Login at Kodi Home

semolina · (This post was last modified: 2015-09-04, 18:57 by semolina.)

I am experiencing reproducible system crashes when playing back audio files on my AMD machine installed using the guide by fritsch ( found here ). While initially the files play back fine (for a few minutes), they will eventually bring down the whole system. Before crashing the sound turns into something that sounds like a spiral of the last played second or so repeating over and over again and getting faster (myabe it could also be described as an echo). The files are perfectly OK. I can play them from any other device connecting to the crashing machine's library without error.

My logs:
X.org log
Kodi log
dmesg output

The machine is connected to a TV via HDMI which is also used for playback of audio.
Ubuntu 14.04 server kept up to date.
Kodi version is 15.1 from the official repo, crashes happened also before updating from 15.0
Crashes seem to be skin independent.

**fritsch** · 2015-09-04, 20:40

crashlog please and also remove next.aired script before

semolina · 2015-09-05, 00:57

Sorry, no crashlogs. I have enabled debugging mode now.

semolina · (This post was last modified: 2015-09-06, 14:04 by semolina.)

What about these lines from dmesg (link posted above):

[ 9.204909] [Hardware Error]: System Fatal error.
[ 9.238334] [Hardware Error]: CPU:0 (16:0:1) MC4_STATUS[Over|UE|-|PCC|AddrV|-|-]: 0xf600000000070f0f
[ 9.271880] [Hardware Error]: MC4_ADDR: 0x00000000c098de00

They coincided with the crash I had yesterday.

That's how the kernel.log shows it:
Sep 4 18:16:51 Bibo kernel: [ 9.204902] MCE: In-kernel MCE decoding enabled.
Sep 4 18:16:51 Bibo kernel: [ 9.204909] [Hardware Error]: System Fatal error.
Sep 4 18:16:51 Bibo kernel: [ 9.238334] [Hardware Error]: CPU:0 (16:0:1) MC4_STATUS[Over|UE|-|PCC|AddrV|-|-]: 0xf600000000070f0f
Sep 4 18:16:51 Bibo kernel: [ 9.271880] [Hardware Error]: MC4_ADDR: 0x00000000c098de00
Sep 4 18:16:51 Bibo kernel: [ 9.273427] lp0: using parport0 (interrupt-driven).
Sep 4 18:16:51 Bibo kernel: [ 9.304471] [Hardware Error]: MC4 Error (node 0): Watchdog timeout due to lack of progress.

Does this possibly point to a bad RAM module?
Answer: maybe not - ran memtest without errors...

semolina · (This post was last modified: 2015-09-06, 14:18 by semolina.)

Still no crash logs, but maybe a hint from the kernel.log:

Today again a crash (at 13.36/37).
dmesg output
in the kernel.log the output that follows directly the last lines of the above says:
Sep 6 13:00:03 Bibo kernel: [ 16.193565] Bluetooth: RFCOMM ver 1.11
Sep 6 13:36:53 Bibo kernel: [ 2226.333095] radeon 0000:00:01.0: ring 0 stalled for more than 10000msec
Sep 6 13:36:53 Bibo kernel: [ 2226.333116] radeon 0000:00:01.0: GPU lockup (waiting for 0x000000000001b676 last fence id 0x000000000001b675 on ring 0)
Sep 6 13:36:54 Bibo kernel: [ 2227.710472] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:36:55 Bibo kernel: [ 2228.721766] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:36:56 Bibo kernel: [ 2229.733063] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:36:57 Bibo kernel: [ 2230.744358] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:36:58 Bibo kernel: [ 2231.755660] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:36:59 Bibo kernel: [ 2232.766961] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:37:00 Bibo kernel: [ 2233.778265] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:37:01 Bibo kernel: [ 2234.789573] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
Sep 6 13:37:02 Bibo kernel: [ 2235.800866] [drm:uvd_v1_0_start] *ERROR* UVD not responding, trying to reset the VCPU!!!
This is when the machine reboots
Sep 6 13:37:39 Bibo kernel: [ 0.000000] Initializing cgroup subsys cpuset

I know it is against the rules pasting snippets of logs, but in this case I think it may be helpful...

Here is the Kodi.log
Of course it starts after the crash, but maybe it contains more info than the last one, since I enabled debugging with verbosity for audio and video.

**fritsch** · 2015-09-06, 14:24

Can you try to add the wsnipex mesa ppa please? And upgrade to that?

Also kernel upgrade might be useful.

semolina · 2015-09-06, 15:50

Just installed mesa from the wsnipex ppa.

What kernel version would you recommend?

semolina · 2015-09-06, 16:01

Now there is a crash log.
http://paste.ubuntu.com/12295263/ Kodi-crashlog

**fritsch** · 2015-09-06, 16:02

Try 4.1.6

**fritsch** · 2015-09-06, 16:02

sudo apt-get install gdb and crash it again, please

semolina · 2015-09-06, 19:21

I have been trying for 3 hours now to let it crash again - without success ;-)
So maybe the new Kernel saved me.
I will report here if crashes reappear, but I hope that I won't have to.
Thanks again!