> <@trungnt2910:matrix.org> ``` > /Data> strace -c /Data/dotnet/dotnet --version > 8.0.100-preview.6.23303.2 > > Time % Usecs Calls Usecs/call Syscall > ------ ---------- ------- ---------- -------------------- > 51.59 1374087 7096 193 _kern_get_clock > 26.63 709385 4263 166 _kern_set_memory_protection > 5.37 142989 82 1743 _kern_map_file > 4.15 110445 30 3681 _kern_debug_output > 3.95 105204 652 161 _kern_mutex_unlock > 2.49 66350 493 134 _kern_read_stat > 1.57 41705 509 81 _kern_get_next_image_info > 0.54 14421 144 100 _kern_read > 0.38 10101 92 109 _kern_open > 0.37 9966 63 158 _kern_normalize_path > 0.37 9884 28 353 _kern_mutex_switch_lock > 0.37 9874 61 161 _kern_resize_area > 0.31 8299 75 110 _kern_fcntl > 0.18 4848 51 95 _kern_close > 0.18 4748 20 237 _kern_spawn_thread > 0.12 3104 20 155 _kern_resume_thread > [some entries omitted] > ``` > > Haiku's so slow... So slow that if I optimize `hyclone_server`'s servercall handling with a thread pool it might beat Haiku performance...