fix: data race on collision query profiling counters

queryTimeMs and queryCallCount on WMORenderer and M2Renderer were plain
mutable doubles/uint32s written by getFloorHeight (dispatched on async
threads from CameraController) and read by the main thread. This is
undefined behavior per C++ — thread sanitizer would flag it. Changed to
std::atomic with relaxed ordering (adequate for diagnostics) and updated
QueryTimer to use atomic fetch_add/compare_exchange.
This commit is contained in:
Kelsi 2026-03-29 21:26:11 -07:00
parent 3dd1128ecf
commit bbb560f93c
3 changed files with 23 additions and 12 deletions

View file

@ -5,6 +5,7 @@
#include <vulkan/vulkan.h>
#include <vk_mem_alloc.h>
#include <glm/glm.hpp>
#include <atomic>
#include <memory>
#include <unordered_map>
#include <unordered_set>
@ -531,9 +532,10 @@ private:
std::unordered_map<uint32_t, size_t> instanceIndexById;
// Collision scratch buffers are thread_local (see m2_renderer.cpp) for thread-safety.
// Collision query profiling (per frame).
mutable double queryTimeMs = 0.0;
mutable uint32_t queryCallCount = 0;
// Collision query profiling — atomic because getFloorHeight is dispatched
// on async threads from camera_controller while the main thread reads these.
mutable std::atomic<double> queryTimeMs{0.0};
mutable std::atomic<uint32_t> queryCallCount{0};
// Persistent render buffers (avoid per-frame allocation/deallocation)
struct VisibleEntry {