[FFmpeg-soc] swscale ideas

Michael Niedermayer michaelni at gmx.at
Thu Jul 10 12:23:06 CEST 2008


On Thu, Jul 10, 2008 at 01:40:26PM +0900, Luca Barbato wrote:
> Michael Niedermayer wrote:
> > Hi
> > 
> > Here are some ideas about swscale work
> > 
> > * the dither code depends on non constant globals, it should use the dither
> >   variables in the context
> 
> MMX dither code mangles global vars, shouldn't.

?


> 
> MMX code has variables in the context not ifdeffed, shouldn't.

?


Code _changes_ global vars and breaks multi threaded apps. <--- that should
be fixed.



> 
> I'd rather have the mmx code shuffled away completely (c.f. nodame 
> branch) 

Your git repo is a mess, ive mentioned it already in private mail but
commits like that are not reviewable:

    ppc now working

 Makefile              |    6 +-
 internal_bfin.S       |  607 -------------------------------
 ppc/swscale_altivec.c |   12 +-
 ppc/swscale_altivec.h |   53 +++
 swscale.c             |   17 +-
 swscale_bfin.c        |   94 -----
 yuv2rgb_altivec.c     |  960 -------------------------------------------------
 yuv2rgb_bfin.c        |  207 -----------
 8 files changed, 67 insertions(+), 1889 deletions(-)

(and thats just the last commit from when i pulled last time)


> before addressing those gory details. Having the swscale C 
> implementation clean is a better starting point than trying to address 
> the x86 hacks first.

We have reached the middle of SOC, you HAD time to do that, you failed.
I do expect a little more than moving asm code around for the google
payment!

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Many that live deserve death. And some that die deserve life. Can you give
it to them? Then do not be too eager to deal out death in judgement. For
even the very wise cannot see all ends. -- Gandalf
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-soc/attachments/20080710/ca364c37/attachment.pgp>


More information about the FFmpeg-soc mailing list