[fortran] PR108434 - [12/13/14/15 Regression] ICE in class_allocatable, at fortran/expr.cc:5000

  Hi Harald, hi all,

As of today, Gerhard Steinmetz has no fewer than 33 regressions to his name
out of a total of 54 for fortran and libgfortran. It's time that some of
these bugs are swatted, I think :-)

As well as this PR, 106946 seems to have fixed itself and I have fixes for
102333 and 96087 waiting for this one to be OK'd.

The attached is the patch posted on the PR, modified for a translator
friendly error message as requested by Harald.

Regtests OK. OK for trunk and backporting to 13- and 14-branches?

Paul
  

On 1/10/25 9:19 AM, Paul Richard Thomas wrote:
> Hi Harald, hi all,
> 
> As of today, Gerhard Steinmetz has no fewer than 33 regressions to his 
> name out of a total of 54 for fortran and libgfortran. It's time that 
> some of these bugs are swatted, I think :-)
> 
> As well as this PR, 106946 seems to have fixed itself and I have fixes 
> for 102333 and 96087 waiting for this one to be OK'd.
> 
> The attached is the patch posted on the PR, modified for a translator 
> friendly error message as requested by Harald.
> 
> Regtests OK. OK for trunk and backporting to 13- and 14-branches?
> 
> Paul
> 
> 

OK Paul, thanks for fix.

Jerry
  

On Fri, Jan 10, 2025 at 05:19:34PM +0000, Paul Richard Thomas wrote:
> 
> As of today, Gerhard Steinmetz has no fewer than 33 regressions to his name
> out of a total of 54 for fortran and libgfortran. It's time that some of
> these bugs are swatted, I think :-)
> 

When I was much more active in all things gfortran,
I found Gerhard's bug reports to be some of the best.
Small, self-contained code snippets made debugging
easiers.

Seems PR 77415 is fixed.  I've converted the codes
in the PR to testcases.

diff --git a/gcc/testsuite/gfortran.dg/pr77415_1.f90 b/gcc/testsuite/gfortran.dg/pr77415_1.f90
new file mode 100644
index 00000000000..00488009044
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr77415_1.f90
@@ -0,0 +1,6 @@
+integer function f()
+   f(g) = 0       ! { dg-error "Statement function" }
+contains
+   integer function g()
+   end
+end
diff --git a/gcc/testsuite/gfortran.dg/pr77415_2.f90 b/gcc/testsuite/gfortran.dg/pr77415_2.f90
new file mode 100644
index 00000000000..858c8223b84
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr77415_2.f90
@@ -0,0 +1,8 @@
+! { dg-do compile }
+function f()
+   f(g) = 0       ! { dg-error "Statement function" }
+contains
+   function g()
+   end
+end
+
  

On Fri, Jan 10, 2025 at 05:19:34PM +0000, Paul Richard Thomas wrote:
> 
> As of today, Gerhard Steinmetz has no fewer than 33 regressions to his name
> out of a total of 54 for fortran and libgfortran. It's time that some of
> these bugs are swatted, I think :-)
> 

This patch fixes PR71844.  As the error message indicates,
the source-expr in 'allocate(x, source=null())' cannot
be null().

diff --git a/gcc/fortran/resolve.cc b/gcc/fortran/resolve.cc
index dab0c3af601..538917fe56a 100644
--- a/gcc/fortran/resolve.cc
+++ b/gcc/fortran/resolve.cc
@@ -8965,6 +8965,13 @@ resolve_allocate_expr (gfc_expr *e, gfc_code *code, bool *array_alloc_wo_spec)
   gfc_component *c;
   bool t;
 
+  /* source-expr in either SOURCE= or MODE= cannot be NULL().  */
+  if (code->expr3 && code->expr3->expr_type == EXPR_NULL)
+    {
+      gfc_error ("Source-expr at %L cannot be NULL()", &code->expr3->where);
+      goto failure;
+    }
+
   /* Mark the utmost array component as being in allocate to allow DIMEN_STAR
      checking of coarrays.  */
   for (ref = e->ref; ref; ref = ref->next)
diff --git a/gcc/testsuite/gfortran.dg/pr71844.f90 b/gcc/testsuite/gfortran.dg/pr71844.f90
new file mode 100644
index 00000000000..af990f32fbb
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr71844.f90
@@ -0,0 +1,10 @@
+!
+! { dg-do compile }
+!
+program p
+   class(*), allocatable :: x, y
+   character(:), allocatable :: z
+   allocate (x, source=null())   ! { dg-error "cannot be NULL" }
+   allocate (y, mold=null())     ! { dg-error "cannot be NULL" }
+   allocate (character(*) :: z)  ! { dg-error "Incompatible allocate-object" }
+end
  

On Fri, Jan 10, 2025 at 05:19:34PM +0000, Paul Richard Thomas wrote:
> 
> As of today, Gerhard Steinmetz has no fewer than 33 regressions to his name
> out of a total of 54 for fortran and libgfortran. It's time that some of
> these bugs are swatted, I think :-)
> 

PR 70949 appears to have been fixed at some point
in the past.  The following patch converts Gerhard's
code into testcases.

diff --git a/gcc/testsuite/gfortran.dg/pr70949_1.f90 b/gcc/testsuite/gfortran.dg/pr70949_1.f90
new file mode 100644
index 00000000000..91cd18069fc
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr70949_1.f90
@@ -0,0 +1,27 @@
+!
+! { dg-do run}
+!
+program p
+
+   type t1
+   end type
+
+   type t2
+      type(t1), pointer :: q
+   end type
+
+   type(t1), pointer :: a
+   type(t2) :: c
+
+   allocate(a)
+   c%q => a
+   if (.not. associated(a, f(c))) stop 1
+
+   contains
+
+      function f(x) result (z)
+         type(t2), intent(in) :: x
+         class(t1), pointer :: z
+         z => x%q
+      end function f
+end
diff --git a/gcc/testsuite/gfortran.dg/pr70949_2.f90 b/gcc/testsuite/gfortran.dg/pr70949_2.f90
new file mode 100644
index 00000000000..eb064b6fa80
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/pr70949_2.f90
@@ -0,0 +1,27 @@
+!
+! { dg-do run}
+!
+program p
+
+   type t1
+   end type
+
+   type t2
+      type(t1), pointer :: q
+   end type
+
+   type(t1), pointer :: a
+   type(t2) :: c
+
+   allocate(a)
+   c%q => a
+   if (.not. associated(a, f(c))) stop 1
+
+   contains
+
+      function f(x) result (z)
+         type(t2), intent(in) :: x
+         type(t1), pointer :: z
+         z => x%q
+      end function f
+end