CEngine::Frame() tends to return out early if we are ahead of frame time, and therefore need to sleep. In this particular engine, CEngine::Frame() returns true if it had actually executed an engine frame, else it returns false.
Therefore, we end up calling 'NvAPI_D3D_Sleep()' multiple times a frame, and thus causing a major loss in performance. Now we check if we have actually ran a frame (incrementing reflex frame if so). If the frame was not incremented, but 'GFX_RunLowLatencyFrame()' was called, it will not call 'NvAPI_D3D_Sleep()'.
Code has been tested and appears to work on my system (never called twice or more per frame). Needs more testing on other systems to make sure its good.
Use CUtlVector, and remove every copy caused by passing vectors by value. CUtlVector does not support copying. Also removed all extraneous std::string copies caused by calling itoa instead of std::to_string, or std::stoll, etc. All features have been tested and work as designed.
Removed all extraneous copies by adding the class 'InterfaceReg' which will construct a new interface, and link it to the engine's static register. The Source Engine macro 'EXPOSE_INTERFACE_FN' will help utilizing this. The game module from the plugin is not obtained through the process environment block, so the executable is no longer sensitive to names.
The implementation in the engine always took/returned signed 64bit size types, but I made a mistake when reversing the virtual function table. All types have been changed to what they should be (mostly signed 64bit, very few unsigned), and usage in-SDK has been adjusted accordingly.
Move all classes deriving from ConCommandBase to a single file, and split out CCommand, CCvar, CCvarUtilities etc to their own files. This makes it possible to use CCommand and stuff in external tools without linker errors/warnings.
CConnectedNetConsoleData was allocated and deallocated manually, but std::vector is able to do it on its own. The vector type has been swapped out with CUtlVector as this removed the need of having to cast size types to a 32bit int to make it more compatible with the other sdk/engine types without throwing compiler warnings.
Added public interface to ICVarIteratorInternal, this class also deals with the deletion of the iterator. This fixes the problem of cvars not showing up that have been registered after 'CModAppSystemGroup::Create()' has been called. Currently, no cvar is registered after it. This will/could change when the plugin system is utilized.
Additional changes:
- Cvar 'con_suggestion_flags_realtime' has been removed, as we no longer keep an early copy anymore; real time flags are always shows from now on.
- Fixed a bug in 'CConsole::FindFromPartial()' where we would return in the loop when 'm_vSuggest.size() >= con_suggestion_limit->GetInt()', instead of breaking out, which skips the element sorting process.
The 3 versions are now:
- cfg/server/startup.bin
- cfg/client/startup.bin
- cfg/startup.bin
The last one is used by the GameSDK. Bumped header version to 6.
Added:
- FCVAR_ACCESSIBLE_FROM_THREADS [1<<25]
- FCVAR_STUDIO_SYSTEM [1<<26]
- FCVAR_SERVER_FRAME_THREAD [1<<27]
- FCVAR_PLATFORM_SYSTEM [1<<31]
This commit also removes the 'FCVAR_MATERIAL_THREAD_MASK' from the ConVarFlags class (which is a helper struct for displaying flags on the developer console), as this is a mask and would therefore be marked as such in case a single bit in the mask matches, and all these bits are already added to the list.
CClient array is part of the CPlayer class. The array has been added, filling up most of the 'unknown' gap. Vtable override functions have been added as the array is static, and therefore requires an implementation of the vtable, these are just stubs however, since the vtable pointer points to the implementation in the engine.
