Eliminated one frame memory copy and one clear frame operation in rendering chain, 20-70% higher framerate

Author: ProjectPhysX

README.md

@@ -137,6 +137,15 @@ The fastest and most memory efficient lattice Boltzmann CFD software, running on
   - fixed compiler warnings on Android
   - fixed `make.sh` failing on some systems due to nonstandard interpreter path
   - fixed that `make` would not compile with multiple cores on some systems
+- [v2.15](https://github.com/ProjectPhysX/FluidX3D/releases/tag/v2.15) (09.04.2024) [changes](https://github.com/ProjectPhysX/FluidX3D/compare/v2.14...v2.15) (framerate boost)
+  - eliminated one frame memory copy and one clear frame operation in rendering chain, for 20-70% higher framerate on both Windows and Linux
+  - enabled `g++` compiler optimizations for faster startup and higher rendering framerate
+  - fixed bug in multithreaded sanity checks
+  - fixed wrong unit conversion for thermal expansion coefficient
+  - fixed density to pressure conversion in LBM units
+  - fixed bug that raytracing kernel could lock up simulation
+  - fixed minor visual artifacts with raytracing
+  - fixed that console sometimes was not cleared before `INTERACTIVE_GRAPHICS_ASCII` rendering starts
 
 </details>
 

src/graphics.cpp

@@ -335,8 +335,8 @@ void draw_line_label(const int x0, const int y0, const int x1, const int y1, con
 		}
 	}
 }
-void draw_bitmap(const int* bitmap) {
-	std::copy(bitmap, bitmap+(int)camera.width*(int)camera.height, camera.bitmap);
+void draw_bitmap(int* bitmap) {
+	std::swap(camera.bitmap, bitmap); // swap pointers instead of memory copy
 }
 
 void draw_pixel(const float3& p, const int color) {
@@ -471,7 +471,7 @@ void update_frame(const double frametime) {
 	main_label(frametime);
 	SetBitmapBits(frameDC, 4*(int)camera.width*(int)camera.height, camera.bitmap);
 	BitBlt(displayDC, 0, 0, (int)camera.width, (int)camera.height, memDC, 0, 0, SRCCOPY); // copy back buffer to front buffer
-	camera.clear_frame(); // clear frame
+	//camera.clear_frame(); // clear frame
 }
 LRESULT CALLBACK WndProc(HWND window, UINT message, WPARAM wParam, LPARAM lParam) {
 	if(message==WM_DESTROY) {
@@ -618,10 +618,11 @@ int key_linux(const int keycode) {
 }
 void update_frame(const double frametime) {
 	main_label(frametime);
+	x11_image->data = (char*)camera.bitmap; // camera.bitmap pointer might have been changed, so update x11_image->data pointer here
 	updating_frame = true;
 	XPutImage(x11_display, x11_window, x11_gc, x11_image, 0, 0, 0, 0, camera.width, camera.height);
 	updating_frame = false;
-	camera.clear_frame(); // clear frame
+	//camera.clear_frame(); // clear frame
 	if(!camera.lockmouse) XWarpPointer(x11_display, x11_window, x11_window, 0, 0, camera.width, camera.height, camera.width/2, camera.height/2); // center cursor
 }
 void hide_cursor() {
@@ -766,7 +767,7 @@ void update_frame(const double frametime) {
 	print_video_dither(&image, textwidth, textheight);
 	print(alignr(textwidth, to_string(textwidth)+"x"+to_string(textheight)+" "+alignr(4, to_int(1.0/frametime))+"fps"));
 	show_console_cursor(true);
-	camera.clear_frame(); // clear frame
+	//camera.clear_frame(); // clear frame
 }
 void input_detection() {
 	while(running) {

src/graphics.hpp

@@ -157,10 +157,8 @@ class Camera {
 		if(!lockmouse) update_rotation(mouse_x, mouse_y);
 	}
 	void clear_frame() {
-		for(uint i=0u; i<width*height; i++) {
-			bitmap[i] = GRAPHICS_BACKGROUND_COLOR;
-			zbuffer[i] = min_int;
-		}
+		std::fill(bitmap, bitmap+width*height, GRAPHICS_BACKGROUND_COLOR); // faster than "for(uint i=0u; i<width*height; i++) bitmap[i] = GRAPHICS_BACKGROUND_COLOR;"
+		std::fill(zbuffer, zbuffer+width*height, min_int); // faster than "for(uint i=0u; i<width*height; i++) zbuffer[i] = min_int;"
 	}
 	float data(const uint i) const { // returns all camera data required for rendering
 		switch(i) {
@@ -332,7 +330,7 @@ extern bool key_1, key_2, key_3, key_4, key_5, key_6, key_7, key_8, key_9, key_0
 
 void set_light(const uint i, const float3& p);
 
