Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
* [PATCH] gdb: remove attempted type de-duplication when building gdb-index
@ 2025-10-02  9:36 Andrew Burgess
  2025-10-16 18:11 ` [PATCHv2] " Andrew Burgess
  0 siblings, 1 reply; 10+ messages in thread
From: Andrew Burgess @ 2025-10-02  9:36 UTC (permalink / raw)
  To: gdb-patches; +Cc: Andrew Burgess

This commit removes the attempted de-duplication of types when
building the gdb-index.  This commit is the natural extension of this
earlier commit:

  commit aef36dee93bf194cb0b976a4ae49a79a736a188d
  Date:   Sun Aug 13 14:08:06 2023 +0200

      [gdb/symtab] Don't deduplicate variables in gdb-index

Which removed the de-duplication of variables.  It is worth reading
the earlier commit as all the justifications for that patch also
apply to this one.

Currently, when building the gdb-index we sort the type entries,
moving declarations to the end of the entry list, and non-declarations
to the front, sorting non-declarations by CU offset.

We then emit the first entry for any given type name.

There are two problems with this.

First, a non-declaration entry could be a definition, but it could
also be a typedef.  Now sure, a typedef is a definition, but not
necessarily a useful one.

If we have a header file that contains:

  typedef struct foo_t foo_t;

And a CU which makes use of 'foo_t', then the CU will include both a
typedef and a type declaration.  The target of the typedef will be the
declaration.

If we have two CUs, one which only sees the above typedef and
declaration, and another which sees the typedef and an actual type
definition, then the final list of entries for this type's name will
be:

  1. A typedef entry that points at the declaration.
  2. A typedef entry that points at the definition.
  3. A definition.
  4. A declaration.

Now (4) will get sorted to the end of the entry list.  But the order
of (1), (2), and (3) will depend on the CU offset.  If the CU which
containing the typedef and declaration has the smallest offset,
then (1) will be sorted to the front of the list of entries for this
type name.  Due to the de-duplication code this means that only (1)
will be added to the gdb-index.

Now, after GDB starts and parses the index, if a user references
'foo_t' then GDB looks in the index and finds just (1).  GDB loads the
CU containing (1) and find both the typedef and the declaration.  But
GDB does not find the full definition.  As a result GDB will display
'foo_t' as an incomplete type.

This differs from the behaviour when no index is used.  With no index
GDB expands every CU containing the 'foo_t' type, which means that GDB
finds the type's definition, and prints the type correctly.

We could solve this problem by marking typedefs as a distinct
sub-category of types, just as we do with declarations.  Then we could
sort definitions to the front of the list, then typedefs, and finally,
declarations after that.  This would, I think, mean that we always
prefer emitting a definition for a type, which would resolve the first
problem.  But wouldn't, I think, solve the second problem.

The second problem is that the Python API can be used to query type
symbols.  As such, GDB needs to be able to find all the CUs which
contain a given type.  Especially as it is possible that a type might
be defined differently within different CUs.

NOTE: Obviously a program doing this would need to be mindful of the
One Definition Rule, but so long as the type doesn't escape outside of
a single CU then reusing a type name isn't, as I understand it,
wrong.  And even if it is, the fact that it compiles, and could be a
source of bugs, means (in my opinion) that GDB should handle this case
to enable debugging of it.

Given all of the above, I think that any attempt to remove type
entries from the gdb-index is unsafe and can result in GDB behaving
differently when using the gdb-index when compared to using no index.

The solution is to remove the de-duplication code, which is what this
patch does.

Now that we no longer need to sort declarations to the end of the
entry list, I've removed all the code related to the special use of
GDB_INDEX_SYMBOL_KIND_UNUSED5 (which is how we marked declarations),
this cleans things up a little bit.

