Hello,

Here is a DMESG kernel dump that I'm getting everytime I boot the Odroid HC2 with a 5.4.x kernel:

 

Also random crashes related to memory.

Spoiler

[ +15.195670] 8<--- cut here ---
[  +0.001580] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007017] pgd = 632a7f01
[  +0.002682] [00000000] *pgd=00000000
[  +0.027685] 8<--- cut here ---
[  +0.001585] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007017] pgd = 632a7f01
[  +0.002681] [00000000] *pgd=00000000
[  +0.026593] 8<--- cut here ---
[  +0.001589] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007024] pgd = 448f724b
[  +0.002677] [00000000] *pgd=00000000
[  +0.028227] 8<--- cut here ---
[  +0.002157] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007017] pgd = 05c074b5
[  +0.002680] [00000000] *pgd=00000000
[  +0.027714] 8<--- cut here ---
[  +0.001591] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007021] pgd = 7fb9cff9
[  +0.002676] [00000000] *pgd=00000000
[  +0.022340] 8<--- cut here ---
[  +0.001586] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007020] pgd = 4ece155c
[  +0.002682] [00000000] *pgd=00000000
[  +0.029920] 8<--- cut here ---
[  +0.001590] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007016] pgd = 4ece155c
[  +0.002681] [00000000] *pgd=00000000
[  +0.024843] 8<--- cut here ---
[  +0.001588] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007021] pgd = 4ece155c
[  +0.002677] [00000000] *pgd=00000000
[  +0.038428] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = 12fe33c6
[  +0.002680] [00000000] *pgd=00000000
[  +0.080285] 8<--- cut here ---
[  +0.001593] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007015] pgd = 196c4997
[  +0.002685] [00000000] *pgd=00000000
[  +0.192258] 8<--- cut here ---
[  +0.001754] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007015] pgd = 4ece155c
[  +0.002682] [00000000] *pgd=00000000
[  +0.025531] 8<--- cut here ---
[  +0.001582] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = d4a0de92
[  +0.002683] [00000000] *pgd=00000000
[  +0.024739] 8<--- cut here ---
[  +0.001590] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007013] pgd = daa5c755
[  +0.002680] [00000000] *pgd=00000000
[  +0.020426] 8<--- cut here ---
[  +0.001588] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007012] pgd = daa5c755
[  +0.002682] [00000000] *pgd=00000000
[  +0.023073] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007018] pgd = d4a0de92
[  +0.002674] [00000000] *pgd=00000000
[  +0.024059] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = d4a0de92
[  +0.002680] [00000000] *pgd=00000000
[  +0.022939] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007015] pgd = 617cfa5c
[  +0.002681] [00000000] *pgd=00000000
[  +0.023886] 8<--- cut here ---
[  +0.001590] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007019] pgd = 33b29230
[  +0.002678] [00000000] *pgd=00000000
[  +0.024048] 8<--- cut here ---
[  +0.001591] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007021] pgd = d4a0de92
[  +0.002678] [00000000] *pgd=00000000
[  +0.068901] 8<--- cut here ---
[  +0.001592] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = cbe8dee6
[  +0.002681] [00000000] *pgd=00000000
[  +0.246169] 8<--- cut here ---
[  +0.001593] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007013] pgd = eba26a5a
[  +0.002681] [00000000] *pgd=00000000
[  +0.024084] 8<--- cut here ---
[  +0.001586] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007013] pgd = cbe8dee6
[  +0.002682] [00000000] *pgd=00000000
[  +0.021854] 8<--- cut here ---
[  +0.001586] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = cbe8dee6
[  +0.002682] [00000000] *pgd=00000000
[  +0.027139] 8<--- cut here ---
[  +0.001592] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007010] pgd = cbe8dee6
[  +0.002681] [00000000] *pgd=00000000
[  +0.031291] 8<--- cut here ---
[  +0.001600] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007002] pgd = 1513e724
[  +0.002681] [00000000] *pgd=00000000
[  +0.027216] 8<--- cut here ---
[  +0.001589] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = 1513e724
[  +0.002680] [00000000] *pgd=00000000
[  +0.024718] 8<--- cut here ---
[  +0.001590] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007007] pgd = 0b9959dc
[  +0.002682] [00000000] *pgd=00000000
[  +0.029158] 8<--- cut here ---
[  +0.001605] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007009] pgd = 1513e724
[  +0.002679] [00000000] *pgd=00000000
[  +0.023788] 8<--- cut here ---
[  +0.001584] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007016] pgd = 0b9959dc
[  +0.002679] [00000000] *pgd=00000000
[  +0.079814] 8<--- cut here ---
[  +0.001853] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007013] pgd = 34b00026
[  +0.002680] [00000000] *pgd=00000000
[  +0.562546] 8<--- cut here ---
[  +0.001606] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007007] pgd = a3ed1d58
[  +0.002681] [00000000] *pgd=00000000
[  +0.031487] 8<--- cut here ---
[  +0.001598] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007010] pgd = a3ed1d58
[  +0.002680] [00000000] *pgd=00000000
[  +0.027269] 8<--- cut here ---
[  +0.001599] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007015] pgd = a3ed1d58
[  +0.002680] [00000000] *pgd=00000000
[  +0.022020] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007015] pgd = 8c57ccf9
[  +0.002682] [00000000] *pgd=00000000
[  +0.023431] 8<--- cut here ---
[  +0.001593] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007018] pgd = 8c57ccf9
[  +0.002680] [00000000] *pgd=00000000
[  +0.028660] 8<--- cut here ---
[  +0.001594] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007012] pgd = a3ed1d58
[  +0.002679] [00000000] *pgd=00000000
[  +0.023420] 8<--- cut here ---
[  +0.001590] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007014] pgd = 161fb286
[  +0.002681] [00000000] *pgd=00000000
[  +0.023969] 8<--- cut here ---
[  +0.001595] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007007] pgd = 8cacda11
[  +0.002680] [00000000] *pgd=00000000
[  +0.027205] 8<--- cut here ---
[  +0.001591] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007016] pgd = 8cacda11
[  +0.002680] [00000000] *pgd=00000000
[  +0.070703] 8<--- cut here ---
[  +0.001598] Unhandled fault: imprecise external abort (0x1406) at 0x00000000
[  +0.007012] pgd = 8cacda11
[  +0.002681] [00000000] *pgd=00000000

