https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88486
--- Comment #12 from Steve Kargl <kargl at gcc dot gnu.org> --- (In reply to Mikael Morin from comment #11) > (In reply to Mikael Morin from comment #10) > > It seems gfc_conv_string_length changes the string length backend_decl of x > > from *_x to a new variable (D.4740 with my out of date tree), so that the > > reassignment code updates that variable instead of the original one _x. > > Adjusting the condition at the beginning of gfc_conv_string_length doesn't > seem to work. I must be missing something. Hi Mikael, If the 2nd assignment is replaced with "x = ['^' // [x(1),x(2)] // '^']", then the correct answer is generated. So, it seems the reference to the whole array "x" seems to get in the way of updating the length for the lhs.
