[MPlayer-dev-eng] [RFC] libavutil/x86_cpu.h

[Reimar Döffinger]

On Wed, Oct 06, 2010 at 10:51:30PM +0200, Michael Niedermayer wrote:
> On Fri, Aug 06, 2010 at 09:06:00PM +0200, Reimar Döffinger wrote:
> > On Fri, Aug 06, 2010 at 06:28:17PM +0200, Diego Biurrun wrote:
> > > On Fri, Aug 06, 2010 at 06:03:33PM +0200, Reimar Döffinger wrote:
> > > > On Fri, Aug 06, 2010 at 11:40:05AM +0200, Diego Biurrun wrote:
> > > > > On Fri, Aug 06, 2010 at 11:23:25AM +0200, Nico Sabbi wrote:
> > > > > > Carl Eugen Hoyos ha scritto:
> > > > > >> Reimar Döffinger<Reimar.Doeffinger<at>  gmx.de>  writes:
> > > > > >>
> > > > > >>> I am not particularly in favour of supporting such a build configuration,
> > > > > >>> also anyone who removes libavutil can still copy that file themselves...
> > > > > >>>      
> > > > > >> +1
> > > > > >
> > > > > > I'm more in favor of forcing ffmpeg to export all needed headers,  
> > > > > > however private they may be
> > > > > 
> > > > > As I already said, for x86_cpu.h this will not be possible.  So you
> > > > > will have to choose between copying it and keeping libavutil in the
> > > > > source tree.
> > > > 
> > > > I suspect that it might be possible to export the parts of it we
> > > > actually need, I just don't like it either since the REG_* all
> > > > would need to be renamed.
> > > 
> > > Feel free to attempt this, but when I raised the topic on IRC, it came
> > > to end quickly and I don't expect this to go differently on the ml:
> > > 
> > > [11:42:09] <DonDiego> what are the chances of libavutil/x86_cpu.h becoming an installed header?
> > > [11:42:17] * DonDiego guesses 0
> > > [12:13:55] * mru wonders what diego wants with that header
> > > [12:18:42] * kshishkov suspects it may be for MPlayer
> > > [12:19:07] * kshishkov also suspect that if it is so, that means MPlayer is broken
> > > [12:19:14] <thresh> mplayer is notorious for using uninstalled headers
> > > [12:19:24] <mru> that header can never be installed
> > > [12:19:31] <mru> it uses compiler-specific tests
> > 
> > I'd like to point out that Måns says the same thing almost reflexively,
> > and some headers have become installed where he said the same thing.
> > But of course just mindlessly installing it is probably not going to
> > be accepted.
> 
> id like to point out that, as mans has left ffmpeg there is nothing preventing
> us to install this header. Especially the parts that dont depend on the
> compiler but just the HW.

Honestly I consider the issue mostly that someone should do the work
to split it properly.
I'd expect Måns to have it accepted as well, at least if nobody gave him
a hint it had anything to do with MPlayer :-)