Fortran : ICE in gfc_validate_kind PR95586

Message ID 846da5e1-6924-0fa3-03b9-b9472d502de9@codethink.co.uk
State New
Headers show
Series
  • Fortran : ICE in gfc_validate_kind PR95586
Related show

Commit Message

Mark Eggleston June 17, 2020, 6:36 a.m.
Please find attached fix for PR95586.

OK to commit and backport?

Proposed commit message:

Fortran  :  ICE in gfc_validate_kind PR95586

Report syntax error for invalid letter-spec in IMPLICIT statements
for derived types and not an ICE.

Original patch by Steve Kargl.  Added test cases based on those
provided by G. Steinmetz  in the PR.

2020-06-17  Steven G. Kargl  <kargl@gcc.gnu.org>

gcc/fortran/

     PR fortran/95586
     * dec.l (gfc_match_implicit): Only perform else branch if
     the type spect is not BT_DERIVED.

2020-06-17  Mark Eggleston <markeggleston@gcc.gnu.org>

gcc/testsuite/

     PR fortran/95586
     * gfortran.dg/pr95586_1.f90: New test.
     * gfortran.dg/pr95586_2.f90: New test.

-- 
https://www.codethink.co.uk/privacy.html

Comments

Uros Bizjak via Gcc-patches June 20, 2020, 10:05 a.m. | #1
Hi Mark,

> OK to commit


OK for master.

Since, according to the PR, this is not a regression and this
is an ICE on invalid, I don't think a backport is needed.  On
the other hand, the patch is simple enough that I am quite
certain that it will not hurt.

So, I'll leave that up to you.

Best regards

	Thomas

Patch

From aefef0f66763aa436ce30f046cccf1ac41285954 Mon Sep 17 00:00:00 2001
From: Mark Eggleston <markeggleston@gcc.gnu.org>
Date: Wed, 10 Jun 2020 16:18:23 +0100
Subject: [PATCH] Fortran  :  ICE in gfc_validate_kind PR95586

Report syntax error for invalid letter-spec in IMPLICIT statements
for derived types and not an ICE.

Original patch by Steve Kargl.  Added test cases based on those
provided by G. Steinmetz  in the PR.

2020-06-17  Steven G. Kargl  <kargl@gcc.gnu.org>

gcc/fortran/

	PR fortran/95586
	* dec.l (gfc_match_implicit): Only perform else branch if
	the type spect is not BT_DERIVED.

2020-06-17  Mark Eggleston  <markeggleston@gcc.gnu.org>

gcc/testsuite/

	PR fortran/95586
	* gfortran.dg/pr95586_1.f90: New test.
	* gfortran.dg/pr95586_2.f90: New test.
---
 gcc/fortran/decl.c                      |  2 +-
 gcc/testsuite/gfortran.dg/pr95586_1.f90 |  9 +++++++++
 gcc/testsuite/gfortran.dg/pr95586_2.f90 | 11 +++++++++++
 3 files changed, 21 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/gfortran.dg/pr95586_1.f90
 create mode 100644 gcc/testsuite/gfortran.dg/pr95586_2.f90

diff --git a/gcc/fortran/decl.c b/gcc/fortran/decl.c
index c8a98537e87..c27cfacf2e4 100644
--- a/gcc/fortran/decl.c
+++ b/gcc/fortran/decl.c
@@ -4827,7 +4827,7 @@  gfc_match_implicit (void)
       /* Last chance -- check <TYPE> <SELECTOR> (<RANGE>).  */
       if (ts.type == BT_CHARACTER)
 	m = gfc_match_char_spec (&ts);
-      else
+      else if (ts.type != BT_DERIVED)
 	{
 	  m = gfc_match_kind_spec (&ts, false);
 	  if (m == MATCH_NO)
diff --git a/gcc/testsuite/gfortran.dg/pr95586_1.f90 b/gcc/testsuite/gfortran.dg/pr95586_1.f90
new file mode 100644
index 00000000000..009eb3e88f0
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr95586_1.f90
@@ -0,0 +1,9 @@ 
+! { dg-do compile }
+!
+
+program test
+  implicit type(t) (1)  ! { dg-error "Syntax error" }
+  type t
+  end type
+end program
+
diff --git a/gcc/testsuite/gfortran.dg/pr95586_2.f90 b/gcc/testsuite/gfortran.dg/pr95586_2.f90
new file mode 100644
index 00000000000..81ea6e9fa56
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr95586_2.f90
@@ -0,0 +1,11 @@ 
+! { dg-do compile }
+!
+
+program test
+  integer, parameter :: n1 = 1
+  implicit type(t) (n1)  ! { dg-error "Syntax error" }
+  type t
+  end type
+end program
+
+
-- 
2.11.0