]> git.sesse.net Git - ffmpeg/log
ffmpeg
16 years agoRemove delayed_output_pic, I do not understand what this variable was good for.
Michael Niedermayer [Sat, 19 Jul 2008 02:53:57 +0000 (02:53 +0000)]
Remove delayed_output_pic, I do not understand what this variable was good for.
It is simpler and might be faster now.

Originally committed as revision 14289 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDrop out of order pics even if bitstream_restriction_flag is set.
Michael Niedermayer [Sat, 19 Jul 2008 02:44:17 +0000 (02:44 +0000)]
Drop out of order pics even if bitstream_restriction_flag is set.

Originally committed as revision 14288 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSet avctx->coded_frame in RoQ encoder. At some point in
Vitor Sessak [Sat, 19 Jul 2008 02:43:23 +0000 (02:43 +0000)]
Set avctx->coded_frame in RoQ encoder. At some point in
the SVN history this became mandated. Fix issue 548.

Originally committed as revision 14287 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake the ugly hack which uses an unused entry in the internal buffer
Michael Niedermayer [Sat, 19 Jul 2008 00:32:15 +0000 (00:32 +0000)]
Make the ugly hack which uses an unused entry in the internal buffer
array actually use a unused one, so it does work.

Originally committed as revision 14286 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake ffmpeg -t X with X<=0 transcode nothing.
Stefano Sabatini [Fri, 18 Jul 2008 22:23:29 +0000 (22:23 +0000)]
Make ffmpeg -t X with X<=0 transcode nothing.

Originally committed as revision 14285 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agominor simplification of qdm2_decode_fft_packets()
Michael Niedermayer [Fri, 18 Jul 2008 21:16:59 +0000 (21:16 +0000)]
minor simplification of qdm2_decode_fft_packets()

Originally committed as revision 14284 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake frames non reference by default, so they are disposed of automatically
Michael Niedermayer [Fri, 18 Jul 2008 20:59:17 +0000 (20:59 +0000)]
Make frames non reference by default, so they are disposed of automatically
in MPV_frame_start() if we break out due to an error at a random place.
Fixes issue334

Originally committed as revision 14283 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCorrecting displayed muxing overhead and video stream size.
Michael Niedermayer [Fri, 18 Jul 2008 13:52:23 +0000 (13:52 +0000)]
Correcting displayed muxing overhead and video stream size.
Partially fixes issue348.

Originally committed as revision 14282 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSupport reading large metadata.
Michael Niedermayer [Fri, 18 Jul 2008 12:03:21 +0000 (12:03 +0000)]
Support reading large metadata.
fixes issue187

Originally committed as revision 14281 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake bitstream_* fields unsigned.
Michael Niedermayer [Fri, 18 Jul 2008 11:39:41 +0000 (11:39 +0000)]
Make bitstream_* fields unsigned.

Originally committed as revision 14280 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoOnly realloc() bitstream buffer when the needed size increased,
Michael Niedermayer [Fri, 18 Jul 2008 11:38:53 +0000 (11:38 +0000)]
Only realloc() bitstream buffer when the needed size increased,
this is needed to prevent loosing bitstream data with large metadata.

Originally committed as revision 14279 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoIf metadata has been parsed goto end instead of trying to
Michael Niedermayer [Fri, 18 Jul 2008 11:08:15 +0000 (11:08 +0000)]
If metadata has been parsed goto end instead of trying to
decorrelate and output somehing nonexistng.

