On Mon, Dec 06, 2004 at 10:21:36PM -0000, John Harvey wrote:
> Chris
>       There is still one occurrence of pci_dma_sync_single in ivtv-osd.c.
> Next time you make a new release we could do with replacing it for 2.6.8 and
> newer. I assume you don't need a patch for this?

That's only in 0.3.1 right, found it in there but making sure in case I'm
missing it :-).

Thanks,
Chris
> 
> John
> 
> > -----Original Message-----
> > From: [EMAIL PROTECTED] [mailto:ivtv-devel-
> > [EMAIL PROTECTED] On Behalf Of Chris Kennedy
> > Sent: 21 November 2004 15:12
> > To: [EMAIL PROTECTED]
> > Subject: Re: [ivtv-devel] dma_sync_single fix needed and [patch]
> > mod_param, ivtv-0.2.0-rc2t & 2.6.10-rc2
> > 
> > 
> > Yup, that's what happened, the sync call now is one way and for the _cpu
> > and _device ownership of DMA.  Interesting, the old way was bad but the
> > only way (and worked), but seems possibly the change in the kernel and
> > our one sided work around way very bad.  So this is great, thanks, will
> > apply it right now, although it only pertains to people with 2.6.8+ it
> > seems, everything before then should be correct still.
> > 
> > Thanks,
> > Chris
> > 
> > On Sun, Nov 21, 2004 at 06:42:36AM -0500, Robert Hardy wrote:
> > > On Sun, 21 Nov 2004, Robert Hardy wrote:
> > > >I've been testing ivtv under 2.6.10-rc2. Unfortunately it won't even
> > > >compile, although the attached patch brings it closer. The 1st problem
> > was
> > > >that the mod_param_array syntax seems to have changed (sigh what ever
> > > >happened to a "stable" kernel series?....) Please find attached a
> > backwards
> > > >compatible fix for the first problem. Please commit to 0.2.0-rc2
> > series.
> > > >
> > > >The 2nd problem is related to more changes to dma_sync_single family of
> > > >functions and I now think we may be improperly using those functions.
> > i.e.
> > > >we can't blindly replace pci_dma_sync_single w/ dma_sync_single_cpu and
> > > >expect things to work.
> > > >
> > > >I have no idea what the timeline for release but IMHO we should try to
> > fix
> > > >this before release of 0.2.0.
> > > >
> > > >My understand of the code isn't the best, but I think I can figure out
> > the
> > > >appropriate changes.
> > > >
> > > >Here are the relevant documentation point in case some else wants to
> > fix
> > > >it before I get to it:
> > > >linux-2.6.9/Documentation/DMA-mapping.txt around 539.
> > > >http://lwn.net/Articles/75781/
> > > >http://lwn.net/Articles/75780/
> > >
> > > With the attached patch, I think I have now fixed both problems. Please
> > > apply
> > > the attached patch to ivtv-0.2.0-rc2 INSTEAD of my patch from the quoted
> > > message above.
> > >
> > > I'm no expert on DMA calls but I believe I've properly selected either
> > > pci_dma_sync_single_for_cpu or pci_dma_sync_single_for_device to replace
> > > calls to pci_dma_sync_single as appropriate. All these "replacements"
> > only
> > > affect kernels which need them i.e. 2.6.8 and later so this should all
> > be
> > > backwards compatible.
> > >
> > > FYI this quote from the URL below defines the difference between the
> > calls:
> > > "dma_sync_single_for_cpu() gives ownership of the DMA buffer back to the
> > > processor. After that call, driver code can read or modify the buffer,
> > but
> > > the device should not touch it. A call to dma_sync_single_for_device()
> > is
> > > required to allow the device to access the buffer again."
> > > quoted from http://lwn.net/Articles/75780/
> > >
> > > The new code still works on 2.6.9. I can pause live TV in Mythtv.
> > > Everything runs very smoothly.
> > >
> > > I should point out I am not using the PVR-350 decoder. I do my playback
> > via
> > > NVidia SVideo out. It would be good idea if someone could double check
> > the
> > > patch doesn't break decoding. It is highly unlikely that it would.
> > >
> > > The code builds, depmods and modprobes on 2.6.10-rc2-bk5. I'm about to
> > try
> > > it on that kernel here. I'll let you know how 2.6.10-rc2-bk5 testing
> > goes...
> > >
> > > Regards,
> > > Rob
> > >
> > > --
> > > ---------------------"Happiness is understanding."----------------------
> > > Robert Hardy, B.Eng Computer Systems                  C.E.O. Webcon Inc.
> > > rhardy <at> webcon <dot> ca    GPG Key available          (613) 276-7327
> > 
> > 
> > 
> > --
> > ---
> >  Chris Kennedy / [EMAIL PROTECTED]
> >   Engineer KMOS-TV/KTBG-FM
> >   Broadcasting Services Department
> >   Central Missouri State University
> > 
> > 
> > -------------------------------------------------------
> > This SF.Net email is sponsored by: InterSystems CACHE
> > FREE OODBMS DOWNLOAD - A multidimensional database that combines
> > robust object and relational technologies, making it a perfect match
> > for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8
> > _______________________________________________
> > ivtv-devel mailing list
> > [EMAIL PROTECTED]
> > https://lists.sourceforge.net/lists/listinfo/ivtv-devel
> 
> 
> 
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now. 
> http://productguide.itmanagersjournal.com/
> _______________________________________________
> ivtv-devel mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/ivtv-devel

-- 
---
 Chris Kennedy / [EMAIL PROTECTED]
  Engineer KMOS-TV/KTBG-FM
  Broadcasting Services Department
  Central Missouri State University


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/
_______________________________________________
ivtv-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/ivtv-devel

Reply via email to