Openelec Build for OPI PC and 2 now with HW decoding

slaine · Published in 2016-3-15 17:29:06

jernej replied at 2016-3-15 04:01
I think I found the problem with stuttering. Please, test this OPiPC image and report: https://mega. ...

I think the situation is better than with image 2016/03/09. But unfortunately video is not jitter free. I think that h.265 video is smoother than h.264 where the jitter appear cca every 10 sec. I also think, that deinterlacing in PVR does not work very well.

cbm80 · Published in 2016-3-16 00:17:19

Regarding booting (or not booting) from 2GB card I'm attaching log from serial port.

U-Boot SPL 2016.03-rc2 (Mar 14 2016 - 20:40:26)
DRAM: 1024 MiB
Trying to boot from MMC

U-Boot 2016.03-rc2 (Mar 14 2016 - 20:40:26 +0100) Allwinner Technology

CPU: Allwinner H3 (SUN8I)

I2C: ready

DRAM:  1 GiB

MMC: SUNXI SD/MMC: 0

*** Warning - bad CRC, using default environment

In: serial

Out: serial

Err: serial

Net: No ethernet found.

starting USB...

No controllers found

Hit any key to stop autoboot:  2 ... 1 ... 0

switch to partitions #0, OK

mmc0 is current device

Scanning mmc 0:1...

Found U-Boot script /boot.scr

reading /boot.scr

329 bytes read in 25 ms (12.7 KiB/s)

## Executing script at 43100000

gpio: pin PL10 (gpio 298) value is 1

gpio: pin PG11 (gpio 203) value is 1

reading script.bin

36236 bytes read in 211 ms (167 KiB/s)

reading KERNEL

6133520 bytes read in 28084 ms (212.9 KiB/s)

## Booting kernel from Legacy Image at 42000000 ...

Image Name: Linux-3.4.110

Image Type: ARM Linux Kernel Image (uncompressed)

Data Size: 6133456 Bytes = 5.8 MiB

Load Address: 40008000

Entry Point:  40008000

Verifying Checksum ... OK

Loading Kernel Image ... OK

Using machid 0x1029 from environment

Starting kernel ...

[sun8i_fixup]: From boot, get meminfo:

.Start:.0x40000000

.Size:.1024MB

ion_carveout reserve: 160m@0 256m@0 130m@1 200m@1

ion_reserve_common: ion reserve: [0x70000000, 0x80000000]!

jernej · Published in 2016-3-16 02:59:58

cbm80 replied at 2016-3-15 17:17
Regarding booting (or not booting) from 2GB card I'm attaching log from serial port.

This seems to be some error in the kernel. Such problem ususally occurs when there is some problem with early driver. Probably earlyprintk or something similar would reveal more.

jernej · Published in 2016-3-16 03:00:58

slaine replied at 2016-3-15 10:29
I think the situation is better than with image 2016/03/09. But unfortunately video is not jitter ...

Are you saying that H264 has a problem and H265 not? I had different experience...

thc013 · Published in 2016-3-16 05:26:26

running it for 2 days now and different codecs etc tvheadend and no framedrops onlly when you you do something in kodi like enable subs or put up the codec rate overlay you see a framedrop or 2

fast and funny find of the problem

jernej · Published in 2016-3-16 05:38:03

thc013 replied at 2016-3-15 22:26
running it for 2 days now and different codecs etc tvheadend and no framedrops onlly when you you do ...

Thanks for the report.

Very frustrating thing.

slaine · Published in 2016-3-16 15:30:12

jernej replied at 2016-3-16 03:00
Are you saying that H264 has a problem and H265 not? I had different experience...

No both codec has problem, but it seems that h.265 is more fluent.

pmos69 · Published in 2016-3-16 19:52:52

jernej replied at 2016-3-11 23:44
ok, what I found out currently:
git commit 92926ea2c6b44b6c7e19b34c3c726599648239ae
- RSS feeds work ...

Jernej,

Were there any commits after 92926ea that changed CPU settings?
I checked out and built 92926ea and, when idle, the cpu temp rises very, very slowly up to 65º C and stays around there.
(cat/sys/devices/virtual/thermal/thermal_zone1/temp)
I'm using a cpu heatsink. If I use a mini fan the temps keep a lot lower, but I hate the noise.

Latter builds all, again when idle, make the CPU temp rise very fast to 70ºC and cores start throtlingdown/shutting down fast, as seen in dmesg.

jernej · Published in 2016-3-17 01:11:42

slaine replied at 2016-3-16 08:30
No both codec has problem, but it seems that h.265 is more fluent.

Do you have any sample video (e.g. few megs) to demonstrate this? After more testing, I can say that it works for me.

jernej · Published in 2016-3-17 01:17:27

pmos69 replied at 2016-3-16 12:52
Jernej,

Were there any commits after 92926ea that changed CPU settings?

Only one commit could affect that: https://github.com/jernejsk/Open ... 04c9c853ffc9e255eee

Care to check and test which setting caused this?