Rom1v gl filter part4

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

For convenience, use a local variable to reference the vlc_gl_api.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

This avoids to rewrite consecutive cleanups on each error case.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

The name of arrays variable should use the plural form for clarity.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

The purpose of this callback is to load sampler data (bind textures,
load uniforms, etc.) to be used for the OpenGL draw call.

In addition, the name "load" also allows to possibly add an "unload"
function later (to reset bindings).

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

This makes explicit that these functions are set by the sampler
implementation and must be called by the user of sampler.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

If GetUniformLocation() returns -1, then this is an implementation
error.

Assert instead of reporting the error to the caller.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

Hide sampler implementation details from renderer.

This will allow to expose only what OpenGL filters need in the public
API.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

For now, only the renderer dumps its shaders depending on the verbose
level.

This simplifies vlc_gl_renderer_New(), and paves the way to convert the
renderer to an OpenGL filter.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

The code in fragment_shaders.c is part of the implementation of the
sampler: it generates the fragment shader to expose a GLSL function
"vlc_texture(vec2 coords)".

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 4 years ago and Alexandre Janniaux committed 4 years ago

The function vlc_gl_renderer_Draw() (almost) just called
DrawWithShaders(). There is no need for a separate function.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Hugo Beauzée-Luyssen authored 4 years ago and Thomas Guillem committed 4 years ago

Signed-off-by: Thomas Guillem <thomas@gllm.fr>

Add references to the functions creating tracks and tracklists from the delete
functions.

To keep compability with the old track API since VLC's ports are still using it.

This will be removed when the old track API is removed.

akrug authored 4 years ago and Hugo Beauzée-Luyssen committed 4 years ago

Browse again with increased "StartingIndex" and adapted "RequestCount".
    Solves
    #21381 Panasonic Viera returns maximal 20 items on uPnP
    #22496 DLNA/UPnP - Panasonic recorder 12 records limit

Fixes #15876

Signed-off-by: Felix Paul Kühne <felix@feepk.net>
Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

When AVI_IndexLoad gets invoked multiple time on the same file, we end
up overwriting the previous index table.

The timer API is not supported in Winstore builds so switch to something that
works for all.

The timer (thread based) is created once and armed/disarmed when needed.

This adds the possibility to pause the player callback, read the event, and
resume the player callback.

Since the mock:// duration is 5seconds.

This debug code assert that a track should not not deleted with
libvlc_media_track_delete() when it comes from a list.

Use the string identifier from track events.

That will be used by player tracks.

This common opaque struct will be used by the media and the media_player.

And use a trackpriv struct to hold the specific audio/video/sub part, removing
an extra alloc usage.

This is what's used to provide the thumbnails sizes through libvlc

Close() is called if any error occurs in Open() so it should be safe from
uninitialized fields.

refs #24842

The avcodec-fast option, if enabled, sets the AV_CODEC_FLAG2_FAST
option. Using AV_CODEC_FLAG2_FAST can cause use of unsafe code in
avcodec which could lead to crashes.

Backport a UNICODE fix that was merged after the GnuTLS code freeze.

Forbid call to GetHandleInformation(). fcntl never goes in that code section in
Winstore builds. It's only used for linux random number generation.

Do not allow calling SecureZeroMemory which is forbidden which is forbidden
and not available as an inline function in some cases.

This is the proper counterpart to windowsapp.

Now that we have a proper Docker image to build it:
registry.videolan.org/vlc-debian-llvm-uwp:20200603145315

A recent mingw64 8 (unreleased) is needed to make use of this. It's available
in our Docker images and in msys2 (although it's using msvcrt so it will
probably fail to link properly)

The forced -lwindowsappcompat is added like the other LDFLAGS in configure.ac.

The other UCRT builds might work in win7.

-luuid is needed as we use IID_IUnknown

Add missing $(AM_LIBADD) messing up the order of libraries

Fixes this warning with clang10:
incompatible pointer types passing 'u_long *' (aka 'unsigned long *') to parameter of type 'uint32_t *' (aka 'unsigned int *')

Similar to what is done in access/smb2.c

As long as at least one client is active, we need to keep trying to
read/write to drain the RX buffers or fill the TX buffers.

