• 1
  • 66
  • 67
  • 68(current)
  • 69
  • 70
  • 189
How to Install XBMC PVR Xvba for AMD/Nvidia/Intel GPUs
(2012-08-06, 08:56)fritsch Wrote: @freakwenc:
I updated the howto slightly, as the legacy driver is stable now. You don't have to copy back the signature file. Be aware you have to start with the first Box and the continue with the seconds box after Step4. Also the Codebox is scrollable.

You can also join #irc and ping wsnipex or me.

Edit: Can you tell us a bit about your soundsetup? Receiver connected? HDMI, Optical, Analog? I think your sound is configured wrong.

Code:
16:59:44 T:3075537776    INFO: CAESinkALSA - ALSA: pcm_hw.c:587:(snd_pcm_hw_start) SNDRV_PCM_IOCTL_START failed (-32): Broken pipe

My device is odd but it works! I have the 5.1 PCI sound card using analog to outputting 3x 3.5mm speaker jacks to a cheap pc 5.1 speaker setup. as I say it's odd besauce i need to set my audio to output through optical to get the sound through my speakers otherwise it fails to play back any sound. Because I have got a work around I've never thought about it. This is also true with the standard/master xbmc so not a bug in this branch it is possibly a bug with the sound driver side of it.

I'm trying the updated guide now and will let you know if it helps.
@fritsch BIOS reports GPU at around 34C. XBMC reports 49C. I have my CPU (a8) fan at level 5. I can bump it up to see if that helps. No issues since you helped me out Saturday, even over Samba. I was able to reproduce video stutter on the local file by changing audio settings during video playback. Changing them back and rebooting got rid of the stutter. I haven't really had a chance to do much with it since then though.
(2012-08-06, 16:38)fritsch Wrote: ...
@Pyromaniac:
Do you have some rough numbers on "when it is too hot"?

I'm not sure anymore, but I think it was something round 60-65°C when it started to stutter and showing artifacts.

I don't know if I've mentioned it here in the thread, the MoBo I am using is: ASrock E350M1/USB3
Maybe this Info is somehow useful.
Different sensors module give different readouts.
This is on my E350M1 (older, USB2 model) after running "stress -c2" for a couple of minutes:

Code:
xbmc@xbmc:~$ stress -c 2
stress: info: [30767] dispatching hogs: 2 cpu, 0 io, 0 vm, 0 hdd
^C
xbmc@xbmc:~$ sensors
k10temp-pci-00c3
Adapter: PCI adapter
temp1:        +71.2°C  (high = +70.0°C)
                       (crit = +75.0°C, hyst = +72.0°C)

nct6775-isa-0290
Adapter: ISA adapter
Vcore:        +0.70 V  (min =  +0.00 V, max =  +1.74 V)
in1:          +0.89 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
AVCC:         +3.38 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
+3.3V:        +3.36 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in4:          +1.39 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in5:          +1.86 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in6:          +1.69 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
3VSB:         +3.44 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
Vbat:         +3.39 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
fan1:           0 RPM  (min =    0 RPM, div = 128)
fan2:        4687 RPM  (min =    0 RPM, div = 8)  ALARM
fan3:           0 RPM  (min =    0 RPM, div = 64)
fan4:           0 RPM  (div = 64)
SYSTIN:       +49.0°C  (high =  +0.0°C, hyst =  +0.0°C)  ALARM  sensor = thermistor
CPUTIN:       +52.5°C  (high = +80.0°C, hyst = +75.0°C)  sensor = thermistor
AUXTIN:       -11.0°C  (high = +80.0°C, hyst = +75.0°C)  sensor = thermistor
cpu0_vid:    +0.000 V

xbmc@xbmc:~$ DISPLAY=:0 /usr/bin/aticonfig --od-gettemperature

Default Adapter - AMD Radeon HD 6300 series Graphics
                  Sensor 0: Temperature - 62.00 C

BIOS shows something like nct6775-isa-0290.

Had a similar problem with wrong temp on my E350. Somehow k10temp module gave temps 10-20 °C more than w83627ehf (nct6775). Check with sensors-detect if it can find some other module to use for more accurate readout.
@v8rumble

Hi i wrote a full tutorial base on xbmc xvba and this forum, i also use a desktop version because i wanted to be able to use some features like integrating a web browser inside XBMC which is little bit harder without a desktop.

So the answer is yes, absolutely

http://youresuchageek.blogspot.fr/2012/0...linux.html
Please support my Blog, XBMC and Linux related :-)
http://youresuchageek.blogspot.com
@fritsch
Hi, you said it was not necessary to copy anymore /etc/ati/control, AMD updated the driver ? Could you confirm this ?
Please support my Blog, XBMC and Linux related :-)
http://youresuchageek.blogspot.com
(2012-08-05, 08:30)fritsch Wrote: @freakwenc:
I don't really understand what your usecase is. You are writing about an "old pc". Can you explain it more in detail please? Does tvheadend stop working?
@Balthazar:
As we talked in IRC yesterday. Did the local files, you copied directly on the htpc work? I suspect a samba problem.

