Commit Graph

24 Commits

Author SHA1 Message Date
gpoirier 68a1de942c moves invariant code (chiefly MMX register initialization) out of loops as well as eliminating some memory accesses within the inner loop.
Patch by Zuxy Meng < zuxy POIS meng AH gmail POIS com >
Original thread:
Date: Mon, 12 Jun 2006 00:31:53 -0700 (PDT)
Subject [MPlayer-dev-eng] [PATCH] Loop invariant motion in libvo/osd_template.c


git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@18686 b3059339-0415-0410-9bf9-f77b7e298cf2
2006-06-12 11:14:10 +00:00
reimar 2f5698c9cb Fix MMX accelerated RGB24 OSD, fixes "ugly OSD with -vo gl2".
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@16484 b3059339-0415-0410-9bf9-f77b7e298cf2
2005-09-13 21:41:44 +00:00
aurel 798d2d1337 adapting existing mmx/mmx2/sse/3dnow optimizations so they work on x86_64
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@13721 b3059339-0415-0410-9bf9-f77b7e298cf2
2004-10-21 11:55:20 +00:00
nplourde d2ad2147cc draw alpha for uyvy
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@12517 b3059339-0415-0410-9bf9-f77b7e298cf2
2004-05-31 15:13:35 +00:00
alex c68f67e2e0 removed obsolete my_profile
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@10517 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-08-03 18:33:28 +00:00
michael 9c160c32b5 big endian yellow fix?
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@9961 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-04-21 17:24:29 +00:00
michael 7d74522cdd workaround lack of -fomit-frame-pointer
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5140 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-16 17:12:19 +00:00
atmos4 6afa7d6988 mangling in libvo
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4246 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-19 04:44:49 +00:00
michael a1bb772c18 yuy2 in C color bugfix patch from Artur Zaprzala <artur.zaprzala@talex.com.pl>
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@3432 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-12-10 13:14:28 +00:00
michael 2e4b5d0908 runtime cpu detection
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@3143 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-26 21:12:15 +00:00
atmos4 c184876b1a Compile fix.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2851 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-12 02:01:24 +00:00
michael 6778e88bf5 p2/p3 bgr32 version (20%faster)
yv12 and yuy2 versions in mmx


git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2847 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-11 22:14:13 +00:00
nick 615471a069 Use new logic suggested by Michael Niedermayer
K7:
Non-MMX stuff
    total=299997790
Old-MMX stuff:
    total=159120928
New-MMX stuff:
    total=80925337


git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2844 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-11 17:14:57 +00:00
nick c5e3ac66ac Cloning 32 stuff to 24
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2840 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-11 16:09:19 +00:00
nick 1ffb6bfa2b Minor speedup for K6-2, K7, P3
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2836 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-11 15:35:11 +00:00
nick 7035adb0bd Enable MMX stuff
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
2001-11-11 14:42:10 +00:00
nick f4730e10c1 Extract parallelism from OSD stuff + MMX2 optimization.
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
2001-11-11 11:18:50 +00:00
michael 20d9b4ff45 bugfix
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
2001-11-11 01:14:09 +00:00
michael 751324d0d3 .balign or we´ll align by 64kb on some architectures
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2801 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-11-10 20:39:23 +00:00
nick 672e0016c9 Code cleanup - emms is not required when MMX block is commented out.
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
2001-11-10 18:40:49 +00:00
michael 028ee068e4 slow mmx & not so slow asm versions (outcommented)
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@2579 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-10-30 22:35:02 +00:00
arpi_esp ba8f720ce0 FAST_OSD option to disable font outline antialiasing
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@948 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-06-02 16:02:38 +00:00
arpi_esp 07bdafcbe7 to be sure in that header is okey
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@623 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-04-24 20:03:13 +00:00
arpi_esp 2f46205ad3 OSD alpha renderers moved to osd.c
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@327 b3059339-0415-0410-9bf9-f77b7e298cf2
2001-04-10 02:29:38 +00:00