ffmpeg.git
7 years agoARM: NEON put/avg_pixels8/16 cosmetics
Mans Rullgard [Thu, 1 Dec 2011 15:01:44 +0000 (15:01 +0000)]
ARM: NEON put/avg_pixels8/16 cosmetics

This makes whitespace and register names consistent with
the style used in more recent code.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoARM: add remaining NEON avg_pixels8/16 functions
Mans Rullgard [Thu, 1 Dec 2011 05:07:09 +0000 (05:07 +0000)]
ARM: add remaining NEON avg_pixels8/16 functions

7 years agoARM: clean up NEON put/avg_pixels macros
Mans Rullgard [Thu, 1 Dec 2011 05:04:32 +0000 (05:04 +0000)]
ARM: clean up NEON put/avg_pixels macros

Although this adds a few lines, the macro calls are less convoluted.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agofate: split acodec-pcm into individual tests
Justin Ruggles [Wed, 30 Nov 2011 21:51:01 +0000 (16:51 -0500)]
fate: split acodec-pcm into individual tests

this removes 2 redundant tests for pcm in mkv.
we can add the coverage back in later as fate-lavf tests if needed.

7 years agoswscale: #include "libavutil/mathematics.h"
Reinhard Tartler [Thu, 1 Dec 2011 17:48:33 +0000 (18:48 +0100)]
swscale: #include "libavutil/mathematics.h"

this file uses the M_PI macro since
4e74187db2f5db52f88729efc662df9d6bc763e1, so include the correct header
directly.

Signed-off-by: Reinhard Tartler <siretart@tauware.de>
7 years agopmpdec: don't use deprecated av_set_pts_info.
Anton Khirnov [Thu, 1 Dec 2011 16:28:36 +0000 (17:28 +0100)]
pmpdec: don't use deprecated av_set_pts_info.

7 years agorv34: align temporary block of "dct" coefs
Mans Rullgard [Thu, 1 Dec 2011 00:18:38 +0000 (00:18 +0000)]
rv34: align temporary block of "dct" coefs

This is needed for optimised transforms.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoAdd PlayStation Portable PMP format demuxer
Reimar Döffinger [Sat, 26 Mar 2011 15:25:10 +0000 (16:25 +0100)]
Add PlayStation Portable PMP format demuxer

Not yet complete, for demuxing AAC the AAC header must be generated
manually.
Possibly the decoder could accept the header as extradata to simplify
this.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agoproto: Realign struct initializers
Martin Storsjö [Thu, 1 Dec 2011 09:53:57 +0000 (11:53 +0200)]
proto: Realign struct initializers

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agoproto: Use .priv_data_size to allocate the private context
Martin Storsjö [Thu, 1 Dec 2011 09:44:21 +0000 (11:44 +0200)]
proto: Use .priv_data_size to allocate the private context

This simplifies the open functions by avoiding one function
call that needs error checking, reducing the amount of
extra bulk code.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agommsh: Properly clean up if the second ffurl_alloc failed
Martin Storsjö [Thu, 1 Dec 2011 09:48:03 +0000 (11:48 +0200)]
mmsh: Properly clean up if the second ffurl_alloc failed

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agortmp: Clean up properly if the handshake failed
Martin Storsjö [Thu, 1 Dec 2011 09:34:06 +0000 (11:34 +0200)]
rtmp: Clean up properly if the handshake failed

This prevents memory leaks if this function returns an error.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agomd5proto: Remove the get_file_handle function
Martin Storsjö [Thu, 1 Dec 2011 09:30:03 +0000 (11:30 +0200)]
md5proto: Remove the get_file_handle function

The private data pointer isn't a file handle, this protocol
doesn't have any file handle to return.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agoapplehttpproto: Use the close function if the open function fails
Martin Storsjö [Thu, 1 Dec 2011 09:07:35 +0000 (11:07 +0200)]
applehttpproto: Use the close function if the open function fails

