[Bug fortran/87045] pointer to array of character
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87045 anlauf at gcc dot gnu.org changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #5 from anlauf at gcc dot gnu.org --- Fixed on trunk.
[Bug fortran/87045] pointer to array of character
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87045 --- Comment #4 from anlauf at gcc dot gnu.org --- Author: anlauf Date: Wed Mar 13 21:33:27 2019 New Revision: 269664 URL: https://gcc.gnu.org/viewcvs?rev=269664=gcc=rev Log: 2019-03-13 Harald Anlauf PR fortran/87045 * trans-expr.c (gfc_trans_pointer_assignment): Move check for same string length so that we do not get false errors for deferred length. PR fortran/87045 * gfortran.dg/pr87045.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/pr87045.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog
[Bug fortran/87045] pointer to array of character
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87045 anlauf at gcc dot gnu.org changed: What|Removed |Added CC||anlauf at gcc dot gnu.org --- Comment #3 from anlauf at gcc dot gnu.org --- Patch that moves the check submitted: https://gcc.gnu.org/ml/fortran/2019-03/msg00050.html
[Bug fortran/87045] pointer to array of character
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87045
--- Comment #2 from Harald Anlauf ---
With trunk it appears essential to have at least -fcheck=bounds
as option, otherwise the testcase passes for me.
The dump-tree for the critical code part (with -fcheck=bounds):
p = t;
p.span = (integer(kind=8)) SAVE_EXPR ;
if ((integer(kind=8)) (.p != .t))
{
_gfortran_runtime_error_at (&"At line 8 of file pr87045.f90"[1]{lb: 1
sz: 1}, &"Unequal character lengths (%ld/%ld) in %s"[1]{lb: 1 sz: 1}, .p, .t,
&"pointer assignment"[1]{lb: 1 sz: 1});
}
.p = .t;
Has the check been put into the wrong place?
[Bug fortran/87045] pointer to array of character
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87045 Thomas Koenig changed: What|Removed |Added Keywords||wrong-code Status|UNCONFIRMED |NEW Last reconfirmed||2018-08-25 CC||tkoenig at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #1 from Thomas Koenig --- Confirmed; valgrind shows uninitialized variables. That seems to go back to 4.8 at least, so it does not look like a regression.