I've also renamed some of the functions away from minimize, now that
there's no minimization being done.
---
 gdb/NEWS                                      |   5 +
 gdb/dwarf2/index-write.c                      |  80 ++-------
 .../gdb.base/gdb-index-many-types-1.c         |  30 ++++
 .../gdb.base/gdb-index-many-types-2.c         |  56 +++++++
 .../gdb.base/gdb-index-many-types-3.c         |  56 +++++++
 .../gdb.base/gdb-index-many-types.exp         | 155 ++++++++++++++++++
 gdb/testsuite/gdb.base/gdb-index-many-types.h |  39 +++++
 .../gdb.base/gdb-index-many-types.py          |  54 ++++++
 8 files changed, 405 insertions(+), 70 deletions(-)
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types-1.c
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types-2.c
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types-3.c
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types.exp
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types.h
 create mode 100644 gdb/testsuite/gdb.base/gdb-index-many-types.py

diff --git a/gdb/NEWS b/gdb/NEWS
index b632cde7e0d..0afc95f4c3c 100644
--- a/gdb/NEWS
+++ b/gdb/NEWS
@@ -14,6 +14,11 @@
   this flag is used gdbserver will not escape special shell characters
   within the inferior arguments.
 
+* GDB now adds all type symbols to the .gdb_index section.  The index
+  version number has not increased as a consequence of this change.
+  This fixes an issue where GDB could fail to find a type when relying
+  on the index.  Any existing indexes should be regenerated.
+
 * New targets
 
 GNU/Linux/MicroBlaze (gdbserver) microblazeel-*linux*
diff --git a/gdb/dwarf2/index-write.c b/gdb/dwarf2/index-write.c
index 37e66a59c16..80109017818 100644
--- a/gdb/dwarf2/index-write.c
+++ b/gdb/dwarf2/index-write.c
@@ -55,7 +55,7 @@
 #define DW2_GDB_INDEX_SYMBOL_KIND_SET_VALUE(cu_index, value) \
   do { \
     gdb_assert ((value) >= GDB_INDEX_SYMBOL_KIND_TYPE \
-		&& (value) <= GDB_INDEX_SYMBOL_KIND_UNUSED5); \
+		&& (value) < GDB_INDEX_SYMBOL_KIND_UNUSED5); \
     GDB_INDEX_SYMBOL_KIND_SET_VALUE((cu_index), (value)); \
   } while (0)
 
@@ -184,9 +184,8 @@ struct symtab_index_entry
      of this name.  */
   std::vector<offset_type> cu_indices;
 
-  /* Minimize CU_INDICES, sorting them and removing duplicates as
-     appropriate.  */
-  void minimize ();
+  /* Sort CU_INDICES.  */
+  void sort ();
 };
 
 /* The symbol table.  This is a power-of-2-sized hash table.  */
@@ -198,16 +197,16 @@ struct mapped_symtab
   }
 
   /* If there are no elements in the symbol table, then reduce the table
-     size to zero.  Otherwise call symtab_index_entry::minimize each entry
+     size to zero.  Otherwise call symtab_index_entry::sort each entry
      in the symbol table.  */
 
-  void minimize ()
+  void minimize_and_sort ()
   {
     if (m_element_count == 0)
       m_data.resize (0);
 
     for (symtab_index_entry &item : m_data)
-      item.minimize ();
+      item.sort ();
   }
 
   /* Add an entry to SYMTAB.  NAME is the name of the symbol.  CU_INDEX is
@@ -417,69 +416,19 @@ mapped_symtab::add_index_entry (const char *name, int is_static,
 /* See symtab_index_entry.  */
 
 void
-symtab_index_entry::minimize ()
+symtab_index_entry::sort ()
 {
   if (name == nullptr || cu_indices.empty ())
     return;
 
