduce this issue anymore. Needs more testing to say for sure if it is fixed or not.

Edit: This is still an issue, and for now it is uncertain what causes it.

[zador-blood-stained]

Same errors/bugs as thread starter. Tried ALL legacy and mainline which everyone of them ends up with Segmentation Faults. PINE64+.. Desktop versions are out of the question here as every images i have tried on Armbian and the official site (with android) ends up with either bootloops or segmentation errors.

I suspect you have HW issues (or HW tolerance issues) since for other people these and other official images work fine. We can't help with those, your best option is contacting the seller for a replacement or a refund.

[zador-blood-stained]

I tried contact support but they said it could be power issue though i have checked that it's not the case..

How did you check this? The only reliable way is powering a board via GPIO pins with something like a lab power supply or ATX power supply.

The one thing I know - people tested different Armbian images and those images worked for them, but my board, for example, can't boot from SD anymore, and before that I had similar "Call trace" issues.

[zador-blood-stained]

By using a stepup converter, 3.7v to 5v https://www.adafruit.com/product/2465 with 18650 Li-Po batteries.

Connected to the GPIO or to the microUSB port?

[zador-blood-stained]

The original issue (soft lockup) was reproduced on longsleep's images, so this can be closed and upstream issue should be used for tracking the status of this bug: https://github.com/longsleep/build-pine64-image/issues/51

[pi-r-p]

(just a comment on this closed issue, because it is one of the first google result of pine64 cpu stuck) I also had this random boot error, really annoying for a headless system. For people looking for a quick and dirty solution : Recompile your kernel with two options set in Kernel hacking section of make menuconfig :

• Panic (Reboot) On Hard Lockups
• Panic (Reboot) On Soft Lockups

It won't fix the CPU stuck error, but it reboots a few seconds later. Not a fix, just a workaround.

[igorpecovnik]

We are getting a modern 4.19.y kernel: https://forum.armbian.com/topic/8795-next-lts-kernel-419y-allwinner-a10-a20-a64-h2-h3-h5-h6-debugging-party/ ... Don't have Pine64 but many other A64 based boards act much better.