For left HFYU prediction, we predict from the buffer buf+1 using 8- or
16-byte reads. This means that aligning the buffer by 16 bytes is in
itself not sufficient, because if the width itself is 16- or 8-byte
aligned, the buffer will not be padded, and thus a read of size 16 at
buf+1 will overflow boundaries at the right edge. Padding the buffer by
1 byte is sufficient to not overflow its boundaries.
Fixes bug 342.
This makes add_hfyu_left_prediction_sse4() handle sources that are not
16-byte aligned in its own function rather than by proxying the call to
add_hfyu_left_prediction_ssse3(). This fixes a crash on Win64, since the
sse4 version clobberes xmm6, but the ssse3 version (which uses MMX regs)
does not restore it, thus leading to XMM clobbering and RSP being off.
Fixes bug 342.
Some calculations were changed in b6a3849 to use mmsize, which was not correct
for the AVX version, which uses INIT_YMM and therefore has mmsize == 32.
Fixes Bug 341.
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
This completes the conversion of h264dsp to yasm; note that h264 also
uses some dsputil functions, most notably qpel. Performance-wise, the
yasm-version is ~10 cycles faster (182->172) on x86-64, and ~8 cycles
faster (201->193) on x86-32.
Without this, cglobal will expand "z" to "zh" to access the high byte
in a register's word, which causes a name collision with the ZH(x) macro
further up in this file.
Mixing yasm and inline asm is a bad idea, since if either yasm or inline
asm is not supported by your toolchain, all of the asm stops working.
Thus, better to use either one or the other alone.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Write out the NAL decoding loops in full so that they are easier
to parse for a preprocessor without it having to be aware of macros
or other such things in C code.
This also makes the code more readable.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
The actual result of the check isn't used anywhere (since we
use this function unconditionally within #ifdef _WIN32), but it
makes sure we explicitly link to shell32 (which is linked in
implictly on mingw).
Signed-off-by: Martin Storsjö <martin@martin.st>
This function is only available in the headers if unconditionally
targeting a version >= XP. It is not strictly necessary, since we
try loading these functions dynamically at runtime on windows in
the fallback, but this makes things a bit faster and more
straightforward.
On mingw32, this function isn't visible by default, while it is
on mingw64 (on both 32 and 64 bit).
Signed-off-by: Martin Storsjö <martin@martin.st>
This allows compiling and running these tests on systems lacking a built-
in version of getopt(), such as MSVC.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
This is required for CommandLineToArgvW. Normally, shellapi.h is included
implicitly by windows.h, but if we define WIN32_LEAN_AND_MEAN (or some of
the other earlier headers have included windows.h with that option),
windows.h doesn't include this one.
Thus explicitly include the headers we really need, for clarity and
compatibility.
Also replace x>>av_log2(sizeof(..)) + 1 by x/sizeof(..). The +1 is
probably meant to emulate av_log2_ceil(sizeof(..)) in cases where ".."
is not a power of two.
intfloat.h is a public header, and is now (since a1245d5ca) included
by mathematics.h, which many external callers include.
This fixes building third party applications that include
mathematics.h in a language that doesn't support designated
initalizers.
Signed-off-by: Martin Storsjö <martin@martin.st>
Some compilers, MSVC among them, don't recognize the divisions by
zero as meaning infinity/nan.
These macros should, according to the standard, expand to constant
expressions, but this shouldn't matter for our usage.
Signed-off-by: Martin Storsjö <martin@martin.st>
Mingw headers provide similar defines already (unconditional #defines,
without any #undef or #ifdef around it), while MSVC doesn't have
them.
Signed-off-by: Martin Storsjö <martin@martin.st>
io.h is required for open and _wopen, and fcntl.h is required for
the O_CREAT flag. On mingw, fcntl.h is included by os_support.h (and
the mingw fcntl.h includes io.h), but include it explicitly here
since this implementation requires it.
Also move the #undef open up. open must not be defined to ff_win32_open
while including the headers that declare the open function. On mingw,
this happened in os_support.h before open was redirected.
Signed-off-by: Martin Storsjö <martin@martin.st>
It is included for the open/read/write/close functions. On
MSVC, where this header does not exist, the same functions
are provided by io.h, which is already included.
On windows, these functions are provided by io.h. Make sure
io.h is included if it exists, regardless of the setmode
function.
Signed-off-by: Martin Storsjö <martin@martin.st>
unistd.h is used for open/read/close, but if this header does not
exist, there's probably no use in trying to open /dev/*random
at all.
Signed-off-by: Martin Storsjö <martin@martin.st>
unistd.h used to be required for gethostname. On windows, gethostname
is provided by winsock2.h. Now network.h includes both unistd.h and
winsock2.h if they exist.
Signed-off-by: Martin Storsjö <martin@martin.st>
Apparently, some build environments require dxva.h even for dxva2,
while others lack this header entirely. Including it conditionally
allows building in both cases.
Signed-off-by: Martin Storsjö <martin@martin.st>
Since the errno.h values don't match the error codes that winsock
returns, map the winsock error codes to the errno ones, to make
sure explicit checks against AVERROR(x) match.
Signed-off-by: Martin Storsjö <martin@martin.st>
Apparently this include is needed on some systems for building the
poll fallback (for the timeval struct for select?), but it isn't
available on all systems. Thus only include it if it exists.
Signed-off-by: Martin Storsjö <martin@martin.st>
This is required for isatty, which exists on MSVC and is found by
configure, but is provided by io.h instead of unistd.h.
Signed-off-by: Martin Storsjö <martin@martin.st>
The check uses check_func_header, since this function is
conditionally available depending on the targeted MSVCRT
version.
Signed-off-by: Martin Storsjö <martin@martin.st>
Defining restrict results - for some compilers - in changing other
uses of the restrict keyword also, e.g. __declspec(restrict) gets
changed to __declspec(__restrict) on MSVC. This causes compilation
failures. Therefore, using a private namespace macro instead is
more reliable and robust.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This prevents gcc from assuming that contents of it may have changed
between calls to vp56_range_get_prob(), thus preventing countless (and
unnecessary) movs. Decoding of sintel trailer goes from (avg+SG) 9.796
+/- 0.003 to 9.635 +/- 0.010.
Prevents subsequent overreads when these numbers are used as indices
in arrays.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
Wrong bit depth can lead to invalid rowsize values, which crashes the
decoder further down.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Progressive data is allocated later in decode_sof(), not allocating
that data leads to NULL dereferences.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
This prevents sample_rate/data_length from going negative, which
caused various crashes and undefined behaviour further down.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Ronald S. Bultje