I don't know why Michael Niedermayer disabled it but:
my benchmarks of 25 frames for TIMER+SUBTITLE (measured by RDTSC):
Non-MMX stuff:
total=306142159
MMX stuff:
total=159534150
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2834 b3059339-0415-0410-9bf9-f77b7e298cf2
I've found that mplayer's measuring is not precise :(
Here my test with using RDTSC:
Old stuff:
rd_tsc: 774377
rd_tsc: 765985
rd_tsc: 265309
New CPU optimized stuff:
rd_tsc: 661154
rd_tsc: 641317
rd_tsc: 222448
New MMX2 optimized stuff:
rd_tsc: 269544
rd_tsc: 329189
rd_tsc: 173110
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2824 b3059339-0415-0410-9bf9-f77b7e298cf2
nick why did u remove the #if 0 around that code without testing it?
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2808 b3059339-0415-0410-9bf9-f77b7e298cf2
Special notes for Michael Niedermayer:
Are you still here?
If you don't like for(cond;cond;cond) C-constructions and prefer
asm ones: "jb 1b" then use .align 16 pseudo assembler instructions
else loops rather will be not aligned on correct boundary.
(16 it's for K7 for pent should be 8).
Your parts have a lot such lacks.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2799 b3059339-0415-0410-9bf9-f77b7e298cf2
albeu
atmos4
michael
nick
arpi_esp