| Sat, 29 Sep 2007 22:31:18 +0000 |
lorenm |
sse2 version of compute_autocorr().
libavcodec
|
| Sun, 26 Aug 2007 16:10:53 +0000 |
ramiro |
CONFIG_7REGS has been renamed to HAVE_7REGS
libavcodec
|
| Sun, 26 Aug 2007 12:34:22 +0000 |
michael |
workaround gcc bug, untested as my gcc is not complaining
libavcodec
|
| Sun, 26 Aug 2007 01:11:02 +0000 |
michael |
avoid overflow in the 3rd lifting step, this now needs mmx2 at minimum
libavcodec
|
| Sat, 25 Aug 2007 15:20:56 +0000 |
michael |
update mmx code to latest snow changes
libavcodec
|
| Sat, 25 Aug 2007 03:00:51 +0000 |
michael |
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
libavcodec
|
| Tue, 24 Jul 2007 08:54:56 +0000 |
aurel |
help some gcc version to optimize out those functions
libavcodec
|
| Tue, 10 Jul 2007 20:27:50 +0000 |
aurel |
cosmetics: indentation
libavcodec
|
| Tue, 10 Jul 2007 20:23:08 +0000 |
aurel |
Avoid linking with h263.c functions when the relevant codecs
libavcodec
|
| Sun, 08 Jul 2007 23:15:00 +0000 |
mru |
make arguments to ssd_int8_vs_int16() const
libavcodec
|
| Sun, 20 May 2007 05:07:44 +0000 |
zuxy |
Remove incorrect comment; MMX2 is preferred over 3DNow! on Athlon
libavcodec
|
| Fri, 18 May 2007 08:18:56 +0000 |
zuxy |
3DNow! and SSSE3 optimization to QNS DSP functions; use pmulhrw/pmulhrsw instead of pmulhw
libavcodec
|
| Wed, 16 May 2007 23:23:45 +0000 |
aurel |
better separation of vp3dsp functions from dsputil_mmx.c
libavcodec
|
| Wed, 16 May 2007 09:51:45 +0000 |
diego |
Add libavcodec to compiler include flags in order to simplify header
libavcodec
|
| Mon, 14 May 2007 14:28:13 +0000 |
takis |
Make vp3dsp*.c compilation optional.
libavcodec
|
| Sat, 12 May 2007 02:41:25 +0000 |
lorenm |
factor sum_abs_dctelem out of dct_sad, and simd it.
libavcodec
|
| Sat, 12 May 2007 01:16:06 +0000 |
lorenm |
sse2 & ssse3 versions of hadamard. unroll and inline diff_pixels.
libavcodec
|
| Wed, 09 May 2007 01:46:33 +0000 |
lorenm |
cosmetics: remove code duplication in hadamard8_diff_mmx
libavcodec
|
| Tue, 08 May 2007 17:55:56 +0000 |
lorenm |
cosmetics: remove duplicate transpose macro
libavcodec
|
| Sat, 07 Apr 2007 14:10:02 +0000 |
diego |
typos
libavcodec
|
| Fri, 30 Mar 2007 19:15:31 +0000 |
lorenm |
mmx 16-bit ssd. 2.3x faster svq1 encoding.
libavcodec
|
| Sat, 24 Feb 2007 11:58:52 +0000 |
diego |
Fix wrong conditional, Snow decoding, not encoding, was SIMD-accelerated.
libavcodec
|
| Tue, 30 Jan 2007 10:31:34 +0000 |
takis |
Add the const specifier as needed to reduce the number of warnings.
libavcodec
|
| Tue, 14 Nov 2006 03:18:09 +0000 |
mru |
revert bad checkin
libavcodec
|
| Tue, 14 Nov 2006 03:12:29 +0000 |
mru |
move some CFLAGS settings away from config.* writing section
libavcodec
|
| Sun, 12 Nov 2006 03:34:12 +0000 |
mru |
adding more static keywords
libavcodec
|
| Fri, 03 Nov 2006 02:03:56 +0000 |
michael |
optimize H264_DEBLOCK_P0_Q0
libavcodec
|
| Sat, 14 Oct 2006 17:04:50 +0000 |
diego |
Put libmpeg2 IDCT functions under CONFIG_GPL, fixes link failure
libavcodec
|
| Wed, 11 Oct 2006 23:17:58 +0000 |
diego |
Rename ABS macro to FFABS.
libavcodec
|
| Sat, 07 Oct 2006 15:30:46 +0000 |
diego |
Change license headers to say 'FFmpeg' instead of 'this program/this library'
libavcodec
|
| Thu, 05 Oct 2006 00:23:24 +0000 |
diego |
Switch idct_mmx_xvid.c from GPL to LGPL as permitted by the
libavcodec
|
| Sun, 01 Oct 2006 21:25:17 +0000 |
lorenm |
approximate qpel functions: sacrifice some quality for some decoding speed. enabled on B-frames with -lavdopts fast.
