Subject: Re: [linux-audio-dev] Re: Sync Issues (was Re: External MIDI Sync using OSS/Free)
From: Roger Larsson (roger.larsson_AT_skelleftea.mail.telia.com)
Date: la loka 30 1999 - 21:04:14 EDT
Hi,
Benno I guess you mean something like "PCI Latency Timer (PCI Clocks)"
I have it as a BIOS option, mine is set to 32 with a 30 MHz PCI
(180 MHz CPU) it should give 1 us latency.
It can be configured up to 248 that should give about 8 us latency.
If I understand the 440FX documentation correctly no PCI device is
allowed to hold the PCI bus a longer time...
Strange.
/RogerL
Benno Senoner wrote:
>
> On Fri, 29 Oct 1999, Roger Larsson wrote:
> >>
> > Or you can try to find out what part of your video card driver that
> > destroys latency and correct that.
> >
> > You can find out the problematic spot with my patch:
> > latency-profiling-2.2.10-r6.patch [for Linux 2.2.10]
> >>
> > The results are messages in 'dmesg', latency time and instruction
> > pointers
> > are printed. You then decode the instruction pointers to routine name.
> >
> > /RogerL
>
> Not quite correct: actually it's not a driver's fault !
> In most cases the audio dropout while doing heavy GFX is caused by PCI busmaster
> transfers by the gfx card that hogs the entire PCI bus for SEVERAL milliseconds,
> in order to provide better gfx throughput.
> On some cards you can avoid this problem by setting a special variable into
> your XF86Config (no sample handy at the moment) , which slows down the gfx
> performance a bit but gives you better lowlatency performance.
>
> On some other cards the only solution is to throw your gfx card out of the
> nearest window.
> :-)
>
> PS: Benjamin, I stopped the crossposting , at least for this mail.
> :-)
>
> Benno.
--The Internet interprets Windows as damage, and routes around it.
Roger Larsson Skellefteċ Sweden
This archive was generated by hypermail 2b28 : pe maalis 10 2000 - 07:28:00 EST