Skip to content

Latest commit

 

History

History
266 lines (191 loc) · 9.41 KB

README.md

File metadata and controls

266 lines (191 loc) · 9.41 KB
Raw

Exercises: HowTo

  • Each exercise is in its own directory and referred to in the following by the name of the directory.
  • Almost each exercise contains a solution.
  • We tried to provide more material than many participants can finish during the sessions. Participants are welcome to finish the more complicated ones in self study, but questions regarding the solutions are welcome in the lecture and especially during the following hands-on sessions.
  • For each day, the exercises are given in order in which they should be done in

Getting the code

git clone https://github.com/hsf-training/cpluspluscourse.git
cd cpluspluscourse/exercises

Recommended setup

Required

  • decent C++ editor
  • any C++ compiler supporting C++20
  • C++17 compiler will do for most of the exercises though
  • git for getting/managing your code

Good to have

  • gdb to debug your problems
  • valgrind, kcachegrind, cppcheck, clang-format and clang-tidy for corresponding exercises
  • g++ or clang++ as C++ compiler for sanitizer exercises

More setup for the advanced course

C++ and python specific needs

  • python3, libpython3-dev
  • ctypes, matplotlib, numpy python packages

A word on timing

  • several exercises ask you to "time" things
  • here it's always sufficient to use the time command line:
time ./racing
real 0m0.022s
user 0m0.011s
sys  0m0.021s
  • and just take the first number (real time)

How to test your setup

  • Please run check_setup.sh to check your setup on Linux / Mac. The optional tools are not required for the essentials course.
  • go to exercises/hello
  • follow the README.md

How to compile and run?

cd exercises/hello
make
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:.
./hello

Optionally: check for valgrind/cppcheck

  • valgrind --tool=callgrind ./hello; kcachegrind
  • cppcheck .

Anatomy of an exercise

Each exercise is in a subdirectory with mainly 4 sets of files

Instructions

  • Each exercise comes with a set of instructions in the course slides
  • Most of them also with a README.md in the exercises subdir
  • And with instructions in the code

*.hpp and *.cpp files

  • the code to understand and fix/complete
  • you never start from scratch

Makefile / cmake

  • prepared Makefile for easy compilation
  • make and make clean are available
  • CMake is also supported:
    • Create a build directory such as mkdir build
    • cd build
    • Configure using cmake ..
    • Compile e.g. using make

solution subdirectory

  • the solution to the exercise (make solution)
  • please do not use before trying!

Examples of working environments

Using lxplus

  • lxplus9 at CERN is perfectly suitable for the exercises
  • it has all the needed compilers/tools
  • run e.g.
ssh lxplus9.cern.ch
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:.
git clone https://github.com/hsf-training/cpluspluscourse.git
cd cpluspluscourse/exercises/hello
make
./hello

Using OS/X

  • Install xcode and the command line tools
  • Compiling, running, debugging will work, but a few tools such as valgrind are not available

Using Windows

There are multiple options if you bring a Windows machine to the course:

  1. Use the Windows Subsystem for Linux (WSL)
  2. Use an ssh client and connect to lxplus
  3. Use Visual Studio and the native Windows compiler (MSVC)

We advise to use either WSL or lxplus, so a non-Windows experience. This way, you can follow all exercises, especially the ones on tools, which are mostly Linux focused. Also, we can ensure that mentors can help you in case you get stuck. Using lxplus is probably the easiest, but if you have a really hard time using the terminal, you can go for WSL and use Visual Studio Code.

WSL

Important: if you have VirtualBox installed on Windows do not try to install WSL2, due to this issue. It will either not work or run very slow. Instead, we recommend you use VirtualBox directly to set up a suitable virtualised Linux environment. Or use lxplus.

The procedure is described in detail here.

Windows 10 (version 2004 or higher)

The installation of WSL is as easy as running (with Administrator privileges):

wsl --install
# and reboot

This will activate WSL 2 and install Ubuntu 22.04. If you prefer a different Linux distribution, run:

wsl --list --online
wsl --install -d <Distro>

Older Windows 10 versions

On older Windows versions, the procedure is more involved and documented here.

It's lightly recommended to use WSL 2 as the Linux compatibility is better, but everything in the C++ course should work on either the WSL 1 or 2 backends.

Setup your Linux

Start your Linux inside a terminal by calling either wsl or bash. When you start for the first time you'll be asked to create a user account. Then you can login and install the packages which are necessary for the tutorial:

sudo -s # will ask for your password
apt update
apt upgrade
apt install git g++ gdb valgrind cppcheck cmake clang-format clang-tidy python3

Verify that the tools are there by running:

$ g++ --version
g++-11 (Ubuntu 11.2.0-19ubuntu1) 11.2.0
Copyright (C) 2021 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

You can install an editor inside the WSL system itself, but we recommend to use VS Code in Windows natively.

Connect to lxplus from Windows

Recent versions of Windows 10 (version 1809 or higher) come with an ssh client. If that does not work, you can install a third-party ssh client for Windows. We have good experience with https://www.putty.org/.

Open Windows Terminal, or PowerShell if you run an older version of Windows. Use ssh lxplus9.cern.ch to connect to lxplus. Follow the lxplus guide from here.

If you are having troubles with the terminal based environment, you can also try to use VS Code natively on your Windows to connect to lxplus. We have not tested this seriously yet, but it should work.

Visual Studio

If you want to stay on Windows, make sure you have Visual Studio and cmake installed. Visual Studio comes in a few distributions, some requiring a paid license. The Community edition is free for academic use or for working on open source software. During installation of Visual Studio, select the Desktop development with C++ workload.

Visual Studio Community: https://visualstudio.microsoft.com/vs/community/ CMake: https://cmake.org/download/

For the exercises, you will use the standard CMake workflow as described in the course slides (create a build directory, run cmake, etc.). CMake will create a Visual Studio solution file .sln in the build directory, which you can open to start working on an example within the Visual Studio IDE.

Visual Studio Code

For developing on Windows natively, but compiling and running on a remote Linux system (including WSL), we recommend using Visual Studio Code (VS Code). VS Code is a more lightweight code editor and not to be confused with the full-blown IDE Visual Studio. Please install VS Code from this website.

For remote development you will also need the following extensions to VS Code:

  • WSL
  • Remote - SSH
  • C/C++ Extension Pack

You can install them from the extension menu (Ctrl + Shift + X).

There are two ways to launch VS Code for working with WSL:

  1. GUI first: Launch VS Code on Windows. Run from the command palette (Ctrl + Shift + P), Remote-WSL New Window. VS Code will now begin a session running inside the WSL Linux on your system. You can click on Clone Git Repository... or, if you already cloned the GitHub repository, click Open Folder.... You can also open a terminal with bash via the menu Terminal -> New Terminal. Use the newly opened terminal to type commands for building, debugging, etc.
  2. Terminal first: Launch bash or wsl either via a powershell or from Windows Terminal (if installed). Navigate to your cloned GitHub repository, or git clone it if you have not done it yet. Inside the <repo>/exercises or any <repo>/exercises/example directory, run code .. This will launch VS Code on Windows, already connected to your WSL and set to the right folder. Use your initial terminal to type commands for building, debugging etc.

More documentation on using WSL with Visual Studio Code can be found here.

Experimental (we have not tested this seriously): You can also use VS Code to work on lxplus. Run from the command palette (Ctrl + Shift + P), Remote-SSH Connect to host.... Then enter <user>@lxplus9.cern.ch, then your password. The first connection takes a while for VS Code to setup its server on lxplus.