MOD-14130 Fix build failures on focal cross-compilation and x86 with old binutils

[ofiryanai]

Summary

This PR fixes three separate build issues affecting VectorSimilarity on different platforms:

1. SVE benchmark compilation bug (aarch64)

Files: tests/benchmark/spaces_benchmarks/bm_spaces_int8.cpp, bm_spaces_uint8.cpp

Copy-paste bug: #ifdef OPT_SVE blocks incorrectly referenced sve2_supported (only defined in OPT_SVE2 block) instead of sve_supported.

2. Cross-compilation architecture detection

File: src/VecSim/spaces/CMakeLists.txt

Used CMAKE_HOST_SYSTEM_PROCESSOR (the machine running CMake) instead of CMAKE_SYSTEM_PROCESSOR (the target architecture). This caused x86 AVX512 code to be compiled when cross-compiling for ARM.

3. AVX512-FP16/BF16 build failure with old binutils

File: cmake/x86_64InstructionFlags.cmake

On systems with modern GCC (13+) but old binutils (e.g., Ubuntu 20.04 focal with binutils 2.34):

• CMake's CHECK_CXX_COMPILER_FLAG(-mavx512fp16 ...) passes (GCC accepts the flag)
• But the assembler fails with Error: no such instruction: 'vfmadd132ph ...'

Fix: Detect binutils version at configure time and disable:

• CXX_AVX512BF16 if binutils < 2.34
• CXX_AVX512FP16 if binutils < 2.38

Jira

https://redislabs.atlassian.net/browse/MOD-14130

---

Pull Request opened by Augment Code with guidance from the PR author

---

Pull Request opened by Augment Code with guidance from the PR author

---

Note

Low Risk
Build/benchmark-only changes with straightforward conditional logic; main risk is mis-detecting binutils/CPU and unintentionally enabling/disabling optimized code paths.

Overview
Fixes cross-compilation build breakages by switching VecSim spaces’ architecture gating from CMAKE_HOST_SYSTEM_PROCESSOR to CMAKE_SYSTEM_PROCESSOR, preventing x86-only AVX sources from being compiled when targeting ARM.

Improves x86 AVX512 feature detection on Linux by checking the installed binutils/assembler version and disabling CXX_AVX512BF16 (<2.34) and CXX_AVX512FP16 (<2.38) when assembly support is too old.

Fixes AArch64 benchmark compilation by using sve_supported (not sve2_supported) inside #ifdef OPT_SVE blocks in the INT8/UINT8 space benchmarks.

^{Written by Cursor Bugbot for commit dcde91f. This will update automatically on new commits. Configure here.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.09%. Comparing base (a68ba7d) to head (dcde91f).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #912   +/-   ##
=======================================
  Coverage   97.09%   97.09%           
=======================================
  Files         129      129           
  Lines        7500     7500           
=======================================
  Hits         7282     7282           
  Misses        218      218           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.