summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
...
* lavfi/dnn: Async Support for TensorFlow BackendShubhanshu Saxena2021-08-10
| | | | | | | | | | | | | | | | | | | This commit enables async execution in the TensorFlow backend and adds function to flush extra frames. The async execution mechanism executes the TFInferRequests on a separate thread which is joined before the next execution of same TFRequestItem/while freeing the model. The following is the comparison of this mechanism with the existing sync mechanism on TensorFlow C API 2.5 CPU variant. Async Mode: 4m32.846s Sync Mode: 5m17.582s The above was performed on super resolution filter using SRCNN model. Signed-off-by: Shubhanshu Saxena <shubhanshu.e01@gmail.com>
* lavfi/dnn_backend_tf: TFInferRequest Execution and DocumentationShubhanshu Saxena2021-08-10
| | | | | | | This commit adds a function for execution of TFInferRequest and documentation for functions related to TFInferRequest. Signed-off-by: Shubhanshu Saxena <shubhanshu.e01@gmail.com>
* lavfi/dnn: Common Function to Get Async Result in DNN BackendsShubhanshu Saxena2021-08-10
| | | | | | | This commits refactors the get async result function for common use in all three backends. Signed-off-by: Shubhanshu Saxena <shubhanshu.e01@gmail.com>
* lavfi/dnn: Add Async Execution Mechanism and DocumentationShubhanshu Saxena2021-08-10
| | | | | | | | | This commit adds an async execution mechanism for common use in the TensorFlow and Native backends. This commit also adds the documentation of typedefs and functions in the async module for common use in DNN backends. Signed-off-by: Shubhanshu Saxena <shubhanshu.e01@gmail.com>
* configure: [loongson] adjust MMI check in configureJin Bo2021-08-10
| | | | | | | | | | | After standardizing the use of 'pxor' in commit 'ebedd26', FFmpeg build failed with upstream compiler, for 'pxor' is not supported in time. This patch helps to workaround the build failure by checking whether 'pxor' is supported during configuration, if not, MMI will be disabled. Reviewed-by: yinshiyou-hf@loongson.cn Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* movenc: Don't try to fix the fragment end duration if none will be writtenMartin Storsjö2021-08-10
| | | | | | | | | If autoflushing on a new packet (e.g. due to the frag_every_frame flag being set), there's no samples to be written in the new fragment, so we can't overwrite the track duration in order to make it line up with the next packet to be written. Signed-off-by: Martin Storsjö <martin@martin.st>
* lavfi/qsvvpp: do not mix up FFmpeg and SDK error codeHaihao Xiang2021-08-09
| | | | | | | | The function ff_qsvvpp_filter_frame should return a FFmpeg error code if there is an error. However it might return a SDK error code without this patch. Reviewed-by: Soft Works <softworkz@hotmail.com>
* avformat/paf: read frame rate from header @0x88Gregory Montoir2021-08-10
| | | | Signed-off-by: Gregory Montoir <cyx@users.sourceforge.net>
* avcodec/nvenc: always set qp cr/cb offsetsTimo Rothenpieler2021-08-09
|
* aarch64: h264dsp: Fix indentation of some functions to match the restMartin Storsjö2021-08-08
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* aarch64: h264dsp: Remove unnecessary sign extensionsMartin Storsjö2021-08-08
| | | | | | | | | | These became unnecessary when the stride arguments were changed from int to ptrdiff_t in bc26fe89275c267d169b468356c82ee59874407d (0576ef466d8a631326d1d0a5ec2e4c4c81d25353) and d5d699ab6e6f8a8290748d107416fd5c19757a1b (aa844dc46f93182a63ec0b53267d19e7342c79b9). Signed-off-by: Martin Storsjö <martin@martin.st>
* avcodec/mpeg12dec: report error when picture type is unknown and err_detect ↵Marton Balint2021-08-08
| | | | | | | | is EXPLODE Also split error message to error and warning. Signed-off-by: Marton Balint <cus@passwd.hu>
* avformat/mxfdec: make MXFMetadataSet part of all metadata setsMarton Balint2021-08-08
| | | | | | | The code expects every kind of metadata set to start with the generic metadata set attributes. Signed-off-by: Marton Balint <cus@passwd.hu>
* FATE: add a test for sliced scalingAnton Khirnov2021-08-08
|
* tools/venc_data_dump: factor out demux/decode codeAnton Khirnov2021-08-08
| | | | It can be shared with other simple demux/decode tools.
