[siracozmen01]

Problem

v0.0.53's check_nodiscard.py reports two unmarked bool functions in the shared maul/renamed/src/cpu.h: @P@IsCpuBackendSupported and @P@CheckCpu. #124 marked the other renamed files but missed these, because the @P@ placeholder hid their names from its scan. maul2d, maul3d and maul-anim carry cpu.h and cannot pass the check without editing a family file.

Change

cpu.h includes @LIB@/base.h and marks both declarations @PP@_NODISCARD. A public base.h that also declares them (all three libraries do, unmarked) keeps its own declaration; the mark on this later one makes a dropped result warn inside the library.

Verification

Synced from this branch into scratch trees of maul2d, maul3d and maul-anim:

• src/cpu.h passes clang-format --dry-run -Werror, and src/cpu.c builds with gcc-14 and clang-20 at -std=c23 -Wall -Wextra -Wpedantic -Werror;
• tools/check_nodiscard.py no longer reports cpu.h;
• a call that drops maniCheckCpu's result warns under gcc-14;
• every call of the two functions in the three libraries (src, tests, bench) uses the result.

Closes #125

[siracozmen01]

Review at ada6ddc: ready. Read, not run.

No cycle. cpu.h now includes @LIB@/base.h. base.h includes only <stdbool.h>, <stddef.h> and <stdint.h> (maul2d's at its sync head, the same shape in the family), never a src/ header, and both have include guards, so the include cannot loop.

No ordering problem.

• cpu.c includes cpu.h first, as before.
• For maul2d, maul3d and maul-anim, base.h's unmarked @P@_API bool @P@IsCpuBackendSupported(void); now comes before cpu.h's marked redeclaration, and before the definition in cpu.c. A later declaration may add [[nodiscard]], as the body's gcc-14 and clang-20 runs show.
• On Windows, a redeclaration without __declspec(dllexport) after one with it keeps the export.

No module rule reached. check_modules.py reads only includes without a slash ("([^"/]+)\.h"), so "@LIB@/base.h" is a public header there, as its comment allows. maul2d's cpu: line needs no new edge.

One effect to know about. cpu.c's definitions now see base.h's @P@_API declaration in the three libraries where it exists. In a shared build that gives @P@IsCpuBackendSupported default visibility (or dllexport) through its definition's own translation unit. That is what @P@_API asks for, but it may change the shared library's export list where the definition was compiled without seeing it. Each library's check_exports test in its sync PR will show it. I did not build a shared tree.

The marks are right: every call in the three libraries uses the result, per the body, and a dropped one now fails at -Werror.