mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Mark Knecht <markknecht@gmail.com>
To: Lee Revell <rlrevell@joe-job.com>
Cc: linux-kernel@vger.kernel.org, Ingo Molnar <mingo@elte.hu>
Subject: Re: 2.6.14-rc5-rt3 - `IRQ 8'[798] is being piggy
Date: Fri, 21 Oct 2005 15:25:18 -0700	[thread overview]
Message-ID: <5bdc1c8b0510211525v62212d33j84491cfc687bd200@mail.gmail.com> (raw)
In-Reply-To: <1129923883.17709.11.camel@mindpipe>

On 10/21/05, Lee Revell <rlrevell@joe-job.com> wrote:
> On Fri, 2005-10-21 at 11:52 -0700, Mark Knecht wrote:
> > On 10/21/05, Lee Revell <rlrevell@joe-job.com> wrote:
> > > On Fri, 2005-10-21 at 10:40 -0700, Mark Knecht wrote:
> > > > On 10/21/05, Mark Knecht <markknecht@gmail.com> wrote:
> > > > > Hi,
> > > > >    Maybe I'm catching something here? Maybe not - no xruns as of yet,
> > > > > but I've never seen these messages before. Kernel config attached.
> > > > >
> > > > >    dmesg has filled up with these messages:
> > > > >
> > >
> > > This isn't a real problem.  You enabled CONFIG_RTC_HISTOGRAM.  Don't do
> > > that.
> > >
> > > Lee
> >
> >
> > Right, but the 'piggy' messages are a real prblem, aren't they?
>
> No I don't think so.  CONFIG_RTC_HISTOGRAM is a hack, designed to work
> with a specific test program that runs SCHED_FIFO and poll()s on the
> RTC.  VLC apparently poll()s on the RTC but does not run SCHED_FIFO.  So
> of course there will be delays.
>
> Now that the kernel has good soft realtime support and non-root RT
> scheduling, these apps really need to adopt a correct soft RT design
> like JACK.  AFAICT they don't even bother to try to get SCHED_FIFO for
> the time-sensitive rendering threads.  I can't even get totem-xine (the
> default "Sound and Movie Player" for Gnome) to keep the audio and video
> in sync.  mplayer only plays smoothly if I run it at nice -10.  Etc.
>
> The Linux kernel is pretty good for RT these days but compared to OSX or
> even Windows the apps are a joke.
>
> Lee

Lee,
   Indeed, you are correct. Apparently under character devices I had
turned on the RTC histogram feature. With that off I am not only
getting the maximum latency values we expect.

   Now I'll have to let it run for hours/days to see if I catch any
info on these xruns, should I get another rash of them. Yesterday they
came only after about 14 hours of work. I had none the previous two
days.

Thanks,
Mark

  reply	other threads:[~2005-10-21 22:25 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-10-21 17:03 Mark Knecht
2005-10-21 17:40 ` Mark Knecht
2005-10-21 18:45   ` Lee Revell
2005-10-21 18:52     ` Mark Knecht
2005-10-21 19:44       ` Lee Revell
2005-10-21 22:25         ` Mark Knecht [this message]
2005-10-21 23:28           ` Lee Revell
2005-10-22  3:44             ` Ingo Molnar

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5bdc1c8b0510211525v62212d33j84491cfc687bd200@mail.gmail.com \
    --to=markknecht@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=rlrevell@joe-job.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®