[PATCH 0/9] Cleanups in global_symbol_searcher

[PATCH 0/9] Cleanups in global_symbol_searcher

[Simon Marchi]

I had a look at global_symbol_searcher while reviewing a recent patch,
and I had a bit of trouble understanding how it worked.  I made a few
changes to try to make it clearer.

Simon Marchi (9):
  gdb: add symbol_search_kind enum for global_symbol_searcher
  gdb: factor out regexp compilation from global_symbol_searcher::search
  gdb: simplify insertion in
    global_symbol_searcher::add_matching_symbols
  gdb: use a switch on m_kind in add_matching_symbols
  gdb: make global_symbol_searcher::is_suitable_msymbol non-static
  gdb: update comment of global_symbol_searcher::expand_symtabs
  gdb: add matching helpers to global_symbol_searcher
  gdb: pass result containers by reference in global_symbol_searcher
  gdb: fix some comments in global_symbol_searcher

 gdb/mi/mi-symbol-cmds.c |  36 ++--
 gdb/python/python.c     |   2 +-
 gdb/symtab.c            | 458 +++++++++++++++++++++++-----------------
 gdb/symtab.h            | 105 +++++----
 4 files changed, 348 insertions(+), 253 deletions(-)


base-commit: ccf0f68e72ea0a5628097b980a04e1a8262781df
-- 
2.55.0

[PATCH 1/9] gdb: add symbol_search_kind enum for global_symbol_searcher

[Simon Marchi]

While reading the global_symbol_searcher code, something that was not
clear to me was that despite accepting a domain_search_flags,  searches
are mutually exclusive between types, variables and functions.

One snippet of code that really confused me was (now removed by a
previous commit):

		  if ((kind & SEARCH_FUNCTION_DOMAIN) != 0
		      ? (find_compunit_symtab_for_pc
			 (msymbol->value_address (objfile)) == NULL)
		      : (lookup_symbol_in_objfile_from_linkage_name
			 (objfile, msymbol->linkage_name (),
			  SEARCH_VFT)
			 .symbol == NULL))

I kept thinking: what if you search for both functions and variables,
this will take the "function" path, this can't work.  The answer is that
it's not supposed to happen.

This patch adds the symbol_search_kind enum and uses it in
global_symbol_searcher (and a few caller functions), making it clear
that you search just one kind at a time.

The new enum has a certain overlap with enum domain_search_flag, but I
think that having the more scoped enum makes global_symbol_searcher
easier to understand, by specifying clearly what it needs to handle.

Change-Id: I4a3f4752c7be51a4da5831f9780d21e6e7271b38
---
 gdb/mi/mi-symbol-cmds.c |  36 +++++++------
 gdb/python/python.c     |   2 +-
 gdb/symtab.c            | 112 ++++++++++++++++++++++------------------
 gdb/symtab.h            |  33 ++++++++----
 4 files changed, 106 insertions(+), 77 deletions(-)

diff --git a/gdb/mi/mi-symbol-cmds.c b/gdb/mi/mi-symbol-cmds.c
index 58f2feef5533..30b8ec495e8f 100644
--- a/gdb/mi/mi-symbol-cmds.c
+++ b/gdb/mi/mi-symbol-cmds.c
@@ -71,7 +71,7 @@ mi_cmd_symbol_list_lines (const char *command, const char *const *argv,
    results.  */
 
 static void
-output_debug_symbol (ui_out *uiout, domain_search_flags kind,
+output_debug_symbol (ui_out *uiout, symbol_search_kind kind,
 		     struct symbol *sym, int block)
 {
   ui_out_emit_tuple tuple_emitter (uiout);
@@ -80,7 +80,8 @@ output_debug_symbol (ui_out *uiout, domain_search_flags kind,
     uiout->field_unsigned ("line", sym->line ());
   uiout->field_string ("name", sym->print_name ());
 
-  if ((kind & (SEARCH_FUNCTION_DOMAIN | SEARCH_VAR_DOMAIN)) != 0)
+  if (kind == symbol_search_kind::FUNCTION
+      || kind == symbol_search_kind::VARIABLE)
     {
       string_file tmp_stream;
       type_print (sym->type (), "", &tmp_stream, -1);
@@ -111,7 +112,7 @@ output_nondebug_symbol (ui_out *uiout, const bound_minimal_symbol &msymbol)
    and then prints the matching [m]symbols in an MI structured format.  */
 
 static void
-mi_symbol_info (domain_search_flags kind, const char *name_regexp,
+mi_symbol_info (symbol_search_kind kind, const char *name_regexp,
 		const char *type_regexp, bool exclude_minsyms,
 		size_t max_results)
 {
@@ -188,7 +189,7 @@ parse_max_results_option (const char *arg)
    Processes command line options from ARGV and ARGC.  */
 
 static void
-mi_info_functions_or_variables (domain_search_flags kind,
+mi_info_functions_or_variables (symbol_search_kind kind,
 				const char *const *argv, int argc)
 {
   size_t max_results = SIZE_MAX;
@@ -215,7 +216,7 @@ mi_info_functions_or_variables (domain_search_flags kind,
   while (1)
     {
       const char *cmd_string
-	= ((kind == SEARCH_FUNCTION_DOMAIN)
+	= ((kind == symbol_search_kind::FUNCTION)
 	   ? "-symbol-info-functions" : "-symbol-info-variables");
       int opt = mi_getopt (cmd_string, argc, argv, opts, &oind, &oarg);
       if (opt < 0)
@@ -251,7 +252,7 @@ using module_symbol_search_iterator
 static module_symbol_search_iterator
 output_module_symbols_in_single_module_and_file
 	(struct ui_out *uiout, module_symbol_search_iterator iter,
-	 const module_symbol_search_iterator end, domain_search_flags kind)
+	 const module_symbol_search_iterator end, symbol_search_kind kind)
 {
   /* The symbol for the module in which the first result resides.  */
   const symbol *first_module_symbol = iter->first.symbol;
@@ -286,7 +287,7 @@ output_module_symbols_in_single_module_and_file
 static module_symbol_search_iterator
 output_module_symbols_in_single_module
 	(struct ui_out *uiout, module_symbol_search_iterator iter,
-	 const module_symbol_search_iterator end, domain_search_flags kind)
+	 const module_symbol_search_iterator end, symbol_search_kind kind)
 {
   gdb_assert (iter->first.symbol != nullptr);
   gdb_assert (iter->second.symbol != nullptr);
@@ -314,7 +315,7 @@ output_module_symbols_in_single_module
    command line options passed to the MI command.  */
 
