Bug 27336 - [dwarf-5] abort on info addr in gdb.cp/temargs.exp
Summary: [dwarf-5] abort on info addr in gdb.cp/temargs.exp
Status: RESOLVED FIXED
Alias: None
Product: gdb
Classification: Unclassified
Component: symtab (show other bugs)
Version: HEAD
: P2 normal
Target Milestone: 11.1
Assignee: Not yet assigned to anyone
URL:
Keywords:
Depends on:
Blocks: 27453
  Show dependency treegraph
 
Reported: 2021-02-03 14:17 UTC by Tom de Vries
Modified: 2021-02-24 22:59 UTC (History)
1 user (show)

See Also:
Host:
Target:
Build:
Last reconfirmed:
Project(s) to access:
ssh public key:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Tom de Vries 2021-02-03 14:17:06 UTC
With test-case gdb.cp/temargs.exp on target board unix/gdb:debug_flags=-gdwarf-5 I run into:
...
(gdb) PASS: gdb.cp/temargs.exp: continue to breakpoint: continue to fourth breakpoint for temargs
ptype T^M
type = double^M
(gdb) PASS: gdb.cp/temargs.exp: test type of T in templ_m
print I^M
$10 = 23^M
(gdb) PASS: gdb.cp/temargs.exp: test value of I in templ_m
info addr I^M
ERROR: GDB process no longer exists
GDB process exited with wait status 32286 exp19 0 0 CHILDKILLED SIGABRT SIGABRT
UNRESOLVED: gdb.cp/temargs.exp: test address of I in templ_m
...
Comment 1 Tom de Vries 2021-02-03 15:28:34 UTC
Bisects to commit 529908cbd0a "Remove DW_UNSND".
Comment 2 Tom de Vries 2021-02-03 16:51:08 UTC
So, we got a DW_AT_decl_file:
...
 <1><221>: Abbrev Number: 4 (DW_TAG_structure_type)
    <222>   DW_AT_name        : (indirect string, offset: 0x2e7): Base<double, 23, (& a_global), &S::f>
    <226>   DW_AT_byte_size   : 1
    <226>   DW_AT_decl_file   : 1
    <226>   DW_AT_decl_line   : 30
    <227>   DW_AT_sibling     : <0x299>
...
with form DW_FORM_implicit_const:
...
   4      DW_TAG_structure_type    [has children]
    DW_AT_name         DW_FORM_strp
    DW_AT_byte_size    DW_FORM_implicit_const: 1
    DW_AT_decl_file    DW_FORM_implicit_const: 1
    DW_AT_decl_line    DW_FORM_data1
    DW_AT_sibling      DW_FORM_ref4
    DW_AT value: 0     DW_FORM value: 0
...
which fails here:
...
      attr = dwarf2_attr (die,
                          inlined_func ? DW_AT_call_file : DW_AT_decl_file,
                          cu);
      if (attr != nullptr && attr->form_is_unsigned ())
...
because DW_FORM_implicit_const encodes a signed LEB128, so attr->form_is_unsigned () return false.
Comment 3 Tom de Vries 2021-02-03 18:18:12 UTC
Tentative fix:
...
diff --git a/gdb/dwarf2/attribute.h b/gdb/dwarf2/attribute.h
index aded129764a..baf93903d19 100644
--- a/gdb/dwarf2/attribute.h
+++ b/gdb/dwarf2/attribute.h
@@ -70,7 +70,7 @@ struct attribute
      form.  */
   LONGEST as_signed () const
   {
-    gdb_assert (form == DW_FORM_sdata || form == DW_FORM_implicit_const);
+    gdb_assert (form_is_signed ());
     return u.snd;
   }
 
@@ -92,6 +92,28 @@ struct attribute
     return u.unsnd;
   }
 
+  /* Return true if the value is nonnegative.  Requires that that
+     reprocessing not be needed.  */
+  bool is_nonnegative () const
+  {
+    if (form_is_unsigned ())
+      return true;
+    if (form_is_signed ())
+      return as_signed () >= (LONGEST)0;
+    return false;
+  }
+
+  /* Return the nonnegative value.  Requires that that reprocessing not be
+     needed.  */
+  ULONGEST as_nonnegative () const
+  {
+    if (form_is_unsigned ())
+      return as_unsigned ();
+    if (form_is_signed ())
+      return (ULONGEST)as_signed ();
+    gdb_assert (false);
+  }
+
   /* Return non-zero if ATTR's value is a section offset --- classes
      lineptr, loclistptr, macptr or rangelistptr --- or zero, otherwise.
      You may use the as_unsigned method to retrieve such offsets.
@@ -147,6 +169,9 @@ struct attribute
   /* Check if the attribute's form is an unsigned integer form.  */
   bool form_is_unsigned () const;
 