(2012-08-06, 16:33)Pyromaniac Wrote:
(2012-08-04, 22:07)Balthazar Wrote: ...It runs ok for a while, but then I start to get really jerky video. Does anyone have any ideas?...

I've had a similar problem, the reason was: the cpu/gpu was too hot.

Many thanks to @fritsch for helping me out over on the IRC channel Saturday. After reinstalling everything I have nice smooth 1080p playback using the settings provided in the first post. I do think I have a couple of problems. One, local files play really nice, no stuttering or jerkiness. Files on my server streaming over wired ethernet were jerky on Saturday, but I haven't had any issues since then. I reinstalled the TV Guide add-on to let me play xmltv streams from my HDhomerun, but those are still terribly jerky. If I browse to the xmltv stream file directly through videos the stream plays fine. I think I started messing things up by opening settings during the stream playback and "tweaking" settings and just got my settings all f-ed up. Then none of my other files would playback right.

For right now I am sticking with things the way they are. I'm going out of town for a week and my wife will kick my butt if I take off and leave the HTPC all jacked up. I am concurrently trying to get the HDhomerun to work with TVheadend, but that is a whole different story (right @pyromaniac!). I am looking forward to going through @guilimxm's blog, especially changing the boot splash, but for right now I am sorted. I have been keeping an eye on temperatures, but they seem pretty stable.

BTW - @guilmxm: @fritsch is on holiday right now, but I am pretty sure you're right about copying the /etc/ati/control. I think all the other devs are on holiday too, but maybe they could confirm that when they get back.
@fritsch

I've tried the "updated/fixed" tutorial and still got hardware not found so I experimented and copied the control file back and it found the graphics card but when I complete the tutorial (legacy one) x server fails to work. i get
Code:
FEH.PY: cannot connect to X server
This is also like when i upgrade to ubuntu 12.04 LTS i've looked into that and this in the Xorg.0.log and both come up with Illegal instruction at address #x########




(2012-08-06, 23:07)guilmxm Wrote: @fritsch
Hi, you said it was not necessary to copy anymore /etc/ati/control, AMD updated the driver ? Could you confirm this ?

only for the legacy driver. The standard driver still needs that
@freakwenc last try:

Code:
sudo service lightdm stop
sudo apt-get remove --purge fglrx* nvidia*
now follow the driver howto http://forum.xbmc.org/showthread.php?tid...pid1138272 for the legacy driver and reboot.

if it still does not work, run http://dl.dropbox.com/u/55395638/collectsysteminfo.sh and post the pastebin link.
(2012-08-07, 09:26)wsnipex Wrote: only for the legacy driver. The standard driver still needs that

@freakwenc last try:

code...

if it still does not work, run http://dl.dropbox.com/u/55395638/collectsysteminfo.sh and post the pastebin link.

I've tried that but still get the
Code:
aticonfig: No supported adapters detected
error when I get to step 7.

here is the collectsysteminfo.log.sh http://pastebin.com/KHHQyHuR

Edit:

just noticed it has not displayed any driver info and when I run your script I get
Code:
collectsysteminfo.sh: 31: [[: not found
collectsysteminfo.sh: 31: [[: not found
collectsysteminfo.sh: 31: [[: not found
collectsysteminfo.sh: 63: [[: not found
collectsysteminfo.sh: 70: [[: not found
Error: unable to open display :0.0
FEH.py: cannot connect to X server :0.0
collectsysteminfo.sh: 80: [[: not found
collectsysteminfo.sh: 83: [[: not found
collectsysteminfo.sh: 86: [[: not found
collectsysteminfo.sh: 99: [[: not found
collectsysteminfo.sh: 102: [[: not found
collectsysteminfo.sh: 132: [[: not found
@wsnipex
ok thks :-)
Please support my Blog, XBMC and Linux related :-)
http://youresuchageek.blogspot.com
@freakwenc: if you do this via ssh, you might need to do:
Code:
export DISPLAY=:0
before you can run aticonfig.
(2012-08-06, 11:31)cirkator Wrote: @ wsnipex: would you recommend to always purge pulseaudio when using a full install of eg. Ubuntu?

Anything else that should be removed or added when using a full install?

Any thoughts on this?
@cirkator: yes, unless you need it Smile my xbmc packages are build without pulse
(2012-08-07, 20:38)wsnipex Wrote: @freakwenc: if you do this via ssh, you might need to do:
Code:
export DISPLAY=:0
before you can run aticonfig.

1 step ahead, done it and then tried on the machine itself as well.

I've given up, this is only a test platform until the mele a1000 is usable so don't worry too much. if anyone else gets the bug as well then I'll try again then if I'm still using this pc.

  • 1
  • 66
  • 67
  • 68(current)
  • 69
  • 70
  • 189

Logout Mark Read Team Forum Stats Members Help
How to Install XBMC PVR Xvba for AMD/Nvidia/Intel GPUs23