* fftools/cmdutils: Remove unused cpu_countAndreas Rheinhardt2021-08-08
| | | | | | Added in cf12a478b206cd107343827426a05aedb83816bc, but not used. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avcodec/qsv: Remove unused ff_qsv_level_to_mfx()Andreas Rheinhardt2021-08-08
| | | | | | Unused since 00d0a4aa9eda8553113e51556123c46648a5f31b. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avfilter/asrc_flite: Don't define an object by accidentAndreas Rheinhardt2021-08-08
| | | | | | | | The flite filter apparently only wanted to declare a struct, but mistakenly also defined an unused and zero-initialized element with external linkage. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avcodec/h264dsp, h264idct: Fix lengths of array parametersAndreas Rheinhardt2021-08-08
| | | | | | Fixes many -Warray-parameter warnings from GCC 11. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avcodec/aacps: Fix length of array in function definitionAndreas Rheinhardt2021-08-08
| | | | | | | | | | hybrid2_re() has a parameter declared as "const INTFLOAT filter[8]", although the actual argument for said parameter only has seven elements; the code itself only uses seven elements, so change the parameter. Fixes a -Wstringop-overread warning with GCC 11. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avfilter/vf_vif: Make ff_compute_vif2 staticAndreas Rheinhardt2021-08-08
| | | | | | Only used here; the vif.h header can then be removed. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avfilter/vf_vif: Fix mismatch in number of array elementsAndreas Rheinhardt2021-08-08
| | | | | | | | | | | The function definition used float *data_buf[14], although there are only 13 elements (and only 13 are used); the declaration used 13. Given that the type will be converted to float **data_buf anyway, this is not in violation of the C specs, but nevertheless a bug. GCC 11 has a new warning for this -Warray-parameter. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avfilter/avf_showspectrum: fix another case of frames queuing upPaul B Mahol2021-08-08
|
* avcodec/nvenc: make aware of SDK 11.1 driver requirementsTimo Rothenpieler2021-08-08
|
* avcodec/nvenc: add chroma qp offset supportRicardo Monteiro2021-08-08
| | | | Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
* avcodec/ass_split: Rename ff_ass_split_dialog2->ff_ass_split_dialogAndreas Rheinhardt2021-08-08
| | | | | Reviewed-by: James Almer <jamrial@gmail.com> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avcodec/ass_split: Remove unused ff_ass_split_dialogue()Andreas Rheinhardt2021-08-07
| | | | | | | Unused since 1f63665ca567fbc49fa80166d468a822c2999efa. Reviewed-by: James Almer <jamrial@gmail.com> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avcodec/acelp_vectors: Add missing bracketsAndreas Rheinhardt2021-08-07
| | | | | | | | | | | | | | Before 3793caa5e2d1d16ed45771574b2ffc932497cfcf the code was "if (...) do { ... } while (...);". After said commit this became "if (...) av_assert0(...); do { ... } while (...);", i.e. the loop is always executed. This commit changes the logic to what it was before said commit. Notice that the condition is always true in FATE, so no changes are necessary there. This fixes a -Wmisleading-indentation warning from GCC 11. Reviewed-by: Michael Niedermayer <michael@niedermayer.cc> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* libavformat/mov: Change log line from debug to traceSoft Works2021-08-07
| | | | | | It's creating too much noise, even for debug Signed-off-by: softworkz <softworkz@hotmail.com>
* ffprobe: remove references to frame->pkt_ptsJames Almer2021-08-07
| | | | | | The field was removed during the last major bump. Signed-off-by: James Almer <jamrial@gmail.com>
* avfilter/avf_showspectrum: improve dBFS scale legendPaul B Mahol2021-08-07
| | | | | Make it more intuitive looking and correct for non-log scaling. Add option to set upper limit of input audio value in dBFS.
