[FFmpeg-cvslog] avformat/utils: Fix type mismatch

ffmpeg | branch: master | Michael Niedermayer | Sun Nov 27 03:39:20 2016 +0100| [a06e84b56e936ff3ca090f53d81f9cbc3514e0e0] | committer: Michael Niedermayer avformat/utils: Fix type mismatch Signed-off-by: Michael Niedermayer > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=a06e84b

[FFmpeg-cvslog] libschroedingerdec: don't produce empty frames

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sun Nov 13 22:59:47 2016 +0100| [70ca4ce17a0a56118998afb81374d5b6d287182c] | committer: Andreas Cadhalpun libschroedingerdec: don't produce empty frames They are not valid and can cause problems/crashes for API users. Reviewed-by: Michael Nie

[FFmpeg-cvslog] escape124: reject codebook size 0

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Wed Nov 9 00:38:50 2016 +0100| [dc821d42a2fd79ac64161f50decf5922e431445a] | committer: Andreas Cadhalpun escape124: reject codebook size 0 It causes a cb_depth of 32, leading to assertion failures in get_bits. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Tue Nov 8 00:42:23 2016 +0100| [3148d1c25f229e272a9242bfd81eacd6ce3ff716] | committer: Andreas Cadhalpun matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header The code assumes that s->streams[0] is valid. Reviewed-by: M

[FFmpeg-cvslog] dvbsubdec: fix division by zero in compute_default_clut

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Tue Nov 8 22:32:42 2016 +0100| [2260c0776ac664af2d4b09e177d5f1d6fc7864d7] | committer: Andreas Cadhalpun dvbsubdec: fix division by zero in compute_default_clut This problem was introduced in commit 4b90dcb8493552c17a811c8b1e6538dae4061f9d.

[FFmpeg-cvslog] smacker: limit recursion depth of smacker_decode_bigtree

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sat Nov 19 14:21:11 2016 +0100| [8a7b2fbf6f4d902b7ac245ceae7ee1ced5ff9545] | committer: Andreas Cadhalpun smacker: limit recursion depth of smacker_decode_bigtree This fixes segmentation faults due to stack-overflow caused by too deep recursio

[FFmpeg-cvslog] ppc: pixblockdsp: do unaligned block accesses correctly again

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Wed Nov 2 21:28:49 2016 +0100| [a94f846e2dbac02191551fc0198282ac57a8ffe5] | committer: Andreas Cadhalpun ppc: pixblockdsp: do unaligned block accesses correctly again This was broken by the following Libav commit: 4c387c7 ppc: dsputil: do una

[FFmpeg-cvslog] libopusdec: default to stereo for invalid number of channels

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Mon Nov 14 21:41:45 2016 +0100| [56b120630fd2cd39134fc25cb41caea0b620df7c] | committer: Andreas Cadhalpun libopusdec: default to stereo for invalid number of channels This fixes an out-of-bounds read if avc->channels is 0. Reviewed-by: Michae

[FFmpeg-cvslog] mpegts: prevent division by zero

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Mon Nov 7 23:37:59 2016 +0100| [b3ac458a5a207794dd32a06b66bcdd3ccf5c8b39] | committer: Andreas Cadhalpun mpegts: prevent division by zero Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit 1bbb18fe8

[FFmpeg-cvslog] proresdec_lgpl: explicitly check coff[3] against slice_data_size

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Wed Nov 9 23:49:46 2016 +0100| [5c55f9881ed032ff83515f4e39216dee97685a5a] | committer: Andreas Cadhalpun proresdec_lgpl: explicitly check coff[3] against slice_data_size The implicit checks via v_data_size and a_data_size don't work in the ca

[FFmpeg-cvslog] icodec: fix leaking pkt on error

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Tue Nov 8 23:53:52 2016 +0100| [8a56b31e7cd1af1e3de59bfe3e79d5d7171404d4] | committer: Andreas Cadhalpun icodec: fix leaking pkt on error Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit 467eece1b

[FFmpeg-cvslog] icodec: correctly check avio_read return value

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Tue Nov 8 23:29:28 2016 +0100| [71fa32bbb7478a865ba9e955393766f733fdda50] | committer: Andreas Cadhalpun icodec: correctly check avio_read return value It can read less than the requested amount, in which case buf contains uninitialized data,

[FFmpeg-cvslog] filmstripdec: correctly check image dimensions

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sun Nov 13 18:22:12 2016 +0100| [d8ec9e97b93eddd1b44e22356e3e242649476cc9] | committer: Andreas Cadhalpun filmstripdec: correctly check image dimensions This prevents a division by zero in read_packet. Reviewed-by: Paul B Mahol Signed-off-by

