75da766724fcd37b7c8859e42513a454453468c2
2 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
46232a5ad1
|
Give libccd a static arena so libakgl still allocates nothing
libakgl does not call malloc at runtime. That is stated in seven places in the manual and is AGENTS.md's second standing rule, and every object comes from a fixed array in akgl/heap.h. libccd's EPA path does not know that: it builds its expanding polytope out of realloc and free, and ccdGJKPenetration documents a -2 return for when that fails. The alternative was to compile only the three files MPR needs and leave EPA out of the build. That works and it puts a copy of somebody else's source list in this repository, where it rots silently on the next submodule bump. This instead compiles all of libccd and points its allocator at a bump allocator over static BSS, which keeps the promise literally true -- and keeps EPA available rather than amputated, for the day a precise contact manifold is worth having. The arena is reset at the top of each query rather than freed block by block, so the lifetime is one narrowphase call, `free` is a no-op, and allocation is a pointer bump. Exhaustion returns NULL, which libccd already handles by unwinding to -2, which becomes AKGL_ERR_COLLISION naming the high-water mark -- a loud failure with a number in it rather than a silently missed collision, which would be a floor an actor falls through reported as success. One GJK/EPA box pair costs 7,264 bytes, measured and printed by the suite. The 64 KB ceiling is that with about nine times headroom, and the high-water mark is reported so the next reader can re-derive it rather than trust this line. The redirect lives in src/ccd_arena_shim.h, injected with -include, and not in CMake's COMPILE_DEFINITIONS. It was in COMPILE_DEFINITIONS first, and that is a mistake worth recording: CMake cannot carry a function-like macro through a -D, so it dropped __CCD_ALLOC_MEMORY without a diagnostic. libccd went on calling the C library's realloc while the shim's `free` quietly discarded the results -- strictly worse than doing nothing, and invisible, because it leaks rather than crashes. What caught it was insisting the test prove the wiring rather than the outcome. The first version asserted the arena balanced back to zero after a query, which turned out to be the wrong assertion for a different reason -- free is a no-op by design, so it cannot balance -- but a test that had merely checked "two boxes collide" would have passed throughout, against an allocator nothing was using. The suite now asserts what is actually provable: that a query allocates from the arena at all, and that the process survives, since glibc aborts when the real free(3) is handed a pointer it never issued. Also here: AKGL_ERR_COLLISION, with the name registered -- tests/error.c asserts the band and the names agree, and it caught the missing one immediately. -fvisibility=hidden and CCD_STATIC_DEFINE keep every ccd* symbol out of libakgl.so's dynamic table, which `nm -D` confirms is empty; AGENTS.md records a shipped defect where an exported `renderer` was preempted by a same-named symbol elsewhere, and a game linking a system libccd would hit exactly that. Co-Authored-By: Claude Code <noreply@anthropic.com> Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
|||
|
9b124f2e27
|
Migrate to the libakerror 1.0.0 status registry
libakerror 1.0.0 replaced the consumer-sized status-name array with a private registry and made status-code ownership explicit and enforced. AKERR_MAX_ERR_VALUE and __AKERR_ERROR_NAMES are gone, and the registry entry points raise akerr_ErrorContext * instead of returning int. See deps/libakerror/UPGRADING.md. The break was not only source-level. libakgl's codes sat at AKERR_LAST_ERRNO_VALUE + 18 through + 22, and 1.0.0 claimed exactly those five offsets for its own AKERR_STATUS_* registry codes, so every AKGL_ERR_* was aliasing a libakerror status. HANDLE(e, AKGL_ERR_LOGICINTERRUPT) at physics.c:222 would have swallowed a foreign-name refusal. - Move the band to AKERR_FIRST_CONSUMER_STATUS (256) as fixed offsets, so a libc that grows an errno cannot move the codes, and add AKGL_ERR_OWNER, AKGL_ERR_LIMIT and AKGL_ERR_COUNT to describe it. - Reserve the range and register the names through the owned entry points, PASS-ing each: these are AKERR_NOIGNORE, and the old akerr_name_for_status calls discarded failure silently. - Drop the AKERR_MAX_ERR_VALUE=256 compile definition. - Guard on AKERR_FIRST_CONSUMER_STATUS in include/akgl/error.h, which now includes <akerror.h> so the guard is reliable. The embedded build is fine, but the find_package path can pick up a stale installed header, and 1.0.0 has an soname, so that pairing is an ABI mismatch rather than a compile problem. Same guard libakstdlib already carries. Registration also moves out of akgl_heap_init into a new akgl_error_init in src/error.c. It was in the heap pool's initializer only because that was the first thing akgl_game_init called, and the upgrade turned five fire-and-forget name calls into a library-wide ownership claim that can fail. That placement was hiding a defect: game.c raises AKGL_ERR_SDL when SDL_CreateMutex fails, five lines before akgl_heap_init ran, so the earliest error path in the library was guaranteed to print "Unknown Error". akgl_error_init is now the first statement in akgl_game_init. Callers that drive subsystems directly must call akgl_error_init first; it is idempotent, so ordering it precisely is not required. The eleven test suites that relied on akgl_heap_init to name their statuses now call it explicitly, or their failure messages would have degraded to "Unknown Error". Add tests/error.c: assert every code reads back its registered name, that the name table and AKGL_ERR_COUNT agree, that a foreign owner is refused with AKERR_STATUS_NAME_FOREIGN and AKERR_STATUS_RANGE_OVERLAP, and that repeating the init is a no-op. That last one is a live constraint, not a triviality -- libakerror treats only an identical reservation as a repeat, so a subset or superset raises. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