 static void
-mi_info_module_functions_or_variables (domain_search_flags kind,
+mi_info_module_functions_or_variables (symbol_search_kind kind,
 				       const char *const *argv, int argc)
 {
   const char *module_regexp = nullptr;
@@ -341,7 +342,7 @@ mi_info_module_functions_or_variables (domain_search_flags kind,
   while (1)
     {
       const char *cmd_string
-	= ((kind == SEARCH_FUNCTION_DOMAIN)
+	= ((kind == symbol_search_kind::FUNCTION)
 	   ? "-symbol-info-module-functions"
 	   : "-symbol-info-module-variables");
       int opt = mi_getopt (cmd_string, argc, argv, opts, &oind, &oarg);
@@ -383,7 +384,7 @@ void
 mi_cmd_symbol_info_functions (const char *command, const char *const *argv,
 			      int argc)
 {
-  mi_info_functions_or_variables (SEARCH_FUNCTION_DOMAIN, argv, argc);
+  mi_info_functions_or_variables (symbol_search_kind::FUNCTION, argv, argc);
 }
 
 /* Implement -symbol-info-module-functions command.  */
@@ -392,7 +393,8 @@ void
 mi_cmd_symbol_info_module_functions (const char *command,
 				     const char *const *argv, int argc)
 {
-  mi_info_module_functions_or_variables (SEARCH_FUNCTION_DOMAIN, argv, argc);
+  mi_info_module_functions_or_variables (symbol_search_kind::FUNCTION,
+					 argv, argc);
 }
 
 /* Implement -symbol-info-module-variables command.  */
@@ -401,7 +403,8 @@ void
 mi_cmd_symbol_info_module_variables (const char *command,
 				     const char *const *argv, int argc)
 {
-  mi_info_module_functions_or_variables (SEARCH_VAR_DOMAIN, argv, argc);
+  mi_info_module_functions_or_variables (symbol_search_kind::VARIABLE,
+					 argv, argc);
 }
 
 /* Implement -symbol-inf-modules command.  */
@@ -444,7 +447,8 @@ mi_cmd_symbol_info_modules (const char *command, const char *const *argv,
 	}
     }
 
-  mi_symbol_info (SEARCH_MODULE_DOMAIN, regexp, nullptr, true, max_results);
+  mi_symbol_info (symbol_search_kind::MODULE, regexp, nullptr, true,
+		  max_results);
 }
 
 /* Implement -symbol-info-types command.  */
@@ -487,8 +491,8 @@ mi_cmd_symbol_info_types (const char *command, const char *const *argv,
 	}
     }
 
-  mi_symbol_info (SEARCH_TYPE_DOMAIN | SEARCH_STRUCT_DOMAIN, regexp, nullptr,
-		  true, max_results);
+  mi_symbol_info (symbol_search_kind::TYPE, regexp, nullptr, true,
+		  max_results);
 }
 
 /* Implement -symbol-info-variables command.  */
@@ -497,5 +501,5 @@ void
 mi_cmd_symbol_info_variables (const char *command, const char *const *argv,
 			      int argc)
 {
-  mi_info_functions_or_variables (SEARCH_VAR_DOMAIN, argv, argc);
+  mi_info_functions_or_variables (symbol_search_kind::VARIABLE, argv, argc);
 }
diff --git a/gdb/python/python.c b/gdb/python/python.c
index d51cd83966be..467317b7996e 100644
--- a/gdb/python/python.c
+++ b/gdb/python/python.c
@@ -848,7 +848,7 @@ gdbpy_rbreak (PyObject *self, PyObject *args, PyObject *kw)
       minsyms_p = cmp;
     }
 
-  global_symbol_searcher spec (SEARCH_FUNCTION_DOMAIN, regex);
+  global_symbol_searcher spec (symbol_search_kind::FUNCTION, regex);
 
   /* The "symtabs" keyword is any Python iterable object that returns
      a gdb.Symtab on each iteration.  If specified, iterate through
diff --git a/gdb/symtab.c b/gdb/symtab.c
index 27070f74f83e..708df078e2ea 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4698,11 +4698,31 @@ treg_matches_sym_type_name (const compiled_regex &treg,
   return treg.exec (printed_sym_type_name.c_str (), 0, NULL, 0) == 0;
 }
 
