On Saturday 17 June 2006 17:26, Brian Bustin wrote:
> > Date: Sat, 17 Jun 2006 09:52:40 +0200
> > From: Hans Verkuil <[EMAIL PROTECTED]>
> > Subject: Re: [ivtv-devel] Potential fix for DMA problems
> > To: Discussion list for development of the IVTV driver
> > <[email protected]>
> > Message-ID: <[EMAIL PROTECTED]>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > On Saturday 17 June 2006 02:39, Brian Bustin wrote:
> >> I made 3 changes recently that seem to have made all the
> >> difference in the world. The system has been up for a little over
> >> 17 hours and it seems pretty solid.
> >>
> >> The three changes that I made were:
> >> 1. In General Settings I removed Optimize for size
> >> * I had not realized that this option was selected in
> >> the first place. It very clearly notes that it may cause problems
> >> with certain compilers.
> >
> > Which gcc version do you have? (gcc --version) If you have a recent
> > gcc
> > I don't think this is the cause of the problem.
> >
> >> 2. In Bus Options and then PCI Support I removed Unordered IO
> >> Mapping
> >> * I had also not realized that this was selected. The
> >> help text clearly notes that it is experimental code, and will
> >> only work properly for drivers that have been coded for specific
> >> platforms.
> >
> > I don't have that option in my vanilla kernel. Seems to be a patch
> > gentoo added. Actually, this sounds like something that can cause
> > problems. What does it do?
> >
> >> 3. I am now using the kernelspace ondemand cpu frequency scaler
> >> instead of using cpudyn
> >
> > I personally am running powernowd and I never had any problems with
> > that
> > cpu scaler.
> >
> >> System:
> >> Athlon 64 3000+
> >> 1 Gig of Ram
> >> ASUS A8N-SLI
> >> SATA hard drives using libata
> >> 2 PVR-150
> >> 1 HD3000
> >> Gentoo Linux
> >> IVTV 0.6.2
> >> kernel: linux-2.6.16-gentoo-r9
> >> MythTV 0.19
> >>
> >> There is more information about this on my blog:
> >> http://www.morph3ous.net/2006/06/16/solution-to-ivtv-drivers-ivtv0
> >>- warning-enc-0-dma-error-0x0000000b/
> >>
> >> I hope that this is helpful.
> >
> > Certainly. Can you also post your INIT IVTV messages from the
> > kernel log?
> >
> > Thanks,
> >
> > Hans
>
> I apologize in advance for sending 3 emails. I realized that the
> other 2 emails did not get connected with this thread. Must be
> because I subscribe to the digests and actually created a new message
> with the same subject thinking it would work.
>
> I have gcc version:
> gcc (GCC) 3.4.5 (Gentoo 3.4.5, ssp-3.4.5-1.0, pie-8.7.9)
Hmm, it could be that this compiler has problems with size
optimizations. I'm using gcc-4.1.1.
>
> The unordered IO mapping gives this information in its help text:
> CONFIG_UNORDERED_IO:
>
> Use unordered stores to access IO memory mappings in device
> drivers. Still very experimental. When a driver works on
> IA64/ppc64/pa- risc it should
> work with this option, but it makes the drivers behave
> differently from i386. Requires that the driver writer used memory
> barriers properly.
Sounds dangerous.
It would be nice if you can tell which of these two options are the
culprit. If it is the UNORDERED_IO, then I may have to check for memory
barriers.
Hans
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