[FFmpeg-cvslog] cavsdec: unref frame before referencing again

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Oct 20 22:14:22 2016 +0200| [51ff17d6b993fed3eaeddf7e718391f905eec047] | committer: Andreas Cadhalpun cavsdec: unref frame before referencing again This fixes asserts (from commit 13aae8) in av_frame_ref and av_frame_move_ref. Reviewed-by

[FFmpeg-cvslog] mpegaudio_parser: don't return AVERROR_PATCHWELCOME

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Mon Nov 7 01:16:14 2016 +0100| [f964046c58fb4ce4aaa31e5ed196f9336e03481b] | committer: Andreas Cadhalpun mpegaudio_parser: don't return AVERROR_PATCHWELCOME The API does not allow returning AVERROR codes. It triggers an assert in av_parser_p

[FFmpeg-cvslog] avformat: prevent triggering request_probe assert in ff_read_packet

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Wed Oct 19 19:23:49 2016 +0200| [e14da0578c85721ac6f59ddcd9fb4c778bfc5d09] | committer: Andreas Cadhalpun avformat: prevent triggering request_probe assert in ff_read_packet If probe_codec is called with pkt == NULL, it sets probe_packets to 0

[FFmpeg-cvslog] mss2: only use error correction for matching block counts

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Nov 24 23:57:46 2016 +0100| [028c87be95dd85329f5c68d78e152d859afec7e9] | committer: Andreas Cadhalpun mss2: only use error correction for matching block counts This fixes a heap-buffer-overflow in ff_er_frame_end when decoding mss2 with co

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference in mxf_read_packet_old

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Nov 17 22:53:51 2016 +0100| [8f27508f1cad0adca6486de9265290a6de4e4357] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference in mxf_read_packet_old Metadata streams have priv_data set to NULL. Reviewed-by: Josh de Kock Sign

[FFmpeg-cvslog] smvjpegdec: make sure cur_frame is not negative

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Nov 10 22:09:03 2016 +0100| [f76947fd567f20287bc8ca8e692a59026bac19e7] | committer: Andreas Cadhalpun smvjpegdec: make sure cur_frame is not negative This fixes a heap-buffer-overflow detected by AddressSanitizer. Reviewed-by: Michael Nie

[FFmpeg-cvslog] mpeg12dec: unref discarded picture from extradata

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Oct 20 22:51:55 2016 +0200| [c5fb9df38aac5ff5e6c136f2bb7d946822ba4cc9] | committer: Andreas Cadhalpun mpeg12dec: unref discarded picture from extradata Otherwise another frame gets referenced into picture, triggering an assert (from commit

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sat Nov 5 00:17:53 2016 +0100| [0cc619e0d7567894bc0f92de963be050c3679492] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference Metadata streams have priv_data set to NULL. Reviewed-by: Michael Niedermayer Signed-off-by: Andre

[FFmpeg-cvslog] softfloat: decrease MIN_EXP to cover full float range

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Fri Nov 25 00:26:51 2016 +0100| [b45e112bbdcc44a96907662136e49514a204543a] | committer: Andreas Cadhalpun softfloat: decrease MIN_EXP to cover full float range floats are not necessarily normalized, so a normalized softfloat needs MIN_EXP lowe

[FFmpeg-cvslog] libschroedingerdec: fix leaking of framewithpts

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sun Nov 13 23:10:06 2016 +0100| [7552f6fc1bbb5d3a7f1d4c1e9ed89ae1f2e2299b] | committer: Andreas Cadhalpun libschroedingerdec: fix leaking of framewithpts Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from c

[FFmpeg-cvslog] sbgdec: prevent NULL pointer access

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Thu Nov 10 22:21:20 2016 +0100| [22cd4aa22116a365d611091916a3eff74e21ec3f] | committer: Andreas Cadhalpun sbgdec: prevent NULL pointer access Reviewed-by: Josh de Kock Signed-off-by: Andreas Cadhalpun (cherry picked from commit dbefbb61b785c

[FFmpeg-cvslog] Update Changelog

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sun Nov 27 00:47:03 2016 +0100| [970781f5f29d3f7c9b1238b7ac6d94d70048b2f0] | committer: Andreas Cadhalpun Update Changelog > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=970781f5f29d3f7c9b1238b7ac6d94d70048b2f0 --- Changelog | 2