-  /* We sort the indexes in a funny way: GDB_INDEX_SYMBOL_KIND_UNUSED5
-     is always sorted last; then otherwise we sort by numeric value.
-     This ensures that we prefer the definition when both a definition
-     and a declaration (stub type) are seen.  */
+  /* Sort the indexes numerically.  */
   std::sort (cu_indices.begin (), cu_indices.end (),
 	     [] (offset_type vala, offset_type valb)
 	       {
-		 auto kinda = GDB_INDEX_SYMBOL_KIND_VALUE (vala);
-		 auto kindb = GDB_INDEX_SYMBOL_KIND_VALUE (valb);
-		 if (kinda != kindb)
-		   {
-		     /* Declaration sorts last.  */
-		     if (kinda == GDB_INDEX_SYMBOL_KIND_UNUSED5)
-		       return false;
-		     if (kindb == GDB_INDEX_SYMBOL_KIND_UNUSED5)
-		       return true;
-		   }
 		 return vala < valb;
 	       });
   auto from = std::unique (cu_indices.begin (), cu_indices.end ());
   cu_indices.erase (from, cu_indices.end ());
-
-  /* Rewrite GDB_INDEX_SYMBOL_KIND_UNUSED5.  This ensures that a type
-     declaration will be deleted by the subsequent squashing step, if
-     warranted.  */
-  for (auto &val : cu_indices)
-    {
-      gdb_index_symbol_kind kind = GDB_INDEX_SYMBOL_KIND_VALUE (val);
-      if (kind != GDB_INDEX_SYMBOL_KIND_UNUSED5)
-	continue;
-
-      offset_type newval = 0;
-      DW2_GDB_INDEX_CU_SET_VALUE (newval, GDB_INDEX_CU_VALUE (val));
-      DW2_GDB_INDEX_SYMBOL_STATIC_SET_VALUE
-	(newval, GDB_INDEX_SYMBOL_STATIC_VALUE (val));
-      DW2_GDB_INDEX_SYMBOL_KIND_SET_VALUE (newval,
-					   GDB_INDEX_SYMBOL_KIND_TYPE);
-
-      val = newval;
-    }
-
-  /* We don't want to enter a type more than once, so
-     remove any such duplicates from the list as well.  When doing
-     this, we want to keep the entry from the first CU -- but this is
-     implicit due to the sort.  This choice is done because it's
-     similar to what gdb historically did for partial symbols.  */
-  gdb::unordered_set<offset_type> seen;
-  from = std::remove_if (cu_indices.begin (), cu_indices.end (),
-			 [&] (offset_type val)
-    {
-      gdb_index_symbol_kind kind = GDB_INDEX_SYMBOL_KIND_VALUE (val);
-      if (kind != GDB_INDEX_SYMBOL_KIND_TYPE)
-	return false;
-
-      val &= ~GDB_INDEX_CU_MASK;
-      return !seen.insert (val).second;
-    });
-  cu_indices.erase (from, cu_indices.end ());
 }
 
 /* A form of 'const char *' suitable for container keys.  Only the
@@ -1307,16 +1256,7 @@ write_cooked_index (cooked_index *table,
 	       || entry->tag == DW_TAG_enumerator)
 	kind = GDB_INDEX_SYMBOL_KIND_VARIABLE;
       else if (tag_is_type (entry->tag))
-	{
-	  /* If we added a type declaration, we want to note this
-	     fact for later, because we don't want a type declaration
-	     to cause the real definition to be omitted from the
-	     index.  GDB_INDEX_SYMBOL_KIND_UNUSED5 is used here, but
-	     rewritten later before the index is written.  */
-	  kind = ((entry->flags & IS_TYPE_DECLARATION) == 0
-		  ? GDB_INDEX_SYMBOL_KIND_TYPE
-		  : GDB_INDEX_SYMBOL_KIND_UNUSED5);
-	}
+	kind = GDB_INDEX_SYMBOL_KIND_TYPE;
       else
 	kind = GDB_INDEX_SYMBOL_KIND_OTHER;
 
@@ -1458,7 +1398,7 @@ write_gdbindex (dwarf2_per_bfd *per_bfd, cooked_index *table,
 
   /* Now that we've processed all symbols we can shrink their cu_indices
      lists.  */
-  symtab.minimize ();
+  symtab.minimize_and_sort ();
 
   data_buf symtab_vec, constant_pool;
 