+/* Return the domain search flags matching symbols of kind KIND.  */
+
+static domain_search_flags
+to_search_flags (symbol_search_kind kind)
+{
+  switch (kind)
+    {
+    case symbol_search_kind::VARIABLE:
+      return SEARCH_VAR_DOMAIN;
+    case symbol_search_kind::FUNCTION:
+      return SEARCH_FUNCTION_DOMAIN;
+    case symbol_search_kind::TYPE:
+      return SEARCH_TYPE_DOMAIN | SEARCH_STRUCT_DOMAIN;
+    case symbol_search_kind::MODULE:
+      return SEARCH_MODULE_DOMAIN;
+    }
+
+  gdb_assert_not_reached ("invalid symbol_search_kind");
+}
+
 /* See symtab.h.  */
 
 bool
 global_symbol_searcher::is_suitable_msymbol
-	(const domain_search_flags kind, const minimal_symbol *msymbol)
+	(symbol_search_kind kind, const minimal_symbol *msymbol)
 {
   switch (msymbol->type ())
     {
@@ -4710,12 +4730,12 @@ global_symbol_searcher::is_suitable_msymbol
     case mst_bss:
     case mst_file_data:
     case mst_file_bss:
-      return (kind & SEARCH_VAR_DOMAIN) != 0;
+      return kind == symbol_search_kind::VARIABLE;
     case mst_text:
     case mst_file_text:
     case mst_solib_trampoline:
     case mst_text_gnu_ifunc:
-      return (kind & SEARCH_FUNCTION_DOMAIN) != 0;
+      return kind == symbol_search_kind::FUNCTION;
     default:
       return false;
     }
@@ -4727,7 +4747,6 @@ bool
 global_symbol_searcher::expand_symtabs
 	(objfile *objfile, const std::optional<compiled_regex> &preg) const
 {
-  domain_search_flags kind = m_kind;
   bool found_func_msymbol_without_debug_info = false;
 
   auto do_file_match = [&] (const char *filename, bool basenames)
@@ -4748,7 +4767,7 @@ global_symbol_searcher::expand_symtabs
      },
      NULL,
      SEARCH_GLOBAL_BLOCK | SEARCH_STATIC_BLOCK,
-     kind);
+     to_search_flags (m_kind));
 
   /* Here, we search through the minimal symbol tables for functions that
      match, and force their symbols to be read.  The symbol will then be found
@@ -4764,7 +4783,7 @@ global_symbol_searcher::expand_symtabs
      symtab.  When no file names were given the caller unconditionally rescans
      the minimal symbols for SEARCH_VAR_DOMAIN.  */
   if (m_filenames.empty ()
-      && (kind & SEARCH_FUNCTION_DOMAIN) != 0)
+      && m_kind == symbol_search_kind::FUNCTION)
     {
       for (minimal_symbol *msymbol : objfile->msymbols ())
 	{
@@ -4773,7 +4792,7 @@ global_symbol_searcher::expand_symtabs
 	  if (msymbol->created_by_gdb)
 	    continue;
 
-	  if (is_suitable_msymbol (kind, msymbol))
+	  if (is_suitable_msymbol (m_kind, msymbol))
 	    {
 	      if (!preg.has_value ()
 		  || preg->exec (msymbol->natural_name (), 0,
@@ -4804,7 +4823,7 @@ global_symbol_searcher::add_matching_symbols
 	 const std::optional<compiled_regex> &treg,
 	 std::set<symbol_search> *result_set) const
 {
-  domain_search_flags kind = m_kind;
+  domain_search_flags domain = to_search_flags (m_kind);
 
   /* Add matching symbols (if not already present).  */
   for (compunit_symtab &cust : objfile->compunits ())
@@ -4832,7 +4851,7 @@ global_symbol_searcher::add_matching_symbols
 					 m_filenames, false))))
 		continue;
 
-	      if (!sym->matches (kind))
+	      if (!sym->matches (domain))
 		continue;
 
 	      if (preg.has_value () && preg->exec (sym->natural_name (), 0,
@@ -4845,7 +4864,7 @@ global_symbol_searcher::add_matching_symbols
 		   && !treg_matches_sym_type_name (*treg, sym)))
 		continue;
 
-	      if ((kind & SEARCH_VAR_DOMAIN) != 0)
+	      if (m_kind == symbol_search_kind::VARIABLE)
 		{
 		  if (sym->loc_class () == LOC_UNRESOLVED
 		      /* LOC_CONST can be used for more than
@@ -4882,8 +4901,6 @@ global_symbol_searcher::add_matching_msymbols
 	(objfile *objfile, const std::optional<compiled_regex> &preg,
 	 std::vector<symbol_search> *results) const
 {
-  domain_search_flags kind = m_kind;
-
   for (minimal_symbol *msymbol : objfile->msymbols ())
     {
       QUIT;
@@ -4891,7 +4908,7 @@ global_symbol_searcher::add_matching_msymbols
       if (msymbol->created_by_gdb)
 	continue;
 
-      if (is_suitable_msymbol (kind, msymbol))
+      if (is_suitable_msymbol (m_kind, msymbol))
 	{
 	  if (!preg.has_value ()
 	      || preg->exec (msymbol->natural_name (), 0,
@@ -4899,7 +4916,7 @@ global_symbol_searcher::add_matching_msymbols
 	    {
 	      /* For functions we can do a quick check of whether the
 		 symbol might be found via find_pc_symtab.  */
-	      if ((kind & SEARCH_FUNCTION_DOMAIN) == 0
+	      if (m_kind != symbol_search_kind::FUNCTION
 		  || (find_compunit_symtab_for_pc
 		      (msymbol->value_address (objfile)) == NULL))
 		{
@@ -5006,12 +5023,12 @@ global_symbol_searcher::search () const
      minimal symbol, as we assume that a minimal symbol does not have a
      type.  */
   if ((found_func_msymbol_without_debug_info
-       || (m_filenames.empty () && (m_kind & SEARCH_VAR_DOMAIN) != 0))
+       || (m_filenames.empty () && m_kind == symbol_search_kind::VARIABLE))
       && !m_exclude_minsyms
       && !treg.has_value ())
     {
-      gdb_assert ((m_kind & (SEARCH_VAR_DOMAIN | SEARCH_FUNCTION_DOMAIN))
-		  != 0);
+      gdb_assert (m_kind == symbol_search_kind::VARIABLE
+		  || m_kind == symbol_search_kind::FUNCTION);
       for (objfile &objfile : current_program_space->objfiles ())
 	if (!add_matching_msymbols (&objfile, preg, &result))
 	  break;
@@ -5143,7 +5160,7 @@ print_msymbol_info (bound_minimal_symbol msymbol)
 
 static void
 symtab_symbol_info (bool quiet, bool exclude_minsyms,
-		    const char *regexp, domain_enum kind,
+		    const char *regexp, symbol_search_kind kind,
 		    const char *t_regexp, int from_tty)
 {
   const char *last_filename = "";
@@ -5152,11 +5169,7 @@ symtab_symbol_info (bool quiet, bool exclude_minsyms,
   if (regexp != nullptr && *regexp == '\0')
     regexp = nullptr;
 
