Based on patch from Mozzhuhin Andrey <nopscmn at gmail.com>
Add a table based matrix8 multiplication implementation. The algorithm
does not do any clipping so we need to make sure we never call this on
input that might need to be clipped. In general, this algorithm is
2 times faster than the orc optimized one and would be chosen for all
RGB -> YUV conversions and some YUV->YUV and RGB->RGB conversions.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732186
When we are using the fast linear resampler, use the ->inc to calculate
the first and last pixel we need so that we can do vertical resampling
on the right amount of pixels.
CLAMP checks both if value is '< 0' and '> max'. Value will never be a negative
number since it in an unsigned integer. Removing that check and only checking
if it is bigger than max and setting it appropriately.
CID #1271606
Only activate the scaler fastpath for x2 up and downscale when the
scaler method is respectively nearest and linear because that is what
those fastpaths really implement.
Add support for alpha. Make it possible to copy, set and multiply the
alpha value of a frame during conversion.
Set the border alpha to 0xff by default.
Go over some of the fastpaths and add alpha handling.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=745006
Make a convenience function that combines 2 scalers to perform a 2d
scale. This removes quite a bit of overhead in method calls when doing a
typical scale and it also can reuse a piece of unused memory in the
vertical scaler.
Use the 2d scaler in video-converter and remove the other scalers and
temp memory.
Only merge scalers for selected formats.
Use nearest neighbour scaling for chroma when doing nearest neighbour
for the luma.
Also fastpath GRAY16_OE in nearest neighbour.
configure parameters correctly for packed fastpath.
Small performance tweaks for RGB and friends.
Add, but ifdef out, alternative nearest neighbour scaling, it is slower
than the current table based version.
Use memcpy instead of orc_memcpy because it is measurably faster.
Fix YUY2 and friends vertical scaling.
video-scaler.c:1331:14: error: variable 'func' is used uninitialized whenever 'if' condition is false
[-Werror,-Wsometimes-uninitialized]
} else if (bits == 16) {
^~~~~~~~~~
video-scaler.c:1348:3: note: uninitialized use occurs here
func (scale, src_lines, dest, dest_offset, width, n_elems);
^~~~
video-scaler.c:1331:10: note: remove the 'if' if its condition is always true
} else if (bits == 16) {
^~~~~~~~~~~~~~~~
video-scaler.c:1260:27: note: initialize the variable 'func' to silence this warning
GstVideoScalerVFunc func;
^
= NULL
video-converter.c:3406:12: error: implicit conversion from enumeration type 'GstVideoFormat' to different
enumeration type 'GstFormat' [-Werror,-Wenum-conversion]
format = convert->fformat[plane];
~ ^~~~~~~~~~~~~~~~~~~~~~~
video-converter.c:3413:44: error: implicit conversion from enumeration type 'GstFormat' to different enumeration
type 'GstVideoFormat' [-Werror,-Wenum-conversion]
gst_video_scaler_horizontal (h_scaler, format,
~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
video-converter.c:3471:12: error: implicit conversion from enumeration type 'GstVideoFormat' to different
enumeration type 'GstFormat' [-Werror,-Wenum-conversion]
format = convert->fformat[plane];
~ ^~~~~~~~~~~~~~~~~~~~~~~
video-converter.c:3487:42: error: implicit conversion from enumeration type 'GstFormat' to different enumeration
type 'GstVideoFormat' [-Werror,-Wenum-conversion]
gst_video_scaler_vertical (v_scaler, format, lines, d + out_x, i,
~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
video-converter.c:3551:12: error: implicit conversion from enumeration type 'GstVideoFormat' to different
enumeration type 'GstFormat' [-Werror,-Wenum-conversion]
format = convert->fformat[plane];
~ ^~~~~~~~~~~~~~~~~~~~~~~
video-converter.c:3569:46: error: implicit conversion from enumeration type 'GstFormat' to different enumeration
type 'GstVideoFormat' [-Werror,-Wenum-conversion]
gst_video_scaler_horizontal (h_scaler, format,
~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
video-converter.c:3577:42: error: implicit conversion from enumeration type 'GstFormat' to different enumeration
type 'GstVideoFormat' [-Werror,-Wenum-conversion]
gst_video_scaler_vertical (v_scaler, format, lines, d + out_x, i,
~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
In function gst_video_scaler_vertical() the bits variable is always
set to either 8 or 16 in every possible format. No need to initialize it.