[FFmpeg-cvslog] pnmdec: make sure v is capped by maxval

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Wed Nov 9 01:09:35 2016 +0100| [0b948b1b8d100491934f00bc6d72009de0c99b05] | committer: Andreas Cadhalpun pnmdec: make sure v is capped by maxval Otherwise put_bits can be called with a value that doesn't fit in the sample_len, causing an asse

[FFmpeg-cvslog] softfloat: handle -INT_MAX correctly

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Sun Nov 13 20:52:02 2016 +0100| [ccda73a711ba0d194e97fac476f9676183281ae7] | committer: Andreas Cadhalpun softfloat: handle -INT_MAX correctly This is similar to commit 9ac61e73d0843ec4b83f4e3d47eded73234e406e. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] diracdec: check return code of get_buffer_with_edge

ffmpeg | branch: release/2.8 | Andreas Cadhalpun | Fri Nov 4 19:00:17 2016 +0100| [0e8c44076d313c2addc26c2e03c9e6430fa8cecf] | committer: Andreas Cadhalpun diracdec: check return code of get_buffer_with_edge If it fails, buffers aren't allocated, causing NULL pointer dereferencing. Reviewed-

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference in mxf_read_packet_old

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Nov 17 22:53:51 2016 +0100| [1c282152c1c7ca23c388ecb04e9997030fbba19a] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference in mxf_read_packet_old Metadata streams have priv_data set to NULL. Reviewed-by: Josh de Kock Sign

[FFmpeg-cvslog] diracdec: check return code of get_buffer_with_edge

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Fri Nov 4 19:00:17 2016 +0100| [dcc8d2418acac6539e2533fc046f3d00f1c0c333] | committer: Andreas Cadhalpun diracdec: check return code of get_buffer_with_edge If it fails, buffers aren't allocated, causing NULL pointer dereferencing. Reviewed-

[FFmpeg-cvslog] libopusdec: default to stereo for invalid number of channels

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Mon Nov 14 21:41:45 2016 +0100| [dbad79248704d4b7cfcabbe46a734c51c1ada9a0] | committer: Andreas Cadhalpun libopusdec: default to stereo for invalid number of channels This fixes an out-of-bounds read if avc->channels is 0. Reviewed-by: Michae

[FFmpeg-cvslog] pgssubdec: only set w/h/linesize when allocating data

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Nov 9 23:23:16 2016 +0100| [5d2f1ffef1f26e4d66c8e63ab1101da179af11e6] | committer: Andreas Cadhalpun pgssubdec: only set w/h/linesize when allocating data Rects with positive w/h/linesize but no data are invalid. Reviewed-by: Petri Hintu

[FFmpeg-cvslog] interplayacm: increase bitstream buffer size by AV_INPUT_BUFFER_PADDING_SIZE

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Oct 30 21:18:20 2016 +0100| [aca7f5f0607789b26b1b5d0c2dbe3ae39a229806] | committer: Andreas Cadhalpun interplayacm: increase bitstream buffer size by AV_INPUT_BUFFER_PADDING_SIZE This fixes out-of-bounds reads by the bitstream reader. Rev

[FFmpeg-cvslog] sbgdec: prevent NULL pointer access

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Nov 10 22:21:20 2016 +0100| [80b85300ae0b8c6c1bb813a4390d94e3ea87b1da] | committer: Andreas Cadhalpun sbgdec: prevent NULL pointer access Reviewed-by: Josh de Kock Signed-off-by: Andreas Cadhalpun (cherry picked from commit dbefbb61b785c

[FFmpeg-cvslog] mpeg12dec: unref discarded picture from extradata

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Oct 20 22:51:55 2016 +0200| [30d542d55ddbcb2edb0f1351e1ecb6ad286b37ff] | committer: Andreas Cadhalpun mpeg12dec: unref discarded picture from extradata Otherwise another frame gets referenced into picture, triggering an assert (from commit

[FFmpeg-cvslog] softfloat: handle -INT_MAX correctly

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Nov 13 20:52:02 2016 +0100| [48496e4d4fadd01ff3ce45d2540e8cf75ac5cd89] | committer: Andreas Cadhalpun softfloat: handle -INT_MAX correctly This is similar to commit 9ac61e73d0843ec4b83f4e3d47eded73234e406e. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] smacker: limit recursion depth of smacker_decode_bigtree

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sat Nov 19 14:21:11 2016 +0100| [48d24cca1308d7cc78aa99827b5449a8577a68a2] | committer: Andreas Cadhalpun smacker: limit recursion depth of smacker_decode_bigtree This fixes segmentation faults due to stack-overflow caused by too deep recursio

