C++ With the Command Line

Author: Many

Contributors: Benjamin Qi, Hankai Zhang, Anthony Wang, Nathan Wang, Nathan Chen, Michael Lan, Arpan Banerjee

OS-specific instructions for installing and running C++ via the command line.

Command Line Basics


what compiling a simple program looks like

This section is not complete.

Any help would be appreciated! Just submit a Pull Request on Github.

hm, anything simpler / interactive (and free)?


Learn Enough


William Shotts


Should be mostly the same as Linux ...

Open the Terminal application and familiarize yourself with some basic commands. Upgrade to zsh if you haven't already.

Jim Hoskins
Rahul Saigal

keyboard shortcuts / terminal commands

Armin Briegel


Installing GCC

USACO (and most contests) use GCC's g++ to compile and run your code. You'll need g++ specifically to use the #include <bits/stdc++.h> header file; see Running Code Locally for details.

On Linux

GCC is usually preinstalled on most Linux distros. You can check if it is installed with

whereis g++

If it is not preinstalled, you can probably install it using your distro's package manager.

On Windows

There are multiple ways to install C++ command line tools on windows, depending on your needs. To find more information, go to Running Code Locally with Windows.

On Mac

  1. Install XCode command line tools.

    xcode-select --install

    If you previously installed these you may need to update them:

    softwareupdate --list # list updates
    softwareupdate -i -a # installs all updates

    After this step, clang should be installed (try running clang --version in Terminal).

  2. Install Homebrew.

  3. Install gcc with Homebrew.

    brew install gcc

    According to this if brew doesn't seem to finish for a long time then

    brew install gcc --force-bottle

    probably suffices.

  4. You should be able to compile with g++-#, where # is the version number (e.g., 10). Running the following command

    g++-10 --version

    should display something like this:

    g++-10 (Homebrew GCC 10.2.0_2) 10.2.0
    Copyright (C) 2020 Free Software Foundation, Inc.
    This is free software; see the source for copying conditions.  There is NO
  5. If you want to be able to compile with just g++, write a shell alias! Put the following lines into your shell's rc file (~/.bashrc if you use bash, and ~/.zshrc if you use zsh).

    alias g++=g++-10

    Once you do so, g++ --version should now output the same thing as g++-10 --version.

    Note: avoid overriding the system g++ with symlinking or hard-linking as that will almost surely cause problems. Don't worry if you don't know what those terms mean.

C++ with the Command Line

Basics of Compiling & Running

Consider a simple program such as the following, which we'll save in name.cpp.

#include <iostream>
using namespace std;
int main() {
int x; cin >> x;
cout << "FOUND " << x << "\n";

It's not hard to compile & run a C++ program. First, open up Powershell on Windows, Terminal on Mac, or your distro's terminal in Linux. We can compile name.cpp into an executable named name with the following command:

g++ name.cpp -o name

Then we can execute the program:


If you type some integer and then press enter, then the program should produce output. We can write both of these commands in a single line:

g++ name.cpp -o name && ./name

Note that && ensures that ./name only runs if g++ name.cpp -o name finishes successfully.

Redirecting Input & Output

If you want to read standard input from inp.txt, use the following:

./name < inp.txt

If you want to write standard output to out.txt, then use the following:

./name > out.txt

They can also be used in conjunction, as shown below:

./name < inp.txt > out.txt

See Input & Output for how to do file input and output within the program.

Compiler Options (aka Flags)

Use compiler flags to change the way GCC compiles your code. Usually, we use something like the following in place of g++ name.cpp -o name:

g++ -std=c++17 -O2 name.cpp -o name -Wall
  • -O2 tells g++ to compile your code to run more quickly while increasing compilation time (see here).
  • -std=c++17 allows you to use features that were added to C++ in 2017. USACO recently upgraded from C++11 to C++17.
  • -Wall checks your program for common errors. See Debugging for more information.

You should always compile with these flags.


Adding Shortcuts

For Users of Linux & Windows

The process is similar for Linux. If you're on Windows, you can use an IDE to get these shortcuts, or you can install WSL (mentioned above).

Retyping the flags above can get tedious. You should define shortcuts so you don't need to type them every time!

First, create your .zshrc if it doesn't already exist.

touch ~/.zshrc

Open your .zshrc with a text editor.

open ~/.zshrc

or some text editor (ex. sublime text with subl).

subl ~/.zshrc

You can add aliases and functions here, such as the following to compile and run C++ on Mac.

co() { g++ -std=c++17 -O2 -o "${1%.*}" $1 -Wall; }
run() { co $1 && ./${1%.*} & fg; }

Now you can easily compile and run name.cpp from the command line with co name.cpp && ./name or run name.cpp. Note that all occurrences of $1 in the function are replaced with name.cpp, while ${1%.*} removes the file extension from $1 to produce name.

