Commit Graph

49 Commits

Author SHA1 Message Date
Christophe GISQUET
34454c761f SBR DSP x86: implement SSE sbr_sum_square_sse
The 32bits targets have been compiled with -mfpmath=sse for proper reference.
sbr_sum_square C  /32bits: 82c (unrolled)/102c
               C  /64bits: 69c (unrolled)/82c
               SSE/32bits: 42c
               SSE/64bits: 31c

Use of SSE4.1 dpps to perform the final sum is slower.
Not unrolling to perform 8 operations in a loop yields 10 more cycles.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2012-02-23 15:50:06 -08:00
Ronald S. Bultje
7e4d9d5d45 win64: add a XMM clobber test configure option.
This will be useful to test more aggressively for failures to mark XMM
registers as clobbered in Win64 builds, and prevent regressions thereof.

Based on a patch by Ramiro Polla <ramiro.polla@gmail.com>
2012-02-02 12:00:48 -08:00
Christophe Gisquet
e5c9de2ab7 rv40: x86 SIMD for biweight
Provide MMX, SSE2 and SSSE3 versions, with a fast-path when the weights are
multiples of 512 (which is often the case when the values round up nicely).

*_TIMER report for the 16x16 and 8x8 cases:
C:
9015 decicycles in 16, 524257 runs, 31 skips
2656 decicycles in 8, 524271 runs, 17 skips
MMX:
4156 decicycles in 16, 262090 runs, 54 skips
1206 decicycles in 8, 262131 runs, 13 skips
MMX on fast-path:
2760 decicycles in 16, 524222 runs, 66 skips
995 decicycles in 8, 524252 runs, 36 skips
SSE2:
2163 decicycles in 16, 262131 runs, 13 skips
832 decicycles in 8, 262137 runs, 7 skips
SSE2 with fast path:
1783 decicycles in 16, 524276 runs, 12 skips
711 decicycles in 8, 524283 runs, 5 skips
SSSE3:
2117 decicycles in 16, 262136 runs, 8 skips
814 decicycles in 8, 262143 runs, 1 skips
SSSE3 with fast path:
1315 decicycles in 16, 524285 runs, 3 skips
578 decicycles in 8, 524286 runs, 2 skips

This means around a 4% speedup for some sequences.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-01-30 23:58:25 +01:00
Diego Biurrun
91bafb52ae x86: Give RV40 init file a more suitable name. 2012-01-30 23:58:24 +01:00
Ronald S. Bultje
59f474b49d png: convert DSP functions to yasm. 2012-01-29 18:47:50 -08:00
Ronald S. Bultje
e92003514d png: move DSP functions to their own DSP context. 2012-01-29 08:11:18 -08:00
Christophe GISQUET
3faa303a47 rv34: DC-only inverse transform
When decoding coefficients, detect whether the block is DC-only, and take
advantage of this knowledge to perform DC-only inverse transform.

This is achieved by:
- first, changing the 108x4 element modulo_three_table into a 108 element
  table (kind of base4), and accessing each value using mask and shifts.
- then, checking low bits for 0 (as they represent the presence of higher
  frequency coefficients)

Also provide x86 SIMD code for the DC-only inverse transform.

Signed-off-by: Kostya Shishkov <kostya.shishkov@gmail.com>
2012-01-12 09:52:33 +01:00
Vitor Sessak
39df0c434c mpegaudiodec: optimized iMDCT transform
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2012-01-08 17:40:55 -08:00
Diego Biurrun
30bbd5cbc0 x86: conditionally compile dnxhd encoder optimizations 2011-12-19 13:54:10 +01:00
Diego Biurrun
88b9735753 build: conditionally compile x86 H.264 chroma optimizations 2011-12-14 11:58:45 +01:00
Ronald S. Bultje
e3f530feca prores: idct sse2/sse4 optimizations.
~3.0-3.5x as fast as original C version, 1.6x as fast overall.
2011-10-11 07:50:48 -07:00
Kostya Shishkov
d241f51e0f Move RV3/4-specific DSP functions into their own context
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2011-08-11 16:07:15 -07:00
Daniel Kang
9bfa5363da H.264: Add x86 assembly for 10-bit H.264 qpel functions.
Mainly ported from 8-bit H.264 qpel.