Originally committed as revision 14278 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix a mem leak in vc1_decode_frame().
Erik Hovland [Fri, 18 Jul 2008 07:15:50 +0000 (07:15 +0000)]
Fix a mem leak in vc1_decode_frame().
Patch by Erik Hovland erik hovland org

Originally committed as revision 14277 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix a mem leak in av_find_stream_info().
Erik Hovland [Fri, 18 Jul 2008 07:13:56 +0000 (07:13 +0000)]
Fix a mem leak in av_find_stream_info().
Patch by Erik Hovland erik hovland org

Originally committed as revision 14276 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agointreadwrite: support DEC compiler __unaligned type qualifier
Måns Rullgård [Fri, 18 Jul 2008 01:18:59 +0000 (01:18 +0000)]
intreadwrite: support DEC compiler __unaligned type qualifier

Originally committed as revision 14275 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMPEGTS: Improve probe function
Måns Rullgård [Fri, 18 Jul 2008 01:07:17 +0000 (01:07 +0000)]
MPEGTS: Improve probe function

When a sync byte is found, check that transport_error_indicator is zero,
and adaptation_field_control is valid (non-zero).

Originally committed as revision 14274 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocosmetics, remove space
Baptiste Coudurier [Fri, 18 Jul 2008 00:24:31 +0000 (00:24 +0000)]
cosmetics, remove space

Originally committed as revision 14273 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoReturn max score when ftyp atom is encountered.
Baptiste Coudurier [Fri, 18 Jul 2008 00:23:37 +0000 (00:23 +0000)]
Return max score when ftyp atom is encountered.

Originally committed as revision 14272 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake C code in yuv2yuv1() do accurate rounding, this could be split
Michael Niedermayer [Fri, 18 Jul 2008 00:09:09 +0000 (00:09 +0000)]
Make C code in yuv2yuv1() do accurate rounding, this could be split
depending on SWS_ACCURATE as well if someone wants.

Originally committed as revision 27323 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoSimplify: do not overuse pointer aritmetic
Vitor Sessak [Thu, 17 Jul 2008 23:48:53 +0000 (23:48 +0000)]
Simplify: do not overuse pointer aritmetic

Originally committed as revision 14271 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoindent
Michael Niedermayer [Thu, 17 Jul 2008 23:43:48 +0000 (23:43 +0000)]
indent

Originally committed as revision 27322 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoForgotten accurate rounding function YSCALEYUV2YV121_ACCURATE.
Michael Niedermayer [Thu, 17 Jul 2008 23:42:06 +0000 (23:42 +0000)]
Forgotten accurate rounding function YSCALEYUV2YV121_ACCURATE.

Originally committed as revision 27321 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agosimplify yuv2yuv1()
Michael Niedermayer [Thu, 17 Jul 2008 23:35:36 +0000 (23:35 +0000)]
simplify yuv2yuv1()

Originally committed as revision 27320 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoDeclare parameters of pred() that could be const as such
Vitor Sessak [Thu, 17 Jul 2008 23:02:52 +0000 (23:02 +0000)]
Declare parameters of pred() that could be const as such

Originally committed as revision 14270 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAnother simplification
Vitor Sessak [Thu, 17 Jul 2008 23:01:31 +0000 (23:01 +0000)]
Another simplification

Originally committed as revision 14269 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify
Vitor Sessak [Thu, 17 Jul 2008 22:59:53 +0000 (22:59 +0000)]
Simplify

Originally committed as revision 14268 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify
Vitor Sessak [Thu, 17 Jul 2008 22:42:21 +0000 (22:42 +0000)]
Simplify

Originally committed as revision 14267 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDeclare function parameters as const
Vitor Sessak [Thu, 17 Jul 2008 22:39:29 +0000 (22:39 +0000)]
Declare function parameters as const

Originally committed as revision 14266 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago10l forgot SWS_BILINEAR
Michael Niedermayer [Thu, 17 Jul 2008 22:30:02 +0000 (22:30 +0000)]
10l forgot SWS_BILINEAR

Originally committed as revision 27318 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoEnsure that exactly one scaler algo is used.
Michael Niedermayer [Thu, 17 Jul 2008 22:01:18 +0000 (22:01 +0000)]
Ensure that exactly one scaler algo is used.

