This together with adjusting the inclusion define allows for the
build to not fail with latest Vulkan-Headers that contain the
stabilized Vulkan AV1 decoding definitions.
Compilation fails currently as the AV1 header is getting included
via hwcontext_vulkan.h -> <vulkan/vulkan.h> -> vulkan_core.h, which
finally includes vk_video/vulkan_video_codec_av1std.h and the decode
header, leading to the bundled header to never defining anything
due to the inclusion define being the same.
This fix is imperfect, as it leads to additional re-definition
warnings for things such as
VK_STD_VULKAN_VIDEO_CODEC_AV1_DECODE_SPEC_VERSION. , but it is
not clear how to otherwise have the bundled version trump the
actually standardized one for a short-term compilation fix.
A lot of changes and fixes to channel layout parsing, notably
- get rid of dynamic allocation of channel positions
- signal unimplemented speaker positions as unknown instead of failure, but
warn the user about it
- native order, and that a single channel only appears once was always assumed
for less than 64 channels, obviously this was incorrect
Signed-off-by: Marton Balint <cus@passwd.hu>
It should be available in all relevant modern compilers and will allow
us to use features like anonymous unions.
Note that stdatomic.h is still emulated on MSVC, as current versions
require the /experimental:c11atomics, and do not support
ATOMIC_VAR_INIT() anyway.
According to ITU-T H.265 7.4.2.1 this byte sequence should not appear in a
NAL unit but in practice in rare cases it seems it does, possibly due to buggy
encoders. Other players like VLC and Quicktime seem to be fine with it.
Currently when this sequence is found it is treated as if the next start code
has been found and the NAL unit gets truncated.
This change limits the code to only look for first start code 0x0000001 or
first escape 0x000003.
Sadly i can't share the original source file with the issue but the first
80 bytes of the NAL unit looks like this:
│00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f│0123456789abcdef│
0x00000│00 00 00 01 02 01 d0 bc 57 a1 b8 44 70 01 00 0b│........W..Dp...│
0x00010│80 2e 00 c2 6c ec 3e b9 e3 03 fb 91 2e d2 43 cb│....l.>.......C.│
0x00020│1d 2c 00 00 02 00 02 00 5c 93 72 6f 31 76 18 00│.,......\.ro1v..│
0x00030│08 38 aa b1 4c 33 3f fd 08 cb 77 9b d4 3c db 02│.8..L3?...w..<..│
0x00040│a2 04 73 15 75 de 3b c4 67 c0 8f ca ad 31 f1 99│..s.u.;.g....1..│
Signed-off-by: Anton Khirnov <anton@khirnov.net>
I generated a DXV2 file with an interesting alpha channel using
Adobe Media Encoder 2015 and compared decoding it using Resolume Alley
and ffmpeg. Similarly to DXV3 files, Alley expects premultiplied alpha
and ffmpeg matches its decoding more closely when it does the same.
Reference file: https://connorworley.com/dxv2-dxt5.mov
Existing FATE tests for DXV2 files do not cover this change.
Signed-off-by: Connor Worley <connorbworley@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
This change decouples the frame dimensions from avctx, which is useful
for DXV decoding, and fixes incorrect behavior in the existing
implementation.
Tested with `make fate THREADS=7` and
`make fate THREADS=7 THREAD_TYPE=slice`.
Signed-off-by: Connor Worley <connorbworley@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
We currently mostly do not empty the MMX state in our MMX
DSP functions; instead we only do so before code that might
be using x87 code. This is a violation of the System V i386 ABI
(and maybe of other ABIs, too):
"The CPU shall be in x87 mode upon entry to a function. Therefore,
every function that uses the MMX registers is required to issue an
emms or femms instruction after using MMX registers, before returning
or calling another function." (See 2.2.1 in [1])
This patch does not intend to change all these functions to abide
by the ABI; it only does so for ff_simple_idct_mmx, as this
function can by called by external users, because it is exported
via the avdct API. Without this, the following fragment will
assert (on x86_32, as ff_simple_idct_mmx is not used on x64):
int16_t __attribute__ ((aligned (16))) block[64];
AVDCT *dct = avcodec_dct_alloc();
dct->idct_algo = FF_IDCT_AUTO;
avcodec_dct_init(dct);
dct->idct(block);
av_assert0_fpu();
[1]: https://raw.githubusercontent.com/wiki/hjl-tools/x86-psABI/intel386-psABI-1.1.pdf
Reviewed-by: Kieran Kunhya <kierank@obe.tv>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The PPS should be used instead of the SPS to get the current picture's
dimensions. Using the SPS can cause issues if the resolution changes
mid-sequence. In particular, it was leading to invalid memory accesses
if the resolution decreased.
