Fix v_round and enable unit tests for scalable universal intrinsic 64F type.

[hanliutong]

This may be a legacy issue from the previous PR #24325. I don't quite remember why the float 64 part of the unit test was not enabled at that time.

Whatever, this patch enables the unit tests for scalable 64F type , and makes the necessary modifications to the RVV backend to make the tests pass.

This patch is compiled by GCC 14 and LLVM 17 &18, and tested on QEMU and k230.

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

• I agree to contribute to the project under Apache 2 License.
• To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
• The PR is proposed to the proper branch
• There is a reference to the original bug report and related work
• There is accuracy test, performance test and test data in opencv_extra repository, if applicable
  Patch to opencv_extra has the same branch name.
• The feature is well documented and sample code can be built with the project CMake

[hanliutong]

About #24746:

• The root cause of the problem was an incorrect implementation of v_round in RVV backend, which has now been fixed.

-- return vfncvt_x(vset(vlmul_ext_f64m2(vfadd(a, 1e-6, VTraits<v_float64>::vlanes())), 1, b), VTraits<v_float32>::vlanes());
++ return vfncvt_x(vset(vlmul_ext_f64m2(a), 1, b), VTraits<v_float32>::vlanes());

• The indirect cause is that our unit test for Universal Intrinsic was

1. not enabled for scalable backends on 64F type previously, fixed by:

-- #if CV_SIMD_64F
++ #if (CV_SIMD_64F || CV_SIMD_SCALABLE_64F)

2. not covered testing of v_round(v_f64, v_f64), fixed by

++ TheTest & test_round_pair_f64()
++ {
++ ...
++ }

[mshabunin]

It seems LLVM 17 doesn't support vcreate intrinsic, I've got an error:

/opencv/modules/core/include/opencv2/core/hal/intrin_rvv_scalable.hpp:1540:25: error: use of undeclared identifier '__riscv_vcreate_v_f64m1_f64m2'
 1540 |     vfloat64m2_t temp = __riscv_vcreate_v_f64m1_f64m2(a0, a1);
      |                         ^

Toolchain: LLVM 17.0.6
CMake options (excerpt):

    -DRISCV_CLANG_BUILD_ROOT=${llvm_root} \
    -DRISCV_GCC_INSTALL_ROOT=${gcc_path} \
    -DCMAKE_TOOLCHAIN_FILE=/opencv/platforms/linux/riscv64-clang.toolchain.cmake \
    -DBUILD_SHARED_LIBS=OFF \
    -DWITH_OPENCL=OFF \
    -DRISCV_RVV_SCALABLE=ON \
    -DCPU_BASELINE=RVV \
    -DCPU_BASELINE_REQUIRE=RVV \
    -DCMAKE_BUILD_TYPE=Release

My attemt to recreate: https://godbolt.org/z/f9oh77dqM (clang 18 works). Perhaps intrinsics version check should be >0.12? - it doesn't seem to help.

UPD: yes, it has been added in clang 18 - llvm/llvm-project#70355. Since intrinsics version has not been bumped we need to check clang version if we want to use this intrinsic.

[hanliutong]

Bug reproduce. Seem like the "if code" and the "then code" generate the same code, so maybe we can just not use vctreat and leave it there (if 0) until the intrinsic frozen and compiler fully supports the intrinsic spec.

[mshabunin]

Thank you!