[RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[Tom de Vries]

Hi,

this patch (without test-case for now) generates a warning if
DW_AT_upper_bound or DW_AT_count is defined, but can't be translated.  This
is triggered for current gcc in lto mode for vla test-cases.

F.i.:
...
$ gcc gcc/testsuite/gcc.dg/guality/vla-1.c -g -O2 -flto \
    -DPREVENT_OPTIMIZATION -o ./vla-1.exe
$ ./gdb -batch -ex "set complaints 10" -ex "file ./vla-1.exe"
During symbol reading, Unresolved DW_AT_upper_bound - DIE at 0x337
  [in module vla-1.exe].
...

Good idea?

Any other comments?

Thanks,
- Tom

[gdb/symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

2018-07-18  Tom de Vries  <tdevries@suse.de>

	* dwarf2read.c (read_subrange_type): Warn if DW_AT_upper_bound or
	DW_AT_count can't be translated to a dynamic prop.

---
 gdb/dwarf2read.c | 19 +++++++++++++++++--
 1 file changed, 17 insertions(+), 2 deletions(-)

diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index b7933de49c..3f696ea060 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -17596,10 +17596,11 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
 	       sect_offset_str (die->sect_off),
 	       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
 
-  attr = dwarf2_attr (die, DW_AT_upper_bound, cu);
+  struct attribute *attr_ub, *attr_count;
+  attr = attr_ub = dwarf2_attr (die, DW_AT_upper_bound, cu);
   if (!attr_to_dynamic_prop (attr, die, cu, &high))
     {
-      attr = dwarf2_attr (die, DW_AT_count, cu);
+      attr = attr_count = dwarf2_attr (die, DW_AT_count, cu);
       if (attr_to_dynamic_prop (attr, die, cu, &high))
 	{
 	  /* If bounds are constant do the final calculation here.  */
@@ -17608,6 +17609,20 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
 	  else
 	    high_bound_is_count = 1;
 	}
+      else
+	{
+	  if (attr_ub)
+	    complaint (_("Unresolved DW_AT_upper_bound "
+			 "- DIE at %s [in module %s]"),
+		       sect_offset_str (die->sect_off),
+		       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
+	  if (attr_count)
+	    complaint (_("Unresolved DW_AT_count "
+			 "- DIE at %s [in module %s]"),
+		       sect_offset_str (die->sect_off),
+		       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
+	}
+	
     }
 
   /* Dwarf-2 specifications explicitly allows to create subrange types

Re: [RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[Tom Tromey]

>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:

Tom> this patch (without test-case for now) generates a warning if
Tom> DW_AT_upper_bound or DW_AT_count is defined, but can't be translated.  This
Tom> is triggered for current gcc in lto mode for vla test-cases.

Tom> F.i.:
Tom> ...
Tom> $ gcc gcc/testsuite/gcc.dg/guality/vla-1.c -g -O2 -flto \
Tom>     -DPREVENT_OPTIMIZATION -o ./vla-1.exe
Tom> $ ./gdb -batch -ex "set complaints 10" -ex "file ./vla-1.exe"
Tom> During symbol reading, Unresolved DW_AT_upper_bound - DIE at 0x337
Tom>   [in module vla-1.exe].
Tom> ...

Tom> Good idea?

I think so.

Tom> +	  if (attr_ub)
[...]
Tom> +	  if (attr_count)

The gdb style is to use a check against nullptr in these spots.

Tom

Re: [RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[Tom de Vries]

On Thu, Jul 19, 2018 at 08:40:20AM -0600, Tom Tromey wrote:
> >>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
> 
> Tom> this patch (without test-case for now) generates a warning if
> Tom> DW_AT_upper_bound or DW_AT_count is defined, but can't be translated.  This
> Tom> is triggered for current gcc in lto mode for vla test-cases.
> 
> Tom> F.i.:
> Tom> ...
> Tom> $ gcc gcc/testsuite/gcc.dg/guality/vla-1.c -g -O2 -flto \
> Tom>     -DPREVENT_OPTIMIZATION -o ./vla-1.exe
> Tom> $ ./gdb -batch -ex "set complaints 10" -ex "file ./vla-1.exe"
> Tom> During symbol reading, Unresolved DW_AT_upper_bound - DIE at 0x337
> Tom>   [in module vla-1.exe].
> Tom> ...
> 
> Tom> Good idea?
> 
> I think so.
> 
> Tom> +	  if (attr_ub)
> [...]
> Tom> +	  if (attr_count)
> 
> The gdb style is to use a check against nullptr in these spots.
> 

Updated accordingly, and added test-case.

OK for trunk?

Thanks,
- Tom

[gdb/symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

This patch generates a warning if DW_AT_upper_bound or DW_AT_count is defined,
but can't be translated.  This is triggered for current gcc in lto mode for
vla test-cases.

Build and reg-tested on x86_64-linux.

