[patch, Fortran] Fix PR 71902

[Thomas Koenig]

Hello world,

this patch fixes the performance regression introduced with
the fix for 71783.  It also corrects an error in the logic
in realloc_string_callback by now checking if the experssion
is deferred, instead of checking for the symbol having the
allocatable attribute (which led to false positives).

It also checks for the presence of a substring on the RHS
expression, because the problem cannot happen if there is
no RHS substring.

Regression-tested on trunk.

OK for trunk?  Also OK for a backport?

Regards

	Thomas

2016-08-14  Thomas Koenig  <tkoenig@gcc.gnu.org>

         PR fortran/71902
         * frontend-passes.c (realloc_string_callback):  Check for deferred
         on the expression instead for allocatable on the symbol.  Name 
temporary
         variable "realloc_string".

2016-08-14  Thomas Koenig  <tkoenig@gcc.gnu.org>

         PR fortran/71902
         * gfortran.dg/dependency_47.f90:  New test.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p5.diff
Type: text/x-patch
Size: 1331 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20160814/3feda2b2/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dependency_47.f90
Type: text/x-fortran
Size: 405 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20160814/3feda2b2/attachment-0001.bin>