libavcodec
|
| Wed, 27 Sep 2006 19:54:07 +0000 |
mru |
add some #ifdef CONFIG_ENCODERS/DECODERS
libavcodec
|
| Thu, 14 Sep 2006 22:13:23 +0000 |
aurel |
disable vp3 mmx idct for theora files to avoid artifacts
libavcodec
|
| Thu, 14 Sep 2006 00:38:03 +0000 |
diego |
Remove the LGPL exception clause as discussed on ffmpeg-devel
libavcodec
|
| Tue, 12 Sep 2006 20:58:17 +0000 |
aurel |
Enables back the mmx/sse optimized version of the vp3 idct.
libavcodec
|
| Sat, 02 Sep 2006 23:10:28 +0000 |
mru |
remove redundant declarations
libavcodec
|
| Mon, 28 Aug 2006 09:33:01 +0000 |
lorenm |
mmx implementation of deblocking strength decision.
libavcodec
|
| Sat, 12 Aug 2006 16:37:31 +0000 |
gpoirier |
Support for MacIntel, last part: balign directives
libavcodec
|
| Fri, 11 Aug 2006 18:19:37 +0000 |
lorenm |
vorbis simd tweaks
libavcodec
|
| Thu, 10 Aug 2006 20:24:58 +0000 |
michael |
convert vector_fmul_reverse_sse2 and vector_fmul_add_add_sse2 to sse
libavcodec
|
| Thu, 10 Aug 2006 19:06:25 +0000 |
lorenm |
sse and 3dnow implementations of float->int conversion and mdct windowing.
libavcodec
|
| Wed, 09 Aug 2006 06:37:25 +0000 |
lorenm |
emms -> femms
libavcodec
|
| Tue, 08 Aug 2006 12:03:51 +0000 |
michael |
change vorbis_inverse_coupling_sse2() so it works on sse1 cpus
libavcodec
|
| Thu, 03 Aug 2006 07:09:29 +0000 |
lorenm |
10l, vorbis_inverse_coupling_sse() was really 3dnow
libavcodec
|
| Thu, 03 Aug 2006 03:18:47 +0000 |
lorenm |
sse & sse2 implementations of vorbis channel coupling.
libavcodec
|
| Sat, 29 Jul 2006 08:45:33 +0000 |
stefang |
some MMX optimizations for the CAVS decoder
libavcodec
|
| Wed, 19 Jul 2006 09:49:21 +0000 |
michael |
disable the vp3 mmx and sse2 idcts, their output doesnt match the c idct (tested with -f crc) and the theora spec does not allow different idcts not to mention the difference is quite vissible ...
libavcodec
|
| Wed, 05 Jul 2006 19:31:01 +0000 |
mru |
#define SBUTTERFLY outside CONFIG_ENCODERS
libavcodec
|
| Mon, 03 Jul 2006 10:52:07 +0000 |
lucabe |
Move REG_* macros from libavcodec/i386/mmx.h to libavutil/x86_cpu.h
libavcodec
|
| Wed, 05 Apr 2006 04:13:41 +0000 |
lorenm |
gmc_mmx tweaks
libavcodec
|
| Tue, 04 Apr 2006 09:23:45 +0000 |
lorenm |
mmx implementation of 3-point GMC. (5x faster than C)
libavcodec
|
| Thu, 23 Mar 2006 20:16:36 +0000 |
lorenm |
prefetch pixels for future motion compensation. 2-5% faster h264.
libavcodec
|
| Wed, 22 Mar 2006 22:05:00 +0000 |
lorenm |
added mmx implementation of h264_chroma_mc2
libavcodec
|
| Wed, 22 Mar 2006 12:08:35 +0000 |
gpoirier |
add MMX and SSE versions of ff_snow_inner_add_yblock
libavcodec
|
| Tue, 21 Mar 2006 21:51:07 +0000 |
corey |
snow mmx+sse2 optimizations, part 4
libavcodec
|
| Mon, 20 Mar 2006 22:27:59 +0000 |
gpoirier |
Add the mmx and sse2 implementations of ff_snow_vertical_compose().
libavcodec
|
| Tue, 07 Mar 2006 22:45:56 +0000 |
lorenm |
h264_idct8_add_mmx
libavcodec
|
| Tue, 07 Mar 2006 22:33:32 +0000 |
lorenm |
h264_idct_add only needs mmx1
libavcodec
|
| Fri, 10 Feb 2006 06:55:25 +0000 |
lorenm |
h264: special case dc-only idct. ~1% faster overall
libavcodec
|
| Sun, 05 Feb 2006 13:35:17 +0000 |
diego |
MSVC-compatible __align8/__align16 declaration
libavcodec
|
| Thu, 12 Jan 2006 22:43:26 +0000 |
diego |
Update licensing information: The FSF changed postal address.