-void draw_bitmap(const int* bitmap);
+void draw_bitmap(int* bitmap);
 void draw_label(const int x, const int y, const string& s, const int color);
 void draw_line_label(const int x0, const int y0, const int x1, const int y1, const int color);
 

src/info.cpp

@@ -55,7 +55,7 @@ void Info::print_logo() const {
 	print("|                                  ");                 print("\\  \\ /  /", c);                print("                                  |\n");
 	print("|                                   ");                 print("\\  '  /", c);                 print("                                   |\n");
 	print("|                                    ");                 print("\\   /", c);                 print("                                    |\n");
-	print("|                                     ");                 print("\\ /", c);                 print("               FluidX3D Version 2.14 |\n");
+	print("|                                     ");                 print("\\ /", c);                 print("               FluidX3D Version 2.15 |\n");
 	print("|                                      ");                 print( "'", c);                 print("     Copyright (c) Dr. Moritz Lehmann |\n");
 	print("|-----------------------------------------------------------------------------|\n");
 }
@@ -86,6 +86,7 @@ void Info::print_initialize() {
 	println("| MLUPs   | Bandwidth   | Steps/s   | Current Step      | "+string(steps==max_ulong?"Elapsed Time  ":"Time Remaining")+"      |");
 #else // INTERACTIVE_GRAPHICS_ASCII
 	println("'-----------------'-----------------------------------------------------------'");
+	clear_console();
 #endif // INTERACTIVE_GRAPHICS_ASCII
 	clock.start();
 	allow_rendering = true;

src/info.hpp

@@ -6,6 +6,7 @@ class LBM;
 struct Info { // contains redundant information for console printing
 	LBM* lbm = nullptr;
 	bool allow_rendering = false; // allows interactive redering if true
+	bool allow_labeling = true; // allows drawing label if true
 	double runtime_lbm=0.0, runtime_total=0.0f; // lbm (compute) and total (compute + rendering + data evaluation) runtime
 	double runtime_lbm_timestep_last=1.0, runtime_lbm_timestep_smooth=1.0, runtime_lbm_last=0.0; // for printing simulation info
 	Clock clock; // for measuring total runtime

src/lbm.cpp

@@ -1138,6 +1138,7 @@ int* LBM::Graphics::draw_frame() {
 #endif // GRAPHICS_TRANSPARENCY
 		}
 	}
+	info.allow_labeling = new_frame;
 	return bitmap;
 }
 

src/main.cpp

@@ -56,7 +56,7 @@ void draw_scale(const int field_mode, const int color) {
 	draw_label(ox+min(0, w+7+label_length_max*(FONT_WIDTH)-(int)length(title)*(FONT_WIDTH)), oy-(FONT_HEIGHT)*3/2-6, title, color); // colorbar title
 }
 void main_label(const double frametime) {
-	if(info.allow_rendering) {
+	if(info.allow_rendering&&info.allow_labeling) {
 		info.print_update();
 		const int c = invert(GRAPHICS_BACKGROUND_COLOR);
 		{
@@ -77,7 +77,7 @@ void main_label(const double frametime) {
 			draw_label(ox, oy+i, "Steps "          +alignr(31u, /************************************/ alignr(10u, info.lbm->get_t())+" ("+alignr(5, to_uint(1.0/info.runtime_lbm_timestep_smooth))+" Steps/s)"), c); i+=FONT_HEIGHT;
 			draw_label(ox, oy+i, "FPS "            +alignr(33u, /************************************************************/ alignr(4u, to_uint(1.0/frametime))+" ("+alignr(5u, camera.fps_limit)+" fps max)"), c);
 		}
-		draw_label(2, camera.height-1*(FONT_HEIGHT)-1, "FluidX3D v2.14 Copyright (c) Dr. Moritz Lehmann", c);
+		draw_label(2, camera.height-1*(FONT_HEIGHT)-1, "FluidX3D v2.15 Copyright (c) Dr. Moritz Lehmann", c);
 		if(!key_H) {
 			draw_label(camera.width-16*(FONT_WIDTH)-1, 2, "Press H for Help", c);
 		} else {

src/resource.rc

@@ -2,11 +2,11 @@
 #include "winres.h"
 #undef APSTUDIO_READONLY_SYMBOLS
 
-MAINICON ICON "src/icon.ico"
+MAINICON ICON "icon.ico"
 
 VS_VERSION_INFO VERSIONINFO
-	FILEVERSION 1,0,0,0
-	PRODUCTVERSION 1,0,0,0
+	FILEVERSION 0,0,0,0
+	PRODUCTVERSION 0,0,0,0
 	FILEFLAGSMASK 0x3FL
 	FILEFLAGS 0x0L
 	FILEOS 0x4L
@@ -19,12 +19,12 @@ BEGIN
 		BEGIN
 			VALUE "CompanyName", "ProjectPhysX"
 			VALUE "FileDescription", "FluidX3D"
-			VALUE "FileVersion", "1.0.0.0"
+			VALUE "FileVersion", "0.0.0.0"
 			VALUE "InternalName", "FluidX3D"
-			VALUE "LegalCopyright", "(c) Moritz Lehmann"
+			VALUE "LegalCopyright", "(c) Dr. Moritz Lehmann"
 			VALUE "OriginalFilename", "FluidX3D.exe"
 			VALUE "ProductName", "FluidX3D"
-			VALUE "ProductVersion", "1.0"
+			VALUE "ProductVersion", "v2.15"
 		END
 	END
 	BLOCK "VarFileInfo"