AmirHossein Sojoodi

AmirHossein Sojoodi

PhD student in Computer Engineering at Queen's University

Bash Tips and Tricks

Published:
Updated:
7 minute read

Use “Exit Traps” to make scripts more robust and reliable

You can place any code that you want to be certain to run in this “finish” function.

#!/bin/bash
function finish {
  # Your cleanup code here
}
trap finish EXIT

More info: here

Use Bash Strict Mode (Unless You Love Debugging!)

Add this in your scripts to cause them exit immediately in case of errors, bugs, and exceptions.

#!/bin/bash
set -euxo pipefail

# which is the short version of:

set -e
set -u
set -o pipefail
set -x

Read more from here to understand what each does.

Run a specific command on a specific CPU(s)

# This will run the command on cpu 3:
$ taskset -c 3 command

# This will run the command on cores 64-127:
$ taskset -c 64-127 command

while read url; do axel -n 10 $url; done < myLinks

Redirect the command output

  1. Redirect stdout to one file and stderr to another file:
command > out 2>error
  1. Redirect stdout to a file (>out), and then redirect stderr to stdout (2>&1):
command >out 2>&1
  1. Redirect both to a file (this isn’t supported by all shells, bash and zsh support it, for example, but sh and ksh do not):
command &> out

More detailed methods: (got from here)

          || visible in terminal ||   visible in file   || existing
  Syntax  ||  StdOut  |  StdErr  ||  StdOut  |  StdErr  ||   file   
==========++==========+==========++==========+==========++===========
    >     ||    no    |   yes    ||   yes    |    no    || overwrite
    >>    ||    no    |   yes    ||   yes    |    no    ||  append
          ||          |          ||          |          ||
   2>     ||   yes    |    no    ||    no    |   yes    || overwrite
   2>>    ||   yes    |    no    ||    no    |   yes    ||  append
          ||          |          ||          |          ||
   &>     ||    no    |    no    ||   yes    |   yes    || overwrite
   &>>    ||    no    |    no    ||   yes    |   yes    ||  append
          ||          |          ||          |          ||
 | tee    ||   yes    |   yes    ||   yes    |    no    || overwrite
 | tee -a ||   yes    |   yes    ||   yes    |    no    ||  append
          ||          |          ||          |          ||
 n.e. (*) ||   yes    |   yes    ||    no    |   yes    || overwrite
 n.e. (*) ||   yes    |   yes    ||    no    |   yes    ||  append
          ||          |          ||          |          ||
|& tee    ||   yes    |   yes    ||   yes    |   yes    || overwrite
|& tee -a ||   yes    |   yes    ||   yes    |   yes    ||  append

(*) Bash has no shorthand syntax that allows piping only StdErr to a second command, which would be needed here in combination with tee again to complete the table. If you really need something like that, please look at “How to pipe stderr, and not stdout?” on Stack Overflow for some ways how this can be done e.g. by swapping streams or using process substitution.

Disable the annoying bell in bash

  1. To disable the beep in bash you need to uncomment (or add if not already there) the line set bell-style none in your /etc/inputrc file.

Note: Since it is a protected file you need to be a privileged user to edit it (i.e. launch your text editor with something like sudo <editor> /etc/inputrc).

  1. To disable the beep also in vim you need to add set visualbell in your ~/.vimrc file.

  2. To disable the beep also in less (i.e. also in man pages and when using git diff) you need to add export LESS="$LESS -R -Q" in your ~/.profile file.

Make your prompt look nicer

  • Put this line inside of .bashrc
export PS1="\[\033]0;$TITLEPREFIX:$PWD\007\]\n\[\033[32m\]\u@\h \[\033[35m\]$MSYSTEM \[\033[33m\]\w\[\033[36m\]\[\033[0m\]\n$ "

Display all IP addresses connected to your host

netstat -lantp | grep ESTABLISHED |awk '{print $5}' | awk -F: '{print $1}' | sort -u

