Unlike x86, fmin/fmax are single instructions, not function calls. They
are much much faster than doing a comparison, then branching based on its
results. With this, audiodsp.vector_clipf gets almost twice as fast, and
a properly unrollled version of it gets 4-5x faster, on SiFive-U74.
This is only the low-hanging fruit: FFMIN and FFMAX are presumably
affected as well.
This likely applies to other instruction sets with native IEEE floats,
especially those lacking a conditional select instruction.
This instruction, if aligned on a 4-byte boundary, defines a valid target
("landing pad") for an indirect call or jump. Since this instruction is a
HINT, it is safe to assemble even if not included in the target
instruction set architecture.
The necessary alignment is already provided by the `func` macro. However
this still lacks the ELF attribute to indicate that the zicfilp is supported
in simple mode. This is left for future work as the ELF specification is not
ratified as of yet.
This will also nonobviously require the assembler to support zicfilp,
insofar as the `tail` pseudo-instruction shall clobber T2 (instead of T1) as
its temporary register.
Currently the start of the byte range for each function is aligned to
4 bytes. But this can lead to situations whence the function is preceded
by a 2-byte C.NOP at the aligned 4-byte boundary. Then the first actual
instruction and the function symbol are only aligned on 2 bytes.
This forcefully disables compression for the alignment and the symbol,
thus ensuring that there is no padding before the function.
The B extension was finally ratified in May 2024, encompassing:
- Zba (addresses),
- Zbb (basics) and
- Zbs (single bits).
It does not include Zbc (base-2 polynomials).
configure checks that the assembler supports the B extension (or rather
its constituents) anyway. These macros were dodging sanity checks for
unsupported instructions and nothing else.
The RISC-V B bit manipulation extension was ratified only two months ago.
But it is strictly equivalent to the union of the zba, zbb and zbs
extensions which were defined almost 3 years earlier. Rather than require
new assembler, we can just match the extension name manually and translate
it into its constituent parts.
If __riscv_hwprobe() fails, then the kernel version is presumably too
old. There is not much point falling back to the auxillary vector.
- The Linux kernel requires I, so the flag is always set on Linux, and
run-time detection is unnecessary. Our RISC-V assembler does anyway not
support targets without I.
- Linux can compile with or without F and D, but it cannot perform
run-time detection for them (a kernel with F support will not boot a
processor without F). The run-time detection is thus useless in that
case. Besides F and D extensions are used throughout the C code, so
their run-time detection would not be practical.
- Support for V was added in a later kernel version than riscv_hwprobe(),
so the system call will always be available if the kernel supports V.
The only exception would be vendor kernel forks, but those are known to
haphasardly pretend to support V on systems without actual V support, or
with only pre-ratification binary-incompatible version. Furthermore, a
large chunk of our optimisations require Zba and/or Zbb which cannot be
detected with HWCAP in those kernels.
For what it is worth, OpenJDK already took a similar action. Note that this
keeps AT_HWCAP usage for platforms with neither C run-time <sys/hwprobe.h>
nor kernel <asm/hwprobe.h>, notably kernels other than Linux.
Fixes: CID1604383 Unchecked return value
Fixes: CID1604439 Unchecked return value
Sponsored-by: Sovereign Tech Fund
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: CID1604487 Unchecked return value
Fixes: CID1604494 Unchecked return value
Sponsored-by: Sovereign Tech Fund
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: negation of -2147483648 cannot be represented in type 'int'; cast to an unsigned type to negate this value to itself
Fixes: 68550/clusterfuzz-testcase-minimized-ffmpeg_dem_MXF_fuzzer-6424065930756096
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
I've accidentally used API not available on the checked version.
Additionally check for the SDK to be new enough to even have the
CVImageBufferCreateColorSpaceFromAttachments API to not fail
the build.
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
There's nothing stopping users from writing to such buffers.
Its more accurate to say they are singular, i.e. not duplicated
between multiple submissions.
This can be helpful for global statistics, or error propagation
purposes.
Should make strict compilers happy.