libavcodec
|
| Thu, 22 Dec 2005 01:10:11 +0000 |
diego |
COSMETICS: tabs --> spaces, some prettyprinting
libavcodec
|
| Sat, 17 Dec 2005 18:14:38 +0000 |
diego |
COSMETICS: Remove all trailing whitespace.
libavcodec
|
| Sat, 12 Nov 2005 05:23:25 +0000 |
lorenm |
use sse16_sse2() in nsse
libavcodec
|
| Thu, 27 Oct 2005 06:45:29 +0000 |
lorenm |
faster h264_chroma_mc8_mmx, added h264_chroma_mc4_mmx.
libavcodec
|
| Sun, 09 Oct 2005 23:38:52 +0000 |
lorenm |
h264 mmx weighted prediction. up to 3% overall speedup.
libavcodec
|
| Fri, 30 Sep 2005 02:31:47 +0000 |
lorenm |
sse2 16x16 sum squared diff (306=>268 cycles on a K8)
libavcodec
|
| Wed, 21 Sep 2005 21:17:09 +0000 |
michael |
replace a few mov + psrlq with pshufw, there are more cases which could benefit from this but they would require us to duplicate some functions ...
libavcodec
|
| Sat, 10 Sep 2005 19:30:40 +0000 |
reimar |
Fix compile without CONFIG_GPL, misplaced #endif caused a missing }.
libavcodec
|
| Sat, 10 Sep 2005 17:01:30 +0000 |
michael |
xvids mmx&mmx2 idcts
libavcodec
|
| Tue, 06 Sep 2005 21:25:35 +0000 |
mru |
Kill some compiler warnings. Compiled code verified identical after changes.
libavcodec
|
| Thu, 02 Jun 2005 20:45:35 +0000 |
lorenm |
sort H.264 mmx dsp functions into their own file
libavcodec
|
| Wed, 01 Jun 2005 21:19:00 +0000 |
michael |
adding a few const
libavcodec
|
| Tue, 31 May 2005 22:48:33 +0000 |
lorenm |
MMX for H.264 iDCT (adapted from x264)
libavcodec
|
| Wed, 25 May 2005 08:05:41 +0000 |
hzoli |
MMX code for (put|avg)_h264_chroma_mc8
libavcodec
|
| Wed, 18 May 2005 09:17:22 +0000 |
lorenm |
H.264 deblocking optimizations (mmx for chroma_bS4 case, convert existing cases to 8-bit math)
libavcodec
|
| Tue, 17 May 2005 18:28:40 +0000 |
michael |
porting the mmx&sse2 (sse2 untested) vp3 idcts to the lavc idct API
libavcodec
|
| Tue, 17 May 2005 09:11:48 +0000 |
michael |
disabling vp3 mmx&mmx2 idcts, they must be ported over to the lavc idct API, ill port the vp3 c idct
libavcodec
|
| Sun, 08 May 2005 18:48:19 +0000 |
michael |
fix PIC
libavcodec
|
| Mon, 25 Apr 2005 01:01:41 +0000 |
lorenm |
MMX for H.264 deblocking filter
libavcodec
|
| Mon, 07 Feb 2005 17:09:48 +0000 |
michael |
optimization and gcc 4.0 bug workaround patch by (Martin Drab >drab kepler.fjfi.cvut cz<)
libavcodec
|
| Mon, 11 Oct 2004 02:19:29 +0000 |
michael |
adapting existing mmx/mmx2/sse/3dnow optimizations so they work on x86_64 patch by (Aurelien Jacobs <aurel at gnuage dot org>)
libavcodec
|
| Sat, 25 Sep 2004 23:18:58 +0000 |
michael |
1/2 resolution decoding
libavcodec
|
| Wed, 08 Sep 2004 18:49:12 +0000 |
michael |
h264_qpel8_hv_lowpass_mmx2/3dnow
libavcodec
|
| Wed, 08 Sep 2004 18:25:45 +0000 |
michael |
h264_qpel4_hv_lowpass_mmx2/3dnow
libavcodec
|
| Wed, 08 Sep 2004 00:52:49 +0000 |
michael |
optimization
libavcodec
|
| Tue, 07 Sep 2004 22:56:50 +0000 |
michael |
optimization
libavcodec
|
| Tue, 07 Sep 2004 17:22:37 +0000 |
michael |
optimization
libavcodec
|
| Tue, 07 Sep 2004 01:48:45 +0000 |
michael |
h264 luma motion compensation in mmx2/3dnow
libavcodec
|
| Mon, 06 Sep 2004 03:17:31 +0000 |
michael |