-  domain_search_flags flags = to_search_flags (kind);
-  if (kind == TYPE_DOMAIN)
-    flags |= SEARCH_STRUCT_DOMAIN;
-
-  global_symbol_searcher spec (flags, regexp);
+  global_symbol_searcher spec (kind, regexp);
   spec.set_symbol_type_regexp (t_regexp);
   spec.set_exclude_minsyms (exclude_minsyms);
   std::vector<symbol_search> symbols = spec.search ();
@@ -5166,20 +5179,20 @@ symtab_symbol_info (bool quiet, bool exclude_minsyms,
       const char *classname;
       switch (kind)
 	{
-	case VAR_DOMAIN:
+	case symbol_search_kind::VARIABLE:
 	  classname = "variable";
 	  break;
-	case FUNCTION_DOMAIN:
+	case symbol_search_kind::FUNCTION:
 	  classname = "function";
 	  break;
-	case TYPE_DOMAIN:
+	case symbol_search_kind::TYPE:
 	  classname = "type";
 	  break;
-	case MODULE_DOMAIN:
+	case symbol_search_kind::MODULE:
 	  classname = "module";
 	  break;
 	default:
-	  gdb_assert_not_reached ("invalid domain enum");
+	  gdb_assert_not_reached ("invalid symbol_search_kind");
 	}
 
       if (regexp != NULL)
@@ -5304,7 +5317,7 @@ info_variables_command (const char *args, int from_tty)
     args = nullptr;
 
   symtab_symbol_info
-    (opts.quiet, opts.exclude_minsyms, args, VAR_DOMAIN,
+    (opts.quiet, opts.exclude_minsyms, args, symbol_search_kind::VARIABLE,
      opts.type_regexp.empty () ? nullptr : opts.type_regexp.c_str (),
      from_tty);
 }
@@ -5323,7 +5336,7 @@ info_functions_command (const char *args, int from_tty)
     args = nullptr;
 
   symtab_symbol_info
-    (opts.quiet, opts.exclude_minsyms, args, FUNCTION_DOMAIN,
+    (opts.quiet, opts.exclude_minsyms, args, symbol_search_kind::FUNCTION,
      opts.type_regexp.empty () ? nullptr : opts.type_regexp.c_str (),
      from_tty);
 }
@@ -5366,8 +5379,8 @@ info_types_command (const char *args, int from_tty)
     (&args, gdb::option::PROCESS_OPTIONS_UNKNOWN_IS_OPERAND, grp);
   if (args != nullptr && *args == '\0')
     args = nullptr;
-  symtab_symbol_info (opts.quiet, false, args, TYPE_DOMAIN, nullptr,
-		      from_tty);
+  symtab_symbol_info (opts.quiet, false, args, symbol_search_kind::TYPE,
+		      nullptr, from_tty);
 }
 
 /* Command completer for 'info types' command.  */
@@ -5399,8 +5412,8 @@ info_modules_command (const char *args, int from_tty)
     (&args, gdb::option::PROCESS_OPTIONS_UNKNOWN_IS_OPERAND, grp);
   if (args != nullptr && *args == '\0')
     args = nullptr;
-  symtab_symbol_info (opts.quiet, true, args, MODULE_DOMAIN, nullptr,
-		      from_tty);
+  symtab_symbol_info (opts.quiet, true, args, symbol_search_kind::MODULE,
+		      nullptr, from_tty);
 }
 
 /* Implement the 'info main' command.  */
@@ -5441,7 +5454,7 @@ rbreak_command (const char *regexp, int from_tty)
      because the std::move nullifies file_name.  */
   bool file_name_p = file_name != nullptr;
 
-  global_symbol_searcher spec (SEARCH_FUNCTION_DOMAIN, regexp);
+  global_symbol_searcher spec (symbol_search_kind::FUNCTION, regexp);
   if (file_name_p)
     spec.add_filename (std::move (file_name));
   std::vector<symbol_search> symbols = spec.search ();
@@ -6654,12 +6667,12 @@ static struct cmd_list_element *info_module_cmdlist = NULL;
 
 std::vector<module_symbol_search>
 search_module_symbols (const char *module_regexp, const char *regexp,
-		       const char *type_regexp, domain_search_flags kind)
+		       const char *type_regexp, symbol_search_kind kind)
 {
   std::vector<module_symbol_search> results;
 
   /* Search for all modules matching MODULE_REGEXP.  */
-  global_symbol_searcher spec1 (SEARCH_MODULE_DOMAIN, module_regexp);
+  global_symbol_searcher spec1 (symbol_search_kind::MODULE, module_regexp);
   spec1.set_exclude_minsyms (true);
   std::vector<symbol_search> modules = spec1.search ();
 
@@ -6705,9 +6718,10 @@ search_module_symbols (const char *module_regexp, const char *regexp,
 static void
 info_module_subcommand (bool quiet, const char *module_regexp,
 			const char *regexp, const char *type_regexp,
-			domain_search_flags kind)
+			symbol_search_kind kind)
 {
-  gdb_assert (kind == SEARCH_FUNCTION_DOMAIN || kind == SEARCH_VAR_DOMAIN);
+  gdb_assert (kind == symbol_search_kind::FUNCTION
+	      || kind == symbol_search_kind::VARIABLE);
 