This should clean up leaked memory.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agohttp: Make sure proxyauth is initialized
Martin Storsjö [Thu, 1 Dec 2011 09:24:23 +0000 (11:24 +0200)]
http: Make sure proxyauth is initialized

This string will be passed to ff_http_auth_create_response
even if no proxy is used, resulting in reading uninitialized
memory. The other auth string is always initialized by
av_url_split.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agovble: use dsp.add_hfyu_median_prediction() this allows asm optimizations to be used.
Michael Niedermayer [Wed, 30 Nov 2011 22:28:51 +0000 (17:28 -0500)]
vble: use dsp.add_hfyu_median_prediction() this allows asm optimizations to be used.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agovble: remove flags copy, its not used in any speed relevant code.
Michael Niedermayer [Wed, 30 Nov 2011 22:28:50 +0000 (17:28 -0500)]
vble: remove flags copy, its not used in any speed relevant code.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agovble: merge len and val arrays
Derek Buitenhuis [Wed, 30 Nov 2011 22:28:49 +0000 (17:28 -0500)]
vble: merge len and val arrays

There's no reason to use two arrays for this.

Based off commit 2fea60c60084c4e70d7cef128ea3bca5690ce465
to FFmpeg by Michael Niedermayer.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agovble: remove unused variable len.
Michael Niedermayer [Wed, 30 Nov 2011 22:28:48 +0000 (17:28 -0500)]
vble: remove unused variable len.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agortpdec: Templatize the code for different g726 bitrate variants
Martin Storsjö [Wed, 30 Nov 2011 16:05:06 +0000 (18:05 +0200)]
rtpdec: Templatize the code for different g726 bitrate variants

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agorv40: move loop filter to rv34dsp context
Janne Grunau [Wed, 5 Oct 2011 07:29:38 +0000 (09:29 +0200)]
rv40: move loop filter to rv34dsp context

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agolavf: make av_set_pts_info private.
Anton Khirnov [Tue, 29 Nov 2011 18:28:15 +0000 (19:28 +0100)]
lavf: make av_set_pts_info private.

It's supposed to be called only from (de)muxers.

7 years agortpdec: Add support for G726 audio
Miroslav Slugeň [Mon, 7 Nov 2011 11:13:55 +0000 (12:13 +0100)]
rtpdec: Add support for G726 audio

This requires using a separate init function, since there
isn't necessarily any fmtp lines for this codec, so
parse_sdp_a_line won't be called. Incorporating it with the
alloc function wouldn't do either, since it is called before
the full rtpmap line is parsed (where the sample rate is
extracted).

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agortpdec: Add an init function that can do custom codec context initialization
Martin Storsjö [Tue, 29 Nov 2011 14:51:26 +0000 (16:51 +0200)]
rtpdec: Add an init function that can do custom codec context initialization

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agoavconv: make copy_tb on by default.
Anton Khirnov [Tue, 29 Nov 2011 10:10:31 +0000 (11:10 +0100)]
avconv: make copy_tb on by default.

I.e. on streamcopy set output codec timebase from input stream timebase
(as opposed to input codec timebase). This should be more sane, because
since the stream is not decoded, the input codec tb has no relation to
the timestamps of the copied packets.

7 years agomatroskadec: don't set codec timebase.
Anton Khirnov [Mon, 28 Nov 2011 08:38:32 +0000 (09:38 +0100)]
matroskadec: don't set codec timebase.

It's not supposed to be set outside of lavc.

7 years agormdec: don't set codec timebase.
Anton Khirnov [Mon, 28 Nov 2011 08:17:46 +0000 (09:17 +0100)]
rmdec: don't set codec timebase.

It's not supposed to be set outside of lavc.
Set r_frame_rate instead.

7 years agoavconv: compute next_pts from input packet duration when possible.
Anton Khirnov [Mon, 28 Nov 2011 10:15:06 +0000 (11:15 +0100)]
avconv: compute next_pts from input packet duration when possible.

