Advanced Python Programming

Ignoring locks and sharing resources

Locks are undoubtedly an important tool in synchronization tasks, and in concurrent programming in general. However, if the use of locks leads to an undesirable situation, such as a deadlock, then it is quite natural for us to explore the option of simply not using locks in our concurrent programs. By ignoring locks, our program's resources effectively become shareable among different processes/threads in a concurrent program, thus eliminating the first of the four Coffman conditions: mutual exclusion.

This approach to the problem of deadlock can be straightforward to implement; let us try with the two preceding examples. In the two-lock example, we simply remove the code specifying any interaction with ...