If the format isn't valid it goes to no_func, so there is no need to
handle the case of bits not being 8 or 16.
CID #1268401
Add fastpaths for all planar conversion and scaling.
Improve gray and alpha handling.
Add option to specify the chroma resampler method and set to linear as
default.
video-converter.c:3645:24: error: implicit conversion from enumeration type
'GstFormat' to different enumeration type 'GstVideoFormat'
[-Werror,-Wenum-conversion]
convert->fformat = fformat;
~ ^~~~~~~
video-converter.c:3667:24: error: implicit conversion from enumeration type
'GstFormat' to different enumeration type 'GstVideoFormat'
[-Werror,-Wenum-conversion]
convert->fformat = fformat;
~ ^~~~~~~
video-converter.c:3963:50: error: implicit conversion from enumeration type
'const GstVideoFormat' to different enumeration type 'GstFormat'
[-Werror,-Wenum-conversion]
if (!setup_scale (convert, transforms[i].fformat))
~~~~~~~~~~~ ~~~~~~~~~~~~~~^~~~~~~
In gst_video_resampler_init () if method is GST_VIDEO_RESAMPLER_METHOD_NEAREST
then params.envelope is not initialized but still used later in line 382.
Make sure this variable is initiliazed to avoid undefined behaviour.
CID #1256568
video-converter.c:3073:48: error: implicit conversion from enumeration type 'GstFormat' to different enumeration type 'GstVideoFormat'
[-Werror,-Wenum-conversion]
gst_video_scaler_horizontal (h_scaler, format,
~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
video-converter.c:3081:44: error: implicit conversion from enumeration type 'GstFormat' to different enumeration type 'GstVideoFormat'
[-Werror,-Wenum-conversion]
gst_video_scaler_vertical (v_scaler, format, lines, d, i, out_w);
~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~~
video-converter.c:3137:24: error: implicit conversion from enumeration type 'const GstVideoFormat' to different enumeration type 'GstFormat'
[-Werror,-Wenum-conversion]
convert->fformat = GST_VIDEO_INFO_FORMAT (in_info);
~ ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../../../gst-libs/gst/video/video-info.h:125:43: note: expanded from macro 'GST_VIDEO_INFO_FORMAT'
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../../../gst-libs/gst/video/video-format.h:361:59: note: expanded from macro 'GST_VIDEO_FORMAT_INFO_FORMAT'
~~~~~~~~^~~~~~
video-converter.c:3157:24: error: implicit conversion from enumeration type 'GstVideoFormat' to different enumeration type 'GstFormat'
[-Werror,-Wenum-conversion]
convert->fformat = GST_VIDEO_FORMAT_GRAY8;
Add fast path scaling for YUY2 and other packed YUV formats. Add a new
method to merge the scalers of the Y and UV components into one scaler.
Add faster horizontal 2tap scaler.
See https://bugzilla.gnome.org/show_bug.cgi?id=741987
Add support for scaling of images with pstride == 1. This can be used
to scale individual planes later.
Rework some of the scaling code to take the pstride as a parameter.
CLAMP checks both if value is '< 0' and '> max'. Value will never be a negative
number since it is an unsigned integer. Removing that check and only checking if
it is bigger than max and setting it appropriately.
CID 1256559
CLAMP checks both if n_taps is '< 0' and '> max_taps'. n_taps will never be a
negative number because it is an unsigned integer. Removing that check and only
making sure it isn't set bigger than max.
CID 1256558
Add an option to disable chroma resampling.
Improve the matrix option values so that you can choose to use the input
or output matrix or disable conversion.
Video buffer pool will update video alignment to respect stride alignment
requirement. But haven't updated it to video alignment in configure.
