(i) Priority Inversion is a condition in which low priority task executes while a high priority task waits(blocked) due to resource contentions.

(ii) PI happens because in poorly designed RTFS.

Consider following example:

(iii) Let A, B, C $\to$ 3 process/task.

A- high priority, B- med priority, C- lower priority

Process/task A and task C shares a variable 'X' and access to this variable is sync with use of semaphore.

(iv) Now task C is ready and is picked up for execution by scheduler and acquires the shared variable.

(v) But now Task A is blocked due to unavailability of resource and waiting for the resource being used by task B which has itself pre-empted the task C.

(vi) In this situation, the task A which has high priority level has been effectively inverted to the lower priority level.

This is known as PI.

Bounded inversion: If the inversion will get over in a fixed time boundary, then it is called as bounded priority inversion.

Unbounded inversion: The time for which priorities will be inverted is not bounded then is called as unbounded priority inversion.

Solutions to Priority Inversion

(i) Priority Ceiling

• In this case, the lower priority task has the resource due to which the higher task enters into blocked state, is temporarily given highest(top most) priority of all task.

• This ensures that no lower or other task will be able to acquire the resource in between, since it has become the highest priority task.

• This will ensure No PI take place.

(ii) Priority inheritance

• In this case task having lower priority and accessing shares variable/data is given priority; just above the higher priority(operation remains same as above)