Patch replaces sps->{width,height} with pps->{width,height}. It also
removes sps->{width,height}, as these are no longer used anywhere.
Fixes crash when decoding DVB V&V test sequence
VVC_HDR_UHDTV1_ClosedGOP_Max3840x2160_50fps_HLG10_res_change_without_RPR
Signed-off-by: Frank Plowman <post@frankplowman.com>
This exposes VP8E_SET_SCREEN_CONTENT_MODE option from libvpx.
Co-authored-by: Erik Språng <sprang@webrtc.org>
Signed-off-by: Dariusz Marcinkiewicz <darekm@google.com>
Signed-off-by: James Zern <jzern@google.com>
Make av_tx_init() agree with documentation:
* Real to complex and complex to real DFTs.
* For the float and int32 variants, the scale type is 'float', while for
* the double variant, it's a 'double'. If scale is NULL, 1.0 will be used
* as a default.
Reviewed-by: Lynne <dev@lynne.ee>
Signed-off-by: Peter Ross <pross@xvid.org>
This hack is used to limit the visibility of some AVFilterLink fields to
only certain files. Replace it with the same pattern that is used e.g.
in lavf AVStream/FFStream and avoid exposing these internal fields in a
public header completely.
Makes it robust against adding fields before it, which will be useful in
following commits.
Majority of the patch generated by the following Coccinelle script:
@@
typedef AVOption;
identifier arr_name;
initializer list il;
initializer list[8] il1;
expression tail;
@@
AVOption arr_name[] = { il, { il1,
- tail
+ .unit = tail
}, ... };
with some manual changes, as the script:
* has trouble with options defined inside macros
* sometimes does not handle options under an #else branch
* sometimes swallows whitespace
avfilter_insert_filter() already reports (also with AV_LOG_VERBOSE)
when a filter is auto-inserted.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
fc->ref points to an old VVCFrame, which cannot be used after frame_context_setup.
This prevents crashes in decode_nal_units-->ff_vvc_report_frame_finished.
Signed-off-by: Frank Plowman <post@frankplowman.com>
The avctx passed to avcodec_string() may have unset coded dimensions, as is the
case when called by av_dump_format() where the streams had all the needed
information at the container level, and as such no frames were decoded
internally.
Signed-off-by: James Almer <jamrial@gmail.com>
Besides simplifying address computations (it saves 432B of .text
in hevcdsp.o alone here) it also fixes undefined behaviour that
occurs if mx or my are 0 (happens when the filters are unused)
because they lead to an array index of -1 in the old code.
This happens in the checkasm-hevc_pel FATE-test.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Reviewed-by: Nuo Mi <nuomi2021@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
Otherwise, passing an UNSPECIFIED frame to am MPEG-only filter graph
would trigger insertion of an unnecessary vf_scale filter, which would
perform a memcpy to convert between the two.
This is safe to do because unspecified YUV frames are already
universally assumed to be MPEG range, in particular by swscale.
Currently, this only affects untagged RGB/XYZ/Gray, which get forced to
their corresponding metadata before entering the filter graph. The main
justification for this change, however, is the planned ability to add
automatic promotion of unspecified yuv to mpeg range yuv.
Notably, this change will never allow accidentally cross-promoting
unspecified to jpeg or to a specific YUV matrix, since that is still
bound by the constraints of YUV range negotiation as set up by
query_formats.
Since 7bf1b9b357,
the test produces ordinary \n, yet this is not what the reference
file used for the most time, leading to test failures.
Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
When this filter overrides frame properties, the outgoing frames have
a different YUV colorspace than the incoming ones. This requires
signalling the new colorspace on the outlink, and in particular, making
sure it's *not* set to a common ref with the input - otherwise the point
of this filter would be destroyed.
Untouched fields will continue being passed through, so we don't need to
do anything there.
