275131628d
Remove duplicated dot product code. Use dsputil's
...
scalarproduct instead. Patch by Aurelien Jacobs.
Originally committed as revision 16391 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-29 17:14:30 +00:00
592beefe99
It turned out that RV30 uses motion vectors for forward motion B-frame
...
macroblocks to predict motion vectors for backward motion B-frame macroblocks
and vice versa.
Originally committed as revision 16381 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-29 10:20:03 +00:00
e4f1ec3a4c
Explain the lack of +3/7
...
Originally committed as revision 16380 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-28 23:26:12 +00:00
6735e2c810
Revert r16375: It broke aac decoding.
...
Originally committed as revision 16376 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-28 20:08:39 +00:00
69b38d4dc3
Reduce number of warnings when compiling with icc by two.
...
Originally committed as revision 16375 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-28 19:48:43 +00:00
3a095bee2b
Silence one warning when compiling with icc:
...
warning #188 : enumerated type mixed with another type
if((q->bitrate = determine_bitrate(avctx, buf_size, &buf)) == I_F_Q)
^
Originally committed as revision 16374 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-28 19:44:47 +00:00
2c67c65963
Fix h264 decoding on SSE2 cores with icc compilation.
...
Originally committed as revision 16373 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-28 19:40:13 +00:00
23ae2b68b0
Replace UNALIGNED_STORES_ARE_BAD with !HAVE_FAST_UNALIGNED
...
Originally committed as revision 16364 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 23:12:00 +00:00
edf7c2b26f
simple_idct: use MUL16 and MAC16 from mathops.h
...
Originally committed as revision 16362 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 16:25:36 +00:00
79cb09b2f7
consistency cosmetics: Rename POWERPC identifiers to PPC.
...
Originally committed as revision 16359 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 11:33:26 +00:00
d89eae6f1a
Remove stray code
...
Originally committed as revision 16358 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 11:30:30 +00:00
a6b4448cdf
Cleanup _t types in libavcodec/ppc
...
Originally committed as revision 16357 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 11:21:28 +00:00
b870027a19
Bump minor number for many small improvements, tiff fax decoding,
...
AVCodecContext refs is set during decoding.
Originally committed as revision 16355 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-27 09:22:21 +00:00
f23740d96f
ARM: NEON optimised float_to_int16
...
Originally committed as revision 16352 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 19:52:52 +00:00
92799d3213
Enable faxcompr.
...
Originally committed as revision 16349 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 18:42:26 +00:00
a56f82eaf8
indent
...
Originally committed as revision 16348 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 18:39:24 +00:00
537bf4266f
Simplify vertical mode
...
Originally committed as revision 16347 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 18:39:07 +00:00
eb0094b224
And another sechole ...
...
Originally committed as revision 16346 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 18:14:59 +00:00
268309750b
Fix yet another sechole.
...
Originally committed as revision 16345 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 18:06:29 +00:00
11c5999e20
Fix another sechole.
...
Originally committed as revision 16344 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 17:51:10 +00:00
2a522579aa
Minor simplification.
...
Originally committed as revision 16343 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 17:32:11 +00:00
f78a679177
Remove unused run_start variable.
...
Originally committed as revision 16342 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 17:11:19 +00:00
6f5c00eb9f
Move sync line pointers code to the end of the loop, it is simpler that way.
...
Originally committed as revision 16341 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 17:04:31 +00:00
e3a54b6694
Get rid of check for condition that is always true (run_off < avctx->width).
...
Originally committed as revision 16340 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:59:10 +00:00
43ba8f3545
Get rid of pix_left in the 2d code, it is simpler that way.
...
Originally committed as revision 16339 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:57:35 +00:00
a3636fa033
COSMETICS Part 3 and final of Kenan Gillet's QCELP silence handling patch.
...
Originally committed as revision 16338 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:50:21 +00:00
4f54fb54e3
Part 2 of Kenan Gillet's QCELP silence handling patch.
...
Originally committed as revision 16337 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:49:06 +00:00
6fe68c7b97
Factorize "avctx->width + 2" out to avoid someone mistakenly changing
...
some but not all.
