Jun Zhao [Wed, 22 May 2019 01:41:56 +0000 (09:41 +0800)]
lavfi/lensfun: Use ff_filter_get_nb_threads() get number of threads
ff_filter_get_nb_threads() respect AVFilterContext.nb_threads and
graph->nb_threads both, in most case, we perfer this API than using
ctx->graph->nb_threads directly.
Reviewed-by: Steven Liu <lq@onvideo.cn> Reviewed-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Jun Zhao <barryjzhao@tencent.com>
Xuewei Meng [Sat, 18 May 2019 07:19:18 +0000 (15:19 +0800)]
libavfilter/dnn_native: Add multiple padding methods in dnn native
Add another two padding methods "VALID" and "SAME" as tensorflow,
and keep the existing "SAME_CLAMP_TO_EDGE" method suggested by sr filter.
As "SAME_CLAMP_TO_EDGE"can keep the output with the same size as original input,
and gives a slight better result as mentioned by sr filter.
Reviewed-by: Guo, Yejun <yejun.guo@intel.com> Signed-off-by: Xuewei Meng <xwmeng96@gmail.com> Signed-off-by: Steven Liu <lq@onvideo.cn>
James Almer [Mon, 13 May 2019 02:29:36 +0000 (23:29 -0300)]
avcodec/libdav1d: fine tune thread distribution
As suggested by Ronald, don't map auto threads to frame threads only, and
instead distribute them between frame and tile more efficiently.
Add a new framethreads override option, similar to the tilethreads one.
Fixes: "null pointer dereference" (actually misaligned access but the tools seem not to support this) Fixes: 14551/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_V210_fuzzer-5088609952071680 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The introduced regression caused URL's with @ in them to be parsed incorrectly
which is a bigger issue then not being able to specify the slash character as a
password.
I think there are better ways to fix the original issue, like being able to
specify HTTP username and password as a protocol option, or adding a protocol
option to percent-decode the URL first.
configure: replace 'pr' with printf since busybox does not support pr
This patch is based on https://trac.ffmpeg.org/ticket/5680 provided by
Kylie McClain <somasis@exherbo.org> at Wed, 29 Jun 2016 16:37:20 -0400,
and have some changes.
I'm not sure why this was written the way it was originally. We
initialise the plane addresses correctly in hwcontext_cuda so
why try and play games to calculate the plane offsets directly
in this code?
Philip Langdale [Tue, 14 May 2019 02:15:41 +0000 (19:15 -0700)]
avfilter/vf_scale_cuda: Fix incorrect scaling of > 8bit content
When i converted the filter to use texture objects instead of
texture references, I incorrect dropped the `pixel_size` scaling
factor when setting `pitchInBytes`. `src_pitch` is in pixels and
so must be scaled up.
Lynne [Thu, 2 May 2019 14:07:12 +0000 (15:07 +0100)]
libavutil: add an FFT & MDCT implementation
This commit adds a new API to libavutil to allow for arbitrary transformations
on various types of data.
This is a partly new implementation, with the power of two transforms taken
from libavcodec/fft_template, the 5 and 15-point FFT taken from mdct15, while
the 3-point FFT was written from scratch.
The (i)mdct folding code is taken from mdct15 as well, as the mdct_template
code was somewhat old, messy and not easy to separate.
A notable feature of this implementation is that it allows for 3xM and 5xM
based transforms, where M is a power of two, e.g. 384, 640, 768, 1280, etc.
AC-4 uses 3xM transforms while Siren uses 5xM transforms, so the code will
allow for decoding of such streams.
A non-exaustive list of supported sizes:
4, 8, 12, 16, 20, 24, 32, 40, 48, 60, 64, 80, 96, 120, 128, 160, 192, 240,
256, 320, 384, 480, 512, 640, 768, 960, 1024, 1280, 1536, 1920, 2048, 2560...
The API was designed such that it allows for not only 1D transforms but also
2D transforms of certain block sizes. This was partly on accident as the stride
argument is required for Opus MDCTs, but can be used in the context of a 2D
transform as well.
Also, various data types would be implemented eventually as well, such as
"double" and "int32_t".
Some performance comparisons with libfftw3f (SIMD disabled for both):
120:
22353 decicycles in fftwf_execute, 1024 runs, 0 skips
21836 decicycles in compound_fft_15x8, 1024 runs, 0 skips
128:
22003 decicycles in fftwf_execute, 1024 runs, 0 skips
23132 decicycles in monolithic_fft_ptwo, 1024 runs, 0 skips
384:
75939 decicycles in fftwf_execute, 1024 runs, 0 skips
73973 decicycles in compound_fft_3x128, 1024 runs, 0 skips
640:
104354 decicycles in fftwf_execute, 1024 runs, 0 skips
149518 decicycles in compound_fft_5x128, 1024 runs, 0 skips
768:
109323 decicycles in fftwf_execute, 1024 runs, 0 skips
164096 decicycles in compound_fft_3x256, 1024 runs, 0 skips
960:
186210 decicycles in fftwf_execute, 1024 runs, 0 skips
215256 decicycles in compound_fft_15x64, 1024 runs, 0 skips
1024:
163464 decicycles in fftwf_execute, 1024 runs, 0 skips
199686 decicycles in monolithic_fft_ptwo, 1024 runs, 0 skips
With SIMD we should be faster than fftw for 15xM transforms as our fft15 SIMD
is around 2x faster than theirs, even if our ptwo SIMD is slightly slower.
