WLS with ARMA terms estimation

[PROFI_USER]

Could you pleace comment prohibition of using WLS with ARMA specification. Eviews 7 Help says: "You shold note that weighted estimation is not offered in equations containing ARMA specifications..."
At the same time Eviews enables to estimate pool model with AR terms and Weights (for instance GLS Weights = Cross-Section weights) in the specification.

The second issue i would like to ask is Geleralised Least Squared. Eviews provide Weigted Least Squared. is it possible to estimate GLS model with specified Weights matrix W?
Actually I do not need to estimate W, I just try to estimate Least Squared model with matrix W. I believe it is ease to implement (X' * W-1 * X)-1 * X' * W-1* Y by hand, but I am not sure there is a build-in model in Eviews for this purpose.

Thanks!

[EViews Glenn]

The single equation and pool weighting differ in their effect on the ARMA structure. The GLS weights that are supported in pools correspond only to a within cross-section weight so that the AR process is scaled by the weighting and is not modified by time-series variation in the weights.

What are the characteristics of the weighting matrix? Is W diagonal?

[PROFI_USER]

What are the characteristics of the weighting matrix? Is W diagonal?

Weighting matrix W should be a block diagonal. It is a block matrix which is a square matrix, and having main diagonal blocks square matrices, such that the off-diagonal blocks are zero matrices.
Or it shoud be any block square matrix.

[EViews Glenn]

Nothing built in. If I were you, I'd probably convert to matrices and do it using the matrix operators.