Originally committed as revision 27317 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoRearrange AV_[RW][BL]*() macros
Måns Rullgård [Thu, 17 Jul 2008 18:42:19 +0000 (18:42 +0000)]
Rearrange AV_[RW][BL]*() macros

Originally committed as revision 14265 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago100l, someone used FLT_MIN where -FLT_MAX was meant at a few places in the
Michael Niedermayer [Thu, 17 Jul 2008 17:42:44 +0000 (17:42 +0000)]
100l, someone used FLT_MIN where -FLT_MAX was meant at a few places in the
AVOptions array.

Originally committed as revision 14264 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoReplace LDLATEFLAGS hackery by proper LDFLAGS tests.
Diego Biurrun [Thu, 17 Jul 2008 16:28:48 +0000 (16:28 +0000)]
Replace LDLATEFLAGS hackery by proper LDFLAGS tests.
The original reasons for LDLATEFLAGS are lost in the mists of time.

Originally committed as revision 14263 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoTry to fix FATE after my time_base simplification change.
Michael Niedermayer [Thu, 17 Jul 2008 14:13:44 +0000 (14:13 +0000)]
Try to fix FATE after my time_base simplification change.

Originally committed as revision 14262 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMove up swscale_error target so that regression tests error out immediately.
Diego Biurrun [Thu, 17 Jul 2008 09:30:48 +0000 (09:30 +0000)]
Move up swscale_error target so that regression tests error out immediately.

Originally committed as revision 14261 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFree in avcodec_close() avctx->rc_eq. Fix a memory leak.
Stefano Sabatini [Wed, 16 Jul 2008 22:10:34 +0000 (22:10 +0000)]
Free in avcodec_close() avctx->rc_eq. Fix a memory leak.

Originally committed as revision 14260 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago10000l, revert
Michael Niedermayer [Wed, 16 Jul 2008 17:54:57 +0000 (17:54 +0000)]
10000l, revert
r14254
Log:
The funny memcpyin svq3 generally has src & dst overlapping, so it
should at least be a memmove().
I was tired, they do not overlap.

Originally committed as revision 14259 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoindent
Michael Niedermayer [Wed, 16 Jul 2008 17:51:24 +0000 (17:51 +0000)]
indent

Originally committed as revision 14258 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSVQ3 uses the JPEG variant of YUV 4:2:0.
Michael Niedermayer [Wed, 16 Jul 2008 17:48:47 +0000 (17:48 +0000)]
SVQ3 uses the JPEG variant of YUV 4:2:0.
Fixes issue256

Originally committed as revision 14257 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocosmetics: Fix mxf codec long name.
Diego Biurrun [Wed, 16 Jul 2008 17:20:21 +0000 (17:20 +0000)]
cosmetics: Fix mxf codec long name.

Originally committed as revision 14256 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoprint AVCodecContext address instead of AVClass, better when multiple instances of...
Baptiste Coudurier [Wed, 16 Jul 2008 03:21:23 +0000 (03:21 +0000)]
print AVCodecContext address instead of AVClass, better when multiple instances of same codecs are used, based on r14237

Originally committed as revision 14255 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoThe funny memcpyin svq3 generally has src & dst overlapping, so it
Michael Niedermayer [Wed, 16 Jul 2008 02:10:21 +0000 (02:10 +0000)]
The funny memcpyin svq3 generally has src & dst overlapping, so it
should at least be a memmove().

Originally committed as revision 14254 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoPrint slice num as well.
Michael Niedermayer [Wed, 16 Jul 2008 01:57:28 +0000 (01:57 +0000)]
Print slice num as well.

Originally committed as revision 14253 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agofloat_to_int16_interleave: change src to an array of pointers instead of assuming...
Loren Merritt [Wed, 16 Jul 2008 00:50:12 +0000 (00:50 +0000)]
float_to_int16_interleave: change src to an array of pointers instead of assuming it's contiguous.
this has no immediate effect, but will allow it to be used in more codecs.