7 years agolavf: estimate frame duration from r_frame_rate.
Anton Khirnov [Mon, 28 Nov 2011 09:02:21 +0000 (10:02 +0100)]
lavf: estimate frame duration from r_frame_rate.

If r_frame_rate is set, it should be more reliable for this than either
codec or stream timebase.

7 years agoavconv: update InputStream.pts in the streamcopy case.
Anton Khirnov [Mon, 28 Nov 2011 10:47:02 +0000 (11:47 +0100)]
avconv: update InputStream.pts in the streamcopy case.

This was broken in 2a651b719c309c5e2fc663a5a9d6ca36153ab98f.

7 years agoindeo3: check per-plane data buffer against input buffer bounds.
Aneesh Dogra [Tue, 29 Nov 2011 17:43:35 +0000 (23:13 +0530)]
indeo3: check per-plane data buffer against input buffer bounds.

Fixes : http://bugzilla.libav.org/show_bug.cgi?id=102

Signed-off-by: Alex Converse <alex.converse@gmail.com>
7 years agoavconv: Handle audio sync for non-S16 sample formats.
Alex Converse [Mon, 28 Nov 2011 23:19:39 +0000 (15:19 -0800)]
avconv: Handle audio sync for non-S16 sample formats.

Also fix reporting of the number of samples added/dropped.

7 years agopthread: don't increment index on zero-sized packets.
Aaron Colwell [Mon, 28 Nov 2011 15:23:03 +0000 (07:23 -0800)]
pthread: don't increment index on zero-sized packets.

The next call to decode() will update from an invalid index, which will
either lead to a memcpy() where dest==src (2 threads), or lead to a
crash (>2 threads).

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agoadtsenc: Check frame size.
Alex Converse [Mon, 28 Nov 2011 08:48:53 +0000 (00:48 -0800)]
adtsenc: Check frame size.

Inspired by work from: Michael Niedermayer <michaelni@gmx.at>.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
7 years agotxd: Fix order of operations.
Alex Converse [Mon, 28 Nov 2011 00:17:13 +0000 (16:17 -0800)]
txd: Fix order of operations.

7 years agoAPIchanges: fill in some blanks
Mans Rullgard [Mon, 28 Nov 2011 16:47:03 +0000 (16:47 +0000)]
APIchanges: fill in some blanks

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agotimer: fix misspelling of "decicycles"
Mans Rullgard [Mon, 28 Nov 2011 14:21:48 +0000 (14:21 +0000)]
timer: fix misspelling of "decicycles"

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoEliminate pointless 0/NULL initializers in AVCodec and similar declarations.
Diego Biurrun [Sat, 26 Nov 2011 17:35:26 +0000 (18:35 +0100)]
Eliminate pointless 0/NULL initializers in AVCodec and similar declarations.

7 years agoindeo3: cosmetics
Kostya Shishkov [Mon, 28 Nov 2011 08:33:36 +0000 (09:33 +0100)]
indeo3: cosmetics

Signed-off-by: Diego Biurrun <diego@biurrun.de>
7 years agomd5proto: Fix order of operations.
Alex Converse [Mon, 28 Nov 2011 00:16:44 +0000 (16:16 -0800)]
md5proto: Fix order of operations.

7 years agodca: Replace oversized unused get_bits() with skip_bits_long().
Alex Converse [Sun, 27 Nov 2011 23:48:20 +0000 (15:48 -0800)]
dca: Replace oversized unused get_bits() with skip_bits_long().

7 years agovc1: use an enum for Frame Coding Mode
Luca Barbato [Sat, 26 Nov 2011 12:28:48 +0000 (13:28 +0100)]
vc1: use an enum for Frame Coding Mode

Document it a little and possibly fix a bug in dxva2_vc1.

7 years agodoc: cleanup filter section
Luca Barbato [Fri, 25 Nov 2011 15:03:16 +0000 (16:03 +0100)]
doc: cleanup filter section

Use the @command{} tag when needed and cleanup the examples.

