[MPlayer-cvslog] r26128 - in trunk: configure vidix/Makefile vidix/drivers.c vidix/s3_regs.h vidix/s3_vid.c
Reimar Döffinger
Reimar.Doeffinger at stud.uni-karlsruhe.de
Sat Mar 1 10:39:25 CET 2008
On Fri, Feb 29, 2008 at 11:17:02PM +0100, Benjamin Zores wrote:
> Reimar Döffinger a écrit :
> > On Fri, Feb 29, 2008 at 09:01:29PM +0100, ben wrote:
> >> Author: ben
> >> Date: Fri Feb 29 21:01:28 2008
> >> New Revision: 26128
> >>
> >> Log:
> >> New S3 VIDIX driver.
> >> Provides support for S3 Trio and S3 Virge chipsets.
> >> This deprecates the old Savage driver that worked with latest chips only.
> >> (synchronized with vidix.sf.net r326 and r327)
> >
> > Does not build on PPC due to missing INPORT8/OUTPORT8.
> > Looks like a regression, since the savage driver did at least compile on
> > that...
>
> Hey, you're funny ... who send me the patch and ask me to apply it ?
> I do no more have PowerPC however so can't test it anymore.
Neither do I, but if you want to test compilation, e.g. osuosl.org offer
shell account.
Or you can just bear with my complaints ;-)
> Though, the only PowerPC that I know having a VGA board are old
> Macintosh ones and they all require to have a VGA board with PowerPC
> BIOS which only exist for some ATI/nVidia/Matrox/PowerVR cards.
BIOS is only necessary for the primary card, and even there not if you
can live with e.g. console only via com port.
> There's no way you'll ever be able to use S3 driver on these computers,
> hence, no need to have it compiled.
I am quite sure it is not impossible, and a compiling driver would be
easier to fix than one that does not even work so far.
> Maybe we can simply update configure to prevent bad drivers to compile
> on powerpc.
Not a perfect solution, but it will have to do unless we find someone
who has a PPC, a few left-over cards and wants to hack vidix...
> But in a general way, many VIDIX drivers are using x86 specific code.
> This "Generic OS independant video framework" seems to be more and more
> x86 && linux only.
Or in other words, we are faster and faster getting to the point where
X is better in absolutely all aspects...
Greetings,
Reimar Döffinger
More information about the MPlayer-cvslog
mailing list