   /* Print a header line.  Don't build the header line bit by bit as this
      prevents internationalisation.  */
@@ -6718,12 +6732,12 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 	  if (type_regexp == nullptr)
 	    {
 	      if (regexp == nullptr)
-		gdb_printf ((kind == SEARCH_VAR_DOMAIN
+		gdb_printf ((kind == symbol_search_kind::VARIABLE
 			     ? _("All variables in all modules:")
 			     : _("All functions in all modules:")));
 	      else
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables matching regular expression"
 			" \"%s\" in all modules:")
 		    : _("All functions matching regular expression"
@@ -6734,7 +6748,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 	    {
 	      if (regexp == nullptr)
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables with type matching regular "
 			"expression \"%s\" in all modules:")
 		    : _("All functions with type matching regular "
@@ -6742,7 +6756,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 		   type_regexp);
 	      else
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables matching regular expression "
 			"\"%s\",\n\twith type matching regular "
 			"expression \"%s\" in all modules:")
@@ -6758,7 +6772,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 	    {
 	      if (regexp == nullptr)
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables in all modules matching regular "
 			"expression \"%s\":")
 		    : _("All functions in all modules matching regular "
@@ -6766,7 +6780,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 		   module_regexp);
 	      else
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables matching regular expression "
 			"\"%s\",\n\tin all modules matching regular "
 			"expression \"%s\":")
@@ -6779,7 +6793,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 	    {
 	      if (regexp == nullptr)
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables with type matching regular "
 			"expression \"%s\"\n\tin all modules matching "
 			"regular expression \"%s\":")
@@ -6789,7 +6803,7 @@ info_module_subcommand (bool quiet, const char *module_regexp,
 		   type_regexp, module_regexp);
 	      else
 		gdb_printf
-		  ((kind == SEARCH_VAR_DOMAIN
+		  ((kind == symbol_search_kind::VARIABLE
 		    ? _("All variables matching regular expression "
 			"\"%s\",\n\twith type matching regular expression "
 			"\"%s\",\n\tin all modules matching regular "
@@ -6901,7 +6915,7 @@ info_module_functions_command (const char *args, int from_tty)
     (opts.quiet,
      opts.module_regexp.empty () ? nullptr : opts.module_regexp.c_str (), args,
      opts.type_regexp.empty () ? nullptr : opts.type_regexp.c_str (),
-     SEARCH_FUNCTION_DOMAIN);
+     symbol_search_kind::FUNCTION);
 }
 
 /* Implements the 'info module variables' command.  */
@@ -6920,7 +6934,7 @@ info_module_variables_command (const char *args, int from_tty)
     (opts.quiet,
      opts.module_regexp.empty () ? nullptr : opts.module_regexp.c_str (), args,
      opts.type_regexp.empty () ? nullptr : opts.type_regexp.c_str (),
-     SEARCH_VAR_DOMAIN);
+     symbol_search_kind::VARIABLE);
 }
 
 /* Command completer for 'info module ...' sub-commands.  */
diff --git a/gdb/symtab.h b/gdb/symtab.h
index 6443216fd7f0..a35449fbeeda 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2586,6 +2586,23 @@ struct symbol_search
 				  const symbol_search &sym_b);
 };
 
+/* The kind of symbols searched for by global_symbol_searcher.  */
+
+enum class symbol_search_kind
+{
+  /* Variables, excluding enum constants.  */
+  VARIABLE,
+
+  /* Functions.  */
+  FUNCTION,
+
+  /* Type names.  */
+  TYPE,
+
+  /* Fortran modules.  */
+  MODULE,
+};
+
 /* In order to search for global symbols of a particular kind matching
    particular regular expressions, create an instance of this structure and
    call the SEARCH member function.  */
@@ -2594,7 +2611,7 @@ class global_symbol_searcher
 public:
 
   /* Constructor.  */
-  global_symbol_searcher (domain_search_flags kind,
+  global_symbol_searcher (symbol_search_kind kind,
 			  const char *symbol_name_regexp)
     : m_kind (kind),
       m_symbol_name_regexp (symbol_name_regexp)
@@ -2635,14 +2652,8 @@ class global_symbol_searcher
   /* The set of source files to search in for matching symbols.  */
   std::vector<gdb::unique_xmalloc_ptr<char>> m_filenames;
 
-  /* The kind of symbols are we searching for.
-     VARIABLES_DOMAIN - Search all symbols, excluding functions, type
-			names, and constants (enums).
-     FUNCTIONS_DOMAIN - Search all functions..
-     TYPES_DOMAIN     - Search all type names.
-     MODULES_DOMAIN   - Search all Fortran modules.
-     ALL_DOMAIN       - Not valid for this function.  */
-  domain_search_flags m_kind;
+  /* The kind of symbols we are searching for.  */
+  symbol_search_kind m_kind;
 
   /* Regular expression to match against the symbol name.  */
   const char *m_symbol_name_regexp = nullptr;
@@ -2686,7 +2697,7 @@ class global_symbol_searcher
 			      std::vector<symbol_search> *results) const;
 
   /* Return true if MSYMBOL is of type KIND.  */
-  static bool is_suitable_msymbol (const domain_search_flags kind,
+  static bool is_suitable_msymbol (symbol_search_kind kind,
 				   const minimal_symbol *msymbol);
 };
 
@@ -2703,7 +2714,7 @@ using module_symbol_search = std::pair<symbol_search, symbol_search>;
    within the module.  */
 extern std::vector<module_symbol_search> search_module_symbols
 	(const char *module_regexp, const char *regexp,
-	 const char *type_regexp, domain_search_flags kind);
+	 const char *type_regexp, symbol_search_kind kind);
 
 /* Convert a global or static symbol SYM (based on BLOCK, which should be
    either GLOBAL_BLOCK or STATIC_BLOCK) into a string for use in 'info'
-- 
2.55.0

[PATCH 2/9] gdb: factor out regexp compilation from global_symbol_searcher::search

[Simon Marchi]

Factor out the scopes that compile the regexps, in order to make
global_symbol_searcher::search itself simpler to read.