Originally committed as revision 14252 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agovorbis: reduce memory by reusing buffers
Loren Merritt [Tue, 15 Jul 2008 23:04:28 +0000 (23:04 +0000)]
vorbis: reduce memory by reusing buffers

Originally committed as revision 14251 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agounroll another common case in vorbis_residue_decode
Loren Merritt [Tue, 15 Jul 2008 23:01:15 +0000 (23:01 +0000)]
unroll another common case in vorbis_residue_decode

Originally committed as revision 14250 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoOGG: Add padding at end of Vorbis headers
Siarhei Siamashka [Tue, 15 Jul 2008 23:00:29 +0000 (23:00 +0000)]
OGG: Add padding at end of Vorbis headers

Patch by Siarhei Siamashka <siarhei.siamashka gmail com>

Originally committed as revision 14249 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMove the size of the H264Context delayed_pic array into a #define.
Alexander Strange [Tue, 15 Jul 2008 21:30:04 +0000 (21:30 +0000)]
Move the size of the H264Context delayed_pic array into a #define.
This improves readability by getting rid of some magic numbers.

Originally committed as revision 14248 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake v4l2_read_header() don't free a stream in case of failure when
Stefano Sabatini [Tue, 15 Jul 2008 20:45:42 +0000 (20:45 +0000)]
Make v4l2_read_header() don't free a stream in case of failure when
reading its header. The stream will be freed later in
av_open_input_stream(). Fix a segmentation fault due to a double free
on the same pointer.

Originally committed as revision 14247 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake v4l_read_header() don't free a stream in case of failure when
Stefano Sabatini [Tue, 15 Jul 2008 20:39:20 +0000 (20:39 +0000)]
Make v4l_read_header() don't free a stream in case of failure when
reading its header. The stream will be freed later in
av_open_input_stream(). Fix a segmentation fault due to a double free
on the same pointer.

Originally committed as revision 14246 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: optimised MAC64 and MLS64
Måns Rullgård [Tue, 15 Jul 2008 19:06:32 +0000 (19:06 +0000)]
ARM: optimised MAC64 and MLS64

Originally committed as revision 14245 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: optimised MUL64
Måns Rullgård [Tue, 15 Jul 2008 19:06:29 +0000 (19:06 +0000)]
ARM: optimised MUL64

Originally committed as revision 14244 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: ARMv6 optimised MULH
Måns Rullgård [Tue, 15 Jul 2008 19:06:25 +0000 (19:06 +0000)]
ARM: ARMv6 optimised MULH

Originally committed as revision 14243 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: ARMv6 optimised FASTDIV
Måns Rullgård [Tue, 15 Jul 2008 19:06:18 +0000 (19:06 +0000)]
ARM: ARMv6 optimised FASTDIV

Originally committed as revision 14242 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: disable use of system byteswap.h
Måns Rullgård [Tue, 15 Jul 2008 19:06:09 +0000 (19:06 +0000)]
ARM: disable use of system byteswap.h

Our bswap functions are better than those in glibc, which has
no ARM-optimised versions.

Originally committed as revision 14241 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: fix ARCH_ARM -> ARCH_ARMV4L test in bswap_32()
Måns Rullgård [Tue, 15 Jul 2008 19:06:03 +0000 (19:06 +0000)]
ARM: fix ARCH_ARM -> ARCH_ARMV4L test in bswap_32()

Originally committed as revision 14240 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoARM: ARMv6 optimised bswap_16/32
Måns Rullgård [Tue, 15 Jul 2008 19:05:49 +0000 (19:05 +0000)]
ARM: ARMv6 optimised bswap_16/32

Originally committed as revision 14239 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove dead code in bitstream.h
Måns Rullgård [Tue, 15 Jul 2008 18:43:26 +0000 (18:43 +0000)]
Remove dead code in bitstream.h

Originally committed as revision 14238 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoPrint the address of the AVCodecContext in av_log().
Alexander Strange [Tue, 15 Jul 2008 17:17:35 +0000 (17:17 +0000)]
Print the address of the AVCodecContext in av_log().
This makes debug output much easier to read when the
same codec is open more than once.

