Intel Threading Building Blocks

Deadlock

Deadlock happens when tasks are trying to acquire more than one lock, and each holds some of the locks the other tasks need in order to proceed. More precisely, deadlock happens when:

• There is a cycle of tasks.

• Each task holds at least one lock on a mutex, and is waiting on a mutex for which the next task in the cycle already has a lock.

• No task is willing to give up its lock.

Think of classic gridlock at an intersection. Each car has “acquired” part of the road, but it needs to acquire the road under another car to get through. There are three common ways to avoid deadlock:

• Avoid needing to hold two locks at the same time. Break your program into small actions, each of which can be accomplished while holding a single lock.

• Always acquire locks in the same order. For example, if you have outer container and inner container mutexes, and you need to acquire a lock on one of each, you could always acquire the outer sanctum one first. Another example is to acquire locks in alphabetical order, in a situation where the locks have names. Or if the locks are unnamed, acquire locks in order of the numerical addresses for the mutex.

• Use atomic operations instead of locks, as discussed later in this chapter.