7 years agoindeo3: error out if no motion vector is set.
Aneesh Dogra [Sun, 27 Nov 2011 20:56:52 +0000 (02:26 +0530)]
indeo3: error out if no motion vector is set.

This fixes a crash on a corrupt bitstream (bugzilla #93).

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agox86inc: Flag shufps as an floating-point instruction for the AVX emulation code.
Vitor Sessak [Sun, 27 Nov 2011 14:41:56 +0000 (15:41 +0100)]
x86inc: Flag shufps as an floating-point instruction for the AVX emulation code.

Without this, code like "shufps m0, m1, m2, 0xaa" would not work in CPUs
not supporting SSE2.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agompegaudio: do not use init_static_data() for initializing tables.
Justin Ruggles [Sun, 27 Nov 2011 18:07:19 +0000 (13:07 -0500)]
mpegaudio: do not use init_static_data() for initializing tables.

It is called once for each mpegaudio codec that is registered, which adds
significant overhead.

7 years agomusepack: fix signed shift overflow in mpc_read_packet()
Mans Rullgard [Sun, 27 Nov 2011 10:29:33 +0000 (10:29 +0000)]
musepack: fix signed shift overflow in mpc_read_packet()

Using an unsigned variable avoids problems with overflows.
There is further no need for a 64-bit intermediate here.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agomov: Make format string match variable type.
Alex Converse [Sun, 27 Nov 2011 03:41:54 +0000 (19:41 -0800)]
mov: Make format string match variable type.

7 years agowmavoice: Make format string match variable type.
Alex Converse [Sun, 27 Nov 2011 03:40:34 +0000 (19:40 -0800)]
wmavoice: Make format string match variable type.

7 years agovc1: select interlaced scan table by FCM element
Kostya Shishkov [Sat, 26 Nov 2011 12:23:39 +0000 (13:23 +0100)]
vc1: select interlaced scan table by FCM element

Interlaced videos can contain progressive frames too and now wrong scantable
is selected for them.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agoGeneralize RIFF INFO tag support; support reading INFO tag in wav
Victor Vasiliev [Thu, 24 Nov 2011 20:02:43 +0000 (00:02 +0400)]
Generalize RIFF INFO tag support; support reading INFO tag in wav

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agopthread: track thread existence in a separate variable.
Cheng Sun [Sun, 27 Nov 2011 00:36:50 +0000 (16:36 -0800)]
pthread: track thread existence in a separate variable.

This fixes a compile error on mingw32 when using p->thread
directly (as if it were a pointer) to track thread existence,
because the type is opaque and may be a non-pointer.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agoswscale: fix signed overflow in yuv2mono_X_c_template
Mans Rullgard [Sat, 26 Nov 2011 21:53:00 +0000 (21:53 +0000)]
swscale: fix signed overflow in yuv2mono_X_c_template

As old bits are shifted out of the accumulator, they cause signed
overflows when they reach the end.  Making the variable unsigned fixes
this.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agosnow: fix integer overflows
Mans Rullgard [Sat, 26 Nov 2011 20:59:17 +0000 (20:59 +0000)]
snow: fix integer overflows

The way these values are used, they should have an unsigned type.
A similar change was made for mpegvideo in cb66847.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agosvq1enc: remove stale altivec-related hack
Mans Rullgard [Sat, 26 Nov 2011 20:44:37 +0000 (20:44 +0000)]
svq1enc: remove stale altivec-related hack

altivec.h is no longer (indirectly) included, so this hack is not
needed.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agosnow: fix signed overflow in byte to 32-bit replication
Mans Rullgard [Sat, 26 Nov 2011 20:42:07 +0000 (20:42 +0000)]
snow: fix signed overflow in byte to 32-bit replication

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoadx: rename ff_adx_decode_header() to avpriv_adx_decode_header()
Justin Ruggles [Sat, 26 Nov 2011 22:00:00 +0000 (17:00 -0500)]
adx: rename ff_adx_decode_header() to avpriv_adx_decode_header()

It is used by the ADX decoder, and therefore needs to be exported in order to
work with shared libs.