The goal is to remove the libavcodec/mdct15 code and deprecate the
libavcodec/avfft interface once aarch64 and x86 SIMD code has been ported.
New code throughout the project should use this API.
The implementation passes fate when used in Opus, AAC and Vorbis, and the output
is identical with ATRAC9 as well.
b3b7ba62 introduced undefined behaviour: A (non-modifiable) string
literal has been assigned to a modifiable string; said string was indeed
modified later via av_strtok.
This of course caused compiler warnings because of the discarded
qualifier; these are in particular fixed by this commit.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
avutil: Add missing reference files for pixdesc fate test
Commit cd48318035 added support for NV24 and NV42, including several
fate tests for these formats, but did not include the reference files
for the tests filter-pixdesc-nv24 and filter-pixdesc-nv42. As a result,
these two tests were broken.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Adam Richter [Sun, 12 May 2019 12:03:25 +0000 (05:03 -0700)]
libswcale: Fix possible string overflow in test.
In libswcale/tests/swcale.c, the function fileTest() calls sscanf in
an argument of "%12s" on character srcStr[] and dstStr[], which are
only 12 bytes. So, if the input string is 12 characters, a
terminating null byte can be written past the end of these arrays.
This bug was found by cppcheck.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Philip Langdale [Fri, 10 May 2019 04:02:09 +0000 (21:02 -0700)]
swscale: Add support for NV24 and NV42
The implementation is pretty straight-forward. Most of the existing
NV12 codepaths work regardless of subsampling and are re-used as is.
Where necessary I wrote the slightly different NV24 versions.
Finally, the one thing that confused me for a long time was the
asm specific x86 path that did an explicit exclusion check for NV12.
I replaced that with a semi-planar check and also updated the
equivalent PPC code, which Lauri kindly checked.
Philip Langdale [Tue, 7 May 2019 03:39:39 +0000 (20:39 -0700)]
avutil: Add NV24 and NV42 pixel formats
These are the 4:4:4 variants of the semi-planar NV12/NV21 formats.
These formats are not used much, so we've never had a reason to add
them until now. VDPAU recently added support HEVC 4:4:4 content
and when you use the OpenGL interop, the returned surfaces are in
NV24 format, so we need the pixel format for media players, even
if there's no direct use within ffmpeg.
Separately, there are apparently webcams that use NV24, but I've
never seen one.
avcodec/h264_ps: Check offset_for_non_ref_pic, offset_for_top_to_bottom_field and offset_for_ref_frame
Fixes: signed integer overflow: -2147483648 + -1 cannot be represented in type 'int' Fixes: 14444/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_H264_fuzzer-5675880333967360 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Jun Zhao [Sat, 27 Apr 2019 10:52:57 +0000 (18:52 +0800)]
lavc/libx265: Use avctx->framerate first for frame rate setting
perfer avctx->framerate first than use avctx->time_base when setting
the frame rate to encoder. 1/time_base is not the average frame rate
if the frame rate is not constant, so use avctx->framerate if the
value is not zero.
Jun Zhao [Sat, 27 Apr 2019 10:45:23 +0000 (18:45 +0800)]
lavc/libx264: Use avctx->framerate first for frame rate setting
perfer avctx->framerate first than use avctx->time_base when setting
the frame rate to encoder. 1/time_base is not the average frame rate
if the frame rate is not constant. In this case, we need to setting
avctx->framerate and avctx->time_base both, but avctx->framerate not
equal to 1/(avctx->time_base).
Carl Eugen Hoyos [Fri, 10 May 2019 08:44:34 +0000 (10:44 +0200)]
lavf/avidec: Do not test for bitrate <= INT_MAX.
AVCodecContext->bit_rate is int64_t since 7404f3bd
Unbreaks non-interleaved detection of v210 4k avi files, broken since 0eec40b7. Reported-by: Xavier Càmara, Centre de Conservació i Restauració, Filmoteca de Catalunya
After the last few commits, the functions for writing master elements
with CRC-32 elements didn't really make use of the ebml_master
structure any more, so remove these parameters from the functions.
The only things that still need to be kept are the positions of the
level 1 elements that are written preliminarily and updated later.
