gstreamer-omap:gst-plugins-bad.git
5 years agodri2videosink: destroy context only whe window is destroyed
Vincent Penquerc'h [Fri, 20 Jul 2012 12:16:33 +0000 (12:16 +0000)]
dri2videosink: destroy context only whe window is destroyed

This avoids using the X lock after it was destroyed, as buffers
may be destroyed later, and need to use the context.

5 years agoh264parse: add num-ref-frames in caps
Vincent Penquerc'h [Wed, 18 Jul 2012 11:38:53 +0000 (11:38 +0000)]
h264parse: add num-ref-frames in caps

This is the number of reference frames needed for reordering frames
by presentation time.

5 years agoh264parse: delay setting passthrough mode when waiting to known resolution
Vincent Penquerc'h [Wed, 11 Jul 2012 14:24:16 +0000 (14:24 +0000)]
h264parse: delay setting passthrough mode when waiting to known resolution

This makes a difference on a particular interlaced clip, which looks very
wrong with this patch, and looked even wronger without.

5 years agodri2: link libgstdrm in-tree
Vincent Penquerc'h [Wed, 11 Jul 2012 13:31:36 +0000 (13:31 +0000)]
dri2: link libgstdrm in-tree

5 years agokms: add missing backslash in lib list
Vincent Penquerc'h [Wed, 11 Jul 2012 09:42:26 +0000 (09:42 +0000)]
kms: add missing backslash in lib list

5 years agokmssink: link to our in-tree libgstbasevideo
Vincent Penquerc'h [Tue, 10 Jul 2012 17:50:56 +0000 (17:50 +0000)]
kmssink: link to our in-tree libgstbasevideo

5 years agoh264parse: do not use a buffer after it was unreffed
Vincent Penquerc'h [Tue, 10 Jul 2012 15:31:33 +0000 (15:31 +0000)]
h264parse: do not use a buffer after it was unreffed

In some cases, the frame buffer may be replaced by a new buffer,
making the local cache invalid.

5 years agoh264parse: split interlaced fields in separate packets
Vincent Penquerc'h [Tue, 10 Jul 2012 13:53:09 +0000 (13:53 +0000)]
h264parse: split interlaced fields in separate packets

ducatih264dec requires such fields to be in separate AUs.

5 years agoRevert "h264parse: disable some optimizations with AVC input"
Vincent Penquerc'h [Tue, 10 Jul 2012 13:51:09 +0000 (13:51 +0000)]
Revert "h264parse: disable some optimizations with AVC input"

I now have a fix for the bundling of interlaced NAL units in a
single AU that was tripping ducatih264dec, so we can revert this
stopgap and go back to the faster parsing.

This reverts commit 37257ba09895d567c06e44e1f5c58cb54851fb38.

5 years agopvrvideosink: fix unwanted scaling on cropped videos
Vincent Penquerc'h [Mon, 9 Jul 2012 10:54:57 +0000 (10:54 +0000)]
pvrvideosink: fix unwanted scaling on cropped videos

When a video with crop rectangle was used, the window was created with
the uncropped size, and kept that way after the crop event was received,
leading to a slight scaling of the video (and likely slight aspect ratio
change as well).

This is fixed by resizing the window to the cropped size when it changes.

5 years agoh264parse: wait to get resolution before pushing buffers
Vincent Penquerc'h [Thu, 28 Jun 2012 15:10:08 +0000 (15:10 +0000)]
h264parse: wait to get resolution before pushing buffers

ducatih264dec wants resolution in caps, so we have to wait for it
to become known before pushing the first buffer, so it includes caps
with the resolution.

This fixes prerolling on the disney-sample.mpg test file.

5 years agodrmbufferpool: give this code a separate debug category
Vincent Penquerc'h [Thu, 28 Jun 2012 09:03:59 +0000 (09:03 +0000)]
drmbufferpool: give this code a separate debug category

and add a couple logs while we're at it

5 years agopvrvideosink: recreate the pool on new caps in buffer-alloc
Vincent Penquerc'h [Wed, 27 Jun 2012 18:27:05 +0000 (18:27 +0000)]
pvrvideosink: recreate the pool on new caps in buffer-alloc

When an element downstream calls setcaps, then buffer-alloc,
the setcaps may be delayed, so do not rely on setcaps to reset
the pool before buffer-alloc.

