Setup benchmarh branch to have max performancebranch/simd-math

3 files changed, 14 insertions, 15 deletions

diff --git a/build.sh b/build.sh
index f7cc7fa..a4f0fda 100755
--- a/build.sh
+++ b/build.sh
@@ -14,7 +14,7 @@ build_opts="$build_dir/main"
 lib_path="libs/SDL2 libs/freetype/libfreetype.so"
 link_opts="-L $lib_path -lSDL2"
 
-build_command="clang++ -std=c++11 -g -O0 $include_opts $files $link_opts -o $build_opts" 
+build_command="clang++ -std=c++11 -g -O2 $include_opts $files $link_opts -o $build_opts" 
 
 printf "Building Project...\n"
 printf "$build_command\n\n"
diff --git a/source/main.cpp b/source/main.cpp
index a4d7b2c..8a88231 100755
--- a/source/main.cpp
+++ b/source/main.cpp
@@ -690,8 +690,8 @@ Vec3 get_screen_position_from_percent(GameState state, Vec3 v) {
 
 int main(int argc, char* argv[])
 {
-  u32 scr_width = 1024;
-  u32 scr_height = 768;
+  u32 scr_width = 1920;
+  u32 scr_height = 1080;
 
   if (SDL_Init(SDL_INIT_VIDEO) != 0)
   {
@@ -896,7 +896,7 @@ int main(int argc, char* argv[])
   {
     update_frame_timer(&timer);
     //enforce_frame_rate(&timer, 60);
-    hp_ticks = SDL_GetPerformanceCounter();
+    //hp_ticks = SDL_GetPerformanceCounter();
     cum_math_ticks = 0;
 
     controller.jump = 0;
@@ -1277,8 +1277,7 @@ int main(int argc, char* argv[])
     u32 max_col_ele = max_cq_count/max_row_ele;
     Vec2 screen_render_size = (state.render_scale * state.screen_size)/2.0f;
     Vec2 based_size = Vec2{
-      screen_render_size.x / max_row_ele,
-      screen_render_size.y / max_col_ele
+      16.0f, 16.0f
     };
 #if 1
     for (int i=0;i<max_cq_count;i++) {
@@ -1315,12 +1314,12 @@ int main(int argc, char* argv[])
     renderer->cq_batch_count = 0;
     // render ui text
     
-    u64 new_ticks = SDL_GetPerformanceCounter();
-    r64 hp_time = (r64)(new_ticks - hp_ticks) / (r64) hp_tick_freq;
-    SDL_Log("ticks time: %fs", hp_time);
+    //u64 new_ticks = SDL_GetPerformanceCounter();
+    //r64 hp_time = (r64)(new_ticks - hp_ticks) / (r64) hp_tick_freq;
+    //SDL_Log("ticks time: %fs", hp_time);
 
-    cum_math_time = (r64)cum_math_ticks/(r64)tick_freq;
-    SDL_Log("multiply4mv time: %fs", cum_math_time);
+    //cum_math_time = (r64)cum_math_ticks/(r64)tick_freq;
+    //SDL_Log("multiply4mv time: %fs", cum_math_time);
 
     if (is_collide_x || is_collide_y)
     {
diff --git a/source/math.h b/source/math.h
index 1ac3605..1d7924c 100755
--- a/source/math.h
+++ b/source/math.h
@@ -10,7 +10,7 @@
 
 // @todo:
 // - make everything simd
-#define USE_SSE 1
+#define USE_SSE 0
 
 r32 clampf(r32 x, r32 bottom, r32 top)
 {
@@ -393,7 +393,7 @@ Mat4 subtract4m(Mat4 a, Mat4 b)
 
 Vec4 multiply4mv(Mat4 m, Vec4 v)
 {
-    r64 prev_tick = SDL_GetPerformanceCounter();
+    //r64 prev_tick = SDL_GetPerformanceCounter();
 
     Vec4 res = vec4(0);
 #if USE_SSE
@@ -434,8 +434,8 @@ Vec4 multiply4mv(Mat4 m, Vec4 v)
   res.w += v.w*m.data[3][3];
 #endif
 
-    r64 curr_tick = SDL_GetPerformanceCounter();
-    cum_math_ticks += curr_tick - prev_tick;
+    //r64 curr_tick = SDL_GetPerformanceCounter();
+    //cum_math_ticks += curr_tick - prev_tick;
 
   return res;
 }