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types-1.c b/gdb/testsuite/gdb.base/gdb-index-many-types-1.c
new file mode 100644
index 00000000000..8ddd6fadcfc
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types-1.c
@@ -0,0 +1,30 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2025 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/>.  */
+
+#include "gdb-index-many-types.h"
+
+int
+main ()
+{
+  foo_func_a (0);
+  foo_func_b (0);
+  bar_func_a (0);
+  bar_func_b (0);
+  baz_func_a (0);
+  baz_func_b (0);
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types-2.c b/gdb/testsuite/gdb.base/gdb-index-many-types-2.c
new file mode 100644
index 00000000000..2c0a27f30ba
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types-2.c
@@ -0,0 +1,56 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2025 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/>.  */
+
+#include "gdb-index-many-types.h"
+
+typedef struct foo_t
+{
+  int foo_t_1;
+  int foo_t_2;
+} foo_t;
+
+typedef struct woof_t
+{
+  int woof_t_1;
+  int woof_t_2;
+} woof_t;
+
+static void
+woof_func (woof_t *obj)
+{
+  (void) obj;
+}
+
+void
+foo_func_a (foo_t *obj)
+{
+  woof_func (0);
+  (void) obj;
+}
+
+void
+baz_func_a (baz_t *obj)
+{
+  (void) obj;
+}
+
+void
+bar_func_a (bar_t *obj)
+{
+  woof_func (0);
+  (void) obj;
+}
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types-3.c b/gdb/testsuite/gdb.base/gdb-index-many-types-3.c
new file mode 100644
index 00000000000..54425ed5156
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types-3.c
@@ -0,0 +1,56 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2025 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/>.  */
+
+#include "gdb-index-many-types.h"
+
+typedef struct woof_t
+{
+  double woof_t_3;
+  double woof_t_4;
+} woof_t;
+
+static void
+woof_func (woof_t *obj)
+{
+  (void) obj;
+}
+
+typedef struct bar_t
+{
+  int bar_t_1;
+  int bar_t_2;
+} bar_t;
+
+void
+bar_func_b (bar_t *obj)
+{
+  woof_func (0);
+  (void) obj;
+}
+
+void
+baz_func_b (baz_t *obj)
+{
+  (void) obj;
+}
+
+void
+foo_func_b (foo_t *obj)
+{
+  woof_func (0);
+  (void) obj;
+}
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types.exp b/gdb/testsuite/gdb.base/gdb-index-many-types.exp
new file mode 100644
index 00000000000..debd8193f41
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types.exp
@@ -0,0 +1,155 @@
+# Copyright 2025 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/>.
+
+# Check that adding an index to an executable (both a gdb index and a
+# dwarf-5 index are tested), doesn't prevent GDB from seeing the
+# expected types.
+
+standard_testfile -1.c -2.c -3.c .h
+
+# One of the tests uses this Python file.  The test_* proc checks that
+# GDB supports Python tests.  Some of the other procs don't use this
+# Python file.
+set pyfile [gdb_remote_download host ${srcdir}/${subdir}/${testfile}.py]
+
+if {[build_executable "building" $testfile \
+	 [list $srcfile $srcfile2 $srcfile3]] == -1} {
+    return
+}
+
+# Run 'info types TYPENAME', expect to see an entry from FILENAME for
+# the line matching PATTERN.
+proc check_info_types { testfile typename filename pattern } {
+    with_test_prefix $filename {
+	clean_restart $testfile
+
+	set line_num [gdb_get_line_number $pattern $filename]
+	gdb_test "info types $typename" \
+	    "File \[^\r\n\]+/${filename}:\r\n${line_num}:\[^\r\n\]+.*"
+    }
+}
+
+# Start GDB with FILENAME, and examine some of the types.  This proc
+# might seem to be using clean_restart a little too much, but we need
+# to be really careful here.  As we examine one type, e.g. foo_t, this