[FFmpeg-cvslog] libschroedingerdec: don't produce empty frames

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Nov 13 22:59:47 2016 +0100| [57665e04e22a201d3c56072532c15ce6ba71353a] | committer: Andreas Cadhalpun libschroedingerdec: don't produce empty frames They are not valid and can cause problems/crashes for API users. Reviewed-by: Michael Nie

[FFmpeg-cvslog] lzf: update pointer p after realloc

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Fri Nov 4 22:58:49 2016 +0100| [ef2d91e9c337f50edbc7631485bfec385601f4bb] | committer: Andreas Cadhalpun lzf: update pointer p after realloc This fixes heap-use-after-free detected by AddressSanitizer. Reviewed-by: Luca Barbato Signed-off-b

[FFmpeg-cvslog] interplayacm: validate number of channels

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Oct 30 21:41:11 2016 +0100| [aa32d415275613e406a43d8f091d6981e4e1b872] | committer: Andreas Cadhalpun interplayacm: validate number of channels The number of channels is used as divisor in decode_frame, so it must not be zero to avoid SIGF

[FFmpeg-cvslog] smvjpegdec: make sure cur_frame is not negative

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Nov 10 22:09:03 2016 +0100| [69673d02790f7917d11055236ecfa20c9ff8771a] | committer: Andreas Cadhalpun smvjpegdec: make sure cur_frame is not negative This fixes a heap-buffer-overflow detected by AddressSanitizer. Reviewed-by: Michael Nie

[FFmpeg-cvslog] escape124: reject codebook size 0

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Nov 9 00:38:50 2016 +0100| [e14cc2f1975b608b630725007bda10f49aa096a1] | committer: Andreas Cadhalpun escape124: reject codebook size 0 It causes a cb_depth of 32, leading to assertion failures in get_bits. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Tue Nov 8 00:42:23 2016 +0100| [5801482379cb053c2901da1058e86365aac563a1] | committer: Andreas Cadhalpun matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header The code assumes that s->streams[0] is valid. Reviewed-by: M

[FFmpeg-cvslog] mpegts: prevent division by zero

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Mon Nov 7 23:37:59 2016 +0100| [3d82cebdd2bd5424b0eb37b0e39284da19fa0e5b] | committer: Andreas Cadhalpun mpegts: prevent division by zero Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit 1bbb18fe8

[FFmpeg-cvslog] softfloat: decrease MIN_EXP to cover full float range

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Fri Nov 25 00:26:51 2016 +0100| [88bf1d2749624cf78be0c5a5d74c169ad16ba99c] | committer: Andreas Cadhalpun softfloat: decrease MIN_EXP to cover full float range floats are not necessarily normalized, so a normalized softfloat needs MIN_EXP lowe

[FFmpeg-cvslog] mss2: only use error correction for matching block counts

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Nov 24 23:57:46 2016 +0100| [0496403c08ab35b20490a48aa9e3fdbd4d3bf27d] | committer: Andreas Cadhalpun mss2: only use error correction for matching block counts This fixes a heap-buffer-overflow in ff_er_frame_end when decoding mss2 with co

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sat Nov 5 00:17:53 2016 +0100| [e78d9f3f35e379ab7da729360e3eb5ad39c150fa] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference Metadata streams have priv_data set to NULL. Reviewed-by: Michael Niedermayer Signed-off-by: Andre

[FFmpeg-cvslog] dcstr: fix division by zero

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Oct 20 20:13:54 2016 +0200| [e8ab2bd2ac85ad0b1013f247d35032b7cd7f771e] | committer: Andreas Cadhalpun dcstr: fix division by zero Also check for possible overflows. Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cher

[FFmpeg-cvslog] avformat/icodec: Fix crash probing fuzzed file

ffmpeg | branch: release/3.0 | Mark Harris | Mon Feb 15 23:52:13 2016 -0800| [9375a7d85e8bc78dbb5cc101c37ff7c51f7d9b24] | committer: Andreas Cadhalpun avformat/icodec: Fix crash probing fuzzed file Avoid invalid memory read/crash when frame offset >= 0xfff8. Base64-encoded example: AAABADA

[FFmpeg-cvslog] interplayacm: check for too large b

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Oct 30 20:47:22 2016 +0100| [a1e6daeb1e92abf59e6b5a4229948ca54967f759] | committer: Andreas Cadhalpun interplayacm: check for too large b This fixes out-of-bounds reads. Reviewed-by: Paul B Mahol Signed-off-by: Andreas Cadhalpun (cherry

