[Bug fortran/60795] [4.7/4.8/4.9 Regression] Wrong length when allocating character array

[quantheory at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60795

Sean Santos <quantheory at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |quantheory at gmail dot com

--- Comment #3 from Sean Santos <quantheory at gmail dot com> ---
In comment 2, "y3", "t3", and the second allocation of "s3" are all cases of
PR44672; this is an unsupported feature of F2008.

The original case (comment 0, "s1") may be a duplicate of PR57456.

The case "s2" looks like a duplicate of PR58754, and "s3" and "s4" may be the
same issue. Or they may be related to PR50221.

So there are clearly some issues here, but I think that they are mostly old
issues, i.e. arrays of deferred-length strings have always been somewhat
broken.

I remember something about some intention to address this when implementing the
new array descriptor, though?