* [PATCH,committed] PR fortran/68318 -- increment reference count
@ 2015-11-13 0:18 Steve Kargl
0 siblings, 0 replies; only message in thread
From: Steve Kargl @ 2015-11-13 0:18 UTC (permalink / raw)
To: fortran, gcc-patches
[-- Attachment #1: Type: text/plain, Size: 804 bytes --]
I've committed the attached patch as obvious after
testing on x86_64-*-freebsd.
The short story is that gfortran tracks the number
of ENTRY symbols with a reference count. If an
ENTRY was included in a routine within a MODULE the
reference count was not properly increment. This
patch now does the increment.
As a bonus it fixes a nearby comment that is missing
a space, and changes the only occurence of ++sym->refs
to sym->refs++ for consistency.
2015-11-12 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/68318
* decl.c (get_proc_name): Increment reference count for ENTRY.
While here, fix comment and use postfix ++ for consistency.
2015-11-12 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/68318
* gfortran.dg/pr68318_1.f90: New test.
* gfortran.dg/pr68318_2.f90: Ditto.
--
Steve
[-- Attachment #2: pr68318.diff --]
[-- Type: text/x-diff, Size: 2548 bytes --]
Index: gcc/fortran/decl.c
===================================================================
--- gcc/fortran/decl.c (revision 230064)
+++ gcc/fortran/decl.c (working copy)
@@ -926,6 +926,7 @@ get_proc_name (const char *name, gfc_sym
gfc_find_sym_tree (name, gfc_current_ns, 0, &st);
st->n.sym = *result;
st = gfc_get_unique_symtree (gfc_current_ns);
+ sym->refs++;
st->n.sym = sym;
}
}
@@ -972,7 +973,7 @@ get_proc_name (const char *name, gfc_sym
/* Trap another encompassed procedure with the same name. All
these conditions are necessary to avoid picking up an entry
whose name clashes with that of the encompassing procedure;
- this is handled using gsymbols to register unique,globally
+ this is handled using gsymbols to register unique, globally
accessible names. */
if (sym->attr.flavor != 0
&& sym->attr.proc != 0
@@ -9052,7 +9053,7 @@ gfc_match_final_decl (void)
/* Add this symbol to the list of finalizers. */
gcc_assert (block->f2k_derived);
- ++sym->refs;
+ sym->refs++;
f = XCNEW (gfc_finalizer);
f->proc_sym = sym;
f->proc_tree = NULL;
Index: gcc/testsuite/gfortran.dg/pr68318_1.f90
===================================================================
--- gcc/testsuite/gfortran.dg/pr68318_1.f90 (nonexistent)
+++ gcc/testsuite/gfortran.dg/pr68318_1.f90 (working copy)
@@ -0,0 +1,18 @@
+! { dg-do compile }
+! { dg-options "-O0"
+! PR fortran/68318
+! Original code submitted by Gerhard Steinmetz
+! <gerhard dot steinmetz dot fortran at t-online dot de>
+!
+module m
+ implicit none
+contains
+ subroutine s1
+ entry e ! { dg-error "(2)" }
+ end
+ subroutine s2
+ entry e ! { dg-error "is already defined" }
+ end
+end module
+! { dg-prune-output "Duplicate ENTRY attribute specified" }
+
Index: gcc/testsuite/gfortran.dg/pr68318_2.f90
===================================================================
--- gcc/testsuite/gfortran.dg/pr68318_2.f90 (nonexistent)
+++ gcc/testsuite/gfortran.dg/pr68318_2.f90 (working copy)
@@ -0,0 +1,22 @@
+! { dg-do compile }
+! PR fortran/68318
+! Original code submitted by Gerhard Steinmetz
+! <gerhard dot steinmetz dot fortran at t-online dot de>
+!
+module m1
+ implicit none
+contains
+ subroutine s1
+ entry e
+ end
+end module
+
+module m2
+ use m1 ! { dg-error "(2)" }
+ implicit none
+contains
+ subroutine s2
+ entry e ! { dg-error "is already defined" }
+ end
+end module
+! { dg-prune-output "Cannot change attribute" }
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2015-11-13 0:18 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-11-13 0:18 [PATCH,committed] PR fortran/68318 -- increment reference count Steve Kargl
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).