5 years agopvrvideosink: actually print caps
Vincent Penquerc'h [Wed, 27 Jun 2012 16:25:50 +0000 (16:25 +0000)]
pvrvideosink: actually print caps

5 years agodri2ideosink: fix caps equality check
Vincent Penquerc'h [Wed, 27 Jun 2012 16:23:07 +0000 (16:23 +0000)]
dri2ideosink: fix caps equality check

gst_drm_buffer_pool_check_caps returns FALSE when the caps are not equal.

5 years agopvrvideosink: fix caps equality check
Vincent Penquerc'h [Wed, 27 Jun 2012 16:22:10 +0000 (16:22 +0000)]
pvrvideosink: fix caps equality check

gst_drm_buffer_pool_check_caps returns FALSE when the caps are not equal.

5 years agolibgstdrm: use gst_caps_is_strictly_equal instead of gst_caps_is_equal
Vincent Penquerc'h [Wed, 27 Jun 2012 16:20:41 +0000 (16:20 +0000)]
libgstdrm: use gst_caps_is_strictly_equal instead of gst_caps_is_equal

The latter is, well, not strict enough. In fact, it will be happy to
return TRUE when one caps has interlaced=TRUE, and the other does not
have interlaced at all.

5 years agopvrvideosink: get rid of the dependency on dri2_omap_ws.h
Vincent Penquerc'h [Tue, 26 Jun 2012 11:59:18 +0000 (11:59 +0000)]
pvrvideosink: get rid of the dependency on dri2_omap_ws.h

5 years agodri2videosink: reenable the memcpy on slow path
Vincent Penquerc'h [Mon, 25 Jun 2012 14:56:27 +0000 (14:56 +0000)]
dri2videosink: reenable the memcpy on slow path

This fixes using 2 dri2videosink elements in a pipeline, and does not seem
to be used in the single dri2videosink case. This was apparently a leftover
debugging change.

5 years agodri2videosink: avoid possible aliasing and locale issues
Vincent Penquerc'h [Mon, 25 Jun 2012 12:07:23 +0000 (12:07 +0000)]
dri2videosink: avoid possible aliasing and locale issues

5 years agodri2videosink: do not crash when XOpenDisplay fails
Vincent Penquerc'h [Mon, 25 Jun 2012 10:39:52 +0000 (10:39 +0000)]
dri2videosink: do not crash when XOpenDisplay fails

5 years agodrmbufferpool: fix device leak
Vincent Penquerc'h [Mon, 25 Jun 2012 10:13:46 +0000 (10:13 +0000)]
drmbufferpool: fix device leak

5 years agodri2videosink: fix more leaks
Vincent Penquerc'h [Mon, 25 Jun 2012 10:02:25 +0000 (10:02 +0000)]
dri2videosink: fix more leaks

5 years agodri2videosink: fix caps leak
Vincent Penquerc'h [Mon, 25 Jun 2012 09:19:41 +0000 (09:19 +0000)]
dri2videosink: fix caps leak

5 years agodri2videosink: refcount dri2window
Vincent Penquerc'h [Sun, 24 Jun 2012 13:58:38 +0000 (13:58 +0000)]
dri2videosink: refcount dri2window

I'm not convinced this was wrong in the first place, since the dri2window
was destroying the pool on destruction, and I'm told buffers cannot outlive
the pool, but refcounting makes this certain, though care has to be taken
to avoid ref cycles between dri2window and buffer pool.

5 years agodri2videosink: fix a few leaks, mostly on error
Vincent Penquerc'h [Sun, 24 Jun 2012 13:53:34 +0000 (13:53 +0000)]
dri2videosink: fix a few leaks, mostly on error

5 years agodri2videosink: unset variables before freeing their memory
Vincent Penquerc'h [Sun, 24 Jun 2012 13:52:51 +0000 (13:52 +0000)]
dri2videosink: unset variables before freeing their memory

Fraction variables do not actually leak in practice, but it's good
practice, so let's do it properly.

