[FORTRAN PATCH] PR30400: ANY not accepted as FORALL mask

Tobias Schlüter tobias.schlueter@physik.uni-muenchen.de
Sun Feb 18 17:47:00 GMT 2007


Paul Richard Thomas wrote:
> You are quite right - does the attached do it for you?

This is ok.  I couldn't test it myself due to unrelated issues, but 
there's one corner case which I think we should put into a testcase: a 
mask expression which uses the '==' operator, something like "forall 
(j=1:5, j==2)"  The code in match_forall_iterator seems to be doing the 
right thing when it sees this, but it's worth making sure.  Please also 
cleanup the comment in front of the function, the part about the scalar 
mask is wrong.

- Tobi



More information about the Fortran mailing list