Which will cause user get wrong video alignment.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=741501
This makes sure that the element is in the same state before start() is called
the very first time and every future call after the element was used already.
Also it ensure that we always have a clean state before start(), cleaned the
same way in every case.
The stop() vfunc might mess with some of our fields we have just
reset, which could cause memory leaks or invalid state taken over
to later.
Also the stop() vfunc, or anything called until it from another thread,
might want to be able to use the fields that were just resetted and
become confused because of that.
In the decoder we already had a workaround for things like this happening,
this workaround is not needed anymore.
Allows subclasses to do custom caps query replies.
Also exposes the standard caps query handler so subclasses can just
extend on top of it instead of reimplementing the caps query proxying.
https://bugzilla.gnome.org/show_bug.cgi?id=741263
With the new caps query results the caps returned might have extra fields
that are not required by the decoder (framerate for image decoders) and it
causes a regression making, for example, jpegdec reject caps that don't
have framerates.
The accept-caps implementation will do 2 checks:
1) Do subset check with the template caps, making sure all the required
fields that are present on the template are present on the received caps.
2) Do a intersection check with the result of a caps query, making sure
that downstream can accept the fields in the received caps.
https://bugzilla.gnome.org/show_bug.cgi?id=741263
Refactor the encoder's caps query proxying function to a common place
and use it in the videodecoder to proxy downstream restrictions.
The new function is private to the gstvideo lib.
https://bugzilla.gnome.org/show_bug.cgi?id=741263
Update the new buffer size after alignment in the pool configuration
before calling the parent set_config. This ensures that the parent knows
about the buffer size that we will allocate and makes the size check
work in the release_buffer method.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=741420
This reverts commit 406f32a946.
The problem was apparently that my video-orc.h was not updated and did not
include the prototype for that function. Only a "make clean" caused it to
be regenerated.
Avoid using a constant.
Avoid doing saturated adds, results are not supposed to overflow here.
Rework the C backup function a little in preparation for custom backup
functions in ORC.
See https://bugzilla.gnome.org/show_bug.cgi?id=741015
It will cause the frame to be initialized with inconsistent values that then
later can cause crashes or any other kind of interesting and hard to debug
bugs.
In cases where we just call orc directly this is somewhat
superfluous, but let's do it anyway for consistency. In
other cases the compiler can hopefully use this to optimise
memory access a little.
when using variable taps and when we are limiting the number of taps,
recalculate the lanczos parameters to match the clamped value.
Set the max number of taps to 128
Make a small object to hold a pool of allocated temp lines.
Keep track of how many temp lines each conversion stage needs and use
this to allocate just enough temp lines from the temp lines object. from
the temp lines object.
When dealing with mixed interlaced, setup a scaler and chroma-resampler
for both interlaced and progressive frames and switch between them
depending on the interlace mode of the input frame.
Add an option to limit the number of taps to use in automatic mode. The
problem is that for lanczos, we might use more taps than what we can
handle with the current precision.
Rework the other options a little to make it nicer to set defaults.
Refactor GstVideoInfo init, make function to set default colorimetry.
Call fill_planes after we configure the GstVideoInfo with parameters
from the caps.
The size of the chroma planes for interlaced vertically subsampled
formats needs to be rounded up to 2, we have 2 fields with each
the same anount of chroma lines.
Keep only 1 structure with all matrix information.
Add structure to hold gamma information.
Add more options to control gamma, primaries and color matrix handling.
Add functions to compute transformations to and from XYZ and use this
to convert between primaries.
Merge gamma into the convert to and from RGB stage.
Fix border val.
Simplify the fastpath table, remove unused fields, add some more checks.
Prepare for doing full gamma corrected conversion and scaling by first
splitting the conversions from and to RGB into separate steps.
split scaling in downscaling and upscaling steps to be performed before
and after conversion respectively.
Fix clipping of images that are partially left of the video
surface, they would get clipped on the right side instead of
the left side, because the video unpack functions currently
ignore the x offset parameter. Work around that until that
is implemented.
https://bugzilla.gnome.org/show_bug.cgi?id=739281
In case of overlay being completely or partially outside
the video frame, the offset calculations are not right,
which resulted in the overlay not being displayed as
expected, or crashes due to invalid memory access.