7 years agoavformat: add CRI ADX format demuxer
Justin Ruggles [Mon, 21 Nov 2011 07:34:18 +0000 (02:34 -0500)]
avformat: add CRI ADX format demuxer

7 years agoadx: add an ADX parser.
Justin Ruggles [Mon, 21 Nov 2011 06:49:37 +0000 (01:49 -0500)]
adx: add an ADX parser.

This simplifies the decoder so it doesn't have to process an in-packet header
or handle arbitrary-sized packets. It also fixes decoding of files with large
headers.

7 years agoadx: move header decoding to ADX common code
Justin Ruggles [Mon, 21 Nov 2011 01:16:05 +0000 (20:16 -0500)]
adx: move header decoding to ADX common code

7 years agoadx: calculate the number of blocks in a packet
Justin Ruggles [Sun, 20 Nov 2011 22:53:43 +0000 (17:53 -0500)]
adx: calculate the number of blocks in a packet

7 years agoadx: define and use 2 new macro constants BLOCK_SIZE and BLOCK_SAMPLES
Justin Ruggles [Sun, 20 Nov 2011 22:50:46 +0000 (17:50 -0500)]
adx: define and use 2 new macro constants BLOCK_SIZE and BLOCK_SAMPLES

7 years agoadx: check for unsupported ADX formats
Justin Ruggles [Sun, 20 Nov 2011 22:07:59 +0000 (17:07 -0500)]
adx: check for unsupported ADX formats

7 years agoadx: simplify encoding by using put_sbits()
Justin Ruggles [Sun, 20 Nov 2011 21:49:13 +0000 (16:49 -0500)]
adx: simplify encoding by using put_sbits()

7 years agoadx: calculate correct LPC coeffs
Justin Ruggles [Sun, 20 Nov 2011 19:21:32 +0000 (14:21 -0500)]
adx: calculate correct LPC coeffs

Instead of using fixed coefficients, the correct way is to calculate the
coefficients using the highpass cutoff frequency from the ADX stream header
and the sample rate.

7 years agoadx: use 12-bit coefficients instead of 14-bit to avoid integer overflow
Justin Ruggles [Sun, 20 Nov 2011 19:03:21 +0000 (14:03 -0500)]
adx: use 12-bit coefficients instead of 14-bit to avoid integer overflow

7 years agoadx: simplify adx_decode() by using get_sbits() to read residual samples
Justin Ruggles [Sun, 20 Nov 2011 18:06:32 +0000 (13:06 -0500)]
adx: simplify adx_decode() by using get_sbits() to read residual samples

7 years agoadx: fix the data offset parsing in adx_decode_header()
Justin Ruggles [Sun, 20 Nov 2011 17:57:21 +0000 (12:57 -0500)]
adx: fix the data offset parsing in adx_decode_header()

first 2 bytes are 0x80, 0x00. offset is only 16-bit.
this is according to format descriptions on multimedia wiki and wikipedia.

7 years agoadx: remove unneeded post-decode channel interleaving
Justin Ruggles [Sun, 20 Nov 2011 17:46:41 +0000 (12:46 -0500)]
adx: remove unneeded post-decode channel interleaving

instead interleave channels while decoding

7 years agoadx: validate header values
Justin Ruggles [Sat, 19 Nov 2011 22:51:36 +0000 (17:51 -0500)]
adx: validate header values

7 years agoadx: cosmetics: general pretty-printing and comment clean-up
Justin Ruggles [Sat, 19 Nov 2011 22:46:15 +0000 (17:46 -0500)]
adx: cosmetics: general pretty-printing and comment clean-up