Originally committed as revision 16336 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:38:55 +00:00
621e6525a6
cosmetic (moving pix_left update)
...
Originally committed as revision 16335 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:30:50 +00:00
1d7985072c
Close gaping sechole. That is, a series of run=0 allows arbitrary data to
...
be written over the end of the runs array.
Originally committed as revision 16334 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 16:29:01 +00:00
d02ab57a5f
minor simplification
...
Originally committed as revision 16333 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 15:21:48 +00:00
66909bc4f1
Merge ff_ccitt_unpack_1d and 2d.
...
Originally committed as revision 16332 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 15:05:12 +00:00
d68542f019
Ensure that an invalid run terminates the line.
...
Originally committed as revision 16331 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 14:57:42 +00:00
ec5bdf4989
10l != vs. == (yes, my fault not kostya's).
...
Originally committed as revision 16330 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 14:39:21 +00:00
679f3fa901
indent
...
Originally committed as revision 16329 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:52:07 +00:00
68d48cc592
Simplify zero run handling in put_line().
...
Originally committed as revision 16328 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:51:52 +00:00
bc0f7a9c75
Avoid negation in put_line().
...
Originally committed as revision 16327 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:50:00 +00:00
4e60b64324
Factorize mode switching in put_line().
...
Originally committed as revision 16326 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:48:09 +00:00
eb12536a31
Fix (rewrite) find_group3_syncmarker().
...
Originally committed as revision 16325 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:46:11 +00:00
8d48db314e
CCITT Fax Group 3 and 4 decompression by kostya.
...
Will be activated after a little more reviewing and fixing by me.
Commited now, so as to not lose history of that.
Originally committed as revision 16323 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 13:39:19 +00:00
b8d62672f6
ASAO Nelly Moser implementation moved to a better place on mphq.
...
Originally committed as revision 16321 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 11:00:30 +00:00
5163c0d9a2
s/FFmpeg-devel/ffmpeg-devel/
...
Originally committed as revision 16314 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 08:03:26 +00:00
c1fc70362f
Fix compilation without optimization under 64-bit with x264 deblock asm enabled.
...
Originally committed as revision 16313 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-26 00:19:08 +00:00
760badc1df
ARM: add new h264 idct functions
...
Originally committed as revision 16312 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 23:13:43 +00:00
337e3fd990
Disable usage of ff_h264_idct_add_altivec since AltiVec versions of h264_idct_add16,
...
h264_idct_add16intra, h264_idct_add8 need to be implemented.
Add C version of ff_h264_idct8_dc_add in AltiVec so that ff_h264_idct8_add_altivec
can be used.
Originally committed as revision 16311 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 18:27:49 +00:00
0fa5f24c3b
Change mplayerhq.hu references to ffmpeg.org where appropriate.
...
Originally committed as revision 16310 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 15:49:50 +00:00
ddaf298cee
Fix sample URL.
...
Originally committed as revision 16309 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 15:49:02 +00:00
bb0bfcd5ac
The VC-1/WMV3 decoders depend on ENABLE_MSMPEG4_DECODER.
...
patch by avcoder, ffmpeg gmail com
Originally committed as revision 16308 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 13:26:18 +00:00
1cca8d2423
flatten an array, since gcc fails at optimizing multidimensional arrays
...
h264_idct8_add_c: 780 -> 735 cycles on conroe
Originally committed as revision 16307 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-25 01:20:37 +00:00
802e91466d
export refs
...
Originally committed as revision 16306 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-24 18:39:23 +00:00
53d55e9ea3
Allow decoders to set AVCodecContext.refs.
...
Originally committed as revision 16305 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-24 17:38:27 +00:00
62ea19c061
Optimize pred_pskip_motion()
...
7 cpu cycles faster on pentium dual.
Originally committed as revision 16304 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-24 03:23:38 +00:00
ebfb0449d3
Indent
...
Originally committed as revision 16302 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 22:06:16 +00:00
05d3fd44d7
Integrate get_te0_golomb() calls into the code, this allows some checks
...
to be avoided and the function is pretty small.
3% speedup, though this is probably due to changed inlining and not directly
this change.
