I'm using the video out on my PVR-350 which means I
have no GL capabilities etc. I am not using the
decoder in the card which probably would help but
there are other issues that I do not like if I enable
the decoder. Encoding (MPEG2) should all be handled by
the card though. I'm not sure if I am using XvMC or
not, how do I tell? I don't remember doing any
specific configuration for that. If it wasn't enabled
in the old  RPMS by default and is enabled in the new
RPMS by default that I would suspect that could be my
problem as Xorg CPU is also very high during playback.


I don't remember what the CPU load was like prior to
the update though. The updates should have been very
minor. I think I was running .20-137 and now am
running .20-141. I'm not 100% that the problem is
actually in Myth. It could be another system utility
or library in that yum.log update list that is causing
the problem. Do you have a link to the thread you are
referring to? It might give me some ideas.

Thanks!

--- DM - ATRPMS <[EMAIL PROTECTED]> wrote:

> Increased CPU does seem to be a sporadic side effect
> for some reason, 
> based on a couple posts here and a few more in
> archived on other MythTV 
> oriented lists.  (See my posts from a couple weeks
> ago on this list.) 
> My frontend had significantly increased CPU after
> upgrade (went from 50% 
> to 90+%) while still configured with XvMC.  Since
> CPU loads like that 
> cause lockups on that hardware, this is a problem! 
> I had to completely 
> retweak the playback without using XvMC (which A>
> didn't work before - 
> that's why i went to the trouble of XvMC to begin
> with; and B> better 
> performance is after all the purpose of XvMC) to get
> it working again.
> 
> The backend system (which is a backend/frontend
> combo) had no such 
> issues that I've noticed so far.
> 
> I would be curious to know A> if you are using XvMC
> and B> what your 
> video card is (my problem box is an nVidia).  (Not
> that I'll probably be 
> of much help, but I'm curious.  :)
> 
> I have not paid attention to transcoding jobs, so I
> can't help you 
> there.  You probably have done this already, but if
> not you may want to 
> get a baseline on the server when you are just
> recording (not playing 
> back) to make sure it's not the encoding functions
> that are getting 
> you...  If your encoding usage is higher, the
> backend may have enough 
> CPU to encode it and stream it (i.e. to a frontend
> system), but not 
> encoded it and then decode it again (i.e. watching
> directly on the backend).
> 
> Dan
> 
> Axel Thimm wrote:
> > On Wed, Oct 11, 2006 at 05:30:05PM -0700, Void
> Main wrote:
> >> I built an FC5 MythTV box back on the 16th of
> September and updated
> >> it with all the latest updates and used MythTV
> RPMs from ATrpms.
> > 
> >> I realized the MythTV RPMS had been updated to
> newer version in the
> >> repository so I thought no problem, I'll just
> "yum update" on the
> >> MythTV server and upgrade everything to the
> latest. Well, the remote
> >> frontends seem to still work find with the newest
> MythTV updates but
> >> the main box that servers as both the frontend
> and backend now has
> >> choppy video most noticable when trying to watch
> live TV and seems
> >> to be worst when on screen menus are visible.
> It's like a stutter
> >> every few seconds.
> >>
> >> It is as if the frontend is now taking more CPU
> than it did before
> >> the upgrade. I have also done a couple of
> transcodes since the
> >> upgrade and these also seem slower. I was getting
> over 20 frames per
> >> second when transcoding (~25fps if I recall) and
> now I get less than
> >> 20fps (~17fps).
> > 
> > Is that with or w/o myth running and eating CPU
> time. If it's with
> > it's no surprise, any process will be slower if
> something is eating
> > away all of your CPU. If it's w/o it gets
> interesting, as watching
> > live tv and "transcode" share almost nothing in
> common.
> > 
> >> I am curious if anyone else has run into this
> > 
> > Yes, that would be an important information.
> > 
> > FWIW the updated packages only contain fixes from
> the -fixes
> > branch. I can't really think of something in there
> that starts slowing
> > down your system, though, and indeed there were no
> similar reports
> > yet, but let's see.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

_______________________________________________
atrpms-users mailing list
[email protected]
http://lists.atrpms.net/mailman/listinfo/atrpms-users

Reply via email to