5 years agoWIP: start adding dri2videosink
Rob Clark [Thu, 14 Jun 2012 15:37:06 +0000 (10:37 -0500)]
WIP: start adding dri2videosink

5 years agodrmbufferpool: make it possible to subclass
Rob Clark [Thu, 14 Jun 2012 16:01:37 +0000 (11:01 -0500)]
drmbufferpool: make it possible to subclass

5 years agokmssink: use gst_buffer_replace
Rob Clark [Fri, 15 Jun 2012 01:27:47 +0000 (20:27 -0500)]
kmssink: use gst_buffer_replace

5 years agopvrvideosink: bump rank to PRIMARY+1
Vincent Penquerc'h [Fri, 22 Jun 2012 13:51:06 +0000 (13:51 +0000)]
pvrvideosink: bump rank to PRIMARY+1

We want this to be selected even before xvimagesink.

5 years agopvrvideosink: fix leaks on error paths
Vincent Penquerc'h [Fri, 22 Jun 2012 11:01:46 +0000 (11:01 +0000)]
pvrvideosink: fix leaks on error paths

5 years agopvrvideosink: fix context leak by refcounting it
Vincent Penquerc'h [Thu, 21 Jun 2012 18:26:34 +0000 (18:26 +0000)]
pvrvideosink: fix context leak by refcounting it

5 years agokmssink: round strides to match what gstreamer expects
Vincent Penquerc'h [Thu, 21 Jun 2012 15:04:55 +0000 (15:04 +0000)]
kmssink: round strides to match what gstreamer expects

5 years agodrm: link to libgstdmabuf-0.10
Alessandro Decina [Wed, 20 Jun 2012 10:10:39 +0000 (10:10 +0000)]
drm: link to libgstdmabuf-0.10

5 years agopvrvideosink: port to dmabuf
Alessandro Decina [Thu, 14 Jun 2012 15:01:50 +0000 (15:01 +0000)]
pvrvideosink: port to dmabuf

Use dmabuf to allocate and share buffers. This makes pvrvideosink
much more portable, getting rid of most omap dependencies.
There are still a few rough edges and a couple of dependencies on
omap that need to be taken out.

5 years agoAdd drm/ to gstreamer-plugins-bad-uninstalled.pc.in
Alessandro Decina [Mon, 18 Jun 2012 11:10:14 +0000 (11:10 +0000)]
Add drm/ to gstreamer-plugins-bad-uninstalled.pc.in

5 years agokmssink: move boilerplate decl on top to avoid compiler warning
Alessandro Decina [Mon, 18 Jun 2012 13:54:24 +0000 (13:54 +0000)]
kmssink: move boilerplate decl on top to avoid compiler warning

...plus whitespace fixes

5 years agokmssink: build fixes
Alessandro Decina [Mon, 18 Jun 2012 13:54:15 +0000 (13:54 +0000)]
kmssink: build fixes

5 years agokmssink: do modeset
Rob Clark [Tue, 12 Jun 2012 19:44:46 +0000 (14:44 -0500)]
kmssink: do modeset

Set the mode we actually know what the screen resolution is.  Just
assuming it is the first mode is incorrect and would depend on the
monitor.

We could be a bit more clever and pick the lowest resolution that
is greater than the requested video.  But picking highest resolution
is a safe bet.

5 years agokmssink: don't allow visible buffer to return to pool
Rob Clark [Tue, 12 Jun 2012 18:13:29 +0000 (13:13 -0500)]
kmssink: don't allow visible buffer to return to pool

Since we don't current get a vblank event, just keep the current and
last frame ref'd so they don't get returned to the decoder while they
are potentially visible.

5 years agokmssink: use drm bufferpool and dmabuf to for zero copy
Rob Clark [Tue, 12 Jun 2012 15:19:26 +0000 (10:19 -0500)]
kmssink: use drm bufferpool and dmabuf to for zero copy

Get rid of our internal buffer pool, and instead use GstDRMBufferPool.
Use dmabuf to get handle to underlying buffer object, and qdata to
attach our own private data to the buffer.

This way we don't really care who allocates the buffer, and can deal
in the same way with our own buffers, or any other buffer allocator
that is using dmabuf.

5 years agodrm: add GstDRMBufferPool
Rob Clark [Tue, 12 Jun 2012 14:41:31 +0000 (09:41 -0500)]
drm: add GstDRMBufferPool

A re-usable bufferpool for elements which are using DRM buffers.