Originally committed as revision 16301 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 22:04:34 +00:00
9963b332d1
Use get_ue_golomb_31() where possible, almost all are just in headers
...
though.
Originally committed as revision 16300 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 21:12:19 +00:00
0be514625a
Update golomb decoding tables so that get_*_golomb_31() return some
...
larger values for too large values.
Originally committed as revision 16299 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 21:07:36 +00:00
0e921e97c1
Add get_ue_golomb_31()
...
Originally committed as revision 16298 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 21:05:47 +00:00
2f2893ce9b
indent
...
Originally committed as revision 16296 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 19:12:50 +00:00
d42fc4a8ca
Use the new VLC table for the first non trailing coeff too.
...
Sadly only 5 cycles faster here on pentium dual. So maybe the
complexity is not worth it and this should be reverted ...
Originally committed as revision 16295 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 19:10:46 +00:00
593af7cdda
Optimize esc removal code.
...
Originally committed as revision 16294 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 18:31:44 +00:00
ffedb5c19a
352l: correct calculating number of bits for storing macroblock offset in RV3/4
...
Originally committed as revision 16293 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 18:27:06 +00:00
2d76bf391a
Indent
...
Originally committed as revision 16292 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 17:50:36 +00:00
8140955d39
unified CAVLC level decoding LUT.
...
Quite a bit faster (HPCVMOLQ_BRCM_B.264 was 3% faster here)
Originally committed as revision 16291 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 17:49:38 +00:00
5cb5023c78
Make h264 parser 50% faster.
...
Originally committed as revision 16286 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 02:34:58 +00:00
abb27cfb24
100l, I broke H.264 again, forgot one hunk.
...
Thanks to FATE for finding it.
Originally committed as revision 16285 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 01:11:56 +00:00
e08715d391
Optimize 0 0 0-3 search, 45% faster on pentium dual.
...
Originally committed as revision 16284 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-23 00:38:45 +00:00
11c23b64ac
Revert r16257:
...
Include "libavutil/common.h" where we use llrint() in case ffmpeg's
own llrint() is to be used.
Originally committed as revision 16282 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 22:39:55 +00:00
ec3686e889
Simplify decode_cabac_mb_ref() a little bit, 2 cpu cycles faster on
...
pentium dual.
Originally committed as revision 16279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 17:14:13 +00:00
26695973c7
Indent
...
Originally committed as revision 16278 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 16:56:05 +00:00
b68a455313
inline decode_cabac_mb_type for I & P frames, 9 cycles faster on pentium dual.
...
Originally committed as revision 16277 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 16:46:17 +00:00
1952ac3713
Negate 2 more variables, 1 cpu cycle faster on pentium dual.
...
Originally committed as revision 16276 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 16:20:13 +00:00
03a035e059
Simplify if/else, no speed change
...
Originally committed as revision 16275 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 16:14:06 +00:00
6f3c50f2f9
Negate a few variables, this simplifies the code and makes it 5 cycles faster
...
on pentium dual.
Originally committed as revision 16274 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 16:10:35 +00:00
60c6ba7aea
Simplify ifs(), 8 cpu cycles faster on pentium dual
...
Originally committed as revision 16273 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 15:54:27 +00:00
127a20e3b8
Simplify if(), 3 cpu cycles faster in pentium dual.
...
Originally committed as revision 16272 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 15:53:00 +00:00
a6493a8fbd
Rename libavcodec/i386/ --> libavcodec/x86/.
...
It contains optimizations that are not specific to i386 and
libavutil uses this naming scheme already.
Originally committed as revision 16270 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 09:12:42 +00:00
108c96d5e7
Decode TIFF image only after all tags have been decoded
...
Originally committed as revision 16268 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 06:50:18 +00:00
8eedffde31
Some TIFFs declare -1 for the single strip height
...
Originally committed as revision 16267 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 06:47:51 +00:00
89b4ec5c33
Add monochrome TIFF support
...
Originally committed as revision 16266 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 06:43:14 +00:00
4386ed6922
Calculate line size variable correctly for lower bitdepths and use it for raw data copying
...
Originally committed as revision 16265 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 06:39:31 +00:00
4a2f51de34
Register TIFF tags for CCITT Group 3 and 4 compression options
...