When the overlay rectangle is completely outside,
we need not render the overlay at all.
For partial display of overlay rectangles, src_yoff
was not being calculated, hence it was always clipping
the bottom half of the overlay, By calculating the
src_yoff, now the overlay is clipped properly.
https://bugzilla.gnome.org/show_bug.cgi?id=739281
Make an ORC version of the 2x vertical upsampling code.
Improve unit tests, test chroma up and down sampling.
memset buffer in conversion to make valgrind happy.
Combine multiplies in 4x filters.
Rename conversion functions to make them nicer in orc.
Add ORC versions for various downsampling algorithms
Add unit test chroma resampler
We only need to do the horizontal subsampling on 1 line if we do it
after vertical subsampling and we also avoid doing vertical subsampling
on unused pixels.
Rework the converter, keep track of the conversion steps by chaining the
cache objects together. We can then walk the chain and decide the
optimal allocation pattern.
Remove the free function, we're not going to need this anytime soon.
Keep track of what output line we're constructing so that we can let the
allocator return a line directly into the target image when possible.
Directly read from the source pixels when possible.
We need to allocate the templine with the amount of pixels we are going
to handle, which we only know for the vertical resampler when we are
asked to resample.
Add scaler functions for 16 bits formats.
Rename the scaler functions so that 16bits versions don't look too
weird.
Remove old unused h_2tap functions
Fix v_ntap functions, it was using 1 tap too little.
Rework the way we track the current state of the video through the
different conversion phases and use this to make sure we use the right
format and pstride where needed.
A faster version of 4tap horizontal scaling causes segfaults in ORC
presumably because it uses too many registers so disable it to avoid
crashing in the ORC tests.
video-scaler.c:151:58: error: implicit conversion from enumeration type
'GstVideoScalerFlags' to different enumeration type
'GstVideoResamplerFlags' [-Werror,-Wenum-conversion]
gst_video_resampler_init (&scale->resampler, method, flags, out_size,
~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~
Only apply an offset that is a multiple of the subsampling. To handle
arbitrary offsets in the future, we need to be able to chroma-resample
part of the borders.
Add support for cropping the source and placing the converted image
into a rectangle in the destination frame.
Add an option to add a border and border color.
Add the old ORC functions for nearest and linear. Label them as Low
quality because they are not as accurate but ORC lacks opcodes to
express this for now.
Add a video scaler object build on top of the resampler. It has
implementation to deal with interlaced video as well as horizontal and
vertical scaling functions.
Use a LineCache object to track and process lines between unpack,
upsample, convert, downsample and pack stages. This simplifies the
main core processing function a lot and allows for future additions
easily.
Add support for interlaced formats in chroma up and downsampling.
There are some few but certain conditions where it is possible for the
dest_width to be smaller than x. So we check this before assigning a negative
value to src_width, which is a unsigned and would be promoted to a number that
can segfault videoblend.
https://bugzilla.gnome.org/show_bug.cgi?id=738242
This was never reset when going from PAUSED->READY and resulted
in encoders being not reusable after EOS. They just rejected any
buffer because they received EOS in their previous life.
The flag wasn't used anywhere except for rejecting buffers after
EOS, and this is now handled by GstPad directly.
Move the conversion code used in videoconvert to the video library
and expose a simple but generic API to do arbitrary conversion. It can
currently do colorspace conversion but the plan is to add videoscale to
it as well.
See https://bugzilla.gnome.org/show_bug.cgi?id=732415
Reset last_timestamp_out when applying the output segment
change, to avoid decoder confusion over new timestamp timelines when
a seamless segment change happens.
Move some locks/unlocks to later when they're actually needed.
https://bugzilla.gnome.org/show_bug.cgi?id=734617
This fixes the reverse playback scenario when upstream is not fully
parsing the stream and does not send every keyframe chain separately
with the DISCONT flag on the keyframe.