* avcodec/utils: don't return negative values in av_get_audio_frame_duration()James Almer2021-08-06
| | | | | | | | | | | In some extrme cases, like with adpcm_ms samples with an extremely high channel count, get_audio_frame_duration() may return a negative frame duration value. Don't propagate it, and instead return 0, signaling that a duration could not be determined. Fixes ticket #9312 Signed-off-by: James Almer <jamrial@gmail.com>
* avutil/tx: add a return at the end of non-void functionsJames Almer2021-08-06
| | | | | | Fixes compilation with GCC 11 when configured with --disable-optimizations Signed-off-by: James Almer <jamrial@gmail.com>
* avformat/mxfenc: add a return at the end of non-void functionsJames Almer2021-08-06
| | | | | | Fixes compilation with GCC 11 when configured with --disable-optimizations Signed-off-by: James Almer <jamrial@gmail.com>
* avfilter/vsrc_testsrc: add a return at the end of non-void functionsJames Almer2021-08-06
| | | | | | Fixes compilation with GCC 11 when configured with --disable-optimizations Signed-off-by: James Almer <jamrial@gmail.com>
* fftools/ffmpeg_filter: add a return at the end of non-void functionsJames Almer2021-08-06
| | | | | | Fixes compilation with GCC 11 when configured with --disable-optimizations Signed-off-by: James Almer <jamrial@gmail.com>
* avformat/fifo: Free dict on av_dict_copy() failureAndreas Rheinhardt2021-08-06
| | | | | | | | av_dict_copy() puts the onus on the caller to clean up dst on failure; it can be nonempty if copying a later entry of src fails after having successfully copied an earlier entry. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avformat/ilbc: Use av_get_packetAndreas Rheinhardt2021-08-06
| | | | Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avformat: Use ffio_read_size where appropriateAndreas Rheinhardt2021-08-06
| | | | Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avformat/yuv4mpegenc: Write data generically via AVPixFmtDescriptorAndreas Rheinhardt2021-08-06
| | | | Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
* avformat/adtsenc: return value check for init_get_bits in adts_decode_extradatamaryam ebrahimzadeh2021-08-06
| | | | | | | | As the second argument for init_get_bits (buf) can be crafted, a return value check for this function call is necessary. 'buf' is part of 'AVPacket pkt'. replace init_get_bits with init_get_bits8. Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* avformat/wtvdec: Check for EOF before seeking back in parse_media_type()Michael Niedermayer2021-08-06
| | | | | | | | | Fixes: Infinite loop Fixes: 36311/clusterfuzz-testcase-minimized-ffmpeg_dem_WTV_fuzzer-4889181296918528 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Peter Ross <pross@xvid.org> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* avformat/mpc8: Check first keyframe position for overflowMichael Niedermayer2021-08-06
| | | | | | | | Fixes: signed integer overflow: 9223372036854775791 + 18 cannot be represented in type 'long' Fixes: 36307/clusterfuzz-testcase-minimized-ffmpeg_dem_MPC8_fuzzer-4917863877050368 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* avcodec/exr: Check ac_countMichael Niedermayer2021-08-06
| | | | | | | | Fixes: signed integer overflow: -9223372036854775808 * 2 cannot be represented in type 'long long' Fixes: 36244/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_EXR_fuzzer-6090656186499072 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* tools/target_dec_fuzzer: check max samples in flush loopMichael Niedermayer2021-08-06
| | | | | | | | Fixes: Timeout Fixes: 36020/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_ALS_fuzzer-4774629855068160 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* ffprobe: Rename Audio Service Type 'type' field to 'service_type'Derek Buitenhuis2021-08-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | 677a030b26045acb50353d7954ae984ceefcd807 introduced more printable side data types in ffprobe, however the Audio Service Type side data 'type' field that was introduced aliases an existing field of the same name within the side data array, which can lead to JSON output like: "side_data_list": [ { "side_data_type": "Audio Service Type", "type": 0 }, { "side_data_type": "Stereo 3D", "type": "side by side", "inverted": 1 } ] This, while technically valid JSON, is considered bad practice, since it forces all downstream users to manually parse it and check all types; it makes simple deserialization impossible. Worse, in som loosely type languages, it can lead to silent bugs if exising code assumed it was a different type. As such, rename this second "type" field to "service_type". Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
* avfilter/vf_showinfo: add support for H.274 Film Grain Characteristics frame ↵James Almer2021-08-06
| | | | | | side data Signed-off-by: James Almer <jamrial@gmail.com>
* avcodec/hevc_sei: parse and export Film Grain Characteristics SEI messagesJames Almer2021-08-06
| | | | Signed-off-by: James Almer <jamrial@gmail.com>
* cbs_h265: add support for Film Grain Characteristics SEI messageJames Almer2021-08-06
| | | | Signed-off-by: James Almer <jamrial@gmail.com>
* avcodec/mfenc: remove usage of avctx->pkt_timebaseJames Almer2021-08-06
| | | | | | The field is documented to be for decoding only. Signed-off-by: James Almer <jamrial@gmail.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-31 10:52:33 +0000