Originally committed as revision 16264 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 06:36:10 +00:00
bef05f05e4
Remove a bunch of unused variables.
...
Originally committed as revision 16263 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 00:10:36 +00:00
ebb160a0b0
Use ARCH_X86_32 instead of !ARCH_X86_64, it is more straightforward.
...
Originally committed as revision 16262 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 00:03:30 +00:00
64bfc5845b
Fix incorrectly constructed Dirac parse units that caused A/V sync loss.
...
Fixes issue 694.
patch by Anuradha Suraparaju, anuradha rd.bbc.co uk
Originally committed as revision 16261 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-22 00:01:39 +00:00
c4ff7c53e0
Do not use full include path for headers in the same directory.
...
Originally committed as revision 16260 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 23:56:55 +00:00
c96bd21227
Change mpeg audio parser so it only sets frame_size, channels and bit_rate
...
after it has a few valid frames.
Fixes issue762
Originally committed as revision 16259 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 23:50:16 +00:00
1af34f094e
Include "libavutil/common.h" where we use llrint() in case ffmpeg's
...
own llrint() is to be used.
Originally committed as revision 16257 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 21:53:42 +00:00
ed4c638a07
Fix PTS returned by the demuxers for mpeg4 without a decoder.
...
More precissely fix has_b_frames.
Fixes issue757
Originally committed as revision 16256 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 21:47:29 +00:00
c62da7972b
Move CODEC_FLAG_LOW_DELAY into ff_mpeg4_decode_picture_header().
...
Originally committed as revision 16255 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 21:30:36 +00:00
befc8fe086
Remove useless code.
...
Originally committed as revision 16253 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 17:49:40 +00:00
c212fb0cb1
Only execute clear_blocks() when needed.
...
+0.3% speedup for both aladin & cathedral.
Originally committed as revision 16252 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 15:58:42 +00:00
897c1703d0
AAC: Reindent after last commit
...
Originally committed as revision 16251 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 05:32:55 +00:00
aab5413323
AAC: Fix AAC prediction when used in conjunction with the CPE
...
Patch by Alex Converse (alex converse gmail com)
Originally committed as revision 16250 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 05:28:14 +00:00
99665a21f4
AAC: Don't write an illegal sampling_index in a PCE to the MPEG4AudioConfig
...
member
Patch by Alex Converse (alex converse gmail com)
Originally committed as revision 16249 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-21 05:08:32 +00:00
66c07ca96f
Optimize get_dct8x8_allowed().
...
30 cpu cycles faster on pentium dual.
Originally committed as revision 16248 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-20 23:11:30 +00:00
6beb8b2612
RV30 decoder
...
Originally committed as revision 16247 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-20 18:06:01 +00:00
541e8f3bec
RV30 loop filter
...
Originally committed as revision 16245 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-20 06:09:54 +00:00
3f16ed15df
Cosmetics by Kenan Gillet. Part 1 of 3 of his 'qcelp: silence handling'
...
changeset.
Originally committed as revision 16244 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-20 02:38:34 +00:00
64e74b68a4
Put under #if LIBAVCODEC_VERSION_MAJOR < 53 the deprecated functions
...
av_set_string() and av_set_string2(), they will be dropped at the next
major bump.
Originally committed as revision 16243 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 22:37:56 +00:00
f16dd7e65a
Remove calls to deprecated av_set_string2() with calls to
...
av_set_string3().
Originally committed as revision 16241 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 22:29:30 +00:00
aac8b76983
H.264 loopfilter speed tweaks
...
Originally committed as revision 16240 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 14:49:17 +00:00
a22eff36f9
Port x264 deblocking code to libavcodec.
...
This includes SSE2 luma deblocking code and both MMXEXT and SSE2 luma
intra deblocking code for H.264 decoding. This assembly is available
under --enable-gpl and speeds decoding of Cathedral by 7%.
Originally committed as revision 16239 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 13:45:13 +00:00
6a1846eb4a
Document new idct requirements.
...
Originally committed as revision 16237 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 09:34:18 +00:00
a5805aa9d1
Fix decoding with the plain C idcts of
...
FRExt/HPCAMOLQ_BRCM_B
FRExt/HPCAQ2LQ_BRCM_B
FRExt/HPCVMOLQ_BRCM_B
Originally committed as revision 16236 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 09:21:54 +00:00
bbc888d0c5
a table for upcoming RV30 loop filtering
...
Originally committed as revision 16235 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 06:29:35 +00:00
40c7d0ae3c
Add automatic prefix handling to yasm functions. Does nothing now, but will
...
be useful for porting x264 asm in the future.
Originally committed as revision 16234 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 03:01:08 +00:00
a5b807a6c1
Replace /2 by >>1 in decode_cabac_mb_dqp()
...
3 cpu cycles speed up on pentium dual.
Originally committed as revision 16233 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 02:28:51 +00:00
1aea5d35e5
Simplify ctx update in decode_cabac_mb_dqp().
...
no speed change
Originally committed as revision 16232 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 02:00:33 +00:00
7cfca0dfd8
Simplify ctx calculation in decode_cabac_mb_dqp()
...
no speed change
Originally committed as revision 16231 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 01:50:20 +00:00
d43696309a
Clear FF_INPUT_BUFFER_PADDING_SIZE bytes at the end of NALs in rbsp_buffer.
...
Fixes valgrind uninitialized value warnings at the end of decoding H.264
frames.
Originally committed as revision 16230 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 01:11:52 +00:00
712ca84c21
Move filter_luma_intra into dsputil for later addition of asm.
...
Originally committed as revision 16228 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 00:44:51 +00:00
b9fe706305
Simplify chroma AC in CABAC residual decoding.
...
Originally committed as revision 16227 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 00:43:30 +00:00
8955b66950
Optimize ctx calculation in decode_cabac_mb_mvd(), code by dark shikari.
...
The case for 16x16 blocks becomes 10 cpu cycles faster on pentium dual,
i could not find a speed difference in the case of subblocks though.
Originally committed as revision 16226 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-19 00:05:39 +00:00
17779f39b6
Remove unacceptable NULL pointer hack from mc code.
...
Originally committed as revision 16225 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 23:52:32 +00:00
04618b98e3
Check ref values in CABAC H.264 for validity.
...
Originally committed as revision 16224 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 23:31:10 +00:00
c25ac15a07
Move idct_(dc)add closer to where it is needed.
...
Originally committed as revision 16223 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 18:25:11 +00:00
aebb5d6d96
indent
...
Originally committed as revision 16222 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 18:14:38 +00:00
96465b90a1
Reorder ifs in chroma hl_decode_mb to avoid a duplicate transform_bypass
...
check.
14 cpu cycles speedup on Pentium Dual
Originally committed as revision 16221 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 18:12:59 +00:00
6456d6d87c
s/h->cbp_table[mb_xy]/h->cbp/
...
Originally committed as revision 16220 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 17:42:44 +00:00
04824298a9
Faster CAVLC decoding of trailing_ones. Based on a patch by dark shikari.
...
decode_residual is about 3.3% faster.
Originally committed as revision 16219 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 16:11:19 +00:00
93445d1617
Replace i by trailing_ones, part of a patch by dark shikari.
...
No speed change meassureable with START/STOP_TIMER, but this is needed
for future optimizations.
Originally committed as revision 16218 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 15:51:32 +00:00
c375d87f6f
Remove if() surrounding decode_cabac_mb_type() that can never be true.
...
Originally committed as revision 16217 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 15:32:07 +00:00
c325b5054f
Remove unreachable else clause, found by dark shikari.
...
Originally committed as revision 16216 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 15:21:15 +00:00
dae006d7d7
Remove useless IS_8x8DCT check i forgot, spotted by dark shikari.
...
Originally committed as revision 16215 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 11:17:12 +00:00
2a5a9c2815
10l, svq3 didnt set h->cbp, this broke decoding a little.
...
Originally committed as revision 16214 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 11:01:38 +00:00
efeb298ea0
cosmetics: remove one unneeded space
...
Originally committed as revision 16211 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 06:00:12 +00:00
1eb960352b
Do not calculate idct_dc_add/idct_add when the variables are unused.
...
