tests: Use malloc for OSX/Linux page heap

Switch to malloc and manually aligning the pointer to the page boundary.

mmap is much slower than malloc; this change makes tests ~4x faster.

tests/allocator.cpp

@@ -80,7 +80,9 @@ namespace
 
 	void* allocate_page_aligned(size_t size)
 	{
-		return mmap(0, size + page_size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0);
+		void* result = malloc(size + page_size);
+
+		return reinterpret_cast<void*>(align_to_page(reinterpret_cast<size_t>(result)));
 	}
 
 	void* allocate(size_t size)