7 years agoadx: remove useless comments
Justin Ruggles [Sat, 19 Nov 2011 22:07:46 +0000 (17:07 -0500)]
adx: remove useless comments

7 years agoadx: change short to int16_t
Justin Ruggles [Sat, 19 Nov 2011 22:06:45 +0000 (17:06 -0500)]
adx: change short to int16_t

7 years agoadx: rename struct PREV to ADXChannelState
Justin Ruggles [Sat, 19 Nov 2011 22:05:44 +0000 (17:05 -0500)]
adx: rename struct PREV to ADXChannelState

7 years agodnxhdenc: fix signed overflows
Mans Rullgard [Sat, 26 Nov 2011 20:11:12 +0000 (20:11 +0000)]
dnxhdenc: fix signed overflows

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agodv: simplify bitstream splicing in dv_decode_ac()
Mans Rullgard [Sat, 26 Nov 2011 18:48:16 +0000 (18:48 +0000)]
dv: simplify bitstream splicing in dv_decode_ac()

This is simpler and fixes some overflow checker warnings.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agos3tc: fix shift overflow by using unsigned constant
Mans Rullgard [Sat, 26 Nov 2011 16:51:44 +0000 (16:51 +0000)]
s3tc: fix shift overflow by using unsigned constant

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agosvq1enc: fix signed multiplication overflow
Mans Rullgard [Sat, 26 Nov 2011 16:50:51 +0000 (16:50 +0000)]
svq1enc: fix signed multiplication overflow

This multiplication can overflow the signed range but not the
unsigned.  After right-shifting it will thus fit in the signed
range again.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agosvq1dec: use sign_extend()
Mans Rullgard [Sat, 26 Nov 2011 15:16:29 +0000 (15:16 +0000)]
svq1dec: use sign_extend()

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agorv34: fix signed multiplication overflow
Mans Rullgard [Sat, 26 Nov 2011 14:50:23 +0000 (14:50 +0000)]
rv34: fix signed multiplication overflow

Multiply by unsigned 0x01010101 to replicate bytes into 32-bit
word.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years ago[PATCH] Fix crash when initializing multi-threaded decoding for corrupted file.
Cheng Sun [Sat, 26 Nov 2011 18:45:36 +0000 (10:45 -0800)]
[PATCH] Fix crash when initializing multi-threaded decoding for corrupted file.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
7 years agosol: use AV_RL32 instead of raw pointer cast
Mans Rullgard [Sat, 26 Nov 2011 00:22:28 +0000 (00:22 +0000)]
sol: use AV_RL32 instead of raw pointer cast

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agocook: use a table for xor key rotation
Mans Rullgard [Fri, 25 Nov 2011 20:43:05 +0000 (20:43 +0000)]
cook: use a table for xor key rotation

This is simpler and avoids an invalid shift by 32 in the
aligned case.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoapedec: fix signed integer overflows
Mans Rullgard [Fri, 25 Nov 2011 18:46:35 +0000 (18:46 +0000)]
apedec: fix signed integer overflows

This bit manipulation is equivalent but avoids undefined
shifts and overflows.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoqtrle: simplify 32-bit decode using intreadwrite macros
Mans Rullgard [Fri, 25 Nov 2011 18:16:44 +0000 (18:16 +0000)]
qtrle: simplify 32-bit decode using intreadwrite macros

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agogif: fix invalid signed shifts
Mans Rullgard [Fri, 25 Nov 2011 17:51:10 +0000 (17:51 +0000)]
gif: fix invalid signed shifts

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agortjpeg: simplify get_block() by using get_bits_left.
Reimar Döffinger [Sat, 5 Nov 2011 21:07:22 +0000 (22:07 +0100)]
rtjpeg: simplify get_block() by using get_bits_left.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agortjpeg: check get_block return value for error.
Reimar Döffinger [Sat, 5 Nov 2011 21:26:42 +0000 (22:26 +0100)]
rtjpeg: check get_block return value for error.