Also, make AV_COPY128 use integer operations while at it. Removing the
inclusion of immintrin.h ensures a lot less intrinsic related headers are
included as well, which fixes a clash of defines with some Clang versions.
Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: James Almer <jamrial@gmail.com>
av_executor_execute run the task directly when thread is disabled.
The task can schedule a new task by call av_executor_execute. This
forms an implicit recursive call. This patch removed the recursive
call.
Removed by accident in the previous commits. This makes the code only run when
compiled with GCC and Clang like before. Support for other compilers like msvc
can be added later.
Signed-off-by: James Almer <jamrial@gmail.com>
This has the benefit of removing any SSE -> AVX penalty that may happen when
the compiler emits VEX encoded instructions.
Signed-off-by: James Almer <jamrial@gmail.com>
When called inside a loop, the inline asm version results in one pxor
unnecessarely emitted per iteration, as the contents of the __asm__() block are
opaque to the compiler's instruction scheduler.
This is not the case with intrinsics, where pxor will be emitted once with any
half decent compiler.
This also has the benefit of removing any SSE -> AVX penalty that may happen
when the compiler emits VEX encoded instructions.
Signed-off-by: James Almer <jamrial@gmail.com>
Fixes: CID1591930 Wrong sizeof argument
Sponsored-by: Sovereign Tech Fund
Reviewed-by: Steve Lhomme <robux4@ycbcr.xyz>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: CID1591944 Wrong sizeof argument
Sponsored-by: Sovereign Tech Fund
Reviewed-by: Steve Lhomme <robux4@ycbcr.xyz>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: CID1598558 Resource leak
Sponsored-by: Sovereign Tech Fund
Reviewed-by: Steve Lhomme <robux4@ycbcr.xyz>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: CID1591909 Wrong sizeof argument
Sponsored-by: Sovereign Tech Fund
Reviewed-by: Steve Lhomme <robux4@ycbcr.xyz>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
In addition to the other properties, try to obtain the right
CGColorSpace and set it as well, else it could lead to a CVBuffer
tagged as BT.2020 but with a CGColorSpace indicating BT.709.
Therefore it is essential for consistency to set a colorspace
according to the other values, or if none can be obtained (for example
because the other values are all unspecified) unset it as well.
Fix#10884
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
The documentation was not clear at all what specifically the
function does, so it was left unspecified if it will unset or
not touch attachments it could not map from the AVFrame.
The documentation of the return value was wrong as well.
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
When mapping AVFrame properties to the CVBuffer attachments, it is
necessary to properly delete undefined attachments, else we can
leave incorrect values in there guessed from VideoToolbox for
example, leading to inconsistent results where the AVFrame and
CVBuffer differ in metadata.
Ref #10884
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
Given that a video stream/frame may have only one view or both views coded with
the packing information being unavailable, this commit adds a new type value
AV_STEREO3D_UNSPEC for this purpose.
The most common case for this is container level signaling of Stereo3D video
where the specifics are defined at the bitstream level.
Signed-off-by: James Almer <jamrial@gmail.com>
Before the patch, disable threads support at configure/build time
was the only method to force zero thread in executor. However,
it's common practice for libavcodec to run on caller's thread when
user specify thread number to one. And for WASM environment, whether
threads are supported needs to be detected at runtime. So executor
should support zero thread at runtime.
A single thread executor can be useful, e.g., to handle network
protocol. So we can't take thread_count one as zero thread, which
disabled a valid usercase.
Other libraries take -threads 0 to mean auto. Executor as a low
level utils doesn't do cpu detect. So take thread_count zero as
zero thread, literally.
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
This avoids hardcoding any implementation-specific limitiations as
part of the API, and allows for future expandability.
This also allows API users to more conveniently convert the
values into floats without hardcoding specific conversion constants.
The API was committed a few days ago, so changing this field now
is within the realms of acceptable.
Fei Wang
James Almer
Rémi Denis-Courmont
Niklas Haas
Michael Niedermayer
Zhao Zhili
Marvin Scholz
Lynne
Tong Wu
Marton Balint
J. Dekker
Brad Smith
Cosmin Stejerean