[PATCH][gdb/symtab] Handle DW_AT_string_length with location list

[PATCH][gdb/symtab] Handle DW_AT_string_length with location list

[Tom de Vries]

Hi,

Consider a fortran routine where a string variable s is modified:
...
subroutine f(s)
  character*(*) s
  print *, s
  s(1:3) = 'oof'
  print *, s
end subroutine f
...

When compiling with optimization level -O1 and printing the type of
variable s we get:
...
$ gdb -q -batch outputs/gdb.opt/fortran-string/fortran-string \
  -ex "b f" \
  -ex run \
  -ex "ptype s"
Breakpoint 1 at 0x4006f7: file fortran-string.f90, line 21.

Breakpoint 1, f (s=..., _s=_s@entry=3) at fortran-string.f90:21
21      subroutine f(s)
type = character*1
...
while with -O0 we have instead:
...
type = character (3)
...

The problem is that the type of s is:
...
 <1><2d6>: Abbrev Number: 21 (DW_TAG_string_type)
    <2d7>   DW_AT_string_length: 0xbf (location list)
    <2db>   DW_AT_byte_size   : 4
...
where the DW_AT_string_length is a location list, a case that is not handled
by attr_to_dynamic_prop.

Fix this by handling attr->form_is_section_offset () in attr_to_dynamic_prop.

Tested on x86_64-linux.

The test-case is based on gdb.opt/fortran-string.exp from
https://src.fedoraproject.org/rpms/gdb/raw/f32/f/gdb-archer-vla-tests.patch .
I've updated the copyrights to stretch to 2021.

[ I've tried to create a dwarf assembly test-case for this, but didn't
manage. ]

Co-Authored-By: Jan Kratochvil <jan.kratochvil@redhat.com>

Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=26910

Any comments?

Thanks,
- Tom

[gdb/symtab] Handle DW_AT_string_length with location list

---
 gdb/dwarf2/read.c                        | 24 +++++++++++++++----
 gdb/testsuite/gdb.opt/fortran-string.exp | 41 ++++++++++++++++++++++++++++++++
 gdb/testsuite/gdb.opt/fortran-string.f90 | 32 +++++++++++++++++++++++++
 3 files changed, 93 insertions(+), 4 deletions(-)

diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index cbd9a3012eb..0477a689d54 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -18468,14 +18468,30 @@ attr_to_dynamic_prop (const struct attribute *attr, struct die_info *die,
     }
   else if (attr->form_is_constant ())
     prop->set_const_val (attr->constant_value (0));
-  else
+  else if (attr->form_is_section_offset ())
     {
-      dwarf2_invalid_attrib_class_complaint (dwarf_form_name (attr->form),
-					     dwarf2_name (die, cu));
-      return 0;
+      switch (attr->name)
+	{
+	case DW_AT_string_length:
+	  baton = XOBNEW (obstack, struct dwarf2_property_baton);
+	  baton->property_type = default_type;
+	  fill_in_loclist_baton (cu, &baton->loclist, attr);
+	  prop->set_loclist (baton);
+	  gdb_assert (prop->baton () != NULL);
+	  break;
+	default:
+	  goto invalid;
+	}
     }
+  else
+    goto invalid;
 
   return 1;
+
+ invalid:
+  dwarf2_invalid_attrib_class_complaint (dwarf_form_name (attr->form),
+					 dwarf2_name (die, cu));
+  return 0;
 }
 
 /* See read.h.  */
diff --git a/gdb/testsuite/gdb.opt/fortran-string.exp b/gdb/testsuite/gdb.opt/fortran-string.exp
new file mode 100644
index 00000000000..62de48cf9d7
--- /dev/null
+++ b/gdb/testsuite/gdb.opt/fortran-string.exp
@@ -0,0 +1,41 @@
+# Copyright 2009-2021 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+
+# This file was written by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+# Test GDB can cope with Fortran strings having their length present in a CPU
+# register.  With -O0 the string length is passed on the stack.  To make this
+# test meaningful the follow assertion should pass.  It is not being checked
+# here as the "_s" symbol is compiler dependent:
+#   (gdb) info address _s
+#   Symbol "_s" is a variable in register XX.
+
+standard_testfile .f90
+
+if { [prepare_for_testing "prepare for testing" ${testfile} ${srcfile} \
+	  {debug f90 additional_flags=-O1}] } {
+    return -1
+}
+
+if ![runto f] then {
+    perror "couldn't run to f"
+    continue
+}
+
+gdb_test_no_output "set print frame-arguments all"
+gdb_test "frame" ".*s='foo'.*"
+gdb_test "ptype s" "type = character \\(3\\)"
+gdb_test "p s" "\\$\[0-9\]* = 'foo'"
diff --git a/gdb/testsuite/gdb.opt/fortran-string.f90 b/gdb/testsuite/gdb.opt/fortran-string.f90
new file mode 100644
index 00000000000..0a316b2dc79
--- /dev/null
+++ b/gdb/testsuite/gdb.opt/fortran-string.f90
@@ -0,0 +1,32 @@
+! Copyright 2009-2021 Free Software Foundation, Inc.
+!
+! This program is free software; you can redistribute it and/or modify
+! it under the terms of the GNU General Public License as published by
+! the Free Software Foundation; either version 2 of the License, or
+! (at your option) any later version.
+!
+! This program is distributed in the hope that it will be useful,
+! but WITHOUT ANY WARRANTY; without even the implied warranty of
+! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+! GNU General Public License for more details.
+!
+! You should have received a copy of the GNU General Public License
+! along with this program; if not, write to the Free Software
+! Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+!
+! Ihis file is the Fortran source file for dynamic.exp.
+! Original file written by Jakub Jelinek <jakub@redhat.com>.
+! Modified for the GDB testcase by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+subroutine f(s)
+  character*(*) s
+  print *, s
+  s(1:3) = 'oof'
+  print *, s
+end subroutine f
+
+program main
+  character*3 s;
+  s = 'foo'
+  call f (s)
+end program main

