MSA2 optimizations are attached to MSA macros in generic_macros_msa.h.
It's difficult to do runtime check for them. Remove this part of code
can make it more robust. H264 1080p decoding: 5.13x==>5.12x.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
While Vulkan itself went more or less the way it was expected to go,
libvulkan didn't quite solve all of the opengl loader issues. It's multi-vendor,
yes, but unfortunately, the code is Google/Khronos QUALITY, so suffers from
big static linking issues (static linking on anything but OSX is unsupported),
has bugs, and due to the prefix system used, there are 3 or so ways to type out
functions.
Just solve all of those problems by dlopening it. We even have nice emulation
for it on Windows.
VkPhysicalDeviceLimits.optimalBufferCopyRowPitchAlignment and
VkPhysicalDeviceExternalMemoryHostPropertiesEXT.minImportedHostPointerAlignment are of type
VkDeviceSize (a typedef uint64_t).
VkPhysicalDeviceLimits.minMemoryMapAlignment is of type size_t.
Signed-off-by: James Almer <jamrial@gmail.com>
Reviewed-by: Lynne <dev@lynne.ee>
Originally deprecated in 1296b1f6c0631ab79464e22d48a6a1548450b943;
scheduled again for removal in a991526832.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
This commit adds a pure x86 assembly SIMD version of the FFT in libavutil/tx.
The design of this pure assembly FFT is pretty unconventional.
On the lowest level, instead of splitting the complex numbers into
real and imaginary parts, we keep complex numbers together but split
them in terms of parity. This saves a number of shuffles in each transform,
but more importantly, it splits each transform into two independent
paths, which we process using separate registers in parallel.
This allows us to keep all units saturated and lets us use all available
registers to avoid dependencies.
Moreover, it allows us to double the granularity of our per-load permutation,
skipping many expensive lookups and allowing us to use just 4 loads per register,
rather than 8, or in case FMA3 (and by extension, AVX2), use the vgatherdpd
instruction, which is at least as fast as 4 separate loads on old hardware,
and quite a bit faster on modern CPUs).
Higher up, we go for a bottom-up construction of large transforms, foregoing
the traditional per-transform call-return recursion chains. Instead, we always
start at the bottom-most basis transform (in this case, a 32-point transform),
and continue constructing larger and larger transforms until we return to the
top-most transform.
This way, we only touch the stack 3 times per a complete target transform:
once for the 1/2 length transform and two times for the 1/4 length transform.
The combination algorithm we use is a standard Split-Radix algorithm,
as used in our C code. Although a version with less operations exists
(Steven G. Johnson and Matteo Frigo's "A modified split-radix FFT with fewer
arithmetic operations", IEEE Trans. Signal Process. 55 (1), 111–119 (2007),
which is the one FFTW uses), it only has 2% less operations and requires at least 4x
the binary code (due to it needing 4 different paths to do a single transform).
That version also has other issues which prevent it from being implemented
with SIMD code as efficiently, which makes it lose the marginal gains it offered,
and cannot be performed bottom-up, requiring many recursive call-return chains,
whose overhead adds up.
We go through a lot of effort to minimize load/stores by keeping as much in
registers in between construcring transforms. This saves us around 32 cycles,
on paper, but in reality a lot more due to load/store aliasing (a load from a
memory location cannot be issued while there's a store pending, and there are
only so many (2 for Zen 3) load/store units in a CPU).
Also, we interleave coefficients during the last stage to save on a store+load
per register.
Each of the smallest, basis transforms (4, 8 and 16-point in our case)
has been extremely optimized. Our 8-point transform is barely 20 instructions
in total, beating our old implementation 8-point transform by 1 instruction.
Our 2x8-point transform is 23 instructions, beating our old implementation by
6 instruction and needing 50% less cycles. Our 16-point transform's combination
code takes slightly more instructions than our old implementation, but makes up
for it by requiring a lot less arithmetic operations.
Overall, the transform was optimized for the timings of Zen 3, which at the
time of writing has the most IPC from all documented CPUs. Shuffles were
preferred over arithmetic operations due to their 1/0.5 latency/throughput.
On average, this code is 30% faster than our old libavcodec implementation.
It's able to trade blows with the previously-untouchable FFTW on small transforms,
and due to its tiny size and better prediction, outdoes FFTW on larger transforms
by 11% on the largest currently supported size.
This sadly required making changes to the code itself,
due to the same context needing to be reused for both versions.
The lookup table had to be duplicated for both versions.
This commit refactors the power-of-two FFT, making it faster and
halving the size of all tables, making the code much smaller on
all systems.
This removes the big/small pass split, because on modern systems
the "big" pass is always faster, and even on older machines there
is no measurable speed difference.
av_set_cpu_flags_mask() has been deprecated in the commit which merged
it: 6df42f98746be06c883ce683563e07c9a2af983f; av_parse_cpu_flags() has
been deprecated in 4b529edff8.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Some files currently rely on libavutil/cpu.h to include it for them;
yet said file won't use include it any more after the currently
deprecated functions are removed, so include attributes.h directly.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
av_frame_copy() is allowed to return values >= 0 on success, whereas
the documentation of av_frame_ref() states that the return value is 0 on
success. Ergo the latter must not just return the former's return value.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
av_cpu_count() intends to emit a debug message containing the number of
logical cores when called the first time. The check currently works with
a static volatile int; yet this does not help at all in case of
concurrent accesses by multiple threads. So replace this with an
atomic_int.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
av_adler32_update() is used by av_hash_update() which will be switched
to size_t at the next bump. So it also has to be made to use size_t.
This is also necessary for framecrcenc.c, because the size of side data
will become a size_t, too.
Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
av_bprint_finalize() can still fail even when it has been checked that
the AVBPrint is currently complete: Namely if the string was so short
that it fit into the AVBPrint's internal buffer.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