+# might cause GDB to fully parse a CU, which then means examining
+# bar_t gives the expected result.  When, if we'd first looked for
+# bar_t, then (due to an index bug) we might not have found the
+# correct type definition.
+#
+# The only way we can be sure that an earlier test isn't going to
+# trigger CU expansion is to restart GDB before every query.
+proc run_test { filename } {
+    # Print all the types for which there is only one representation.
+    foreach type { foo_t bar_t baz_t } {
+	clean_restart $filename
+	gdb_test "ptype $type" \
+	    [multi_line \
+		 "type = struct $type {" \
+		 "\\s+int ${type}_1;" \
+		 "\\s+int ${type}_2;" \
+		 "}"]
+    }
+
+    # There are two different versions of woof_t.  For now, when using
+    # `ptype` GDB will just display the first one it finds, which could
+    # legitimately be either.
+    set woof_int_re [multi_line \
+			 "type = struct woof_t {" \
+			 "\\s+int woof_t_1;" \
+			 "\\s+int woof_t_2;" \
+			 "}"]
+    set woof_double_re [multi_line \
+			    "type = struct woof_t {" \
+			    "\\s+double woof_t_3;" \
+			    "\\s+double woof_t_4;" \
+			    "}"]
+    clean_restart $filename
+    gdb_test_multiple "ptype woof_t" "" {
+	-re -wrap $woof_int_re {
+	    pass $gdb_test_name
+	}
+	-re -wrap $woof_double_re {
+	    pass $gdb_test_name
+	}
+    }
+
+    # Check for declarations and definitions of some types.
+    check_info_types $filename foo_t $::srcfile2 "typedef struct foo_t"
+    check_info_types $filename foo_t $::srcfile4 "typedef struct foo_t foo_t;"
+    check_info_types $filename bar_t $::srcfile3 "typedef struct bar_t"
+    check_info_types $filename bar_t $::srcfile4 "typedef struct bar_t bar_t;"
+    check_info_types $filename baz_t $::srcfile4 "typedef struct baz_t"
+    check_info_types $filename woof_t $::srcfile2 "typedef struct woof_t"
+    check_info_types $filename woof_t $::srcfile3 "typedef struct woof_t"
+
+    # Use Python to look for type symbols.
+    if { [allow_python_tests] } {
+	foreach_with_prefix type { foo_t bar_t baz_t } {
+	    clean_restart $filename
+	    gdb_test_no_output "source $::pyfile" "import python scripts"
+	    gdb_test "py-show-type $type" \
+		[multi_line \
+		     "Looking for type '$type':" \
+		     "  Found 2 type symbols" \
+		     "    1: struct $type \\{ int ${type}_1; int ${type}_2; \\}" \
+		     "    2: struct $type \\{ int ${type}_1; int ${type}_2; \\}"]
+	}
+
+	clean_restart $filename
+	gdb_test_no_output "source $::pyfile" "import python scripts"
+	gdb_test "py-show-type woof_t" \
+	    [multi_line \
+		 "Looking for type 'woof_t':" \
+		 "  Found 2 type symbols" \
+		 "    1: struct woof_t \\{ (?:int|double) woof_t_(?:1|3); (?:int|double) woof_t_(?:2|4); \\}" \
+		 "    2: struct woof_t \\{ (?:int|double) woof_t_(?:1|3); (?:int|double) woof_t_(?:2|4); \\}"]
+
+    }
+}
+
+with_test_prefix "no index" {
+    run_test $testfile
+}
+
+# The previous call to 'run_test' will have left GDB active.  Check if
+# BINFILE already has an index.  If it does then we must be running
+# with one of the boardfiles that adds an index.  We could possibly
+# try to remove the index, but for now, just don't run the following
+# parts which rely on adding an index.
+set index_type [get_index_type $binfile "check debug style"]
+if { $index_type ne "cooked" } {
+    unsupported "cannot test without a cooked index"
+    return
+}
+
+foreach_with_prefix index_type { gdb dwarf5 } {
+    set binfile_with_index ${binfile}-idx-${index_type}
+
+    remote_exec build "cp $binfile $binfile_with_index"
+
+    if { $index_type eq "gdb" } {