Originally committed as revision 14237 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago10l, float_to_int16_interleave_sse/3dnow wrote the wrong samples
Loren Merritt [Tue, 15 Jul 2008 04:11:30 +0000 (04:11 +0000)]
10l, float_to_int16_interleave_sse/3dnow wrote the wrong samples

Originally committed as revision 14236 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake flic probe check a few more things.
Michael Niedermayer [Mon, 14 Jul 2008 22:55:07 +0000 (22:55 +0000)]
Make flic probe check a few more things.
fixes issue399

Originally committed as revision 14235 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMove the codec_id overriding at a slightly better place.
Michael Niedermayer [Mon, 14 Jul 2008 20:31:17 +0000 (20:31 +0000)]
Move the codec_id overriding at a slightly better place.

Originally committed as revision 14234 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoH264: fix index range assert() in remove_short_at_index()
Måns Rullgård [Mon, 14 Jul 2008 19:17:13 +0000 (19:17 +0000)]
H264: fix index range assert() in remove_short_at_index()

Originally committed as revision 14233 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove common factors from timebase in av_set_pts_info().
Michael Niedermayer [Mon, 14 Jul 2008 19:11:09 +0000 (19:11 +0000)]
Remove common factors from timebase in av_set_pts_info().
Should fix issue313

Originally committed as revision 14232 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix infinite loop at EOF.
Michael Niedermayer [Mon, 14 Jul 2008 16:18:29 +0000 (16:18 +0000)]
Fix infinite loop at EOF.

Originally committed as revision 14231 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRevert r14218. FFMAX is a macro, so
Vitor Sessak [Mon, 14 Jul 2008 14:41:01 +0000 (14:41 +0000)]
Revert r14218. FFMAX is a macro, so
   sum = FFMAX(1, scalar_product(...));
might calculate the scalar product twice, which is slower.

Originally committed as revision 14230 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoonly animated gif (gifdec.c) was removed)
Ben Littler [Mon, 14 Jul 2008 12:08:48 +0000 (12:08 +0000)]
only animated gif (gifdec.c) was removed)

Originally committed as revision 14229 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agogifdec.c was removed, purge its entry from the Makefile.
Diego Biurrun [Mon, 14 Jul 2008 08:55:07 +0000 (08:55 +0000)]
gifdec.c was removed, purge its entry from the Makefile.

Originally committed as revision 14228 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix decoding of 320x240.ogg.
Michael Niedermayer [Sun, 13 Jul 2008 23:34:29 +0000 (23:34 +0000)]
Fix decoding of 320x240.ogg.

Originally committed as revision 14227 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago-itsscale to scale timestamps
Michael Niedermayer [Sun, 13 Jul 2008 22:55:52 +0000 (22:55 +0000)]
-itsscale to scale timestamps
fixes issue511

Originally committed as revision 14226 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: reindent.
Alexander Strange [Sun, 13 Jul 2008 22:27:50 +0000 (22:27 +0000)]
Cosmetics: reindent.

Originally committed as revision 27279 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoRemove const vector casts.
Alexander Strange [Sun, 13 Jul 2008 22:27:19 +0000 (22:27 +0000)]
Remove const vector casts.
Fixes compilation with altivec and -std=gnu99.

Originally committed as revision 27278 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoremoved gif decoder in r14209
Ben Littler [Sun, 13 Jul 2008 22:21:25 +0000 (22:21 +0000)]
removed gif decoder in r14209

Originally committed as revision 14225 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDrop non-C99-compliant initialization.
Diego Biurrun [Sun, 13 Jul 2008 21:59:58 +0000 (21:59 +0000)]
Drop non-C99-compliant initialization.

Originally committed as revision 14224 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake our flac decoder decode all the data at EOF.
Michael Niedermayer [Sun, 13 Jul 2008 21:51:29 +0000 (21:51 +0000)]
Make our flac decoder decode all the data at EOF.
Fixes issue524

Originally committed as revision 14223 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agotools/trasher: Don't mix srand/rand and srandom/random interface.
Matthieu Castet [Sun, 13 Jul 2008 21:44:00 +0000 (21:44 +0000)]
tools/trasher: Don't mix srand/rand and srandom/random interface.
patch by matthieu castet, castet.matthieu free fr

Originally committed as revision 14222 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agostrdup() the default value for rc_eq in AVCodec when setting it in
Stefano Sabatini [Sun, 13 Jul 2008 21:44:00 +0000 (21:44 +0000)]
strdup() the default value for rc_eq in AVCodec when setting it in
avcodec_get_context_defaults(). The value has to be alloced dinamically
rather than statically since it may be freed for example by
av_set_string2(). Fix a segmentation fault.

Originally committed as revision 14221 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoOn EOF keep calling the decoders until they do not output anything
Michael Niedermayer [Sun, 13 Jul 2008 21:39:18 +0000 (21:39 +0000)]
On EOF keep calling the decoders until they do not output anything
anymore, instead of just once.

Originally committed as revision 14220 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix the logic to access the location of a string to free when setting
Stefano Sabatini [Sun, 13 Jul 2008 21:28:17 +0000 (21:28 +0000)]
Fix the logic to access the location of a string to free when setting
a new value for a string with av_set_string2(). Fix a segmentation
fault.

Originally committed as revision 14219 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify
Vitor Sessak [Sun, 13 Jul 2008 20:41:09 +0000 (20:41 +0000)]
Simplify

Originally committed as revision 14218 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAnother av_clipf() usage
Vitor Sessak [Sun, 13 Jul 2008 20:39:55 +0000 (20:39 +0000)]
Another av_clipf() usage

Originally committed as revision 14217 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDo not declare as double a var that only stores a float
Vitor Sessak [Sun, 13 Jul 2008 20:38:42 +0000 (20:38 +0000)]
Do not declare as double a var that only stores a float

Originally committed as revision 14216 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify: use FFMAX
Vitor Sessak [Sun, 13 Jul 2008 20:36:32 +0000 (20:36 +0000)]
Simplify: use FFMAX

Originally committed as revision 14215 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocheck that csize in ff_lzw_decode_init is < LZW_MAXBITS, <= is not enough and
Reimar Döffinger [Sun, 13 Jul 2008 20:03:57 +0000 (20:03 +0000)]
check that csize in ff_lzw_decode_init is < LZW_MAXBITS, <= is not enough and
might read outside the prefix array

Originally committed as revision 14214 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd av_clipf() function to common.h and use it in ra288.c
Vitor Sessak [Sun, 13 Jul 2008 19:59:44 +0000 (19:59 +0000)]
Add av_clipf() function to common.h and use it in ra288.c

Originally committed as revision 14213 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agodvdsubdec: allow compilation with -DDEBUG
Måns Rullgård [Sun, 13 Jul 2008 19:41:13 +0000 (19:41 +0000)]
dvdsubdec: allow compilation with -DDEBUG

Originally committed as revision 14212 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRTP: use dprintf(), allow compilation with -DDEBUG
Måns Rullgård [Sun, 13 Jul 2008 19:41:10 +0000 (19:41 +0000)]
RTP: use dprintf(), allow compilation with -DDEBUG

Originally committed as revision 14211 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoASF: use dprintf(), allow compilation with -DDEBUG
Måns Rullgård [Sun, 13 Jul 2008 19:41:07 +0000 (19:41 +0000)]
ASF: use dprintf(), allow compilation with -DDEBUG

