The Safe Passwords library is here to make it simpler to manage passwords in memory in a safe manner. In most cases, having passwords linger in your memory is dangerous as someone can eventually get a hold of them.
This library offers a way to clear the passwords once not necessary anymore.
This happens at the time you destroy the safepasswords::string object.
In order to make a memory based password as safe as possible, we want to:
- try to clear all the buffers that included a password in clear
- avoid having the memory holding the password swapped out to disk
- avoid any intermediate buffers if at all possible (directly work with the safepasswords::string instead of std::string)
To prevent Linux from sending data to disk (i.e. when the OS decides to swap out some buffers to disk), the mark our buffers as non-swappable which is done with mlock(3). This functionality requires the buffers to be aligned to a memory page. To do so, we allocate exactly a string buffer of exactly one page. This means all the safepasswords strings are at least 4Kb in size (plus the string and buffer structures).
We want to enhance this by implementing a memory allocation within such buffers (i.e. allocate 16Kb and then allocate the string buffers within that 16Kb of RAM instead of allocating 4K per string).
The generate-password tool can be used to see the list of available
digests or generate passwords from your command line.
The source is covered by the GPL 3 license.
Submit bug reports and patches on github.
This file is part of the snapcpp project.