These positions are stored in the MatroskaMuxContext and
replace the corresponding ebml_master structures.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Improve log messages for blocks
Up until now, a block's relative offset has been reported as the offset
in the log messages output when writing blocks; given that it is
impossible to know the real offset from the beginning of the file at
this point due to the fact that it is not yet known how many bytes will
be used for the containing cluster's length field both the relative
offset in the cluster as well as the offset of the containing cluster
will be reported from now on.
Furthermore, the TrackNumber of the written block has been added to the
log output.
Also, the log message for writing vtt blocks has been brought in line
with the message for normal blocks.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Don't waste bytes writing level 1 elements
Up until now, the length field of most level 1 elements has been written
using eight bytes, although it is known in advance how much space the
content of said elements will take up so that it would be possible to
determine the minimal amount of bytes for the length field. This
commit changes this.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
Fixes intendation, whitespace, a typo and renames a variable
(dyn_bc->cluster_bc) to make its meaning clearer and to bring
it more in line with the naming of similar variables.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Write CRC-32 in non-seekable mode
Given that in both the seekable as well as the non-seekable mode dynamic
buffers are used to write level 1 elements and that now no seeks are
used in the seekable case any more, the two modes can be combined; as a
consequence, the non-seekable mode automatically inherits the ability to
write CRC-32 elements.
There are no differences in case the output is seekable; when it is not
and writing CRC-32 elements is disabled, there can still be minor
differences because before this commit, the EBML ID and length field
were counted towards the cluster size limit; now they no longer are.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Avoid seeking when writing level 1 elements
Up until now, the writing process for level 1 elements (those elements
for which CRC-32 elements are written by default) was this in case the
output was seekable: Write the EBML ID, write an "unkown length" EBML
number of the desired length, then write the element into a dynamic
buffer, then write the dynamic buffer (after possible calculation and
writing of the CRC-element), then seek back to the size element and
overwrite the unknown-size element with the real size. The seeking and
overwriting part has been eliminated by not writing the size initially.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
A Matroska EBML ID can only be maximally four bytes long, so make the
variables denoting EBML IDs uint32_t instead of unsigned int to
better reflect this.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
All places where end_ebml_master_crc32_preliminary are used already
check for whether the output is seekable, so the check in the function
is redundant.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Simplify check for writing CRCs
Up until now, the check for whether to write CRC32 elements was always
mkv->write_crc && mkv->mode != MODE_WEBM. This is equivalent to simply
set write_crc to zero in WebM-mode. And this is what this commit does.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Remove traces of secondary seek head
Up until e7ddafd515dc9826915b739d0b977a63c21e96af the Matroska muxer
wrote a secondary seek head referencing all the clusters. When this
was changed, a (now completely wrong) comment remained and the unique
remaining seek head was still called main_seekhead. This has been
changed.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Don't waste bytes in EBML Header
Up until now the EBML Header length field has been written with eight
bytes, although the EBML Header is always so small that only one byte
is needed for it. This patch saves seven bytes for every Matroska/Webm
file.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
avformat/matroskaenc: Slightly improve size bounds for cues
The upper bounds currently used for determining the size of a CuePoint's
length field can be improved somewhat; as a result, a CuePoint
containing three CueTrackPositions will now only need a size field
with one byte length.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
The earlier code included the size of the BlockGroup's length field and
the EBML ID in the calculation of the size for the payload and ignored
the size of the duration's length field. This meant that Blockgroups
corresponding to packets with size 2^(7n) - 17 - n - i, i = 0,..., n - 1,
n = 1,..., 8 (i.e. 110, 16364, 16365, 2097130..2097132, ...) were written
with length fields that are unnecessarily long.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
At this point, ts already includes the ts_offset so that the relative
time written with the cluster is already given by ts - mkv->cluster_pts.
It is this number that needs to fit into an int16_t.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
libavfilter/dnn: support multiple outputs for tensorflow model
some models such as ssd, yolo have more than one output.
the clean up code in this patch is a little complex, it is because
that set_input_output_tf could be called for many times together
with ff_dnn_execute_model_tf, we have to clean resources for the
case that the two interfaces are called interleaved.
Signed-off-by: Guo, Yejun <yejun.guo@intel.com> Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
libavfilter/dnn: determine dnn output during execute_model instead of set_input_output
Currently, within interface set_input_output, the dims/memory of the tensorflow
dnn model output is determined by executing the model with zero input,
actually, the output dims might vary with different input data for networks
such as object detection models faster-rcnn, ssd and yolo.
This patch moves the logic from set_input_output to execute_model which
is suitable for all the cases. Since interface changed, and so dnn_backend_native
also changes.
In vf_sr.c, it knows it's srcnn or espcn by executing the model with zero input,
so execute_model has to be called in function config_props
Signed-off-by: Guo, Yejun <yejun.guo@intel.com> Signed-off-by: Pedro Arthur <bygrandao@gmail.com>