Some code ported from x264. LGPL ok by author.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2011-07-03 07:43:38 -07:00
Daniel Kang
84e70ef004 h264: Add x86 assembly for 10-bit weight/biweight H.264 functions.
Mainly ported from 8-bit H.264 weight/biweight.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
2011-06-21 15:24:13 +02:00
Daniel Kang
f188a1e0ca H.264: Add x86 assembly for 10-bit MC Chroma H.264 functions.
Mainly ported from 8-bit H.264 MC Chroma.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2011-06-18 07:52:19 -04:00
Daniel Kang
a8d44f9dd5 Add x86 assembly for some 10-bit H.264 intra predict functions.
Parts are inspired from the 8-bit H.264 predict code in Libav.
Other parts ported from x264 with relicensing permission from author.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
2011-06-06 01:31:02 +02:00
Daniel Kang
836f47d34b Add IDCT functions for 10-bit H.264.
Ports the majority of IDCT functions for 10-bit H.264.

Parts are inspired from 8-bit IDCT code in Libav; other parts ported from x264 with relicensing permission from author.

Signed-off-by: Ronald S. Bultje <rbultje@google.com>
2011-05-31 15:02:32 -07:00
Vitor Sessak
3758eb0eb9 dct32: port SSE 32-point DCT to YASM 2011-05-21 17:42:26 +02:00
Mans Rullgard
0b5e44ed29 mpegaudiodsp: fix x86 and ppc makefiles
Signed-off-by: Mans Rullgard <mans@mansr.com>
2011-05-19 16:32:24 +01:00
Jason Garrett-Glaser
9f3d6ca4f1 Port x86 10-bit H.264 deblock asm from x264 2011-05-10 20:02:15 -07:00
Mans Rullgard
a5444fee06 Add CONFIG_AC3DSP symbol to simplify makefiles
Signed-off-by: Mans Rullgard <mans@mansr.com>
2011-03-12 11:35:26 +00:00
Justin Ruggles
dda3f0ef48 Add x86-optimized versions of exponent_min().
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2011-02-10 15:32:47 -05:00
Justin Ruggles
c73d99e672 Separate format conversion DSP functions from DSPContext.
This will be beneficial for use with the audio conversion API without
requiring it to depend on all of dsputil.

Signed-off-by: Mans Rullgard <mans@mansr.com>
2011-02-02 02:44:53 +00:00
Ronald S. Bultje
d0acc2d2e9 Move sse16_sse2() from inline asm to yasm. It is one of the functions causing
Win64/FATE issues.

Originally committed as revision 25136 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-17 01:44:17 +00:00
Ronald S. Bultje
1d16a1cf99 Rename h264_idct_sse2.asm to h264_idct.asm; move inline IDCT asm from
h264dsp_mmx.c to h264_idct.asm (as yasm code). Because the loops are now
coded in asm instead of C, this is (depending on the function) up to 50%
faster for cases where gcc didn't do a great job at looping.

Since h264_idct_add8() is now faster than the manual loop setup in h264.c,
in-asm idct calling can now be enabled for chroma as well (see r16207). For
MMX, this is 5% faster. For SSE2 (which isn't done for chroma if h264.c does
the looping), this makes it up to 50% faster. Speed gain overall is ~0.5-1.0%.

Originally committed as revision 25119 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-14 13:36:26 +00:00
Jason Garrett-Glaser
8acb554aff LGPL SSE2 H.264 iDCT
This leaves no more GPL-only H.264 decoding asm code.

Approved by Loren.

Originally committed as revision 25092 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-10 02:25:12 +00:00
Stefano Sabatini
c6c98d0897 Move mm_support() from libavcodec to libavutil, make it a public
function and rename it to av_get_cpu_flags().

Originally committed as revision 25076 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-08 15:07:14 +00:00
Ronald S. Bultje
2c166c3af1 Port latest x264 deblock asm (before they moved to using NV12 as internal
format), LGPL'ed with permission from Jason and Loren. This includes mmx2
code, so remove inline asm from h264dsp_mmx.c accordingly.

Originally committed as revision 25031 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-03 16:52:46 +00:00
Ronald S. Bultje
a33a2562c1 Rename h264_weight_sse2.asm to h264_weight.asm; add 16x8/8x16/8x4 non-square
biweight code to sse2/ssse3; add sse2 weight code; and use that same code to
create mmx2 functions also, so that the inline asm in h264dsp_mmx.c can be
removed. OK'ed by Jason on IRC.

