Mon, 25 Mar 2002 16:38:13 +0000 |
michael |
10l
|
Mon, 25 Mar 2002 16:35:24 +0000 |
michael |
byte interleaving for mga
|
Sun, 03 Mar 2002 13:33:40 +0000 |
atmos4 |
add mangling
|
Sun, 10 Feb 2002 00:30:24 +0000 |
michael |
rgb24toyv12 in MMX (untested)
|
Mon, 26 Nov 2001 01:15:08 +0000 |
michael |
runtime cpu detection
|
Sun, 11 Nov 2001 22:26:15 +0000 |
michael |
uyvytoyv12 in MMX (untested)
|
Sun, 11 Nov 2001 00:50:40 +0000 |
michael |
yuy2toyv12 bugfix
|
Sat, 10 Nov 2001 23:28:10 +0000 |
arpi |
uyvy->uv12 added
|
Sat, 10 Nov 2001 20:39:23 +0000 |
michael |
.balign or weĦll align by 64kb on some architectures
|
Sat, 10 Nov 2001 19:46:04 +0000 |
michael |
.align 16
|
Wed, 07 Nov 2001 23:17:11 +0000 |
michael |
bugfix
|
Tue, 06 Nov 2001 23:54:41 +0000 |
michael |
rgb32tobgr32 / bgr32torgb32
|
Tue, 06 Nov 2001 17:14:22 +0000 |
nick |
Minor speedup of rgb32to24. (performance is not successful)
|
Tue, 06 Nov 2001 16:35:17 +0000 |
nick |
MMX2, 3DNOW, MMX optimized rgb32(24)to16(15) stuff
|
Tue, 06 Nov 2001 16:21:38 +0000 |
nick |
optimized rgb24to16 stuff
|
Tue, 06 Nov 2001 13:53:43 +0000 |
nick |
rgb24to32 now is faster
|
Tue, 06 Nov 2001 11:22:40 +0000 |
nick |
Move yuv2rgb to postprocess
|
Mon, 05 Nov 2001 19:55:11 +0000 |
michael |
unsigned stuff
|
Mon, 05 Nov 2001 18:50:58 +0000 |
michael |
yuy2toyv12 fixed and speedup
|
Mon, 05 Nov 2001 18:26:49 +0000 |
michael |
fixed yv12toyuy2
|
Mon, 05 Nov 2001 17:56:11 +0000 |
nick |
Fixed rgb32(24)to15 stuff
|
Mon, 05 Nov 2001 17:46:20 +0000 |
nick |
Fixed rgb32(24)to16 stuff, rgb32(24)to15 is still broken
|
Mon, 05 Nov 2001 17:35:28 +0000 |
nick |
Well - old algorithms and new stuff rgb24to16(15)
|
Mon, 05 Nov 2001 08:37:27 +0000 |
nick |
More standards compilance
|
Mon, 05 Nov 2001 02:25:30 +0000 |
michael |
yuy2toyv12 in MMX
|
Mon, 05 Nov 2001 01:21:05 +0000 |
michael |
yv12toyuy2 in MMX
|
Mon, 05 Nov 2001 00:45:55 +0000 |
michael |
yv12 <-> yuy2 in C
|
Sun, 04 Nov 2001 23:04:26 +0000 |
michael |
faster 15to16 bit rgb (the mmx routine is limited by memory speed so there is no difference ): but the c routine is faster
|
Sun, 04 Nov 2001 19:50:29 +0000 |
michael |
palette to bgr24
|
Sun, 04 Nov 2001 19:36:24 +0000 |
michael |
bgr32to16, bgr32to15 (needed for palette stuff)
|
Sun, 04 Nov 2001 11:51:36 +0000 |
nick |
using const modifier
|
Tue, 30 Oct 2001 17:46:07 +0000 |
nick |
cosmetic
|
Mon, 29 Oct 2001 18:28:06 +0000 |
nick |
Finish mmx2, 3dnow optimiz. 15to16 should be tested. Better fix of can't compile
|
Mon, 29 Oct 2001 18:07:40 +0000 |
michael |
cant compile on non x86 bugfix
|
Sun, 28 Oct 2001 17:42:52 +0000 |
nick |
32to24: MMX, MMX2, 3DNOW optimization
|
Sun, 28 Oct 2001 17:15:54 +0000 |
nick |
More elegant solution
|
Sun, 28 Oct 2001 14:03:16 +0000 |
nick |
preprocessor cleanup
|
Sun, 28 Oct 2001 13:59:57 +0000 |
nick |
sfence
|
Sun, 28 Oct 2001 13:06:49 +0000 |
nick |
More accurate prefetching
|
Sun, 28 Oct 2001 12:02:16 +0000 |
nick |
mmx, mmx2, 3dnow optimized 24to32
|
Sun, 28 Oct 2001 11:48:50 +0000 |
nick |
mmx optimized 24to32
|
Sun, 28 Oct 2001 11:00:40 +0000 |
nick |
Prev ver could work only on x86
|
Sun, 28 Oct 2001 10:45:42 +0000 |
nick |
vo_vesa: more rgb2rgb support
|
Sun, 28 Oct 2001 10:28:08 +0000 |
nick |
vo_vesa: more rgb2rgb support
|
Sun, 28 Oct 2001 10:17:57 +0000 |
nick |
vo_vesa: rgb2rgb support
|