bgfx allocator using allocation with alignment;

some more parts of code profiled

src/engine/core/FS/tcp_file_server.cpp

@@ -2,13 +2,14 @@
 
 #include "core/array.h"
 #include "core/free_list.h"
-#include "core/path.h"
-#include "core/stack_allocator.h"
-#include "core/static_array.h"
-#include "core/string.h"
 #include "core/fs/os_file.h"
 #include "core/fs/tcp_file_device.h"
 #include "core/mt/task.h"
+#include "core/path.h"
+#include "core/profiler.h"
+#include "core/stack_allocator.h"
+#include "core/static_array.h"
+#include "core/string.h"
 #include "core/network.h"
 
 
@@ -175,6 +176,7 @@ public:
 
 		while (!quit)
 		{
+			PROFILE_BLOCK("File server operation")
 			int32 op = 0;
 			stream->read(op);
 			switch (op)

src/engine/core/MTJD/scheduler.cpp

@@ -1,7 +1,8 @@
 #include "lumix.h"
 #include "core/MTJD/scheduler.h"
-
 #include "core/MTJD/manager.h"
+#include "core/profiler.h"
+
 
 namespace Lumix
 {
@@ -25,6 +26,7 @@ namespace Lumix
 			{
 				m_data_event.wait();
 
+				PROFILE_BLOCK("Schedule")
 				m_manager.doScheduling();
 			}
 

src/engine/debug/pc/debug.cpp

@@ -381,21 +381,19 @@ void* Allocator::reallocate(void* user_ptr, size_t size)
 
 void* Allocator::allocate_aligned(size_t size, size_t align)
 {
-	ASSERT(false);
-	return nullptr;
+	return m_source.allocate_aligned(size, align);
 }
 
 
 void Allocator::deallocate_aligned(void* ptr)
 {
-	ASSERT(false);
+	m_source.deallocate_aligned(ptr);
 }
 
 
 void* Allocator::reallocate_aligned(void* ptr, size_t size, size_t align)
 {
-	ASSERT(false);
-	return nullptr;
+	return m_source.reallocate_aligned(ptr, size, align);
 }
 
 

src/renderer/culling_system.cpp

@@ -5,6 +5,7 @@
 #include "core/binary_array.h"
 #include "core/free_list.h"
 #include "core/frustum.h"
+#include "core/profiler.h"
 #include "core/sphere.h"
 
 #include "core/mtjd/group.h"
@@ -27,6 +28,7 @@ static void doCulling(int start_index,
 	int64 layer_mask,
 	CullingSystem::Subresults& results)
 {
+	PROFILE_FUNCTION();
 	int i = start_index;
 	ASSERT(results.empty());
 	for (const Sphere *sphere = start; sphere <= end; sphere++, ++i)

src/renderer/render_scene.cpp

@@ -1661,6 +1661,7 @@ public:
 			MTJD::Job* job = MTJD::makeJob(m_engine.getMTJDManager(),
 				[&subinfos, layer_mask, this, &results, subresult_index, &frustum]()
 				{
+					PROFILE_BLOCK("Temporary Info Job");
 					Vec3 frustum_position = frustum.getPosition();
 					const CullingSystem::Subresults& subresults = results[subresult_index];
 					for (int i = 0, c = subresults.size(); i < c; ++i)

src/renderer/renderer.cpp

@@ -89,17 +89,17 @@ struct BGFXAllocator : public bx::AllocatorI
 	}
 
 
-	void* realloc(void* _ptr, size_t _size, size_t, const char*, uint32) override
+	void* realloc(void* _ptr, size_t _size, size_t _alignment, const char*, uint32) override
 	{
-		if (!_ptr) return m_source.allocate(_size);
+		if (!_ptr) return m_source.allocate_aligned(_size, _alignment);
 
 		if (_ptr && _size == 0)
 		{
-			m_source.deallocate(_ptr);
+			m_source.deallocate_aligned(_ptr);
 			return nullptr;
 		}
 
-		return m_source.reallocate(_ptr, _size);
+		return m_source.reallocate_aligned(_ptr, _size, _alignment);
 	}
 
 

src/studio/pc/file_system_watcher.cpp

@@ -1,5 +1,6 @@
 #include "file_system_watcher.h"
 #include "core/MT/task.h"
+#include "core/profiler.h"
 #include "core/string.h"
 
 #include "core/pc/simple_win.h"
@@ -138,6 +139,7 @@ int FileSystemWatcherTask::task()
 	bool finished = false;
 	while (!finished)
 	{
+		PROFILE_BLOCK("Change handling");
 		finished = ReadDirectoryChangesW(m_handle,
 										 m_info,
 										 sizeof(m_info),