 

 

 

Edited April 15, 20206 yr by Igor
use spoilers when adding wall of text

7 hours ago, Belphemur said:

Also random crashes related to memory.

Yes, our tests also show its unstable, use K4.14.y We are desperately trying to fix those problems, but day only has 24h and we can't afford to blow much of our tiny resources on this.

@Igor I'm aware it's not stable, I don't mind testing it.

The problem is, with 4.14.y, I can't run Emby. It keeps crashing for segmentation fault. While on 5.4.y, it runs flawlessly (but the hardware transcoding doesn't work).

 

Is there anything that can be done to help fix this ? I've never worked on kernel before, but I don't mind using some of my free time on this. I enjoy the capabilities of my odroid HC2 and its power, it's a shame that I can't use it fully. I want to help.

5 hours ago, Belphemur said:

I've never worked on kernel before

In that case, there is nothing you can do about.

But we have many tasks where you can add your time and we have it more for things like this. Two days ago I spent a whole day (!), trying to resolve 5.6.y on Odroid XU4. Since I have many other things to do, I also have a family that is far more important than any 3rd party problem, I have no idea when I will be able to afford trying to fix this next time. Perhaps somebody else will, but I don't know that nor anyone can guarantee you. Its a lot of work and its very expensive, when all costs are on us.

@Igor

 

I have a stable 5.4.32 running.
Currently, armbian/build is using the repository of memeka with upstream patches.

 

Hardkernel has a up-to-date branch with 5.4.y with their patches: https://github.com/hardkernel/linux/tree/odroid-5.4.y

 

I can open a PR for odroid-xu4/HC1/HC2 on the armbian/build repository, it's been running for 24h without issues. I'll continue monitor it.

Edited April 20, 20206 yr by Belphemur

3 minutes ago, Belphemur said:

I can open a PR for odroid-xu4/HC1/HC2 on the armbian/build repository, it's been running for 24h without issues. I'll continue monitor it.

OK, make a PR - it can't be worse than what we have now, Memeka build and our desperate attemts to make it stable failed ...

 

I will also join the testings with my devices ... 

@Igor just to keep you in the loop:

 

Still running without any crash, the HardKernel repo is stable for me.
Currently, uptime of 3 days without any memory corruption.

Have you been able to test on your side ?

Edited April 23, 20206 yr by Belphemur

37 minutes ago, Belphemur said:

Have you been able to test on your side ?

 

Thanks.

I was only doing stress tests, nothing long term. All passed so far ... I think @TonyMac32 is crunching numbers with one XU4. Perhaps he already changed to this kernel?

I have not changed yet, I will though, I never got the other kernel to work reliably for more than a day during number crunching.

Sent from my Pixel using Tapatalk

Since I own the HC2, I'm using it as a NAS + Emby on it.

I've a load average: 1.28, 1.47, 1.60 (currently).

 

Also

              total        used        free      shared  buff/cache   available
Mem:          1.9Gi       1.1Gi       102Mi        21Mi       775Mi       824Mi
Swap:         996Mi       715Mi       281Mi

It's running pretty smoothly.

By the way, the swap is the default zram disk with LZO-RLE.

My XU4 with the Armbian kernel is fine with Plex and file server duties. Once I put it on the MC1's and started Einstein@home, it was all over. I'll have to give this a shot.

Sent from my Pixel using Tapatalk