This code only checks hcScale. In practice this is not an issue because
the function pointers should always be identical to hyScale for the same
filter size.
Add an assertion just to make sure this assumption never regresses.
Signed-off-by: Niklas Haas <git@haasn.dev>
Sponsored-by: Sovereign Tech Fund
This corresponds to commit 9278a14cf406f8edb5052c42b83750112bf5b515
in dav1d.
Omitting the C-only functions doesn't speed up benchmarking
anyway (as those has to be benchmarked before we know if we have
any corresponding assembly functions), and being able to benchmark
those functions without corresponding assembly can be valuable in
a number of cases.
Signed-off-by: Martin Storsjö <martin@martin.st>
Share the checkasm_check_pixel macro from hevc_pel in checkasm.h,
to allow other tests to use the same. (To use it in other tests,
those tests need to have a similar setup for high bitdepth pixels,
with a local variable named "bit_depth".)
This simplifies the code for checking the output, and can print
the failing output (including a map of matching/mismatching
elements) if checkasm is run with the -v/--verbose option.
Signed-off-by: Martin Storsjö <martin@martin.st>
There is an issue with the constants used in YUV to YUV range conversion,
where the upper bound is not respected when converting to mpeg range.
With this commit, the constants are calculated at runtime, depending on
the bit depth. This approach also allows us to more easily understand how
the constants are derived.
For bit depths <= 14, the number of fixed point bits has been set to 14
for all conversions, to simplify the code.
For bit depths > 14, the number of fixed points bits has been raised and
set to 18, to allow for the conversion to be accurate enough for the mpeg
range to be respected.
The convert functions now take the conversion constants (coeff and offset)
as function arguments.
For bit depths <= 14, coeff is unsigned 16-bit and offset is 32-bit.
For bit depths > 14, coeff is unsigned 32-bit and offset is 64-bit.
x86_64:
chrRangeFromJpeg8_1920_c: 2127.4 2125.0 (1.00x)
chrRangeFromJpeg16_1920_c: 2325.2 2127.2 (1.09x)
chrRangeToJpeg8_1920_c: 3166.9 3168.7 (1.00x)
chrRangeToJpeg16_1920_c: 2152.4 3164.8 (0.68x)
lumRangeFromJpeg8_1920_c: 1263.0 1302.5 (0.97x)
lumRangeFromJpeg16_1920_c: 1080.5 1299.2 (0.83x)
lumRangeToJpeg8_1920_c: 1886.8 2112.2 (0.89x)
lumRangeToJpeg16_1920_c: 1077.0 1906.5 (0.56x)
aarch64 A55:
chrRangeFromJpeg8_1920_c: 28835.2 28835.6 (1.00x)
chrRangeFromJpeg16_1920_c: 28839.8 32680.8 (0.88x)
chrRangeToJpeg8_1920_c: 23074.7 23075.4 (1.00x)
chrRangeToJpeg16_1920_c: 17318.9 24996.0 (0.69x)
lumRangeFromJpeg8_1920_c: 15389.7 15384.5 (1.00x)
lumRangeFromJpeg16_1920_c: 15388.2 17306.7 (0.89x)
lumRangeToJpeg8_1920_c: 19227.8 19226.6 (1.00x)
lumRangeToJpeg16_1920_c: 15387.0 21146.3 (0.73x)
aarch64 A76:
chrRangeFromJpeg8_1920_c: 6324.4 6268.1 (1.01x)
chrRangeFromJpeg16_1920_c: 6339.9 11521.5 (0.55x)
chrRangeToJpeg8_1920_c: 9656.0 9612.8 (1.00x)
chrRangeToJpeg16_1920_c: 6340.4 11651.8 (0.54x)
lumRangeFromJpeg8_1920_c: 4422.0 4420.8 (1.00x)
lumRangeFromJpeg16_1920_c: 4420.9 5762.0 (0.77x)
lumRangeToJpeg8_1920_c: 5949.1 5977.5 (1.00x)
lumRangeToJpeg16_1920_c: 4446.8 5946.2 (0.75x)
NOTE: all simd optimizations for range_convert have been disabled.
they will be re-enabled when they are fixed for each architecture.
NOTE2: the same issue still exists in rgb2yuv conversions, which is not
addressed in this commit.
