In a previous article, I showed how to use mutexes to prevent race conditions. Condition Variables use mutexes to allow exclusive access to data, but also allow threads to wait for something to happen before they start to do work.

Understanding when Condition Variables are useful is easier with an example. Let’s say we are building a queue system. To keep it simple we will start with 1 producer and 1 consumer. We can make this system safe with a mutex:

I consider dependency management a very important part of a maintainable software project. Once you start writing software that depends on other code, it can quickly become unpractical to do dependency management by hand, so a good system is necessary to make it easy to work on such projects.

Since C++ has been out for a while, I expected to find a very mature ecosystem in this regard. Looking around and reading a little, I found that there is no community agreed way of solving the problem of managing dependencies. One reason for this seems to be that it’s not easy to build a system that helps you do dependency management when there are many ways to build a package.

A while ago I wrote an article about mutexes in C++. I explained how they solve the problem of race conditions by allowing only one thread to access data simultaneously.

A read-write mutex (also known: readers-writer, shared-exclusive, multiple-readers/single-writer, mrsw) is a specialization of a mutex that allows for greater performance in a scenario where reads are more frequent than writes.

Now that I know enough C++ to be able to build something useful, I started looking at tools that will make it possible to create maintainable projects. I’m exploring CMake because it is probably the most popular build tool for C++ projects. It allows you to create platform independent configuration files that can then be translated to the platform of your choice.

Installation

In Ubuntu, you can install CMake using apt-get:

1
sudo apt-get install cmake

For other systems, you can download the binary from CMake’s downloads page.

In a previous article I wrote an article explaining how header files work in C++. Among other things, the article explains how you can split your code in multiple files and link them together to create a single executable. One of the advantages of doing this, is that it allows code to be better organized, by separating different responsibilites in different files.

In this article I want to go one step further and explain how to create a library that can be reused by multiple projects.

lsof stands for List Open Files. It can help us find which process is using a file at a given point in time. The reason lsof is so useful in Unix/Linux systems is that sockets and devices are treated the same way as files (Pretty much everything is considered a file in Unix/Linux).

Running lsof without any arguments will list all open files in the system. If you have a lot of processes working with a lot of files, prepare to wait. The output looks somethins like this:

1
2
3
4
5
6
7
$ sudo lsof
COMMAND     PID   TID            USER   FD      TYPE             DEVICE  SIZE/OFF       NODE NAME
systemd       1                  root  cwd       DIR              253,1      4096          2 /
systemd       1                  root  rtd       DIR              253,1      4096          2 /
systemd       1                  root  txt       REG              253,1   1577264    5374284 /lib/systemd/systemd
systemd       1                  root  mem       REG              253,1     18976    5375835 /lib/x86_64-linux-gnu/libuuid.so.1.3.0
...

In a previous article I showed how to assemble a program using nasm. In this article I’m going to explore different ways to access data and explore some instructions.

Variables

The simplest way do declare variables is by initializing them in the .data segment of a program. The format to define initialized data is:

1
[variable-name]    define-directive    initial-value   [,initial-value]...

An example use:

1
2
3
4
5
6
7
8
9
10
11
section .data
  exit_code dq 0
  sys_call dq 60

section .text
  global _start

_start:
  mov rax, [sys_call]
  mov rdi, [exit_code]
  syscall

The vim command-line mode is the mode we enter when typing Ex commands (:), search patterns (/, ?) or filter commands (!). This mode works a little like insert mode, in that whatever we type, is going to appear in the command line. It is not as powerful as the normal mode, but there are a few combinations we can use to move more efficiently:

• Ctrl+left, Ctrl+right - Using the arrow keys we can move left and right one character at a time. If we press Ctrl together with the left or right keys, we will move one word at a time
• Ctrl+B, Ctrl+E - Move to the beginning and end of the command line, respectively
• Ctrl+W - Deletes the word before the cursor (Only deletes characters at the left of the cursor)

Vim allows us to customize the statusline shown at the bottom of each window. To toggle the statusline visibility, the laststatus option can be used. It can be set to any of these values:

1
2
3
0: never
1: only if there are at least two windows (default)
2: always

If we want to always show the statusline, we can use this command:

1
set laststatus=2

I use grep.vim to grep in multiple files within vim. It usually works pretty well, but I recently needed a little more freedom in how I specify the files in which I want to search, and found that it can’t do what I want. I looked at the code and found that modifying it to do what I want would be hard with the current design, so I decided to create my own plugin.

The plugin I’m going to create is going to be very simple. It is going to provide simple greping functionality with a similar UI to grep.vim, but will allow more freedom on filtering which files to search.