Re: [PATCH][gdb/symtab] Handle DW_AT_string_length with location list

[Simon Marchi]

On 2021-10-14 3:58 p.m., Tom de Vries via Gdb-patches wrote:
> Hi,
>
> Consider a fortran routine where a string variable s is modified:
> ...
> subroutine f(s)
>   character*(*) s
>   print *, s
>   s(1:3) = 'oof'
>   print *, s
> end subroutine f
> ...
>
> When compiling with optimization level -O1 and printing the type of
> variable s we get:
> ...
> $ gdb -q -batch outputs/gdb.opt/fortran-string/fortran-string \
>   -ex "b f" \
>   -ex run \
>   -ex "ptype s"
> Breakpoint 1 at 0x4006f7: file fortran-string.f90, line 21.
>
> Breakpoint 1, f (s=..., _s=_s@entry=3) at fortran-string.f90:21
> 21      subroutine f(s)
> type = character*1

I'm not familiar with fortran, does this notation ("character*1") have a
meaning in fortran?

Again, I don't know anything about fortran, but the change in dwarf2/read.c
looks good to me.

Simon

Re: [PATCH][gdb/symtab] Handle DW_AT_string_length with location list

[Tom de Vries]

On 10/15/21 5:22 PM, Simon Marchi wrote:
> On 2021-10-14 3:58 p.m., Tom de Vries via Gdb-patches wrote:
>> Hi,
>>
>> Consider a fortran routine where a string variable s is modified:
>> ...
>> subroutine f(s)
>>   character*(*) s
>>   print *, s
>>   s(1:3) = 'oof'
>>   print *, s
>> end subroutine f
>> ...
>>
>> When compiling with optimization level -O1 and printing the type of
>> variable s we get:
>> ...
>> $ gdb -q -batch outputs/gdb.opt/fortran-string/fortran-string \
>>   -ex "b f" \
>>   -ex run \
>>   -ex "ptype s"
>> Breakpoint 1 at 0x4006f7: file fortran-string.f90, line 21.
>>
>> Breakpoint 1, f (s=..., _s=_s@entry=3) at fortran-string.f90:21
>> 21      subroutine f(s)
>> type = character*1
> 
> I'm not familiar with fortran, does this notation ("character*1") have a
> meaning in fortran?
> 

I'm not very familiar with fortran either, but AFAIU the meaning is:
character string of length one.

> Again, I don't know anything about fortran, but the change in dwarf2/read.c
> looks good to me.

Thanks for the review.

- Tom

Re: [PATCH][gdb/symtab] Handle DW_AT_string_length with location list

[Tom de Vries]

On 10/15/21 5:22 PM, Simon Marchi wrote:
> On 2021-10-14 3:58 p.m., Tom de Vries via Gdb-patches wrote:
>> Hi,
>>
>> Consider a fortran routine where a string variable s is modified:
>> ...
>> subroutine f(s)
>>   character*(*) s
>>   print *, s
>>   s(1:3) = 'oof'
>>   print *, s
>> end subroutine f
>> ...
>>
>> When compiling with optimization level -O1 and printing the type of
>> variable s we get:
>> ...
>> $ gdb -q -batch outputs/gdb.opt/fortran-string/fortran-string \
>>   -ex "b f" \
>>   -ex run \
>>   -ex "ptype s"
>> Breakpoint 1 at 0x4006f7: file fortran-string.f90, line 21.
>>
>> Breakpoint 1, f (s=..., _s=_s@entry=3) at fortran-string.f90:21
>> 21      subroutine f(s)
>> type = character*1
> 
> I'm not familiar with fortran, does this notation ("character*1") have a
> meaning in fortran?
> 
> Again, I don't know anything about fortran, but the change in dwarf2/read.c
> looks good to me.

I've reviewed the test-case once more, and did the following updates:
- removed the "contributed by" lines in the test-case, since we no
  longer do that.
  ( https://sourceware.org/gdb/wiki/ContributionChecklist#Attribution )
- updated the copyright license version to gplv3, and replaced the
  snailmail pointer to copy of license with the url.

Committed with those changes (too trivial to repost).

Thanks,
- Tom