List the commands you use more often

history | awk '{a[$2]++}END{for(i in a){print a[i] " " i}}' | sort -rn | head

Helpful aliases

  • Put these aliases in .bashrc
alias c='clear'
alias cd..='cd ..'
alias ..='cd ..'
alias ...='cd ../../../'
alias ....='cd ../../../../'
alias .....='cd ../../../../../'
alias .2='cd ../../'
alias .3='cd ../../../'
alias .4='cd ../../../../'
alias .5='cd ../../../../..'
alias mkdir='mkdir -pv'
alias diff='colordiff'
alias h='history'
alias histg="history | grep"
alias now='date +"%T"'
alias nowtime=now
alias nowdate='date +"%d-%m-%Y"'
alias vi='vim'
alias svi='sudo vi'
alias vis='vim "+set si"'
alias edit='vim'
alias ports='netstat -tulanp'
alias su='sudo -i'
alias reboot='sudo /sbin/reboot'
alias poweroff='sudo /sbin/poweroff'
alias halt='sudo /sbin/halt'
alias shutdown='sudo /sbin/shutdown'
alias meminfo='free -m -l -t'
alias psmem='ps auxf | sort -nr -k 4'
alias psmem10='ps auxf | sort -nr -k 4 | head -10'
alias pscpu='ps auxf | sort -nr -k 3'
alias pscpu10='ps auxf | sort -nr -k 3 | head -10'
alias gpumeminfo='grep -i --color memory /var/log/Xorg.0.log'
alias wget='wget -c'
alias df="df -Tha --total"
alias psg="ps aux | grep -v grep | grep -i -e VSZ -e"
alias top="htop"
alias myip="curl http://ipecho.net/plain; echo"
alias DU="du --max-depth=1 -B M |sort -rn"
alias filecount="find . -type f 2> /dev/null | wc -l"

Some useful user config

Put these into the file ~/.inputrc:

# Place this file into your home directory
# none, visible or audible
set bell-style visible

# Ignore case for the command-line-completion functionality
# on:  default on a Windows style console
# off: default on a *nix style console
set completion-ignore-case on

set colored-stats On
set completion-ignore-case On
set completion-prefix-display-length 3
set mark-symlinked-directories On
set show-all-if-ambiguous On
set show-all-if-unmodified On
set visible-stats On

Query the linked libraries of an executable

To see which libraries have been linked to a specific executable, e.g. mpirun

ldd `which mpirun`

Query the executable header, e.g. check RUNPATH entry

readelf -d /home/sojoodi/projects/def-queenspp/sojoodi/OpenMPI-Debug/build/bin/mpirun | grep RUNPATH

LD_PRELOAD Trick

In summary, you can preload the specified shared libraries at runtime to resolve some conflicts happened at link-time, or to overright any previously statically bound libraries, like stubs. For instance, if you want to preload NVIDIA ML library to make sure your program points at the current location at runtime, you can do something like the following:

export LD_PRELOAD=/usr/lib64/nvidia/libnvidia-ml.so
/path/to/executable

To find more information about LD_PRELOAD trick, read here.

You May Also Enjoy

Utilize PTX Just-In-Time (JIT) Compilation in CUDA

4 minute read

Published:

In this post I’ve written about how to utilize PTX Just-In-Time (JIT) compilation in CUDA. PTX is a low-level assembly-like language that is used to represent the GPU code. The PTX code is then compiled to the machine code by the NVIDIA driver at runtime. This process is called Just-In-Time (JIT) compilation. But before I write about how to use PTX JIT compilation, I’ll provide some background on why you might want to use it.

Collection of my dotfiles (public version)

1 minute read

Published:

This is a collection of my dotfiles that I use on my Linux system. I have written about them in various posts, but I thought I can put them all in one place. You can checkout the github repository here, too.

Tmux Tips and Tricks

10 minute read

Published:

(If you are already familiar with tmux, and you are looking for a simple .tmux.conf, go to the end of this post.)

Setup LAMMPS

4 minute read

Published:

Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) is a classical molecular dynamics code that can be used to model atoms or, more generally, as a parallel particle simulator at various scales. The complete documentation of LAMMPS can be found here. In this post, I will provide a guide on how to setup LAMMPS on a Linux machine. My setup is on a cluster with NVIDIA GPUs, UCX, and OpenMPI. Also, we have a built-in module system to load the necessary modules.