Dear all,
as the PR shows, it is likely not a good idea to try to make an
ASSOCIATE variable static when -fno-automatic is specified, so
rather keep it on the stack.
Attached patch regtested on x86_64-pc-linux-gnu. OK for mainline?
Thanks,
Harald
From c29eb3dbe8c541ef83d5fdf12cafa015ed9447ee Mon Sep 17 00:00:00 2001
From: Harald Anlauf
Date: Mon, 6 Feb 2023 20:59:51 +0100
Subject: [PATCH] Fortran: ASSOCIATE variables should not be TREE_STATIC
[PR95107]
gcc/fortran/ChangeLog:
PR fortran/95107
* trans-decl.cc (gfc_finish_var_decl): With -fno-automatic, do not
make ASSOCIATE variables TREE_STATIC.
gcc/testsuite/ChangeLog:
PR fortran/95107
* gfortran.dg/save_7.f90: New test.
---
gcc/fortran/trans-decl.cc| 1 +
gcc/testsuite/gfortran.dg/save_7.f90 | 17 +
2 files changed, 18 insertions(+)
create mode 100644 gcc/testsuite/gfortran.dg/save_7.f90
diff --git a/gcc/fortran/trans-decl.cc b/gcc/fortran/trans-decl.cc
index 7a077801919..f7a7ff607cd 100644
--- a/gcc/fortran/trans-decl.cc
+++ b/gcc/fortran/trans-decl.cc
@@ -742,6 +742,7 @@ gfc_finish_var_decl (tree decl, gfc_symbol * sym)
/* Keep variables larger than max-stack-var-size off stack. */
if (!(sym->ns->proc_name && sym->ns->proc_name->attr.recursive)
&& !sym->attr.automatic
+ && !sym->attr.associate_var
&& sym->attr.save != SAVE_EXPLICIT
&& sym->attr.save != SAVE_IMPLICIT
&& INTEGER_CST_P (DECL_SIZE_UNIT (decl))
diff --git a/gcc/testsuite/gfortran.dg/save_7.f90 b/gcc/testsuite/gfortran.dg/save_7.f90
new file mode 100644
index 000..dfbef12bec0
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/save_7.f90
@@ -0,0 +1,17 @@
+! { dg-do compile }
+! { dg-options "-O2 -fno-automatic" }
+!
+! PR fortran/95107 - do not make associate variables TREE_STATIC
+! Contributed by G.Steinmetz
+
+program p
+ type t
+ real, pointer :: a => null()
+ end type
+ type t2
+ type(t) :: b(1)
+ end type
+ type(t2), save :: x
+ associate (y => x%b)
+ end associate
+end
--
2.35.3
