------- Comment #11 from jvdelisle at gcc dot gnu dot org 2007-09-30 01:58 ------- It turns out that the original patch for this bug is probably what we want. Unfortunately it uncovers a nasty latent bug where an extraneous namelist read is attempted. It only seems to occur with multiple levels of derived types. For example:
&MYNML x(2)%m(1)%ch(2) ='q', , &end Will read correctly with the initial patch for the bug here if the extra comma is placed at the end. This is a bugger since it involves both recursive calls and incrementing of loop specs all within a do - while loop. Still working on it. :) -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33253