[FFmpeg-cvslog] mpegaudio_parser: don't return AVERROR_PATCHWELCOME

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Mon Nov 7 01:16:14 2016 +0100| [e6197a6ce986af0d2f29ed0b649f3ed531a6d66e] | committer: Andreas Cadhalpun mpegaudio_parser: don't return AVERROR_PATCHWELCOME The API does not allow returning AVERROR codes. It triggers an assert in av_parser_p

[FFmpeg-cvslog] rsd: limit number of channels

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Oct 19 23:40:41 2016 +0200| [45b18fbb9a7729c91d69d6359a782a0135b5f2b8] | committer: Andreas Cadhalpun rsd: limit number of channels Negative values don't make sense and too large values can cause overflows. For AV_CODEC_ID_ADPCM_THP this l

[FFmpeg-cvslog] avformat: prevent triggering request_probe assert in ff_read_packet

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Oct 19 19:23:49 2016 +0200| [0009cf348aa85957dbbe262926e432896a628ff3] | committer: Andreas Cadhalpun avformat: prevent triggering request_probe assert in ff_read_packet If probe_codec is called with pkt == NULL, it sets probe_packets to 0

[FFmpeg-cvslog] proresdec_lgpl: explicitly check coff[3] against slice_data_size

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Nov 9 23:49:46 2016 +0100| [416a8a06b977d8b3d113945701d9690abfdd7622] | committer: Andreas Cadhalpun proresdec_lgpl: explicitly check coff[3] against slice_data_size The implicit checks via v_data_size and a_data_size don't work in the ca

[FFmpeg-cvslog] ppc: pixblockdsp: do unaligned block accesses correctly again

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Nov 2 21:28:49 2016 +0100| [087b77741526e701d392c038ae765993d60b5fab] | committer: Andreas Cadhalpun ppc: pixblockdsp: do unaligned block accesses correctly again This was broken by the following Libav commit: 4c387c7 ppc: dsputil: do una

[FFmpeg-cvslog] pnmdec: make sure v is capped by maxval

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Wed Nov 9 01:09:35 2016 +0100| [b32c9941a21ef44b95489c878b4f6a40c077eb47] | committer: Andreas Cadhalpun pnmdec: make sure v is capped by maxval Otherwise put_bits can be called with a value that doesn't fit in the sample_len, causing an asse

[FFmpeg-cvslog] icodec: fix leaking pkt on error

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Tue Nov 8 23:53:52 2016 +0100| [c3307f7e9e1889c15dcaad1247c8212628474bc5] | committer: Andreas Cadhalpun icodec: fix leaking pkt on error Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit 467eece1b

[FFmpeg-cvslog] icodec: add ico_read_close to fix leaking ico->images

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Tue Nov 8 23:54:41 2016 +0100| [05e6606ba98a53f10213b2f8b07afc3e62731c5a] | committer: Andreas Cadhalpun icodec: add ico_read_close to fix leaking ico->images Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked

[FFmpeg-cvslog] libschroedingerdec: fix leaking of framewithpts

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Nov 13 23:10:06 2016 +0100| [4ffd5805af4a8734d238e561daa63c7b3be5eedb] | committer: Andreas Cadhalpun libschroedingerdec: fix leaking of framewithpts Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from c

[FFmpeg-cvslog] cavsdec: unref frame before referencing again

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Thu Oct 20 22:14:22 2016 +0200| [1d439041ece002c6b2e8e3d57c3b389f9fc279b6] | committer: Andreas Cadhalpun cavsdec: unref frame before referencing again This fixes asserts (from commit 13aae8) in av_frame_ref and av_frame_move_ref. Reviewed-by

[FFmpeg-cvslog] icodec: correctly check avio_read return value

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Tue Nov 8 23:29:28 2016 +0100| [3047b0a4a3e749e62e413f8fc96ae2e9fe228477] | committer: Andreas Cadhalpun icodec: correctly check avio_read return value It can read less than the requested amount, in which case buf contains uninitialized data,

[FFmpeg-cvslog] filmstripdec: correctly check image dimensions

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Sun Nov 13 18:22:12 2016 +0100| [e93934e100d41ddb08bb85bf9023ee1716d78718] | committer: Andreas Cadhalpun filmstripdec: correctly check image dimensions This prevents a division by zero in read_packet. Reviewed-by: Paul B Mahol Signed-off-by

[FFmpeg-cvslog] dvbsubdec: fix division by zero in compute_default_clut

ffmpeg | branch: release/3.0 | Andreas Cadhalpun | Tue Nov 8 22:32:42 2016 +0100| [27888d13b8d370b365593312d6755d268a4428f6] | committer: Andreas Cadhalpun dvbsubdec: fix division by zero in compute_default_clut This problem was introduced in commit 4b90dcb8493552c17a811c8b1e6538dae4061f9d.