Originally committed as revision 16210 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 04:13:02 +00:00
62bc966f8f
Remove redundant nnz variable.
...
Originally committed as revision 16209 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 03:04:53 +00:00
0a8ca22f4e
indent
...
Originally committed as revision 16208 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 02:56:44 +00:00
2fd1f0e026
Use the new idct functions (except chroma as it was slower in benchmarks)
...
cathedral +0.5% speed
aladin +0.6% speed [note aladin has been cat-ed 10 times to reduce the influence
of init time]
Speedup also verified via START/STOP_TIMER (difference was very significant
for the changed parts)
Originally committed as revision 16207 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 02:53:18 +00:00
ac22385931
H.264 idct functions that include the chroma, inter luma and intra16 luma loops
...
thus avoiding the calling overhead.
New functions are not yet used.
Originally committed as revision 16206 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 02:36:48 +00:00
49c084a745
Skip non intra luma code when there is no coded luma.
...
0.7% speedup for the cathedral sample.
Originally committed as revision 16203 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 01:10:51 +00:00
00794ed0d9
1.6x faster pred*_vertical_add_c
...
Originally committed as revision 16202 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 01:05:06 +00:00
621561cdf3
Skip chroma handling when there is no coded chroma.
...
0.5% overall speedup for the cathedral sample.
Originally committed as revision 16201 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-18 00:46:54 +00:00
17b17c53f4
add const qualifier to some pointers for input data
...
Originally committed as revision 16195 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 20:17:07 +00:00
bb39171beb
use precalculated mb_pos where available
...
Originally committed as revision 16194 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 20:10:15 +00:00
6c3fca6479
Merge deblocking pattern with CBP for RV30/40 loop filtering
...
Originally committed as revision 16192 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 20:00:48 +00:00
26bc319c1d
Generate correct deblock pattern for RV30
...
Originally committed as revision 16191 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 19:39:56 +00:00
f610fc81c1
Unfortunately RV30 luma thirdpel MC functions cannot be substituted with
...
applying two separate filtering passes, so implement them explicitly.
Originally committed as revision 16190 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 19:37:45 +00:00
6a4583e925
4l: TIFF stores short strings inside tag, do not interpret it is as an offset
...
Fixes issue 753
Originally committed as revision 16189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 17:39:35 +00:00
5f51afd602
add AltiVec implementation of int32_to_float_fmul_scalar
...
Originally committed as revision 16186 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 09:47:06 +00:00
54dab66174
Fix svq3 decoding, is_complex was not initialized.
...
Originally committed as revision 16184 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 03:41:48 +00:00
4080e67c8e
Replace != 0 || check by |
...
3 cpu cycles faster
Originally committed as revision 16183 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 02:53:03 +00:00
ad9ca7e720
Split filter_mb_dir() out of filter_mb().
...
1% overall decoding speed up for cathedral-beta2-400extra-crop-avc.mp4
no speed change for Aladin.mpg
Benchmarks done on Pentium dual
Originally committed as revision 16182 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 02:35:14 +00:00
6c414bb6d4
AC3: fix strict aliasing violation in parser
...
Originally committed as revision 16181 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 02:30:26 +00:00
cb56b44086
AAC: fix strict aliasing violation in parser
...
Originally committed as revision 16180 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 02:30:22 +00:00
a2fc0f6a6d
ARM: replace "armv4l" with "arm"
...
Originally committed as revision 16179 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 00:54:54 +00:00
2600f8c86d
Remove obsolete and misleading comments.
...
Originally committed as revision 16178 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-17 00:39:45 +00:00
ac0623b23c
Fix indention, also do a little vertical alignment of changed lines.
...
Originally committed as revision 16176 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 22:16:11 +00:00
7b1454c439
Implement av_set_string3().
...
Originally committed as revision 16175 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 21:35:35 +00:00
6120a343aa
Factorize 3 multiplications out, code becomes 3 cpu cycles faster.
...
(not significant as thats just per MB)
Originally committed as revision 16174 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 21:08:16 +00:00
1dd488e955
Move ENABLE_SMALL back to the per MB check, as otherwise gcc wont remove
...
the code.