Originally committed as revision 25019 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-01 20:56:16 +00:00
Ronald S. Bultje
14bc1f2485 Split h264dsp_mmx.c (which was #included in dsputil_mmx.c) in h264_qpel_mmx.c,
still #included in dsputil_mmx.c and is part of DSPContext, and h264dsp_mmx.c,
which represents H264DSPContext and is now compiled on its own.

Originally committed as revision 25018 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-09-01 20:48:59 +00:00
Ronald S. Bultje
5929b3a651 Fix vertical align.
Originally committed as revision 25009 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-31 12:32:24 +00:00
Ronald S. Bultje
de1c253bab Split intra prediction initialization (i.e. assigning of function pointers)
into its own file, it doesn't belong in h264dsp_mmx.c (much less so in
dsputil_mmx.c).

Originally committed as revision 24990 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-30 16:34:13 +00:00
Ronald S. Bultje
d0eb5a1174 Move H264 chroma MC from inline asm to yasm. This fixes VP3/5/6 and VC-1
fate failures on Win64.

Originally committed as revision 24989 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-30 16:31:04 +00:00
Ronald S. Bultje
e9f5f020c6 Move VP3 IDCT functions from inline ASM to YASM. This fixes part of the VP3/5/6
issues on Win64.

Originally committed as revision 24988 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-30 16:25:46 +00:00
Ronald S. Bultje
89fa3504ed Move vp6_filter_diag4() x86 SIMD code from inline ASM to YASM. This should
help in fixing the Win64 fate failures.

Originally committed as revision 24922 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-25 13:44:16 +00:00
Ronald S. Bultje
3a0885146c Move vp6_filter_diag4() from DSPContext to VP56DSPContext.
Originally committed as revision 24921 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-25 13:42:28 +00:00
Jason Garrett-Glaser
4a384de5b8 Split h264dsp and h264pred in configure.
Many H.264 derivatives, like RV40 and VP8, use the H.264 prediction functions
but not the weight/loopfilter functions.
This should reduce the size of builds with one of these derivatives but without
H.264 decoding itself.

Originally committed as revision 24741 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-07 23:10:25 +00:00
Eli Friedman
c12d6955e2 H.264: SSE2/SSSE3 weighted prediction asm
Patch by Eli Friedman <eli.friedman at gmail dot com>

Originally committed as revision 24702 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-08-05 00:13:38 +00:00
Vitor Sessak
de4bc44abb Convert deinterlacing MMX code to YASM
Originally committed as revision 24615 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-07-31 14:50:51 +00:00
Loren Merritt
c7b1d9768c relicense h264 deblock sse2 to lgpl
Originally committed as revision 24408 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-07-22 00:39:49 +00:00
David Conrad
faa26db28b MMX/SSE VC1 loop filter
Originally committed as revision 24208 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-07-11 22:53:01 +00:00
Vitor Sessak
881fd7a62f Move SSE optimized 32-point DCT to its own file. Should fix breakage with YASM
disabled.

Originally committed as revision 24078 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-07-06 17:48:23 +00:00
Jason Garrett-Glaser
0178d14fe5 First shot at VP8 optimizations:
- MMXEXT, SSE2 and SSSE3 MC functions
- MMX and SSE4 IDCT dc_add functions

Patch by Jason Garrett-Glaser <darkshikari gmail com> and myself.

Originally committed as revision 23815 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-06-27 02:01:45 +00:00
Måns Rullgård
0912db0206 Make vp8 select h264dsp and use this to pull in mmx intrapred
Originally committed as revision 23790 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-06-25 19:10:08 +00:00
Carl Eugen Hoyos
0c59074868 Fix compilation without --enable-gpl.
Originally committed as revision 23789 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-06-25 19:06:29 +00:00
Jason Garrett-Glaser
4af8cdfc3f 16x16 and 8x8c x86 SIMD intra pred functions for VP8 and H.264
Originally committed as revision 23783 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-06-25 18:25:49 +00:00
Vitor Sessak
bc2b368215 SSE-optimized MP3 floating point windowing functions
Originally committed as revision 23750 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-06-24 07:44:50 +00:00
Måns Rullgård
3bd74e9243 Simplify arch-specific object file lists
Originally committed as revision 22570 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-03-16 21:23:03 +00:00
Måns Rullgård
43f60eba19 Move arch-specific makefile parts into $arch/Makefile
Originally committed as revision 22569 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-03-16 21:22:59 +00:00