10000l fix and use more mmx2/3dnow code for mpeg4 qpel which has been written and commited long time ago but appearently never used, qpel motion compensation is 5% faster
libavcodec
|
| Tue, 08 Jun 2004 02:13:44 +0000 |
michael |
sse8 and nsse in mmx
libavcodec
|
| Tue, 18 May 2004 17:09:46 +0000 |
michael |
some of the warning fixes by (Michael Roitzsch <mroi at users dot sourceforge dot net>)
libavcodec
|
| Tue, 27 Apr 2004 04:06:24 +0000 |
melanson |
move the 0x80 vector outside of the function, thus saving the compiler
libavcodec
|
| Tue, 27 Apr 2004 03:58:06 +0000 |
melanson |
separate out put_signed_pixels_clamped() into its own function and
libavcodec
|
| Mon, 26 Apr 2004 00:20:29 +0000 |
melanson |
reorganize and simplify the VP3 IDCT stuff
libavcodec
|
| Sun, 25 Apr 2004 03:33:30 +0000 |
melanson |
hook up support for SSE2-optimized VP3 IDCT
libavcodec
|
| Mon, 08 Mar 2004 02:18:19 +0000 |
melanson |
use optimized VP3 functions where appropriate
libavcodec
|
| Sun, 29 Feb 2004 22:10:18 +0000 |
michael |
attribute used patch by (mitya at school dot ioffe dot ru (Dmitry Baryshkov))
libavcodec
|
| Fri, 06 Feb 2004 02:12:37 +0000 |
michael |
quantizer noise shaping optimization
libavcodec
|
| Sat, 31 Jan 2004 20:41:07 +0000 |
melanson |
vsad16_mmx2 only applies if encoders are turned on
libavcodec
|
| Thu, 29 Jan 2004 01:15:25 +0000 |
michael |
SSE2 fdct by (Balatoni Denes <pnis at coder dot hu>)
libavcodec
|
| Sat, 10 Jan 2004 16:04:55 +0000 |
michael |
copyright year update of the files i touched and remembered, things look annoyingly unmaintained otherwise
libavcodec
|
| Mon, 05 Jan 2004 22:57:07 +0000 |
michael |
interlaced dct decision cleanup
libavcodec
|
| Tue, 30 Dec 2003 16:07:57 +0000 |
michael |
interlaced motion estimation
libavcodec
|
| Sat, 13 Dec 2003 02:14:18 +0000 |
michael |
Here are juste two added #ifdef CONFIG_ENCODERS to allow
libavcodec
|
| Tue, 02 Dec 2003 22:02:57 +0000 |
michael |
h263_h_loop_filter_mmx
libavcodec
|
| Tue, 02 Dec 2003 20:28:10 +0000 |
michael |
h263_v_loop_filter_mmx
libavcodec
|
| Thu, 23 Oct 2003 23:24:38 +0000 |
alex |
1l
libavcodec
|
| Thu, 23 Oct 2003 13:22:20 +0000 |
michael |
minor mmx2 optimization if the dct
libavcodec
|
| Tue, 14 Oct 2003 04:15:53 +0000 |
melanson |
disable encoders where appropriate (patch courtesy of BERO
libavcodec
|
| Mon, 13 Oct 2003 17:27:30 +0000 |
michael |
mmx2 optimization of huffyuv median encoding
libavcodec
|
| Sat, 21 Jun 2003 01:11:31 +0000 |
michaelni |
the return of the idct with 16bit output by ("Ivan Kalvachev" <ivan at cacad dot com>)
libavcodec
|
| Wed, 07 May 2003 23:32:45 +0000 |
michaelni |
more #ifdef CONFIG_ENCODERS by (Wolfgang Hesseler <qv at multimediaware dot com>)
libavcodec
|
| Mon, 14 Apr 2003 09:58:16 +0000 |
michaelni |
10l (hadamard fix)
libavcodec
|
| Wed, 26 Mar 2003 11:29:45 +0000 |
michaelni |
faster hadamard transform
libavcodec
|
| Mon, 10 Mar 2003 11:22:30 +0000 |
kabi |
* oooooops - sorry for this one - wrong logic
libavcodec
|
| Fri, 07 Mar 2003 13:49:10 +0000 |
kabi |
* support FF_MM_FORCE
libavcodec
|
| Mon, 03 Mar 2003 14:54:00 +0000 |
michaelni |
MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
libavcodec
|
| Tue, 11 Feb 2003 16:57:55 +0000 |
kabi |
* prototypes at one place
libavcodec
|