[FFmpeg-cvslog] rmdec: validate block alignment

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 17 22:46:40 2016 +0100| [e2de6f31c0dbc09033f63a94a3795488065d3b6a] | committer: Andreas Cadhalpun rmdec: validate block alignment This fixes division by zero crashes. Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun

[FFmpeg-cvslog] flvdec: require need_context_update when changing codec id

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Fri Nov 4 21:37:13 2016 +0100| [d0f8741a5a29e6381894fe5eacbeb9145a965b6c] | committer: Andreas Cadhalpun flvdec: require need_context_update when changing codec id Otherwise the codec context and codecpar might disagree on the codec id, trigg

[FFmpeg-cvslog] libopusdec: default to stereo for invalid number of channels

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Mon Nov 14 21:41:45 2016 +0100| [e70caba38480f11043963bb5eb150cfb2eebd41b] | committer: Andreas Cadhalpun libopusdec: default to stereo for invalid number of channels This fixes an out-of-bounds read if avc->channels is 0. Reviewed-by: Michae

[FFmpeg-cvslog] interplayacm: check for too large b

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 30 20:47:22 2016 +0100| [5ede8a9d8c263ff2741bf6a6c54b76287be2af36] | committer: Andreas Cadhalpun interplayacm: check for too large b This fixes out-of-bounds reads. Reviewed-by: Paul B Mahol Signed-off-by: Andreas Cadhalpun (cherry

[FFmpeg-cvslog] interplayacm: validate number of channels

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 30 21:41:11 2016 +0100| [d6fbc7a2daf0311fa7c093d48bcf19d1cf35936a] | committer: Andreas Cadhalpun interplayacm: validate number of channels The number of channels is used as divisor in decode_frame, so it must not be zero to avoid SIGF

[FFmpeg-cvslog] sbgdec: prevent NULL pointer access

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 10 22:21:20 2016 +0100| [312757eb848b96bcc1e0df17312b11a24c4f18d3] | committer: Andreas Cadhalpun sbgdec: prevent NULL pointer access Reviewed-by: Josh de Kock Signed-off-by: Andreas Cadhalpun (cherry picked from commit dbefbb61b785c

[FFmpeg-cvslog] mpeg12dec: unref discarded picture from extradata

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Oct 20 22:51:55 2016 +0200| [facf964d37eae50d1fb5c354e8d4ab897a624e45] | committer: Andreas Cadhalpun mpeg12dec: unref discarded picture from extradata Otherwise another frame gets referenced into picture, triggering an assert (from commit

[FFmpeg-cvslog] pgssubdec: only set w/h/linesize when allocating data

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 9 23:23:16 2016 +0100| [9b506280dd9b95b944676316ae3f8ea5605a7a10] | committer: Andreas Cadhalpun pgssubdec: only set w/h/linesize when allocating data Rects with positive w/h/linesize but no data are invalid. Reviewed-by: Petri Hintu

[FFmpeg-cvslog] interplayacm: increase bitstream buffer size by AV_INPUT_BUFFER_PADDING_SIZE

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 30 21:18:20 2016 +0100| [5a1433b19ab396cfdc7d52d9b479a7828ce5c707] | committer: Andreas Cadhalpun interplayacm: increase bitstream buffer size by AV_INPUT_BUFFER_PADDING_SIZE This fixes out-of-bounds reads by the bitstream reader. Rev

[FFmpeg-cvslog] filmstripdec: correctly check image dimensions

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Nov 13 18:22:12 2016 +0100| [52d8c1e474c3e11fed06cf17e22c1e59406a0ce3] | committer: Andreas Cadhalpun filmstripdec: correctly check image dimensions This prevents a division by zero in read_packet. Reviewed-by: Paul B Mahol Signed-off-by

[FFmpeg-cvslog] proresdec_lgpl: explicitly check coff[3] against slice_data_size

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 9 23:49:46 2016 +0100| [727ec4acc471b167f8af5211e860e2def6d47d02] | committer: Andreas Cadhalpun proresdec_lgpl: explicitly check coff[3] against slice_data_size The implicit checks via v_data_size and a_data_size don't work in the ca

[FFmpeg-cvslog] diracdec: check return code of get_buffer_with_edge

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Fri Nov 4 19:00:17 2016 +0100| [cb0b8182448cd337f8c33cd8686ba0eb04a1] | committer: Andreas Cadhalpun diracdec: check return code of get_buffer_with_edge If it fails, buffers aren't allocated, causing NULL pointer dereferencing. Reviewed-

