[Mplayer-cvslog] CVS: main/DOCS/tech TODO,1.9,1.10

Nick Kurshev nickols_k at mail.ru
Mon Dec 10 18:56:15 CET 2001


Hello, Michael!

On Sun, 9 Dec 2001 20:48:09 +0100 you wrote:

> Hi
> 
> On Sunday 09 December 2001 19:17, Nick Kurshev wrote:
> [...]
> > IMHO such code would be better to move into memcpy_init but memcpy should
> > be replaced with pointer into it ;)
> > Calling through pointer should be faster of such if-else blocks
> no
> try it yourself (attached benchmark compiled with -O4)
> the output for my p3 is
> 51052k cycles needed for 10000000 calls through if/else
> 140573k cycles needed for 10000000 calls through function pointers
And for witch memcpy?
fast_memcpy_MMX2 or fast_memcpy_MMX?
Anyway - it's only in theoretical test but in practic processor won't
to save conditional branch-table for this branch and probably will
predict false branch.
I know only that manuals always suggested to replace conditional jumps
with direct code ;)

If you will accept my idea - I have solutution for self-modification code
without changning logic of mplayer of each your TEMPLATE set.
> [...]
> 
> Michael


Best regards! Nick



More information about the MPlayer-cvslog mailing list