To explain this, let's suppose we have this stream:
0 1 2 3 4 5 6 7 8
K K K
In most circumstances, the upstream parser will chain in the
decoder the buffers in the following order:
6 7 8 3 4 5 0 1 2
D D D
In this case, GstVideoDecoder will flush the parse queue every time
it receives discont (D) and we will eventually get in the output queue:
(flush here) 8 7 6 (flush here) 5 4 3 (flush here) 2 1 0
In case the upstream parser doesn't do this work, though,
GstVideoDecoder will receive the whole stream at once and will flush
the parse queue afterwards:
0 1 2 3 4 5 6 7 8
D
During the flush, it will look backwards for keyframes and will
decode in this order:
6 7 8 3 4 5 0 1 2
This is the same order that it would receive from upstream if
upstream was parsing and looking for the keyframes, only that now
there is no flushing of the output queue in between keyframes,
which will result in the output queue looking like this:
2 1 0 6 5 3 8 7 6
This will confuse downstream obviously and will play incorrectly.
This patch forces the decoder to flush the output queue every time
it picks a new keyframe to decode, so it will end up decoding 6 7 8
and then flushing before picking 3 for decoding, so the output will
get 8 7 6 before 6 5 3 and the video will play back correctly.
https://bugzilla.gnome.org/show_bug.cgi?id=734441
This prevent implementing allocation query, as the format need to be
known in order to determin the size and number of buffers needed.
Note: This may lead to few regressions that will need fixing
https://bugzilla.gnome.org/show_bug.cgi?id=732288
Fix gst_video_decoder_parse_available() to really parse any pending
source data that is still available in the adapter. This is a memory
optimization to avoid expansion of video packed added to the adapter,
but also a fix to EOS condition when the subclass parse() function
ultimately only needed to call into gvd_have_frame() and no additional
source bytes were consumed, i.e. gvd_add_to_frame() is not called.
This situation can occur when decoding H.264 streams in byte-stream/nal
mode for instance. A decoder always requires the next NAL unit to be
parsed so that to determine picture boundaries. When a new picture is
found, no byte is consumed (i.e. gvd_add_to_frame() is not called)
but gvd_have_frame() is called (i.e. priv->current_frame is gone).
Also make sure to avoid infinite loops caused by incorrect subclass
parse() implementations. This can occur when no byte gets consumed
and no appropriate indication (GST_VIDEO_DECODER_FLOW_NEED_DATA) is
returned.
https://bugzilla.gnome.org/show_bug.cgi?id=731974
Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Buffer pool set_config() may return FALSE if requested configuration needed small
changes. Reget the config and try setting it again. This ensure we have a configured
pool if possible.
This should allow for more meaningful errors. Dereferencing NULL
is more useful information than dereferencing a random address
happened to be on the stack.
If gst_video_overlay_rectangle_apply_global_alpha is called with
a rectangle with unsuitable alpha, expanding the alpha plane will
fail, and thus lead to dereferencing a NULL src pointer. It's not
certain this will happen in practice, as the function is static
and callers might ensure suitable alpha before calling, but there
is no apparent explicit such check.
Add prologue asserts for proper alpha to explicitely prevent this.
Coverity 1139707
Videodecoder does late renegotiation, it will wait for the next
buffer before renegotiating its caps and bufferpool. It might happen
that downstream element switched from passthrough to non-passthrough
and sent a reconfigure upstream (that caused this renegotiation).
This downstream element will ask the video sink below for the bufferpool
with an allocation query and will get the same bufferpool that
videodecoder is holding, too.
When renegotiating, if videodecoder deactivates its bufferpool it
might be deactivating the bufferpool that some element downstream
is using and cause the pipeline to fail.
https://bugzilla.gnome.org/show_bug.cgi?id=727498
baseparse will reverse each GOP for us already, so the segment events can
be after our keyframe. Make sure to get it and all other relevant sticky
events before starting to decode.
This was a regression introduced by f52fd7a68, where we started using
the stride to encode the dimensions in tiles. This patch simply updates
offset and size calculation as described in the documentation,
part-mediatype-video-raw.txt.