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference in mxf_read_packet_old

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 17 22:53:51 2016 +0100| [315f1dea84a3865dfaf949f99c9828a5b8dd4bcc] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference in mxf_read_packet_old Metadata streams have priv_data set to NULL. Reviewed-by: Josh de Kock Sign

[FFmpeg-cvslog] pnmdec: make sure v is capped by maxval

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 9 01:09:35 2016 +0100| [a5ba9eab44da13bb0683193e2382f1bfd853a47e] | committer: Andreas Cadhalpun pnmdec: make sure v is capped by maxval Otherwise put_bits can be called with a value that doesn't fit in the sample_len, causing an asse

[FFmpeg-cvslog] smvjpegdec: make sure cur_frame is not negative

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 10 22:09:03 2016 +0100| [eaf79ac2d9c18bce3c1990dbf6722e90d9c788b1] | committer: Andreas Cadhalpun smvjpegdec: make sure cur_frame is not negative This fixes a heap-buffer-overflow detected by AddressSanitizer. Reviewed-by: Michael Nie

[FFmpeg-cvslog] dvbsubdec: fix division by zero in compute_default_clut

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Tue Nov 8 22:32:42 2016 +0100| [5c2e26275cb130293691ed0b335db0a67e8abbcf] | committer: Andreas Cadhalpun dvbsubdec: fix division by zero in compute_default_clut This problem was introduced in commit 4b90dcb8493552c17a811c8b1e6538dae4061f9d.

[FFmpeg-cvslog] mss2: only use error correction for matching block counts

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 24 23:57:46 2016 +0100| [072246993acb9ce22f88e3df697f78dba61fcdd3] | committer: Andreas Cadhalpun mss2: only use error correction for matching block counts This fixes a heap-buffer-overflow in ff_er_frame_end when decoding mss2 with co

[FFmpeg-cvslog] rsd: limit number of channels

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Oct 19 23:40:41 2016 +0200| [13f032abbb26c7b47d700745f7ace05375eae34f] | committer: Andreas Cadhalpun rsd: limit number of channels Negative values don't make sense and too large values can cause overflows. For AV_CODEC_ID_ADPCM_THP this l

[FFmpeg-cvslog] softfloat: decrease MIN_EXP to cover full float range

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Fri Nov 25 00:26:51 2016 +0100| [5d1502d4b68c458522e5a6fc446d5b0e5f88bffb] | committer: Andreas Cadhalpun softfloat: decrease MIN_EXP to cover full float range floats are not necessarily normalized, so a normalized softfloat needs MIN_EXP lowe

[FFmpeg-cvslog] exr: reindent after previous commit

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 16 20:47:35 2016 +0100| [cb936d62664c25909b320ec230b41b5a2b9c9ed3] | committer: Andreas Cadhalpun exr: reindent after previous commit Signed-off-by: Andreas Cadhalpun (cherry picked from commit ce3147eb198770b558acf6c05f33cb807a413707

[FFmpeg-cvslog] ppc: pixblockdsp: do unaligned block accesses correctly again

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 2 21:28:49 2016 +0100| [e3f671b1014095cf299180f4f052a4daaec567f6] | committer: Andreas Cadhalpun ppc: pixblockdsp: do unaligned block accesses correctly again This was broken by the following Libav commit: 4c387c7 ppc: dsputil: do una

[FFmpeg-cvslog] smacker: limit recursion depth of smacker_decode_bigtree

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sat Nov 19 14:21:11 2016 +0100| [53e1493cb5a0977407c7869ec9adf4555ef07c66] | committer: Andreas Cadhalpun smacker: limit recursion depth of smacker_decode_bigtree This fixes segmentation faults due to stack-overflow caused by too deep recursio

[FFmpeg-cvslog] exr: fix out-of-bounds read

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 16 20:46:56 2016 +0100| [71378e7937bd458233b75b4d73f0aabc3ac437fa] | committer: Andreas Cadhalpun exr: fix out-of-bounds read channel_index can be -1. This problem was introduced in commit 2dd7b46132e2801ef34fe1b5c27e0113cdcfa2f9. Re

[FFmpeg-cvslog] libschroedingerdec: don't produce empty frames

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Nov 13 22:59:47 2016 +0100| [89a22d3fbff3db11a0a09e55778ed0e041210327] | committer: Andreas Cadhalpun libschroedingerdec: don't produce empty frames They are not valid and can cause problems/crashes for API users. Reviewed-by: Michael Nie