Originally committed as revision 16173 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 20:43:39 +00:00
89db0bae26
Do not recalculate constant part of is_complex for every MB.
...
Originally committed as revision 16172 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 20:26:22 +00:00
41e4055b63
Support the new lossless mode.
...
Originally committed as revision 16171 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 20:09:09 +00:00
479044cea5
dct-test: add ARM IDCTs
...
Originally committed as revision 16170 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 20:00:07 +00:00
875f31259b
dct-test: add PARTTRANS permutation
...
Originally committed as revision 16169 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 19:59:50 +00:00
d358caed7e
Horizontal and vertical prediction functions needed for new lossless mode
...
of H.264.
Originally committed as revision 16168 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 19:39:01 +00:00
f683ea1cab
Add const qualifiers to decode functions.
...
Originally committed as revision 16165 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 08:02:05 +00:00
7d2cc552ed
Fix warning "format not a string literal and no format arguments".
...
Originally committed as revision 16164 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 07:56:33 +00:00
2169b79c2a
Workaround for shared libs on OS X with Apple gcc 4.0.1
...
Originally committed as revision 16158 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-16 00:47:59 +00:00
ac52613c16
Check extradata is large enough.
...
Patch by Laurent Aimar fenrir \:/ via ecp fr
Originally committed as revision 16157 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 23:01:54 +00:00
301cc4f370
Implement the av_set_number2() internal function, which makes possible
...
to distinguish between a not found option failure and a not valid
value failure.
Originally committed as revision 16156 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:48:10 +00:00
a59d44fdf4
Remove ending period in a log message, which was inconsistent with
...
most other messages.
Originally committed as revision 16155 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:40:28 +00:00
b24298badc
Switch xvmc over to a CONFIG_ option.
...
Originally committed as revision 16153 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:22:51 +00:00
6915cf3854
whitespace cosmetics
...
Originally committed as revision 16152 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:14:58 +00:00
1bf98d19d5
ARM: NEON optimised h264_idct_dc_add
...
Originally committed as revision 16151 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:54 +00:00
c598cf25f4
ARM: NEON optimised h264_idct_add
...
Originally committed as revision 16150 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:51 +00:00
5813e05d08
ARM: NEON optimised H.264 8x8 and 16x16 qpel MC
...
Originally committed as revision 16149 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:47 +00:00
ad74a0f8b3
ARM: NEON optimised H.264 loop filter
...
Originally committed as revision 16148 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:44 +00:00
1cce897aad
ARM: NEON optimised {put,avg}_h264_chroma_mc[48]
...
Originally committed as revision 16147 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:41 +00:00
7eec43289a
ARM: NEON optimised simple_idct
...
Originally committed as revision 16146 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:37 +00:00
569f5a756a
ARM: NEON optimised put_pixels functions
...
Originally committed as revision 16145 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 22:12:32 +00:00
c4687bf607
Avoid invasion of POSIX-reserved _t namespace.
...
Originally committed as revision 16144 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 21:02:17 +00:00
40a19c4434
Fix one peculiar case in Smacker decoding.
...
Patch by Daniel Verkamp ($firstname <at> drv <dot> nu)
Thread [PATCH] Smacker video decoder bug fix
Originally committed as revision 16143 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 17:51:10 +00:00
9506090df9
Avoid POSIX-reserved _t in identifier names.
...
Originally committed as revision 16142 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 10:08:31 +00:00
40f0624121
Enable filter_mb_fast if CODEC_FLAG2_FAST is set.
...
Originally committed as revision 16141 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 01:24:04 +00:00
bf73801bec
vp56: don't reset dimensions to 0 in codec init
...
Originally committed as revision 16140 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-15 00:00:16 +00:00
d0a01fe84b
Move documentation of avcodec_register_all() from allcodecs.c to
...
avcodec.h.
Originally committed as revision 16135 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 21:38:03 +00:00
54c441c860
Remove set_all_opt() and its use in av_set_string2(). Make
...
av_set_string() and av_set_string2() do not accept as parameter the
name of a named constant.
This avoids av_set_string2() and av_set_string() to misbehave when
there is an option and a named constant with the same name.