What is & fg for?

Let prog.cpp denote the following file:

#include <iostream>
#include <vector>
using namespace std;
int main() {
vector<int> v;
cout << v[-1];

According to the resource above, the & fg is necessary for getting zsh on Mac to display crash messages (such as segmentation fault). For example, consider the running the first prog.cpp above with run prog.cpp.

If & fg is removed from the run command above then the terminal displays no message at all. Leaving it in produces the following (ignore the first two lines):

[2] 30594
[2]  - running    ./${1%.*}
zsh: segmentation fault  ./${1%.*}

Measuring Time & Memory Usage

For example, suppose that prog.cpp consists of the following:

#include <bits/stdc++.h>
using namespace std;
const int BIG = 1e7;
int a[BIG];
int main() {
int sum = 0;
for (int i = 0; i < BIG; ++i) sum += a[i];
cout << sum;

Then co prog.cpp && gtime -v ./prog gives the following:

  Command being timed: "./prog"
  User time (seconds): 0.01
  System time (seconds): 0.01
  Percent of CPU this job got: 11%
  Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.22
  Average shared text size (kbytes): 0
  Average unshared data size (kbytes): 0
  Average stack size (kbytes): 0
  Average total size (kbytes): 0
  Maximum resident set size (kbytes): 40216
  Average resident set size (kbytes): 0
  Major (requiring I/O) page faults: 91
  Minor (reclaiming a frame) page faults: 10088
  Voluntary context switches: 3
  Involuntary context switches: 38
  Swaps: 0
  File system inputs: 0
  File system outputs: 0
  Socket messages sent: 0
  Socket messages received: 0
  Signals delivered: 0
  Page size (bytes): 4096
  Exit status: 0

Note that 10710^7 integers require 4107103400004\cdot 10^7\cdot 10^{-3}\approx 40000 kilobytes of memory, which is close to 4021640216 in the above output as expected.

Adjusting Stack Size

Let A.cpp denote the following program:

#include <iostream>
using namespace std;
int res(int x) {
if (x == 200000) return x;
return res(x+1);
int main() {
cout << res(0) << "\n";

If we compile and run this with g++ A.cpp -o A && ./A, this outputs 200000. However, changing 200000 to 300000 gives a segmentation fault. Similarly,

#include <iostream>
using namespace std;
int main() {
int arr[2000000];
cout << arr[0] << "\n";

runs, but changing 2000000 to 3000000 also gives a segmentation fault. This is because the stack size on Mac appears to be limited to 8 megabytes by default.


explanation of the issue

Note that USACO does not have a stack size limit, aside from the usual 256 MB memory limit. Therefore, code that crashes locally due to a stack overflow error may still pass on the USACO servers. To get your code running locally, use one of the methods below.


This matters particularly for contests such as Facebook Hacker Cup where you submit the output of a program you run locally.

Method 1

ulimit -s 65532 will increase the stack size to about 64 MB. Unfortunately, this doesn't work for higher numbers.

Method 2


looks like people complain about FHC every year!

To get around this, we can pass a linker option. According to the manual for ld (enter man ld in Terminal), the option -stack_size size does the following:

Specifies the maximum stack size for the main thread in a program. Without this option a program has a 8MB stack. The argument size is a hexadecimal number with an optional leading 0x. The size should be a multiple of the architecture's page size (4KB or 16KB).

So including -Wl,-stack_size -Wl,0x10000000 as part of your compilation command will set the maximum stack size to 16716^7 bytes \approx 256256 megabytes, which is usually sufficient. However, running the first program above with 200000 replaced by 1e7 still gives an error. In this case, you can further increase the maximum stack size (ex. by adding another 0 to the end).

For Windows

Changing the stack size is relatively easy on Windows. Simply adding -Wl,--stack,268435456 as a part of your compilation flags should do the trick. The 268435456 corresponds to 268435456 bytes, or 256 megabytes. If you are using Windows PowerShell, make sure to wrap it in quotations (like so: "-Wl,--stack,268435456"), since commas are considered to be special characters.

Module Progress:

Join the USACO Forum!

Stuck on a problem, or don't understand a module? Join the USACO Forum and get help from other competitive programmers!