+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: General Support (https://forum.kodi.tv/forumdisplay.php?fid=111)
+---- Forum: Linux (https://forum.kodi.tv/forumdisplay.php?fid=52)
+---- Thread: Linux VAAPI: Nuc, Chromebox, HSW, IVB, Baytrail with Ubuntu 14.04 (/showthread.php?tid=165707)
Pages:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
RE: vaapi-sse4: Deinterlacing Testing - adr3nal1n - 2014-11-06
Hi fritsch,
cat /proc/cpuinfo as follows:
Code:
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 60
model name : Intel(R) Pentium(R) CPU G3240 @ 3.10GHz
stepping : 3
microcode : 0x1c
cpu MHz : 884.710
cache size : 3072 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm pcid sse4_1 sse4_2 movbe popcnt tsc_deadline_timer xsave rdrand lahf_lm abm arat epb pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust erms invpcid xsaveopt
bugs :
bogomips : 6188.32
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:
processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 60
model name : Intel(R) Pentium(R) CPU G3240 @ 3.10GHz
stepping : 3
microcode : 0x1c
cpu MHz : 801.156
cache size : 3072 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
apicid : 2
initial apicid : 2
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm pcid sse4_1 sse4_2 movbe popcnt tsc_deadline_timer xsave rdrand lahf_lm abm arat epb pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust erms invpcid xsaveopt
bugs :
bogomips : 6188.32
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:Hope this information helps.
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
Can you build this kernel: https://github.com/fritsch/linux/tree/gpuhang gpuhang branch? And then test without any kernel params? Make sure to really build the gpuhang branch
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
Does not help: https://bugs.freedesktop.org/show_bug.cgi?id=83677#c79
RE: vaapi-sse4: Deinterlacing Testing - menno - 2014-11-06
My gpu hang on the latest BETA
http://sprunge.us/QBKV
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
Yeah: https://bugs.freedesktop.org/show_bug.cgi?id=83677
We can do absolutely nothing about it, besides the workaround mentioned some posts back.
I just ordered a cubibox. I think I will stop caring completely in the future.
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
(2014-11-06, 14:06)menno Wrote: My gpu hang on the latest BETA
http://sprunge.us/QBKV
There is not a single hang in there.
RE: vaapi-sse4: Deinterlacing Testing - menno - 2014-11-06
Well, that is particular, cause it freezed the whole time wile I generated it. The strange thing is of course that everything else continues to work, audio, ssh etc. I am not sure what log should display the gpu hang...
// edit
here you go
http://sprunge.us/bbUS
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
Yeah - whatever it is, you have another problem as currently discussed. here. Do this: http://forum.kodi.tv/showthread.php?tid=199498&pid=1827074#pid1827074
I am running my system the whole morning and I currently think it's a specific OpenELEC issue introduced by some random bump of random driver. The above will workaround your issue though.
RE: vaapi-sse4: Deinterlacing Testing - menno - 2014-11-06
I see, but your workaround defeats the whole purpose since I went helix to use these hw de-interlacing methods
Thanks, are you interested in my
/sys/class/drm/card0/error
?
Thanks so far!
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
No - I am not interested, cause I _cannot_ do a single thing about it. I reported that bug to the intel folks more than 3 months ago and for now they are not able to fix it. See the bug I linked it was even set to "Highest Blocker" Level: https://bugs.freedesktop.org/show_bug.cgi?id=83677
RE: vaapi-sse4: Deinterlacing Testing - adr3nal1n - 2014-11-06
Hi fritsch,
Sorry for the delayed response,
Been a very long time since I rolled my own kernel, not sure how best to do it on Arch, I'll have to so some reading up. The last time I did this I was running Xebian with kernel 2.4 on a modded original xbox (you may laugh, but it worked quite well as a NAS).
Just reading through the latest posts in this thread and the comments over at the freedesktop bug list. Are you saying this bug is limited only to Celeron and Pentium Haswell chips and does not affect the Core i3, i5 and i7 Haswell chips?
Thanks again for all your help.
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-06
(2014-11-06, 15:34)adr3nal1n Wrote: Hi fritsch,
Sorry for the delayed response,
Been a very long time since I rolled my own kernel, not sure how best to do it on Arch, I'll have to so some reading up. The last time I did this I was running Xebian with kernel 2.4 on a modded original xbox (you may laugh, but it worked quite well as a NAS).
Just reading through the latest posts in this thread and the comments over at the freedesktop bug list. Are you saying this bug is limited only to Celeron and Pentium Haswell chips and does not affect the Core i3, i5 and i7 Haswell chips?
Thanks again for all your help.
It seems to show a whole lot more frequently on Celeron, Pentium HSW gens. My core i3 in the living room did not have a gpu hang for weeks.
Building the kernel is not needed - it does not solve the issue.
RE: vaapi-sse4: Deinterlacing Testing - adr3nal1n - 2014-11-06
Hi fritsch,
Apologies, I have just realised last night's test results using the kernel options suggested are invalid as I had this additional config in a file within /etc/X11/xorg.conf.d/
Code:
Section "Device"
Identifier "Intel Graphics"
Driver "intel"
Option "NoAccel" "True"
EndSectionHave removed the above conf, rebooted and am now re-running the testing with the kernel options you provided.
Will post again with results.
RE: vaapi-sse4: Deinterlacing Testing - adr3nal1n - 2014-11-06
Hi fritsch,
Finished testing again, 1 and a 1/4 hour of a HD movie and then 30 mins of a TV show and no GPU hangs or frame drops with vaapi enabled.
I ran xbmc with debug logging. I have put the log here in case you are interested (you will see a lot of LIRC key presses in there, that was me checking every now and again for frame drops). http://pastebin.com/uZjcXTgB
There were no new entries in dmesg during the whole of my testing.
I did notice this in the systemd journal but I didn't notice any issues on screen and there are no other entries or errors relating to this in the journal around that time. Just thought I'd mention it.
Code:
Nov 06 13:48:58 wolverine systemd-coredump[477]: Process 360 (xbmc.bin) of user 1000 dumped core.Hope this feedback helps in some way.
Hi fritsch,
Actually, have just found the core dump file as mentioned above,
See here http://pastebin.com/injmiLrr
Not sure if it helps but thought I'd include it for completeness.
RE: vaapi-sse4: Deinterlacing Testing - fritsch - 2014-11-08
Please: https://bugs.freedesktop.org/show_bug.cgi?id=83677#c90
Everyone with the hang copy the error file to your hard disk, bzip2 it and post dmesg + the error file to the above bugreport. It seems intel does not realize the impact this bug has at all.
Thanks much. I won't offer support for any intel hang bugs anymore at all until this upstream issue is fixed.
For that long I will closed this Thread out of protest