We still need to poll (with zero timeout) to detect new connections.

Fixes #24824.

In case of TLS, the TCP socket poll state does not necessarily match
the TLS stream state: data may be in the library-side TLS buffers.
So we need to try to read/write regardless of the socket events.

Refs #24824.

Account for actual data read or written, rather than iteration of the
poll loop.

Note: This won't count TLS handshake progress as activity. But if
the client cannot complete the handshake within a time-out interval,
we might as well drop it.

So the caller knows if the client needs to be polled or not.

If the underlying stream failed, there's no point processing an
outstanding request. It's most likely incomplete, and in any case, we
won't be able to send the response.

That corner case only made sense for connections half-closed on read
end. (Even then, it's a little questionable, because half-closed
connections eventually time out if not fully closed.)

(No functional changes)

"passing argument 1 of ‘vlc_obj_malloc’ from incompatible pointer type"
"expected ‘vlc_object_t *’ {aka ‘struct vlc_object_t *’} but argument is of type ‘demux_t *’"

Warnings introduced in

9a55abf4

tired of bogus HLS configs
ref #24817

(same as VLC 3.0 does)

System headers must be included before VLC ones, especially on Windows.

This does not or no longer exist.

Call vlc_fifo_DequeueAllUnlocked() so that the block and byte counters
are correctly reset to zero.

Refs #24811.

(in other words, DCCP)

RTP and RTCP are packet-based. The low-level I/O abstraction needed is
similar but different from that of connection-oriented streams as
provided by `struct vlc_tls'. So this adds a new suitable abstraction
for the RTP plugin(s).

That's what should be used when calling WaitOnAddress.
It's not part of the kernel32.dll anymore.

https://docs.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-waitonaddress

ecc_scalar_random is public in the nettle they embed, but the symbol is not
mangled in the original nettle package, so it's not mangled here either. It
leads to the nettle symbol behind defined twice.

The bug is actually in nettle, not the script that imports nettle in gnutls.

To build native tools it tries to use $CC when cross-compilation is detected,
which is most likely wrong.

This patch already exists in nettle that contains a version of gmp. But it's
not in the upstream version of gmp.

It may link with iphlpapi but be forbidden in UWP.

This test relies on the fact that -Werror-implicit-function-declaration works
in the compiler. And that's why it's done after it's enabled.

It's only used for the --miface option to set the preferred multicast interface.

We know what app we're running.

This version uses msvcrt as the C runtime and not the UCRT which we can't
distribute on Win7 to Win8.1.

Thomas Guillem authored 4 years ago and Hugo Beauzée-Luyssen committed 4 years ago

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

Thomas Guillem authored 4 years ago and Hugo Beauzée-Luyssen committed 4 years ago

This special case is now useless since p_source is always copied to p_dest,
right after in the memmove loop.

Furthermore, this caused an invalid payload size when the buffer was actually
reallocated since the payload calculation was comparing two different buffers
offset.

CVE-2020-13428

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

Thomas Guillem authored 4 years ago and Hugo Beauzée-Luyssen committed 4 years ago

It will test the hxxx_helper API that is only used by hw decoders (that are not
tested). It will also test annexB <-> xvcC conversion, via the
"hxxx-helper-testdec-xvcC" variable.

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

Thomas Guillem authored 4 years ago and Hugo Beauzée-Luyssen committed 4 years ago

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

From Rémi:
"Test plugins should use noinst_*, IMO. With check_*, you would end up with
stale plugins and potentially weird crashes due to ABI breaks if alternating
make all and make check."

Romain Vimont authored 5 years ago and Alexandre Janniaux committed 4 years ago

The renderer is expected to be independant of the input picture format.
Instead, update the sampler directly.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 5 years ago and Alexandre Janniaux committed 4 years ago

Remove the remaining usage of interop from renderer.

Now, the renderer only need to use the sampler.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 5 years ago and Alexandre Janniaux committed 4 years ago

The sampler were owned by the renderer.

Move it to vgl, so that the renderer never have to handle an interop
instance directly.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>

Romain Vimont authored 5 years ago and Alexandre Janniaux committed 4 years ago

For convenience, expose the video format directly in sampler.

Signed-off-by: Alexandre Janniaux <ajanni@videolabs.io>