Also, I found the names "preg" and "treg" very unclear, so replace them
with "name_regex" and "type_regex".

Change-Id: Ie4485b2eb593967069ac635c933df817bc5304e6
---
 gdb/symtab.c | 171 +++++++++++++++++++++++++++++----------------------
 gdb/symtab.h |  43 +++++++------
 2 files changed, 122 insertions(+), 92 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 708df078e2ea..c4f76a494abd 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4745,7 +4745,8 @@ global_symbol_searcher::is_suitable_msymbol
 
 bool
 global_symbol_searcher::expand_symtabs
-	(objfile *objfile, const std::optional<compiled_regex> &preg) const
+	(objfile *objfile,
+	 const std::optional<compiled_regex> &name_regex) const
 {
   bool found_func_msymbol_without_debug_info = false;
 
@@ -4762,8 +4763,8 @@ global_symbol_searcher::expand_symtabs
      &lookup_name_info::match_any (),
      [&] (const char *symname)
      {
-       return (!preg.has_value ()
-	       || preg->exec (symname, 0, NULL, 0) == 0);
+       return (!name_regex.has_value ()
+	       || name_regex->exec (symname, 0, NULL, 0) == 0);
      },
      NULL,
      SEARCH_GLOBAL_BLOCK | SEARCH_STATIC_BLOCK,
@@ -4794,9 +4795,9 @@ global_symbol_searcher::expand_symtabs
 
 	  if (is_suitable_msymbol (m_kind, msymbol))
 	    {
-	      if (!preg.has_value ()
-		  || preg->exec (msymbol->natural_name (), 0,
-				 NULL, 0) == 0)
+	      if (!name_regex.has_value ()
+		  || name_regex->exec (msymbol->natural_name (), 0,
+				       NULL, 0) == 0)
 		{
 		  /* An important side-effect of this lookup function is
 		     to expand the symbol table if msymbol is found, later
@@ -4819,8 +4820,8 @@ global_symbol_searcher::expand_symtabs
 bool
 global_symbol_searcher::add_matching_symbols
 	(objfile *objfile,
-	 const std::optional<compiled_regex> &preg,
-	 const std::optional<compiled_regex> &treg,
+	 const std::optional<compiled_regex> &name_regex,
+	 const std::optional<compiled_regex> &type_regex,
 	 std::set<symbol_search> *result_set) const
 {
   domain_search_flags domain = to_search_flags (m_kind);
@@ -4854,14 +4855,15 @@ global_symbol_searcher::add_matching_symbols
 	      if (!sym->matches (domain))
 		continue;
 
-	      if (preg.has_value () && preg->exec (sym->natural_name (), 0,
-						   nullptr, 0) != 0)
+	      if (name_regex.has_value ()
+		  && name_regex->exec (sym->natural_name (), 0,
+				       nullptr, 0) != 0)
 		continue;
 
 	      if (((sym->domain () == VAR_DOMAIN
 		    || sym->domain () == FUNCTION_DOMAIN)
-		   && treg.has_value ()
-		   && !treg_matches_sym_type_name (*treg, sym)))
+		   && type_regex.has_value ()
+		   && !treg_matches_sym_type_name (*type_regex, sym)))
 		continue;
 
 	      if (m_kind == symbol_search_kind::VARIABLE)
@@ -4898,7 +4900,7 @@ global_symbol_searcher::add_matching_symbols
 
 bool
 global_symbol_searcher::add_matching_msymbols
-	(objfile *objfile, const std::optional<compiled_regex> &preg,
+	(objfile *objfile, const std::optional<compiled_regex> &name_regex,
 	 std::vector<symbol_search> *results) const
 {
   for (minimal_symbol *msymbol : objfile->msymbols ())
@@ -4910,9 +4912,9 @@ global_symbol_searcher::add_matching_msymbols
 
       if (is_suitable_msymbol (m_kind, msymbol))
 	{
-	  if (!preg.has_value ()
-	      || preg->exec (msymbol->natural_name (), 0,
-			     NULL, 0) == 0)
+	  if (!name_regex.has_value ()
+	      || name_regex->exec (msymbol->natural_name (), 0,
+				   NULL, 0) == 0)
 	    {
 	      /* For functions we can do a quick check of whether the
 		 symbol might be found via find_pc_symtab.  */
@@ -4938,65 +4940,84 @@ global_symbol_searcher::add_matching_msymbols
   return true;
 }
 
