Prince Gupta authored 3 years ago and Jean-Baptiste Kempf committed 3 years ago

previously when the image is found in the cache, the generator
was not reset, this caused an invalid image to be loaded

reset the image generator on update request

David authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

OpenJDK builds now exist for arm macOS, and both compilation and execution
of java blurays seem to work fine.

Closes #26023

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

No functional changes.

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

Early fail and leave the function or continue the loop when an error
happens. The code will less look like a tornado and will directly
signal what happens in case of error, without breaking the flow of the
current success path.

Indentation is not changed to ease review, and will be changed in
a subsequent commit.

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

They allow easier error handling in the loop, and will provide the
foundation for much more simplification in the end.

Alexandre Janniaux authored 3 years ago and Rémi Denis-Courmont committed 3 years ago

The new contrib fixes a bunch of reported CVE:

    CVE-2021-30498  (tga export, not concerned)
    CVE-2021-30499  (troff export, not concerned)
    CVE-2021-3410   (canvas.c, v0.99.beta19)
    CVE-2018-20546  (dither.c, v0.99.beta19)
    CVE-2018-20547  (dither.c, v0.99.beta19)
    CVE-2018-20545  (load_image, v0.99.beta19)
    CVE-2018-20548  (load_image, v0.99.beta19)
    CVE-2018-20549  (caca_file_read, 0.99.beta19, not concerned)

See release note[^1] on the github project.

It also allow to migrate the use of the new API instead of libcucul,
which has been merged into libcaca since 0.99.beta15:

    commit f61816ceb7445f8bf818936151554ac060764b39
    Author: Sam Hocevar <sam@hocevar.net>
    Date:   Sat Sep 27 13:12:46 2008 +0000

    Starting refactoring to get rid of libcucul. The initial reason for the
    split is rendered moot by the plugin system: when enabled, binaries do
    not link directly with libX11 or libGL. I hope this is a step towards
    more consisteny and clarity.

It was then completely wiped out by the following commit, which is part
of v0.99.beta20:

    commit 5f0ec215f8c9915ed028324a8ecac8212f68e18d
    Author: Sam Hocevar <sam@hocevar.net>
    Date:   Thu May 3 10:33:30 2018 +0200

    Remove legacy code from 10 years ago.

Check NEWS for details:

    \section news0_99_beta15 Changes between 0.99.beta15 and 0.99.beta14

     - libcucul was merged back into libcaca for more clarity

[^1]: https://github.com/cacalabs/libcaca/releases/tag/v0.99.beta20

Alexandre Janniaux authored 3 years ago and Rémi Denis-Courmont committed 3 years ago

Migrate to the new API instead of libcucul API, which has been merged
into libcaca since 0.99.beta15:

    commit f61816ceb7445f8bf818936151554ac060764b39
    Author: Sam Hocevar <sam@hocevar.net>
    Date:   Sat Sep 27 13:12:46 2008 +0000

    Starting refactoring to get rid of libcucul. The initial reason for the
    split is rendered moot by the plugin system: when enabled, binaries do
    not link directly with libX11 or libGL. I hope this is a step towards
    more consisteny and clarity.

It was then completely wiped out by the following commit, which is part
of v0.99.beta20:

    commit 5f0ec215f8c9915ed028324a8ecac8212f68e18d
    Author: Sam Hocevar <sam@hocevar.net>
    Date:   Thu May 3 10:33:30 2018 +0200

    Remove legacy code from 10 years ago.

Steve Lhomme authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

We no longer build with pthreadGC2 since f1bf4b45

Ronald S. Bultje authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

Fixes #26259 (closed), #26579 and dav1d/#383.

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

override will ensure that the functions have the correct signature
compared to the interface they are implemented against.

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

The class has virtual functions and is calling `delete this;` so it
triggers a warning if the class has no virtual destructors.

    modules/access/decklink.cpp: In member function ‘virtual ULONG {anonymous}::DeckLinkCaptureDelegate::Release()’:
    modules/access/decklink.cpp:268:13: warning: deleting object of polymorphic class type ‘{anonymous}::DeckLinkCaptureDelegate’ which has non-virtual destructor might cause undefined behavior [-Wdelete-non-virtual-dtor]
      268 |             delete this;
          |             ^~~~~~~~~~~

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

Like done by _Generic in the C version, use overloading functions to
define the correct casting behaviour depending on whether the pointer
is already a vlc_object_t or possess a vlc_object_t as ->obj.

This removes the need for listing all the objects which need a casting
case, and non-vlc_object_t objects will fail with the following error:

    include/vlc_objects.h: In instantiation of ‘vlc_object_t* VLC_OBJECT(T*) [with T = {anonymous}::demux_sys_t; vlc_object_t = vlc_object_t]’:
    include/vlc_objects.h:83:18: error: ‘struct {anonymous}::demux_sys_t’ has no member named ‘obj’

or, if there is a obj field which is not a vlc_object_t:

    include/vlc_objects.h: In instantiation of ‘vlc_object_t* VLC_OBJECT(T*) [with T = Open(vlc_object_t*)::foo; vlc_object_t = vlc_object_t]’:
    include/vlc_objects.h:83:18: error: cannot convert ‘Open(vlc_object_t*)::foo::obj*’ to ‘vlc_object_t*’ in return
       83 |     { return &d->obj; }
          |               ~~~^~~
          |                  |
          |                  Open(vlc_object_t*)::foo::obj*

Alexandre Janniaux authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

intf_thread_t wasn't forward-declared and was available only thanks to
vlc_object_cast forward declaration.

This was only used to track the number of planes. We already know that
from the input (or output) picture.

They are only used in a single place nowadays, so adding a layer of
indirection is no longer useful.

So far, there was one common loop for vertical flip, and six loops per
bit depth for each other (non-trivial) transform, or 19 different inner
loops in total via macro expansion.

This keeps the common loop for vertical flip and reduces the other down
to two per bit depth: transposition and horizontal flip.
Antitransposition and the three rectangular rotations are obtained by
composing the previous two transforms with vertical flips before and/or
after the main transform. Those vertical flips are achieved for "free"
by using negative strides.

This is a relic from whence the user-specified transform could be
handled solely by the transform filter whence this code originates.
Now the display modules can accept 4:2:2 chroma subsampling *and*
handle the orientation at the same time and without silently
down-sampling the Cb and Cr components to 4:2:0.

Steve Lhomme authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

After a flush the streamStarted is reset, a new call to MFT_MESSAGE_NOTIFY_START_OF_STREAM
has to be done (ie startStream). It's not necessary to do it on exit.

This is only useful for async MFTs.

Steve Lhomme authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

As it's not always done on exit, we have to release all output the decoder may
have produced.

Steve Lhomme authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

The MFT_MESSAGE_NOTIFY_END_STREAMING actually does some D3D cleaning as well.

+ don't crash if DoRelease was called due to an error.

+ don't reset the input/output types anymore.

Steve Lhomme authored 3 years ago and Hugo Beauzée-Luyssen committed 3 years ago

Sometimes we get that kind of format before the packetizer decides what the
format is.