Makes it a bit more clear where each test belongs.

Suggested by Anton Khirnov.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

This isn't necessary any longer on MSVC 2013 Update 4.
Signed-off-by: Martin Storsjö <martin@martin.st>

This isn't necessary on MSVC 2015 any longer.
Signed-off-by: Martin Storsjö <martin@martin.st>

This is only necessary on MSVC 2010/2012 (and possibly on some
ICL versions). This both avoids an extra hack on newer MSVC
versions, and better documents what the extra compiler option
is used for.
Signed-off-by: Martin Storsjö <martin@martin.st>

The struct definitions in dxva.h, which are necessary in order to
actually use d3d11va, are hidden when WINAPI_FAMILY targets Windows Phone
or WindowsRT.

Building with WINAPI_FAMILY=WINAPI_FAMILY_DESKTOP_APP is disallowed
when targeting ARM. ("Compiling Desktop applications for the ARM
platform is not supported.") So we set _CRT_BUILD_DESKTOP_APP to 0
to tell the runtime not to detect some issues with this mismatching.

The same tweaks to detect if the API is available is done in dxva2_internal.h
when compiling each DXVA2/D3D11VA decoders.
Signed-off-by: Martin Storsjö <martin@martin.st>

Fixes alignment issues and bus errors.
Signed-off-by: Martin Storsjö <martin@martin.st>

It will be shared with the upcoming mpeg2 and hevc decoders.

Declare a named enum, not a variable with anonymous enum type.

init() is called whenever format changes, so current code would leak the
fifo in this case.

This is triggerable with the HEVC decoder. It is unclear yet whether the
bug is in the calling code or the MSDK, but it seems better to check for
this in any case.

Fallback to "link" in the path if the one next to cl is not found.
Signed-off-by: Martin Storsjö <martin@martin.st>

The -nologo- option is the inverse of -nologo, reenabling printing the
"logo" version header. This is useful if the compiler actually is a
wrapper that forces -nologo.
Signed-off-by: Martin Storsjö <martin@martin.st>

This fixes the palette on big endian, broken (or, differing from
little endian) since 57214b2f.
Signed-off-by: Martin Storsjö <martin@martin.st>

Some uses of Snappy require uncompressing to positions within
an existing buffer. Also adds a function to get the uncompressed
length of Snappy data.

this condition breaks reading from the pipe as data_reached variable
have to be set to break while in the asf_read_header just after the Data
Object is found
Signed-off-by: Anton Khirnov <anton@khirnov.net>

A bug was introduced in 97710540 whereby when
frame height wasn't divisible by the number of threads, pixels would be omitted
from the bottom rows during decode.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

A bug was introduced in 6b2b26e7 whereby when
frame height wasn't divisible by the number of threads, pixels would be omitted
from the bottom rows during decode.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

The current code expects query_formats() to be called exactly once, it
will leak if it's not called at all (filter initialized, but never
configured or used) or try to read freed memory if it's called more than
once.
Found-by: James Almer <jamrial@gmail.com>
CC: libav-stable@libav.org

to prevent possible memory leaks
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Found-By: ami_stuff
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

Red and blue channels were decoded in the wrong order.

Found-By: ami_stuff

Drop the global variables with anonymous enum type.

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

return INVALIDDARA if Data Object was not found
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Fixes MSVC compilation.

Checkasm needs to use internal symbols that should not be made public.

Enjoy some cache locality and use less threads.
About the same speedup provided to HAP.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Enjoy some cache locality and use less threads.
About 5x speedup (from 60ms to 12ms to decode a 4k frame).
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

uhd1 and uhd2 would be ambigous.

Avoids a relocation which might end out of range for thumb2.
Reported-By: Ludovic Fauvet <etix@videolan.org>
Bug-Id: https://bugs.webkit.org/show_bug.cgi?id=137022
CC: libav-stable@libav.org