+/* Return the regcomp flags to use for the symbol name and symbol type
+   regexps.  */
+
+static int
+search_regex_cflags ()
+{
+  return REG_NOSUB | (case_sensitivity == case_sensitive_off ? REG_ICASE : 0);
+}
+
+/* See symtab.h.  */
+
+std::optional<compiled_regex>
+global_symbol_searcher::compile_name_regex () const
+{
+  if (m_symbol_name_regexp == nullptr)
+    return {};
+
+  const char *symbol_name_regexp = m_symbol_name_regexp;
+  std::string symbol_name_regexp_holder;
+
+  /* Make sure spacing is right for C++ operators.
+     This is just a courtesy to make the matching less sensitive
+     to how many spaces the user leaves between 'operator'
+     and <TYPENAME> or <OPERATOR>.  */
+  const char *op_end;
+  const char *opname = operator_chars (symbol_name_regexp, &op_end);
+
+  if (*opname)
+    {
+      /* -1 means ok; otherwise number of spaces needed.  */
+      int fix = -1;
+
+      if (c_isalpha (*opname) || *opname == '_' || *opname == '$')
+	{
+	  /* There should 1 space between 'operator' and 'TYPENAME'.  */
+	  if (opname[-1] != ' ' || opname[-2] == ' ')
+	    fix = 1;
+	}
+      else
+	{
+	  /* There should 0 spaces between 'operator' and 'OPERATOR'.  */
+	  if (opname[-1] == ' ')
+	    fix = 0;
+	}
+      /* If wrong number of spaces, fix it.  */
+      if (fix >= 0)
+	{
+	  symbol_name_regexp_holder
+	    = string_printf ("operator%.*s%s", fix, " ", opname);
+	  symbol_name_regexp = symbol_name_regexp_holder.c_str ();
+	}
+    }
+
+  return std::optional<compiled_regex> (std::in_place, symbol_name_regexp,
+					search_regex_cflags (),
+					_("Invalid regexp"));
+}
+
+/* See symtab.h.  */
+
+std::optional<compiled_regex>
+global_symbol_searcher::compile_type_regex () const
+{
+  if (m_symbol_type_regexp == nullptr)
+    return {};
+
+  return std::optional<compiled_regex> (std::in_place, m_symbol_type_regexp,
+					search_regex_cflags (),
+					_("Invalid regexp"));
+}
+
 /* See symtab.h.  */
 
 std::vector<symbol_search>
 global_symbol_searcher::search () const
 {
-  std::optional<compiled_regex> preg;
-  std::optional<compiled_regex> treg;
-
-  if (m_symbol_name_regexp != NULL)
-    {
-      const char *symbol_name_regexp = m_symbol_name_regexp;
-      std::string symbol_name_regexp_holder;
-
-      /* Make sure spacing is right for C++ operators.
-	 This is just a courtesy to make the matching less sensitive
-	 to how many spaces the user leaves between 'operator'
-	 and <TYPENAME> or <OPERATOR>.  */
-      const char *op_end;
-      const char *opname = operator_chars (symbol_name_regexp, &op_end);
-
-      if (*opname)
-	{
-	  int fix = -1;		/* -1 means ok; otherwise number of
-				    spaces needed.  */
-
-	  if (c_isalpha (*opname) || *opname == '_' || *opname == '$')
-	    {
-	      /* There should 1 space between 'operator' and 'TYPENAME'.  */
-	      if (opname[-1] != ' ' || opname[-2] == ' ')
-		fix = 1;
-	    }
-	  else
-	    {
-	      /* There should 0 spaces between 'operator' and 'OPERATOR'.  */
-	      if (opname[-1] == ' ')
-		fix = 0;
-	    }
-	  /* If wrong number of spaces, fix it.  */
-	  if (fix >= 0)
-	    {
-	      symbol_name_regexp_holder
-		= string_printf ("operator%.*s%s", fix, " ", opname);
-	      symbol_name_regexp = symbol_name_regexp_holder.c_str ();
-	    }
-	}
-
-      int cflags = REG_NOSUB | (case_sensitivity == case_sensitive_off
-				? REG_ICASE : 0);
-      preg.emplace (symbol_name_regexp, cflags,
-		    _("Invalid regexp"));
-    }
-
-  if (m_symbol_type_regexp != NULL)
-    {
-      int cflags = REG_NOSUB | (case_sensitivity == case_sensitive_off
-				? REG_ICASE : 0);
-      treg.emplace (m_symbol_type_regexp, cflags,
-		    _("Invalid regexp"));
-    }
+  std::optional<compiled_regex> name_regex = compile_name_regex ();
+  std::optional<compiled_regex> type_regex = compile_type_regex ();
 
   bool found_func_msymbol_without_debug_info = false;
   std::set<symbol_search> result_set;
@@ -5004,13 +5025,15 @@ global_symbol_searcher::search () const
     {
       /* Expand symtabs within objfile that possibly contain matching
 	 symbols.  */
-      found_func_msymbol_without_debug_info |= expand_symtabs (&objfile, preg);
+      found_func_msymbol_without_debug_info
+	|= expand_symtabs (&objfile, name_regex);
 
       /* Find matching symbols within OBJFILE and add them in to the
 	 RESULT_SET set.  Use a set here so that we can easily detect
 	 duplicates as we go, and can therefore track how many unique
 	 matches we have found so far.  */
-      if (!add_matching_symbols (&objfile, preg, treg, &result_set))
+      if (!add_matching_symbols (&objfile, name_regex, type_regex,
+				 &result_set))
 	break;
     }
 
@@ -5025,12 +5048,12 @@ global_symbol_searcher::search () const
   if ((found_func_msymbol_without_debug_info
        || (m_filenames.empty () && m_kind == symbol_search_kind::VARIABLE))
       && !m_exclude_minsyms
-      && !treg.has_value ())
+      && !type_regex.has_value ())
     {
       gdb_assert (m_kind == symbol_search_kind::VARIABLE
 		  || m_kind == symbol_search_kind::FUNCTION);
       for (objfile &objfile : current_program_space->objfiles ())
-	if (!add_matching_msymbols (&objfile, preg, &result))
+	if (!add_matching_msymbols (&objfile, name_regex, &result))
 	  break;
     }
 
diff --git a/gdb/symtab.h b/gdb/symtab.h
index a35449fbeeda..1675b126b19d 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2669,31 +2669,38 @@ class global_symbol_searcher
      of SIZE_MAX, there is no "unlimited".  */
   size_t m_max_search_results = SIZE_MAX;
 
-  /* Expand symtabs in OBJFILE that match PREG, are of type M_KIND.  Return
-     true if any msymbols were seen that we should later consider adding to
-     the results list.  */
-  bool expand_symtabs (objfile *objfile,
-		       const std::optional<compiled_regex> &preg) const;
+  /* Compile M_SYMBOL_NAME_REGEXP, if set.  */
+  std::optional<compiled_regex> compile_name_regex () const;
 
