fix(foundation): use HW_PHYSMEM for physical memory detection on FreeBSD

[PR9000]

What does this PR do?

This PR fixes physical memory detection on FreeBSD by using the standard HW_PHYSMEM sysctl in detect_system_bsd().

On FreeBSD 15.1-STABLE, the current implementation using HW_PHYSMEM64 does not return the expected physical memory size during system initialization. This change routes FreeBSD through the existing HW_PHYSMEM code path already used by OpenBSD.

Changes

• Updated src/foundation/system_info.c.
• Extended the BSD-specific condition:

#if defined(__OpenBSD__)

to:

#if defined(__OpenBSD__) || defined(__FreeBSD__)

so FreeBSD uses HW_PHYSMEM for physical memory detection.

Testing

Tested on FreeBSD 15.1-STABLE (amd64).

Validation performed:

• Successfully built codebase-memory-mcp using:

gmake -f Makefile.cbm cbm

• Successfully ran the CI linters:

gmake -f Makefile.cbm lint-ci CLANG_FORMAT=clang-format21

Result:

=== cppcheck ===
=== clang-format ===
=== NOLINT check ===
=== CI linters passed ===

• Verified that the detected physical memory size is reported correctly after this change.

The test suite was also executed:

gmake -f Makefile.cbm test

However, the test build currently fails during linking with an unrelated AddressSanitizer duplicate symbol error:

duplicate symbol: bsearch
>>> defined at asan_interceptors.cpp
>>> defined at internal/cbm/vendored/grammars/perl/bsearch.h

This failure is independent from this change and occurs in the test infrastructure/linking stage before the tests are executed.

Why this is safe

This is a small, localized, platform-specific change affecting only the FreeBSD BSD implementation path.

Linux, macOS, Windows, and other supported platforms are unaffected. The change reuses the existing HW_PHYSMEM implementation already used for OpenBSD and does not modify the behavior of other operating systems.

Checklist

• Every commit is signed off (git commit -s) — required, CI rejects unsigned commits ([DCO](../DCO), see [CONTRIBUTING.md](../CONTRIBUTING.md))
• Tests pass locally (make -f Makefile.cbm test)
• Lint passes (make -f Makefile.cbm lint-ci)
• New behavior is covered by a test (reproduce-first for bug fixes)

[DeusData]

Thanks for the focused FreeBSD fix and the on-platform validation. FreeBSD's current sys/sysctl.h defines HW_PHYSMEM for this MIB and not HW_PHYSMEM64, so the direction is correct. I have triaged this as a high-priority 0.9.1-rc stability bug. The one-line diff adds no dependency, network, secret, or broader runtime surface.

The remaining blocker is the repository's reproduce-first requirement: the PR currently has no automated guard that fails on main. Because the normal matrix has no FreeBSD runner, please add the smallest testable regression or propose a narrow test seam for the MIB-selection contract before changing CI. Do not add a new workflow or dependency without design approval. The unrelated Perl/ASan duplicate-symbol problem should stay outside this PR.