+	set style ""
+    } else {
+	set style "-dwarf-5"
+    }
+
+    if {[ensure_gdb_index $binfile_with_index $style] != 1} {
+	unsupported "couldn't add $index_type index"
+	return
+    }
+
+    run_test [file tail $binfile_with_index]
+}
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types.h b/gdb/testsuite/gdb.base/gdb-index-many-types.h
new file mode 100644
index 00000000000..688a43cbc1b
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types.h
@@ -0,0 +1,39 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2025 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/>.  */
+
+#ifndef GDB_INDEX_MANY_TYPES_H
+#define GDB_INDEX_MANY_TYPES_H
+
+typedef struct foo_t foo_t;
+typedef struct bar_t bar_t;
+
+extern void foo_func_a (foo_t *obj);
+extern void foo_func_b (foo_t *obj);
+
+extern void bar_func_a (bar_t *obj);
+extern void bar_func_b (bar_t *obj);
+
+typedef struct baz_t
+{
+  int baz_t_1;
+  int baz_t_2;
+} baz_t;
+
+extern void baz_func_a (baz_t *obj);
+extern void baz_func_b (baz_t *obj);
+
+#endif /* GDB_INDEX_MANY_TYPES_H */
diff --git a/gdb/testsuite/gdb.base/gdb-index-many-types.py b/gdb/testsuite/gdb.base/gdb-index-many-types.py
new file mode 100644
index 00000000000..b1de4f9cbed
--- /dev/null
+++ b/gdb/testsuite/gdb.base/gdb-index-many-types.py
@@ -0,0 +1,54 @@
+# Copyright 2025 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/>.
+
+
+class TypeViewer(gdb.Command):
+    """A command which takes a string and looks up types with that name.
+
+    The types are expected to all be structs.  This command prints a
+    basic representation of the struct.  This is only going to work when
+    used with the types defined in the gdb-index-many-types test source
+    files."""
+
+    def __init__(self):
+        super().__init__("py-show-type", gdb.COMMAND_USER)
+
+    def invoke(self, args, from_tty):
+        argv = gdb.string_to_argv(args)
+        if argv[0] == "":
+            raise gdb.GdbError("missing argument")
+        print("Looking for type '" + argv[0] + "':")
+        syms = gdb.lookup_static_symbols(argv[0], gdb.SYMBOL_TYPE_DOMAIN)
+        count = len(syms)
+        print("  Found %d type symbol%s" % (count, "" if count == 1 else "s"))
+        for i, s in enumerate(syms, start=1):
+            t = s.type
+            if t is None:
+                print("    %d: No type." % i)
+            else:
+                fields = "struct " + argv[0] + " {"
+                try:
+                    for f in t.fields():
+                        if len(fields) > 0:
+                            fields = fields + " "
+                        fields = fields + "%s %s;" % (str(f.type), f.name)
+                except:
+                    pass
+                fields = fields + " }"
+
+                print("    %d: %s" % (i, fields))
+
+
+TypeViewer()

base-commit: d9f1578be5a326d34710a38d215e80a43ab910c5
-- 
2.47.1


^ permalink raw reply	[flat|nested] 10+ messages in thread
end of thread, other threads:[~2025-11-13 16:32 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-10-02  9:36 [PATCH] gdb: remove attempted type de-duplication when building gdb-index Andrew Burgess
2025-10-16 18:11 ` [PATCHv2] " Andrew Burgess
2025-10-21 13:45   ` [PATCHv3] " Andrew Burgess
2025-10-22 15:17     ` Tom Tromey
2025-10-22 15:59       ` Andrew Burgess
2025-10-22 16:13         ` Andrew Burgess
2025-10-22 17:21           ` Tom Tromey
2025-11-04 14:23             ` Andrew Burgess
2025-10-22 17:21         ` Tom Tromey
2025-11-13 16:32       ` Andrew Burgess

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox