[mythtv-users] [OT] nvidia 8756: XvMC works!

Steven Adeff adeffs.mythtv at gmail.com
Sat Apr 8 21:32:27 UTC 2006


On 4/8/06, Joe Votour <joevph at yahoo.com> wrote:
> --- Jurgen Kramer <gtm.kramer at inter.nl.net> wrote:
>
> > On Sat, 2006-04-08 at 16:02 -0400, Steven Adeff
> > wrote:
> > > On 4/8/06, Jurgen Kramer <gtm.kramer at inter.nl.net>
> > wrote:
> > > > I just tried the new nvida 8756 drivers from
> > livna on my main PC to see
> > > > if XvMC would finally work....and it works!!
> > > >
> > > > I hope the atrpms packages will be there soon so
> > I can upgrade my
> > > > mythbox.
> > > >
> > > > vd_ffmpeg: XVMC accelerated MPEG2
> > > > trying pixfmt=0
> > > > VDec: vo config request - 1920 x 1080 (preferred
> > colorspace: MPEG1/2
> > > > Motion Compensation and IDCT)
> > > > VDec: using MPEG1/2 Motion Compensation and IDCT
> > as output csp (no 0)
> > > > Movie-Aspect is 1.78:1 - prescaling to correct
> > movie aspect.
> > > > VO: [xvmc] 1920x1080 => 1920x1080 MPEG1/2 Motion
> > Compensation and IDCT
> > > > vo_xvmc: Port 274 grabed
> > > > vo_xvmc: Found matching surface with id=54434449
> > on 274 port at 0
> > > > adapter
> > > > vo_xvmc: Allocated Direct Context
> > > > vo_xvmc: data_blocks allocated
> > > > vo_xvmc: mv_blocks allocated
> > > > vo_xvmc: Motion Compensation context allocated -
> > 8 surfaces
> > > > vo_xvmc: idct=1 unsigned_intra=0
> > > > vo_xvmc: looking for OSD support
> > > >     Subpicture id 0x34344149
> > > > vo_xvmc: OSD support by additional frontend
> > rendering
> > > >
> > > > Jurgen
> > >
> > > Jurgen, what was the issue you had before?
> > > I noticed that I had to run mplayer over and over
> > until it worked
> > > before Myth would start using XvMC, which makes it
> > hard on a dedicated
> > > frontend. If this new version is working right off
> > the bat, that will
> > > be my two problems with the old drivers solved
> > (interlacing and now
> > > possibly XvMC).
> >
> > On my main PC, which has a 6800GS, XvMC was not
> > working at all because
> > of a bug in the driver. On my myth system, which has
> > a 6200, I only got:
> >
> > Error while decoding frame!
> > [mpegvideo_xvmc @ 0x86b2c00]XvMCLoadQMatrix: Error:
> > 2
> > [mpegvideo_xvmc @ 0x86b2c00]XvMCBeginSurface: Error:
> > 2
> > Error while decoding frame!
> > [mpegvideo_xvmc @ 0x86b2c00]XvMCLoadQMatrix: Error:
> > 2
> > [mpegvideo_xvmc @ 0x86b2c00]XvMCBeginSurface: Error:
> > 2
> >
> > Hopefully these errors will be gone as well. I'll
> > wait for the atrpms
> > packages. So far it looks promising, it would save
> > me a CPU upgrade on
> > my myth fe/be system.
> >
> > Jurgen
> >
> > _______________________________________________
> > mythtv-users mailing list
> > mythtv-users at mythtv.org
> >
> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
> >
>
> I just installed it from the ATrpms packages, and it
> pulled in the 2.6.16-1 kernel, which of course means
> an ivtv, alsa and lirc update (at least in the kernel
> modules, anyway).
>
> Unfortunately, XvMC seems broken with OpenGL vsync
> enabled (at least on my version of SVN), and I got a
> hard lockup on my MythTV machine (combined
> backend/frontend) when I tried changing channels while
> watching LiveTV on my FusionHDTV 5 Lite ATSC card.
>
> Given this, I think I'm just going to bite the bullet
> and get a new CPU.  My Athlon64 3200+ (socket 754) is
> almost powerful enough for HD without XvMC (sometimes
> I get the odd prebuffering pause, but only when a
> flurry of disk active happensm like commercial
> flagging or database accesses), and can definitely do
> it with XvMC, but too many problems with it.  Tweaking
> the PCI latencies definitely helped out to get me to
> that point.
>
> I hope that it works for other people, but I think for
> me, an X2 is a better solution.  Too many bugs with
> XvMC and MythTV right now.


the 3200+ should be more than enough for HD without XvMC. You might
want to see whats eating up your CPU.

--
Steve


More information about the mythtv-users mailing list