TODO: remove omap dependencies.. not sure how to handle allocation
since that part still needs a driver specific ioctl, but for the
rest we could probably add some gst drm helpers and use that instead
of libdrm_omap.

5 years agokmssink: share the drm fd with other elements in the pipeline
Alessandro Decina [Thu, 24 May 2012 07:11:07 +0000 (07:11 +0000)]
kmssink: share the drm fd with other elements in the pipeline

Implement the GstVideoContext interface to share the drm fd with
other elements (eg gst-ducati). This change removes the kmssink
dependency on libdce. libdce is still needed by the
sys/kms/gstducati*buffer code, but that will go away soon too.

5 years agokmssink: make sure the connector property is initialized with its default
Alessandro Decina [Wed, 16 May 2012 14:29:04 +0000 (14:29 +0000)]
kmssink: make sure the connector property is initialized with its default

5 years agokmssink: make the connector id a property
Alessandro Decina [Wed, 16 May 2012 14:20:26 +0000 (14:20 +0000)]
kmssink: make the connector id a property

5 years agokmssink: add scale property to optionally render fullscreen
Alessandro Decina [Wed, 16 May 2012 14:07:43 +0000 (14:07 +0000)]
kmssink: add scale property to optionally render fullscreen

5 years agokmssink: center output on screen
Alessandro Decina [Wed, 16 May 2012 13:40:54 +0000 (13:40 +0000)]
kmssink: center output on screen

5 years agokmssink: rework the mode selection code
Alessandro Decina [Wed, 16 May 2012 12:07:20 +0000 (12:07 +0000)]
kmssink: rework the mode selection code

Simplify the mode selection code and take cropping into account.

5 years agokmssink: use dce_deinit to close dev/fd
Alessandro Decina [Wed, 16 May 2012 12:06:16 +0000 (12:06 +0000)]
kmssink: use dce_deinit to close dev/fd

...temporary hack until we still depend on dce

5 years agokmssink: disable preroll to avoid configuring the wrong resolution
Alessandro Decina [Wed, 16 May 2012 12:04:48 +0000 (12:04 +0000)]
kmssink: disable preroll to avoid configuring the wrong resolution

Disable preroll as it's called before GST_CROP_EVENT has been received, making
the sink configure the wrong mode based on padded caps.

5 years agoh264parse: disable some optimizations with AVC input
Alessandro Decina [Tue, 15 May 2012 08:15:37 +0000 (10:15 +0200)]
h264parse: disable some optimizations with AVC input

Don't use the new gst_h264_parser_identify_nalu_unchecked and don't split output
AUs when input AUs end (iow when packetized_last=TRUE). This is a workaround for
some interlaced streams that have both fields in the same AVC access unit, but
that apparently need to be converted to one-AU-per-field in byte-stream,au
output.

5 years agokmssink: hack buffer related code to avoid GType collisions
Alessandro Decina [Fri, 4 May 2012 10:09:35 +0000 (10:09 +0000)]
kmssink: hack buffer related code to avoid GType collisions

5 years agokmssink: consider kms buffers valid drm buffers (eek)
Alessandro Decina [Fri, 4 May 2012 10:08:47 +0000 (10:08 +0000)]
kmssink: consider kms buffers valid drm buffers (eek)

5 years agokms: add kmssink
Alessandro Decina [Fri, 4 May 2012 08:29:03 +0000 (08:29 +0000)]
kms: add kmssink

kmssink is a video sink using the linux kernel mode setting API.
It's useful to render video on linux on non-X setups.

Conflicts:

sys/Makefile.am

5 years agopvrvideosink: Post messages when there is an error
Luciana Fujii Pontello [Tue, 10 Apr 2012 14:40:39 +0000 (10:40 -0400)]
pvrvideosink: Post messages when there is an error

When an irrecoverably error happens, post a message to inform the
application.

5 years agopvrvideosink: fix crop for interlaced streams
Luciana Fujii Pontello [Tue, 6 Mar 2012 14:31:58 +0000 (09:31 -0500)]
pvrvideosink: fix crop for interlaced streams

Doubles the size of crop height for interlaced videos since decoders
send the size of half of the video.

