Extract data from sparse matrix in a parfor loop

Hi, I need to find eigenvectors and eigenvalues for a lot of 100*100 tri-diagonal matrices. they defined as follow:

 A_kk       = a_n  ; %      main diagonal (a_n is some vector)
 A_kk_below = b_n*f; %below main diagonal (b_n is some vector, and f is a number which differ at each matrix)
 A_kk_above = c_n*f; %above main diagonal (c_n is some vector, and f is a number which differ at each matrix)

I'm looking for a fast way to do it so my strategy was to define a large large tridiagonal sparse matrix. Then inside a parfor loop to take part of the sparse matrix and calculate its eigenvectors/values

 N=100;
 Len_f=100;
 a_ln_t=zeros(N,N,Len_f);
 parfor i=1:Len_f
        A_f = full(A((i-1)*N+(1:N),(i-1)*N+(1:N))); % take part of the sparse matrix A
        [EigV,Eig]   = eig(A_f,'vector');           % find eigenvectors/values
        [Eig,iEig]   = sort(Eig);                   % sort eigenvalues
        EigV         = EigV(:,iEig);                % sort eigenvectors
        Eig_f(:,:,i) = EigV;                        % save results in variable
 end

But the problem is that I get a warning when trying to take part of the sparse matrix.

"The entire array or structure "A" is a broadcast variable. This might result in unnecessary communication overhead"

Is there any suggestions on how to avoid this warning and improve this code and perhaps make it faster?

Any input on this matter would be highly appreciated