See the thread: "[PATCH] Remove libavcodec/opt.c:set_all_opt()".
Originally committed as revision 16134 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 21:34:12 +00:00
d1ca187fb6
Document register_avcodec().
...
Originally committed as revision 16132 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 20:04:41 +00:00
7a961a46ba
Make register_avcodec() call avcodec_init().
...
This avoids the possibility to use a registered codec without first
initializing libavcodec, which resulted in unexpected behavior.
Originally committed as revision 16131 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 20:04:00 +00:00
335a761a90
Use a more explicit "codec" rather than "format" as the parameter of
...
register_avcodec().
Originally committed as revision 16130 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 19:36:38 +00:00
1238fa9fb8
indeo3: replace sizeof(struct ...) by sizeof(*var)
...
Originally committed as revision 16128 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 16:41:51 +00:00
39caffeb9a
cosmetics: typo/grammar/wording/whitespace fixes
...
Originally committed as revision 16126 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 15:48:01 +00:00
31c3a4dc7e
Prevent shift overflow, patch by Anders Grönberg, galileo.m2 gmail com.
...
Originally committed as revision 16121 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 12:30:34 +00:00
cac55c9196
Add a few const qualifiers in appropriate places.
...
patch by Anders Grönberg, galileo.m2 gmail com
Originally committed as revision 16120 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 12:24:50 +00:00
70bbd0cee3
undoing revision 16118 for truemotion1.c; commit was premature
...
Originally committed as revision 16119 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 03:41:11 +00:00
02fb2546e8
The POSIX namespace shall be held sacrosanct. To that end,
...
continue eliminating _t from structure names in FFmpeg.
Originally committed as revision 16118 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 03:29:33 +00:00
5e6604490a
avoid POSIX reserved _t suffix
...
Originally committed as revision 16117 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 00:48:16 +00:00
d08ea32c23
Get rid of mmx_t.
...
Originally committed as revision 16116 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-14 00:46:27 +00:00
616deed28f
Do not invade _t POSIX namespace.
...
Originally committed as revision 16109 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 18:16:06 +00:00
bda4e3940a
Replace 'typedef struct ustr_t' by 'struct ustr', _t is POSIX namespace.
...
Originally committed as revision 16099 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 13:34:59 +00:00
db95d3bf50
Add img_convert() to imgconvert.h and #include it from imgresample.c.
...
This fixes an implicit declaration of img_convert() warning.
taken from a patchset from Anders Grönberg, galileo.m2 gmail com
Originally committed as revision 16097 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 12:56:29 +00:00
537db58b6c
Let imgresample altivec code build
...
Originally committed as revision 16093 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 10:53:30 +00:00
7a8f36cccc
Cleanup types
...
Originally committed as revision 16092 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 10:52:34 +00:00
006c8e9e9c
Unbreak imgresample altivec
...
Originally committed as revision 16091 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-13 10:49:38 +00:00
8f993c6725
s/#AVClass/AVClass/, consistent with the rest of the doxy.
...
Originally committed as revision 16085 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-12 23:02:21 +00:00
fe5c7e588d
More POSIX _t namespace cleanup.
...
Originally committed as revision 16082 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-12 16:21:23 +00:00
64adf4f28d
add Altivec implementation of clear_block
...
Originally committed as revision 16078 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-12 09:59:00 +00:00
05dee1b7df
Rename all vorbis encoder related typedefs to not use _t for POSIX compatibility
...
Originally committed as revision 16076 to svn://svn.ffmpeg.org/ffmpeg/trunk
2008-12-12 05:12:51 +00:00
Aurelien Jacobs
Kostya Shishkov
Michael Niedermayer
Carl Eugen Hoyos
Måns Rullgård
Diego Biurrun
Luca Barbato
Reimar Döffinger
Reynaldo H. Verdejo Pinochet
Benoit Fouet
Jason Garrett-Glaser
Guillaume Poirier
avcoder
Loren Merritt
Andreas Öman
Anuradha Suraparaju
Robert Swain
Alex Converse
Stefano Sabatini
Alexander Strange
David Conrad
Laurent Aimar
Daniel Verkamp
Anders Grönberg
Mike Melanson
Oded Shimon