5 years agocamerabin2: record audio with video by default
Luciana Fujii Pontello [Fri, 17 Feb 2012 05:56:02 +0000 (00:56 -0500)]
camerabin2: record audio with video by default

Set audio encoding profile by default, which makes audio recording to be
enabled for video mode.

5 years agovc1parse: disable BDU parsing until it's more mature
Alessandro Decina [Thu, 23 Feb 2012 12:48:57 +0000 (13:48 +0100)]
vc1parse: disable BDU parsing until it's more mature

Parsing and splitting BDUs seems to trigger a few bugs in
downstream decoders so disable it for now.

5 years agoh264parse: flag B frames
Alessandro Decina [Tue, 14 Feb 2012 16:38:20 +0000 (17:38 +0100)]
h264parse: flag B frames

Flag B frames so decoders can decide to drop them if needed

5 years agopvrvideosink: Fix flicker in video borders
Luciana Fujii Pontello [Tue, 7 Feb 2012 23:46:43 +0000 (18:46 -0500)]
pvrvideosink: Fix flicker in video borders

Since pvrvideosink now uses double bufferings, it's necessary to draw
the borders on both buffers to avoid flickering.

5 years agopvrvideosink: Style fixes
Luciana Fujii Pontello [Tue, 7 Feb 2012 23:46:01 +0000 (18:46 -0500)]
pvrvideosink: Style fixes

Fix indentation and style according to gst-indent check.

5 years agovc1parse: set duration on output buffers
Alessandro Decina [Wed, 8 Feb 2012 16:22:08 +0000 (17:22 +0100)]
vc1parse: set duration on output buffers

This has the nice side effect of avoiding hanging at EOS with truncated asf
clips.

5 years agovc1parse: flag bframes
Alessandro Decina [Tue, 31 Jan 2012 14:20:59 +0000 (15:20 +0100)]
vc1parse: flag bframes

5 years agovc1parse: workaround for simple/main codec_data being 5 bytes long
Alessandro Decina [Tue, 31 Jan 2012 10:03:45 +0000 (11:03 +0100)]
vc1parse: workaround for simple/main codec_data being 5 bytes long

Sometimes the codec_data buffer for simple/main pushed by asfdemux is 5 bytes
instead of 4. When that happens, codec_data is still valid but it seems to have
one 0x00 trailing byte. Might be a bug in the demuxer, needs more investigation.

5 years agovc1parse: remove stub stream conversion code
Alessandro Decina [Tue, 31 Jan 2012 10:02:10 +0000 (11:02 +0100)]
vc1parse: remove stub stream conversion code

5 years agovc1parse: add some more debugging to _renegotiate
Alessandro Decina [Tue, 31 Jan 2012 10:00:55 +0000 (11:00 +0100)]
vc1parse: add some more debugging to _renegotiate

5 years agovc1parse: Add VC1 parser plugin
Alessandro Decina [Tue, 31 Jan 2012 09:58:23 +0000 (10:58 +0100)]
vc1parse: Add VC1 parser plugin

Conflicts:

gst/videoparsers/Makefile.am
gst/videoparsers/plugin.c

5 years agocamerabin2: change profile to h264
Luciana Fujii Pontello [Tue, 1 May 2012 21:34:36 +0000 (17:34 -0400)]
camerabin2: change profile to h264

5 years agompegvideoparse: Fix rebase of interlaced field
Luciana Fujii Pontello [Tue, 1 May 2012 21:32:07 +0000 (17:32 -0400)]
mpegvideoparse: Fix rebase of interlaced field

5 years agocamerabin2: Change viewfinderbin to accept yuv
Luciana Fujii Pontello [Wed, 18 Jan 2012 00:29:52 +0000 (19:29 -0500)]
camerabin2: Change viewfinderbin to accept yuv

Viewfinderbin will only work with yuv, which is compatible with
pvrvideosink and omx_camera.

5 years agoomxcamerabinsrc: Request yuv from omx_camera src pad
Luciana Fujii Pontello [Fri, 13 Jan 2012 00:26:05 +0000 (22:26 -0200)]
omxcamerabinsrc: Request yuv from omx_camera src pad

Use yuv instead of yuv-strided in omx_camera src pad and remove
unnecessary stridetransforms.