WASI mssing signal and siglongjmp support. This patch workaround
build error and add simd128 flag. Please note that many tests use
large array on stack, so you need to increase the stack size when
build checkasm, e.g., --extra-ldflags='-Wl,-z,stack-size=10485760'
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
This is a purely cosmetic commit aimed at replacing accesses to
SwsInternal.opts by direct access to SwsContext wherever convenient.
Sponsored-by: Sovereign Tech Fund
Signed-off-by: Niklas Haas <git@haasn.dev>
This is a preliminary step to separating these into a new struct. This
commit contains no functional changes, it is a pure search-and-replace.
Sponsored-by: Sovereign Tech Fund
Signed-off-by: Niklas Haas <git@haasn.dev>
This commit also reduces the number of times ff_sws_init_scale() gets
called (only once per bit depth), and the number of times randomize_buffers()
gets called (only if the function must be checked).
Benchmarks are only performed on bit depths 8 and 16 (since they are
different functions, and not only different constants).
Reduce input sizes to 8 (to test that the function works with widths
smaller than the vector length) and 1920 (raising the largest input
size to improve benchmark results).
And preserve the public SwsContext as separate name. The motivation here
is that I want to turn SwsContext into a public struct, while keeping the
internal implementation hidden. Additionally, I also want to be able to
use multiple internal implementations, e.g. for GPU devices.
This commit does not include any functional changes. For the most part, it is
a simple rename. The only complications arise from the public facing API
functions, which preserve their current type (and hence require an additional
unwrapping step internally), and the checkasm test framework, which directly
accesses SwsInternal.
For consistency, the affected functions that need to maintain a distionction
have generally been changed to refer to the SwsContext as *sws, and the
SwsInternal as *c.
In an upcoming commit, I will provide a backing definition for the public
SwsContext, and update `sws_internal()` to dereference the internal struct
instead of merely casting it.
Sponsored-by: Sovereign Tech Fund
Signed-off-by: Niklas Haas <git@haasn.dev>
Depending on the magnitude of the output values, the potential
errors can be larger.
This fixes errors in the lls tests on x86_32 for some seeds,
observed with GCC 11 (on Ubuntu 22.04, with the distro compiler,
with -m32).
Signed-off-by: Martin Storsjö <martin@martin.st>
The unaligned width test cases fail on i386; we have an assembly
function of rgb24toyv12 which is enabled only within
"#if ARCH_X86_32 && HAVE_7REGS", which seems to fail these new
test cases for unaligned widths.
As that assembly function has existed for a long time in that form,
the issue probably isn't very recent, thus skip testing these cases
for now.
Once the assembly function has been fixed, these test cases can
be readded.
Signed-off-by: Martin Storsjö <martin@martin.st>
Since c0666d8b, rgb24toyv12 is broken for width non-aligned to 16.
Add a simple wrapper to handle the non-aligned part.
Co-authored-by: johzzy <hellojinqiang@gmail.com>
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
When collecting performance information from checkasm it is common
to parse the output for use in graphs to compare vs different
architectures.
Signed-off-by: J. Dekker <jdek@itanimul.li>
F and D extensions are included in all RISC-V application profiles ever
made (so starting from RV64GC a.k.a. RVA20). Realistically they need to be
selected at compilation time.
Currently, there are no consumers for these two flags. If there is ever a
need to reintroduce F- or D-specific optimisations, we can always use
__riscv_f or __riscv_d compiler predefined macros respectively.
This preserves T1 whilst calling the instrumented function. In a Sci-Fi
setting where type-based Control Flow Integrity (CFI) is supported, the
calling code (i.e., the `checkasm` test case) will set T1 to the expected
value of the landing pad label (LPL) of the instrumented function.
The call wrapper will always use LPL zero which is a wild card. We should
preserve the value of T1 at least until the indirect call to the
instrumented function. Of course this is Sci-Fi, because:
1) there is no hardware (or even QEMU) support yet,
2) all our assembler functions currently use LPL zero anyway.
This uses T3 rather than T2 because indirect branches with T2 is reserved
for notionally direct calls made with an indirect call instruction (e.g.
due to GOT indirection), and are exempted from forward-edge CFI checks.
