From bf55715255201002122b15ae2e379fb2c9b23aa1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Tom=C3=A1=C5=A1=20Oberhuber?= <oberhuber.tomas@gmail.com> Date: Fri, 6 Mar 2020 18:02:47 +0100 Subject: [PATCH] Emphasizing the best performance reccomendation. --- .../BuildWithTNL/tutorial_building_applications_with_tnl.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/Tutorials/BuildWithTNL/tutorial_building_applications_with_tnl.md b/Documentation/Tutorials/BuildWithTNL/tutorial_building_applications_with_tnl.md index 87f724bd29..4aad7d741c 100644 --- a/Documentation/Tutorials/BuildWithTNL/tutorial_building_applications_with_tnl.md +++ b/Documentation/Tutorials/BuildWithTNL/tutorial_building_applications_with_tnl.md @@ -29,7 +29,7 @@ g++ -std=c++14 -I${HOME}/.local/include/tnl example-host.cpp -o example-host TNL requires standard C++14 which we enforce with the first parameter `-std=c++14`. Next, we need to tell the compiler the folder with TNL headers. This is done with the flag `-I`. By default, TNL installs into `${HOME}/.local/include/tnl`. You may also replace it just with the path where you have downloaded TNL. TNL is header only library and so it does not require any instalation. Finaly, we just past the source code file `example-host.cpp` using the command-line parameter `-c`. -For the best performance we suggest to add parameters `-DNDEBUG -O3 -funroll-loops`. The first one deactivates assertions in TNL which can significantly slow down your program. +** For the best performance we recommend to add parameters `-DNDEBUG -O3 -funroll-loops`. The source code of TNL contains a lot of assertions which significantly decrease the performance. The parameter `-DNDEBUG` deactivates them. ** ### Compilation with `nvcc` for CUDA <a name="command_line_nvcc"></a> -- GitLab