5 years agopvrvideosink: fix double free freeing ->display_par
Alessandro Decina [Tue, 24 Jan 2012 15:04:54 +0000 (16:04 +0100)]
pvrvideosink: fix double free freeing ->display_par

5 years agompegpsdemux: limit the amount of scanning done on duration queries
Alessandro Decina [Wed, 18 Jan 2012 11:04:52 +0000 (12:04 +0100)]
mpegpsdemux: limit the amount of scanning done on duration queries

Limit the amount of data scanned when looking for PTSes in duration queries as a
failsafe for kinda broken, potentially large files with sparse or no PTSes.

5 years agoh264parse: consider input as packetized only if format=avc is explicitly set in caps
Alessandro Decina [Thu, 12 Jan 2012 13:04:47 +0000 (14:04 +0100)]
h264parse: consider input as packetized only if format=avc is explicitly set in caps

The previous assumption was to consider input as avc if codec_data was set in
caps even if the format field was missing. Apparently the assumption was wrong
as there's at least another non-avc compliant codec_data format out there.

5 years agompegvideoparse: set interlaced field in caps
Alessandro Decina [Fri, 6 Jan 2012 12:38:41 +0000 (13:38 +0100)]
mpegvideoparse: set interlaced field in caps

5 years agopvrvideosink: fix incorrect window coordinates
Rob Clark [Mon, 2 Jan 2012 17:57:38 +0000 (11:57 -0600)]
pvrvideosink: fix incorrect window coordinates

When video is larger than screen resolution, the request to create
window at video size will result in a smaller than requested window.
So get the actual window geometry after the window is created.  This
solves an issue with video appearing "split" with the right half of
the video appearing on the left of the window.

5 years agopvrvideosink: some cleanups
Rob Clark [Mon, 2 Jan 2012 17:56:46 +0000 (11:56 -0600)]
pvrvideosink: some cleanups

5 years agopvrvideosink: fix NPE when closing window
Rob Clark [Mon, 2 Jan 2012 17:12:24 +0000 (11:12 -0600)]
pvrvideosink: fix NPE when closing window

5 years agopvrvideosink: fix potential NPE
Rob Clark [Sun, 1 Jan 2012 03:08:05 +0000 (21:08 -0600)]
pvrvideosink: fix potential NPE

5 years agodvdspu: add interlaced support
Rob Clark [Sun, 1 Jan 2012 03:07:42 +0000 (21:07 -0600)]
dvdspu: add interlaced support

5 years agodvdspu: cleanup debug code
Rob Clark [Sun, 1 Jan 2012 00:35:40 +0000 (18:35 -0600)]
dvdspu: cleanup debug code

Clean up debug code for drawing highlight rects which wasn't even
compiling previously.

TODO: might be nice if this could be controlled by properties or
env variables to make it possible to enable the debug code at
runtime.

5 years agodvdspu: use refcnting instead of explicit copy
Rob Clark [Sat, 31 Dec 2011 21:13:11 +0000 (15:13 -0600)]
dvdspu: use refcnting instead of explicit copy

5 years agopvrvideosink: implement copy vmethod in buffer subclass
Rob Clark [Thu, 29 Dec 2011 06:20:51 +0000 (00:20 -0600)]
pvrvideosink: implement copy vmethod in buffer subclass

When upstream element needs a writeable buffer (ex. dvdspu), it will call
gst_buffer_make_writable(), which will use the copy vmethod to create a
copy of the buffer.  Implement copy vmethod so we don't end up having to
copy back into a bufferpool buffer in order to render.

5 years agopvrvideosink: corruption/jerkiness fixes
Rob Clark [Wed, 28 Dec 2011 08:22:34 +0000 (02:22 -0600)]
pvrvideosink: corruption/jerkiness fixes

1) Use GetDrawableParameters() to get the next client side buffer
(it is double buffered) so we don't immediately try to blit into
the same buffer xorg is reading from.

2) Some related changes to handle invalidated dri2 drawable (such
as window size change or if window is redirected/unredirected by
the window manager).

3) Wait for blits from dst surface before blitting next frame..
this avoids artifacts caused by us writing to a buffer that xorg
is still reading from.  Note that there is a similar fix needed
in the pvr xorg driver to handle the inverse problem.