[FFmpeg-cvslog] softfloat: handle -INT_MAX correctly

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Nov 13 20:52:02 2016 +0100| [d000e66c4f2f6275fc24a424ae85f7e092293347] | committer: Andreas Cadhalpun softfloat: handle -INT_MAX correctly This is similar to commit 9ac61e73d0843ec4b83f4e3d47eded73234e406e. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] cavsdec: unref frame before referencing again

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Oct 20 22:14:22 2016 +0200| [72f1701c92f7f021d00c57f4f928efa719fcd53d] | committer: Andreas Cadhalpun cavsdec: unref frame before referencing again This fixes asserts (from commit 13aae8) in av_frame_ref and av_frame_move_ref. Reviewed-by

[FFmpeg-cvslog] ffmdec: validate codec parameters

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Nov 17 00:04:57 2016 +0100| [b4f42e5c85f589556ed486fc743fbe16f8ed88c8] | committer: Andreas Cadhalpun ffmdec: validate codec parameters A negative extradata size for example gets passed to memcpy in avcodec_parameters_from_context causing

[FFmpeg-cvslog] dcstr: fix division by zero

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Thu Oct 20 20:13:54 2016 +0200| [d77684b85305e9dca9d056c50146f43c08268ac2] | committer: Andreas Cadhalpun dcstr: fix division by zero Also check for possible overflows. Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cher

[FFmpeg-cvslog] libschroedingerdec: fix leaking of framewithpts

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Nov 13 23:10:06 2016 +0100| [f70e9726dcad6dfe13b64b8878ee12a9e41063ed] | committer: Andreas Cadhalpun libschroedingerdec: fix leaking of framewithpts Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from c

[FFmpeg-cvslog] aiffdec: fix division by zero

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 16 22:29:13 2016 +0200| [230c04e3f6d720cc7fa17735b2ef91570417964d] | committer: Andreas Cadhalpun aiffdec: fix division by zero Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit c143a9c96ff9

[FFmpeg-cvslog] icodec: correctly check avio_read return value

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Tue Nov 8 23:29:28 2016 +0100| [c35a140e71710d815bab9581e928b42177feaf7e] | committer: Andreas Cadhalpun icodec: correctly check avio_read return value It can read less than the requested amount, in which case buf contains uninitialized data,

[FFmpeg-cvslog] mxfdec: fix NULL pointer dereference

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sat Nov 5 00:17:53 2016 +0100| [50d34cbf5ac86a42d3273765ce8292a06ae11158] | committer: Andreas Cadhalpun mxfdec: fix NULL pointer dereference Metadata streams have priv_data set to NULL. Reviewed-by: Michael Niedermayer Signed-off-by: Andre

[FFmpeg-cvslog] westwood_aud: prevent division by zero

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 16 22:42:32 2016 +0200| [d4f64a0f545e48224e985b50848153f0b779b8ff] | committer: Andreas Cadhalpun westwood_aud: prevent division by zero Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit bc7

[FFmpeg-cvslog] escape124: reject codebook size 0

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Nov 9 00:38:50 2016 +0100| [1499f65ad42a4f4519f27a7d3b01c55146ce2ad0] | committer: Andreas Cadhalpun escape124: reject codebook size 0 It causes a cb_depth of 32, leading to assertion failures in get_bits. Reviewed-by: Michael Niedermaye

[FFmpeg-cvslog] astdec: fix division by zero

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Sun Oct 16 22:39:47 2016 +0200| [b3991ccd1170d84f5b5d84566b0c78a42724a073] | committer: Andreas Cadhalpun astdec: fix division by zero Reviewed-by: Michael Niedermayer Signed-off-by: Andreas Cadhalpun (cherry picked from commit 9959a52b14bcf

[FFmpeg-cvslog] matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Tue Nov 8 00:42:23 2016 +0100| [c19e9657049d1ac67aee658b2f7ad12ba051b0cd] | committer: Andreas Cadhalpun matroskadec: fix NULL pointer dereference in webm_dash_manifest_read_header The code assumes that s->streams[0] is valid. Reviewed-by: M

[FFmpeg-cvslog] avformat: prevent triggering request_probe assert in ff_read_packet

ffmpeg | branch: release/3.1 | Andreas Cadhalpun | Wed Oct 19 19:23:49 2016 +0200| [d69dc10466e234c0e4ff9cb7e4d0e5d63ceeb357] | committer: Andreas Cadhalpun avformat: prevent triggering request_probe assert in ff_read_packet If probe_codec is called with pkt == NULL, it sets probe_packets to 0

  1   2   >