The vector dequantization has a test in a loop preventing effective SIMD
implementation. By moving it out of the loop, this loop can be DSPized.
Therefore, modify the current DSP implementation. In particular, the
DSP implementation no longer has to handle null loop sizes.
The decode_hf implementations have following timings:
For x86 Arrandale:
C SSE SSE2 SSE4
win32: 260 162 119 104
win64: 242 N/A 89 72
The arm NEON optimizations follow in a later patch as external asm. The
now unused check for the y modifier in arm inline asm is removed from
configure.
Framerate is now a sane rational instead of an integer, and
inputDepth is changed to what it actually is.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Moving cpunop from the HAVE_LIST to the ARCH_EXT_LIST_X86 has the side
effect of enabling it. The semantics of the check have to be changed
from enable if successful to disable if unsuccessful. This was missing
in 2b0bb69997 causing build errors with
nasm.
Fixes dependency file generation with gas-preprocessor.pl and clang.
Flags copied from GCC and tested with Apple's clang from Xcode 5 and
5.1 and clang 3.2, 3.3, 3.4 on Linux.
NEON and VFP are currently mandatory for all ARMv8 profiles. Both are
handled as extensions as far as cpuflags are concerned. This is
consistent with handling x86_64 which always has SSE2, but still
handles it as an extension.
Stack is always 16 byte aligned and clz, 64bit operations and unaligned
memory access are fast in aarch64 mode on ARMv8.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
Previously, if neither of the checks for the closesocket function
succeeded, we still kept winsock2.h and networking in general
enabled.
When targeting the WinRT API subset, the winsock2.h header is
available (making the check for it succeed, giving the impression
that winsock is available), but tests that actually try to use
such a function will fail. In this case, disable the winsock2.h
feature and networking in general, as if the winsock2.h header
test would have failed in the first place.
Signed-off-by: Martin Storsjö <martin@martin.st>
New versions of FreeType have moved the location of their API
header(s) and hide the location behind a macro.
Since the location changes between versions and no other way
to know the location exists, this workaround becomes necessary.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Some pthreads symbols might be present in libc (as shown on various *BSD)
but not all of them, leading to false positives.
Check for the most common compiler flags before the plain symbol check
to avoid known pitfalls.
Directly loads AviSynth through LoadLibrary instead of relying on
Video for Windows, and supports using AvxSynth (via dlopen) to
open scripts on Linux and OS X.
Error messages from AviSynth/AvxSynth are now reported through
av_log and exit, rather than the traditional behavior of generating
an error video that the user would need to watch to diagnose.
The main rewrite was authored by d s <avxsynth.testing@gmail.com>
from the AvxSynth team, with additional contributions by
Oka Motofumi <chikuzen.mo@gmail.com>
Stephen Hutchinson <qyot27@gmail.com>
Diego Biurrun <diego@biurrun.de>
Anton Khirnov <anton@khirnov.net>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