+  /* Check if the attribute's form is a signed integer form.  */
+  bool form_is_signed () const;
+
   /* Check if the attribute's form is a form that requires
      "reprocessing".  */
   bool form_requires_reprocessing () const;
diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index bcac94ed36a..35e6ba0044a 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -21427,10 +21427,10 @@ new_symbol (struct die_info *die, struct type *type, struct 
dwarf2_cu *cu,
       attr = dwarf2_attr (die,
                          inlined_func ? DW_AT_call_file : DW_AT_decl_file,
                          cu);
-      if (attr != nullptr && attr->form_is_unsigned ())
+      if (attr != nullptr && attr->is_nonnegative ())
        {
          file_name_index file_index
-           = (file_name_index) attr->as_unsigned ();
+           = (file_name_index) attr->as_nonnegative ();
          struct file_entry *fe;
 
          if (cu->line_header != NULL)
...
Comment 4 Tom de Vries 2021-02-04 10:08:44 UTC
Conservative fix:
...
diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index 68fc47fb028..932eb7bb11d 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -17243,6 +17243,9 @@ read_array_type (struct die_info *die, struct dwarf2_cu *cu)
        }
     }
 
+  if (type == element_type)
+    return NULL;
+
   /* Understand Dwarf2 support for vector types (like they occur on
      the PowerPC w/ AltiVec).  Gcc just adds another attribute to the
      array type.  This is not part of the Dwarf2/3 standard yet, but a
...
Comment 6 Sourceware Commits 2021-02-24 22:58:46 UTC
The master branch has been updated by Tom de Vries <vries@sourceware.org>:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=aa659cfad638376ab535bc5a7b65c164327e9e21

commit aa659cfad638376ab535bc5a7b65c164327e9e21
Author: Tom de Vries <tdevries@suse.de>
Date:   Wed Feb 24 23:58:42 2021 +0100

    [gdb/symtab] Handle DW_AT_decl_file with form DW_FORM_implicit_const
    
    With test-case gdb.cp/temargs.exp on target board \
    unix/gdb:debug_flags=-gdwarf-5 I run into:
    ...
    (gdb) info addr I^M
    ERROR: GDB process no longer exists
    GDB process exited with wait status 32286 exp19 0 0 CHILDKILLED SIGABRT SIGABRT
    UNRESOLVED: gdb.cp/temargs.exp: test address of I in templ_m
    ...
    
    This is a regression since commit 529908cbd0a "Remove DW_UNSND".
    
    The problem is that this DW_AT_decl_file:
    ...
     <1><221>: Abbrev Number: 4 (DW_TAG_structure_type)
        <222>   DW_AT_name        : Base<double, 23, (& a_global), &S::f>
        <226>   DW_AT_byte_size   : 1
        <226>   DW_AT_decl_file   : 1
        <226>   DW_AT_decl_line   : 30
        <227>   DW_AT_sibling     : <0x299>
    ...
    is not read by this code in new_symbol:
    ....
          attr = dwarf2_attr (die,
                              inlined_func ? DW_AT_call_file : DW_AT_decl_file,
                              cu);
          if (attr != nullptr && attr->form_is_unsigned ())
    ...
    because DW_AT_decl_file has form DW_FORM_implicit_const:
    ...
       4      DW_TAG_structure_type    [has children]
        DW_AT_name         DW_FORM_strp
        DW_AT_byte_size    DW_FORM_implicit_const: 1
        DW_AT_decl_file    DW_FORM_implicit_const: 1
        DW_AT_decl_line    DW_FORM_data1
        DW_AT_sibling      DW_FORM_ref4
        DW_AT value: 0     DW_FORM value: 0
    ...
    which is a signed LEB128, so attr->form_is_unsigned () returns false.
    
    Fix this by introducing new functions is_nonnegative and as_nonnegative, and
    use these instead of form_is_unsigned and as_unsigned.
    
    Tested on x86_64-linux.
    
    gdb/ChangeLog:
    
    2021-02-24  Tom de Vries  <tdevries@suse.de>
    
            PR symtab/27336
            * dwarf2/attribute.c (attribute::form_is_signed): New function
            factored out of ...
            * dwarf2/attribute.h (attribute::as_signed): ... here.
            (attribute::is_nonnegative, attribute::as_nonnegative): New function.
            (attribute::form_is_signed): Declare.
            * dwarf2/read.c (new_symbol): Use is_nonnegative and as_nonnegative
            for DW_AT_decl_file.
Comment 7 Tom de Vries 2021-02-24 22:59:38 UTC
Committed patch, marking resolved-fixed.