Originally committed as revision 14210 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove gif demuxer, the code contains a gif decoder and lzw decoder neither
Michael Niedermayer [Sun, 13 Jul 2008 18:47:24 +0000 (18:47 +0000)]
Remove gif demuxer, the code contains a gif decoder and lzw decoder neither
belongs in a demuxer, furthermore the code might be exploitable.
Fixes issue530

Originally committed as revision 14209 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoUse -vsync 0 unless the user explicitly asked for something
Michael Niedermayer [Sun, 13 Jul 2008 16:19:23 +0000 (16:19 +0000)]
Use -vsync 0  unless the user explicitly asked for something
else when the framerate is above 1000.
Fixes issue259.

Originally committed as revision 14208 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoexploit mdct symmetry
Loren Merritt [Sun, 13 Jul 2008 15:03:58 +0000 (15:03 +0000)]
exploit mdct symmetry
2% faster vorbis on conroe, k8. 7% on celeron.

Originally committed as revision 14207 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agomlpdec: Check for bits left before each read of End-of-Stream indicator and
Ramiro Polla [Sun, 13 Jul 2008 14:59:39 +0000 (14:59 +0000)]
mlpdec: Check for bits left before each read of End-of-Stream indicator and
Substream parity check independently. This allows decoding of streams that
have EOS but don't have Substream parity check.

Originally committed as revision 14206 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agosimplify vorbis windowing
Loren Merritt [Sun, 13 Jul 2008 14:56:01 +0000 (14:56 +0000)]
simplify vorbis windowing

Originally committed as revision 14205 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agobitstream: move put_sbits() from flacenc.c to bitstream.h and use it
Ramiro Polla [Sun, 13 Jul 2008 14:27:48 +0000 (14:27 +0000)]
bitstream: move put_sbits() from flacenc.c to bitstream.h and use it
throughout libavcodec.

Originally committed as revision 14204 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoclip MS ADPCM predictor range to correct [0..6] range, not [0..7]
Reimar Döffinger [Sun, 13 Jul 2008 14:25:31 +0000 (14:25 +0000)]
clip MS ADPCM predictor range to correct [0..6] range, not [0..7]

Originally committed as revision 14203 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoUse bytestream and AV_RL* functions in ADPCM code where possible
Reimar Döffinger [Sun, 13 Jul 2008 14:12:51 +0000 (14:12 +0000)]
Use bytestream and AV_RL* functions in ADPCM code where possible

Originally committed as revision 14202 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix MS ADPCM de- and encoding: first and second sample are swapped.
Reimar Döffinger [Sun, 13 Jul 2008 14:08:28 +0000 (14:08 +0000)]
Fix MS ADPCM de- and encoding: first and second sample are swapped.
Noticeable better decoding of e.g. http://samples.mplayerhq.hu/A-codecs/msadpcm-stereo/dance-b5th.asf

Originally committed as revision 14201 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd parisc* to the list of architectures requiring PIC flags for shared libs.
Diego Biurrun [Sun, 13 Jul 2008 12:40:11 +0000 (12:40 +0000)]
Add parisc* to the list of architectures requiring PIC flags for shared libs.

Originally committed as revision 14200 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agompegaudio: use MAC/MLS macros in ff_mpa_synth_filter
Måns Rullgård [Sun, 13 Jul 2008 12:23:02 +0000 (12:23 +0000)]
mpegaudio: use MAC/MLS macros in ff_mpa_synth_filter

Originally committed as revision 14199 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agomathops: add MAC64, MLS64, and MLS16
Måns Rullgård [Sun, 13 Jul 2008 12:22:57 +0000 (12:22 +0000)]
mathops: add MAC64, MLS64, and MLS16
MAC64: 64 += 32 x 32 (bits)
MLS64: 64 -= 32 x 32
MLS16: 32 -= 16 x 16

Originally committed as revision 14198 to svn://svn.ffmpeg.org/ffmpeg/trunk