Handling Mutually Exclusive variable in background processes

I have two functions where I access a shared variable.

% Create a shared variable using parallel.pool.Constant
sharedVar = parallel.pool.Constant(@() 0); % Initialize shared variable to 0

% Function 1: Increment the shared variable
function incrementSharedVar(sharedVar)
    sharedVar.Value = sharedVar.Value + 1; % Increment shared variable
end

% Function 2: Decrement the shared variable
function decrementSharedVar(sharedVar)
    sharedVar.Value = sharedVar.Value - 1; % Decrement shared variable
end

I run these functions in background.

% Run the increment and decrement functions as background tasks
f1 = parfeval(@incrementSharedVar, 0,sharedVar);
f2 = parfeval(@decrementSharedVar, 0,sharedVar);

% Wait for both tasks to complete
wait([f1, f2]);

How to ensure the mutual exclusion among this shared variable so that one function waits while the other function updates the value?

Answer 1

[Sumukh]

Hi @ARIVARASAN,

To run the functions ”f1” and “f2” as background tasks, the term "backgroundPool" must be passed as the first argument to "parfeval". You can refer to the following documentation to know more about this syntax:

https://www.mathworks.com/help/matlab/ref/parfeval.html#mw_c960825d-a447-4c1f-9009-a6fa9e22720f

The provided code does not change the value of the Constant object "sharedVar" because the value of the Constant object is read-only and cannot be changed once initialized. More details about the Constant object can be found in the following documentation:

https://www.mathworks.com/help/releases/R2024b/matlab/ref/parallel.pool.constant.html

A possible method to provide gaps between execution of the “parfeval” functions “f1” and “f2“ is to use DataQueue object. The DataQueue object can be used to create a queue of function calls by calling “f1” and “f2” one after the other using the “afterEach” command. Kindly refer to the following documentation to know more about the DataQueue object and its function "afterEach":

https://www.mathworks.com/help/parallel-computing/parallel.pool.dataqueue.html

I hope this answers your query.

[ARIVARASAN]

Hey!! Thanks for your reply. I read more about this including the links you provided. I need to execute these functions parallelly in the background where they parallelly edit a variable in a mutually exclusive manner. "parfeval" seems to achieve this, but I can't seem to get my head around how to share a mutually exclusive variable among the baclground processes. This is a bit tricky. Any ideas?