This avoids crashes due to reading out-of-bounds.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agonuv: check per-frame header for validity.
Reimar Döffinger [Sat, 5 Nov 2011 22:13:07 +0000 (23:13 +0100)]
nuv: check per-frame header for validity.

Since it contains dimensions parsing an invalid one has rather
annoying effects.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agonuv: use FFALIGN.
Reimar Döffinger [Sat, 5 Nov 2011 20:11:37 +0000 (21:11 +0100)]
nuv: use FFALIGN.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agoflicvideo: check extradata_size before accessing extradata.
Carl Eugen Hoyos [Wed, 9 Nov 2011 13:04:22 +0000 (14:04 +0100)]
flicvideo: check extradata_size before accessing extradata.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agofraps: check for overread.
Michael Niedermayer [Wed, 9 Nov 2011 18:40:59 +0000 (19:40 +0100)]
fraps: check for overread.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agocook: fix apparent typo in extradata parsing
K.Y.H [Thu, 10 Nov 2011 23:30:39 +0000 (00:30 +0100)]
cook: fix apparent typo in extradata parsing

Signed-off-by: Anton Khirnov <anton@khirnov.net>
7 years agolavc: update doxy to use nondeprecated API.
Anton Khirnov [Fri, 18 Nov 2011 10:16:23 +0000 (11:16 +0100)]
lavc: update doxy to use nondeprecated API.

7 years agoswscale: Readd #define _SVID_SOURCE
Martin Storsjö [Thu, 24 Nov 2011 22:45:16 +0000 (00:45 +0200)]
swscale: Readd #define _SVID_SOURCE

This was removed erroneously in
046f081b46c8479820409cf8f530b988221bd15b. This define still is
necessary for getting MAP_ANONYMOUS defined on linux/glibc,
despite the define reshuffling done in that commit.

Without MAP_ANONYMOUS defined, the mprotect calls for setting the
generated mmx2 scaler code pages executable are left out, causing
crashes if that codepath is chosen.

This patch fixes scaling from 192x144 to 320x240 with
-sws_flags fast_bilinear, which crashes on linux at the
moment.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agoFix av_get_channel_layout_string() for positions >31
Mans Rullgard [Fri, 25 Nov 2011 14:58:50 +0000 (14:58 +0000)]
Fix av_get_channel_layout_string() for positions >31

A value shifted left by >31 needs to have a 64-bit type.
As there are no defined channels in this range, the fix
is purely theoretical at this point, although it does
avoid some invalid shifts triggering the overflow
checker.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agoconfigure: Store vda lib flags in extralibs instead of ldflags
Rafaël Carré [Fri, 25 Nov 2011 00:43:46 +0000 (19:43 -0500)]
configure: Store vda lib flags in extralibs instead of ldflags

This way the needed linking flags end up in libavcodec.pc.

Signed-off-by: Martin Storsjö <martin@martin.st>
7 years agoMake channel layout masks unsigned
Mans Rullgard [Fri, 25 Nov 2011 12:51:57 +0000 (12:51 +0000)]
Make channel layout masks unsigned

It makes more sense for a bit mask to use an unsigned type.
The change should be source and binary compatible on all
supported systems, hence micro version bump.

Fixes a few invalid shifts.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agodca: ARMv6 optimised decode_blockcode()
Mans Rullgard [Sun, 23 Oct 2011 16:39:49 +0000 (17:39 +0100)]
dca: ARMv6 optimised decode_blockcode()

This is a hand-tuned version of the code with impossible parts of
the FASTDIV function ommitted.

2-5% faster overall on Cortex-A8.

Signed-off-by: Mans Rullgard <mans@mansr.com>
7 years agonullenc: drop AVFMT_RAWPICTURE from the flags
Luca Barbato [Thu, 24 Nov 2011 22:03:25 +0000 (23:03 +0100)]
nullenc: drop AVFMT_RAWPICTURE from the flags

This makes the two pass encoding with x264 working with -f null - as
first pass.