2018-07-18  Tom de Vries  <tdevries@suse.de>

	* dwarf2read.c (read_subrange_type): Warn if DW_AT_upper_bound or
	DW_AT_count can't be translated to a dynamic prop.

	* gdb.dwarf2/dw2-unresolved-ub.exp: new file.

---
 gdb/dwarf2read.c                               | 19 +++++++++--
 gdb/testsuite/gdb.dwarf2/dw2-unresolved-ub.exp | 47 ++++++++++++++++++++++++++
 2 files changed, 64 insertions(+), 2 deletions(-)

diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index b7933de49c..92107c0900 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -17596,10 +17596,11 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
 	       sect_offset_str (die->sect_off),
 	       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
 
-  attr = dwarf2_attr (die, DW_AT_upper_bound, cu);
+  struct attribute *attr_ub, *attr_count;
+  attr = attr_ub = dwarf2_attr (die, DW_AT_upper_bound, cu);
   if (!attr_to_dynamic_prop (attr, die, cu, &high))
     {
-      attr = dwarf2_attr (die, DW_AT_count, cu);
+      attr = attr_count = dwarf2_attr (die, DW_AT_count, cu);
       if (attr_to_dynamic_prop (attr, die, cu, &high))
 	{
 	  /* If bounds are constant do the final calculation here.  */
@@ -17608,6 +17609,20 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
 	  else
 	    high_bound_is_count = 1;
 	}
+      else
+	{
+	  if (attr_ub != NULL)
+	    complaint (_("Unresolved DW_AT_upper_bound "
+			 "- DIE at %s [in module %s]"),
+		       sect_offset_str (die->sect_off),
+		       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
+	  if (attr_count != NULL)
+	    complaint (_("Unresolved DW_AT_count "
+			 "- DIE at %s [in module %s]"),
+		       sect_offset_str (die->sect_off),
+		       objfile_name (cu->per_cu->dwarf2_per_objfile->objfile));
+	}
+	
     }
 
   /* Dwarf-2 specifications explicitly allows to create subrange types
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-unresolved-ub.exp b/gdb/testsuite/gdb.dwarf2/dw2-unresolved-ub.exp
new file mode 100644
index 0000000000..e8fcf0414e
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-unresolved-ub.exp
@@ -0,0 +1,47 @@
+# Copyright 2018 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 3 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, see <http://www.gnu.org/licenses/>.
+load_lib dwarf.exp
+
+# This test can only be run on targets which support DWARF-2 and use gas.
+if {![dwarf2_support]} {
+    return 0
+}
+
+if { ![test_compiler_info "gcc-9-*-*"] } {
+    return 0
+}
+
+standard_testfile
+
+if { [prepare_for_testing "failed to prepare" $testfile "${srcdir}/gdb.base/vla-optimized-out.c" \
+	  {debug optimize=-O2 additional_flags=-flto additional_flags=-gstrict-dwarf}] } {
+    return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+
+# From gdb_file_cmd:
+if [is_remote host] {
+    set arg [remote_download host $binfile]
+    if { $arg == "" } {
+	perror "download failed"
+	return -1
+    }
+}
+
+gdb_test_no_output "set complaints 3"
+gdb_test "file $binfile" "DW_AT_upper_bound.*" "Unresolved upper bound"

Re: [RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[Tom Tromey]

>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:

Tom> 2018-07-18  Tom de Vries  <tdevries@suse.de>

Tom> 	* dwarf2read.c (read_subrange_type): Warn if DW_AT_upper_bound or
Tom> 	DW_AT_count can't be translated to a dynamic prop.

This part is fine.

Tom> 	* gdb.dwarf2/dw2-unresolved-ub.exp: new file.

It's great to test the warning, but I think this approach is mostly
testing whether gcc has a particular bug.  The more future-proof way to
do this sort of test would be to use the DWARF assembler to craft a bit
of DWARF that has the problem.

Tom

Re: [RFC][PATCH][symtab] Warn about unresolved DW_AT_upper_bound/DW_AT_count

[Tom de Vries]

On 07/25/2018 08:55 PM, Tom Tromey wrote:
>>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
> 
> Tom> 2018-07-18  Tom de Vries  <tdevries@suse.de>
> 
> Tom> 	* dwarf2read.c (read_subrange_type): Warn if DW_AT_upper_bound or
> Tom> 	DW_AT_count can't be translated to a dynamic prop.
> 
> This part is fine.
> 
> Tom> 	* gdb.dwarf2/dw2-unresolved-ub.exp: new file.
> 
> It's great to test the warning, but I think this approach is mostly
> testing whether gcc has a particular bug.  The more future-proof way to
> do this sort of test would be to use the DWARF assembler to craft a bit
> of DWARF that has the problem.

I see, I'll keep that in mind. For now, I've committed without the
test-case.

Thanks,
- Tom