5 years agopvrvideosink: add navigation
Rob Clark [Tue, 27 Dec 2011 08:33:58 +0000 (02:33 -0600)]
pvrvideosink: add navigation

5 years agopvrvideosink: add filtering
Rob Clark [Tue, 27 Dec 2011 07:05:54 +0000 (01:05 -0600)]
pvrvideosink: add filtering

5 years agodvdspu: some optimizations
Rob Clark [Tue, 27 Dec 2011 18:55:58 +0000 (12:55 -0600)]
dvdspu: some optimizations

Detect invisible pixels, and skip gstspu_vobsub_blend_comp_buffers()
when there are only invisible pixels.  This significantly reduces the
CPU load in cases of DVDs which don't use the clip_rect to exclude
processing for parts of the screen where the video is visible.

5 years agodvdspu: add NV12 support
Rob Clark [Tue, 27 Dec 2011 04:04:04 +0000 (22:04 -0600)]
dvdspu: add NV12 support

5 years agoresindvd: use decodebin2 so needed parsers are plugged
Rob Clark [Mon, 26 Dec 2011 23:14:16 +0000 (17:14 -0600)]
resindvd: use decodebin2 so needed parsers are plugged

5 years agopvrvideosink: fix interlaced top coordinate
Luciana Fujii Pontello [Thu, 5 Jan 2012 13:21:56 +0000 (11:21 -0200)]
pvrvideosink: fix interlaced top coordinate

Top coordinate is not changed the first blit.

Style fixes.

5 years agopvrvideosink: interlaced support
Rob Clark [Tue, 27 Dec 2011 06:16:51 +0000 (00:16 -0600)]
pvrvideosink: interlaced support

5 years agopvrvideosink: a few cleanups..
Rob Clark [Tue, 27 Dec 2011 05:04:42 +0000 (23:04 -0600)]
pvrvideosink: a few cleanups..

Need to avoid caps parsing on every frame, and prefer to detect
case of caps that haven't changed early and bail out.

5 years agopvrvideosink: fix problem with non-ducati buffers
Rob Clark [Tue, 27 Dec 2011 04:50:22 +0000 (22:50 -0600)]
pvrvideosink: fix problem with non-ducati buffers

In some cases dvdspu seems to give us a non-pad-alloc'd buffer.  Don't
crash randomly in these cases.

5 years agoh264parse: avoid pushing SPS/PPS at start only if both are found in stream
Alessandro Decina [Mon, 19 Dec 2011 11:32:26 +0000 (12:32 +0100)]
h264parse: avoid pushing SPS/PPS at start only if both are found in stream

Avoid injecting SPS/PPS nals only if both are already present in the stream.
Fixes some AVC clips that happen to have only PPS nals in stream.

5 years agopvrvideosink: Send application message with crop size
Luciana Fujii Pontello [Mon, 12 Dec 2011 22:21:54 +0000 (20:21 -0200)]
pvrvideosink: Send application message with crop size

Sends the original resolution of the video (not the padded one) as an
application message. This allows applications to use the right
resolution for any purpose, such as calculating aspect ratio themselves
as totem does.

5 years agopvrvideosink: accept -strided video with 2d buffers
Luciana Fujii Pontello [Thu, 8 Dec 2011 22:18:47 +0000 (20:18 -0200)]
pvrvideosink: accept -strided video with 2d buffers

Change pvrvideosink to support -strided video using 2d buffers in the
bufferpool.

5 years agopvrvideosink: consider cropping when resizing for par
Luciana Fujii Pontello [Mon, 21 Nov 2011 02:39:03 +0000 (21:39 -0500)]
pvrvideosink: consider cropping when resizing for par

When configuring pixel-aspect-ratio consider crop event.

5 years agopvrvideosink: respect pixel-aspect-ratio
Luciana Fujii Pontello [Wed, 16 Nov 2011 17:46:51 +0000 (12:46 -0500)]
pvrvideosink: respect pixel-aspect-ratio

Implement pixel-aspect-ratio handling.

5 years agopvrvideosink: initial support for the CROP event
Alessandro Decina [Wed, 21 Sep 2011 11:39:03 +0000 (13:39 +0200)]
pvrvideosink: initial support for the CROP event

Conflicts:

sys/pvr2d/gstpvrvideosink.c
sys/pvr2d/gstpvrvideosink.h