-  /* Add symbols from symtabs in OBJFILE that match PREG, and TREG, and are
-     of type M_KIND, to the results set RESULTS_SET.  Return false if we
-     stop adding results early due to having already found too many results
-     (based on M_MAX_SEARCH_RESULTS limit), otherwise return true.
+  /* Compile M_SYMBOL_TYPE_REGEXP, if set.  */
+  std::optional<compiled_regex> compile_type_regex () const;
+
+  /* Expand symtabs in OBJFILE that match NAME_REGEX, are of type M_KIND.
+     Return true if any msymbols were seen that we should later consider
+     adding to the results list.  */
+  bool expand_symtabs (objfile *objfile,
+		       const std::optional<compiled_regex> &name_regex) const;
+
+  /* Add symbols from symtabs in OBJFILE that match NAME_REGEX, and
+     TYPE_REGEX, and are of type M_KIND, to the results set RESULTS_SET.
+     Return false if we stop adding results early due to having already
+     found too many results (based on M_MAX_SEARCH_RESULTS limit),
+     otherwise return true.
      Returning true does not indicate that any results were added, just
      that we didn't _not_ add a result due to reaching MAX_SEARCH_RESULTS.  */
   bool add_matching_symbols (objfile *objfile,
-			     const std::optional<compiled_regex> &preg,
-			     const std::optional<compiled_regex> &treg,
+			     const std::optional<compiled_regex> &name_regex,
+			     const std::optional<compiled_regex> &type_regex,
 			     std::set<symbol_search> *result_set) const;
 
-  /* Add msymbols from OBJFILE that match PREG and M_KIND, to the results
-     vector RESULTS.  Return false if we stop adding results early due to
-     having already found too many results (based on max search results
-     limit M_MAX_SEARCH_RESULTS), otherwise return true.  Returning true
-     does not indicate that any results were added, just that we didn't
-     _not_ add a result due to reaching MAX_SEARCH_RESULTS.  */
+  /* Add msymbols from OBJFILE that match NAME_REGEX and M_KIND, to the
+     results vector RESULTS.  Return false if we stop adding results early
+     due to having already found too many results (based on max search
+     results limit M_MAX_SEARCH_RESULTS), otherwise return true.  Returning
+     true does not indicate that any results were added, just that we
+     didn't _not_ add a result due to reaching MAX_SEARCH_RESULTS.  */
   bool add_matching_msymbols (objfile *objfile,
-			      const std::optional<compiled_regex> &preg,
+			      const std::optional<compiled_regex> &name_regex,
 			      std::vector<symbol_search> *results) const;
 
   /* Return true if MSYMBOL is of type KIND.  */
-- 
2.55.0

[PATCH 3/9] gdb: simplify insertion in global_symbol_searcher::add_matching_symbols

[Simon Marchi]

std::set::insert does nothing if an equivalent element is already
in the set, so checking with find first is redundant.  Replace find+insert
with emplace.

Change-Id: I869284407cfefb9b2d7ad7305fe04fbdaee25f1f
---
 gdb/symtab.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index c4f76a494abd..4b2c4be8d117 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4880,13 +4880,9 @@ global_symbol_searcher::add_matching_symbols
 	      if (sym->domain () == MODULE_DOMAIN && sym->line () == 0)
 		continue;
 
+	      /* Match, insert if not already in the results.  */
 	      if (result_set->size () < m_max_search_results)
-		{
-		  /* Match, insert if not already in the results.  */
-		  symbol_search ss (block, sym);
-		  if (result_set->find (ss) == result_set->end ())
-		    result_set->insert (ss);
-		}
+		result_set->emplace (block, sym);
 	      else
 		return false;
 	    }
-- 
2.55.0

[PATCH 4/9] gdb: use a switch on m_kind in add_matching_symbols

[Simon Marchi]

add_matching_symbols applies some filters that only make sense for some
kinds of searches: the type regexp for variables and functions, the
LOC_UNRESOLVED / enum constant filter for variables, and the line number
filter for modules.  Some of these check m_kind, others check the
symbol's domain.  Since the earlier sym->matches check only lets through
symbols whose domain corresponds to m_kind, the two are equivalent.

Group these filters in a single switch on m_kind.  Add a comment for
the module filter, which skips module declarations, because it wasn't
obvious to me at first sight.

Change-Id: Idf1b366b4cfd16187848b64075a0067190c111fc
---
 gdb/symtab.c | 33 +++++++++++++++++++--------------
 1 file changed, 19 insertions(+), 14 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 4b2c4be8d117..2bb1cbf1b9ad 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4860,25 +4860,30 @@ global_symbol_searcher::add_matching_symbols
 				       nullptr, 0) != 0)
 		continue;
 
-	      if (((sym->domain () == VAR_DOMAIN
-		    || sym->domain () == FUNCTION_DOMAIN)
-		   && type_regex.has_value ()
-		   && !treg_matches_sym_type_name (*type_regex, sym)))
-		continue;
-
-	      if (m_kind == symbol_search_kind::VARIABLE)
+	      switch (m_kind)
 		{
+		case symbol_search_kind::VARIABLE:
+		  /* LOC_CONST can be used for more than just enums, e.g.,
+		     C++ static const members.  We only want to skip enums
+		     here.  */
 		  if (sym->loc_class () == LOC_UNRESOLVED
-		      /* LOC_CONST can be used for more than
-			 just enums, e.g., c++ static const
-			 members.  We only want to skip enums
-			 here.  */
 		      || (sym->loc_class () == LOC_CONST
-			  && (sym->type ()->code () == TYPE_CODE_ENUM)))
+			  && sym->type ()->code () == TYPE_CODE_ENUM))
 		    continue;
+		  [[fallthrough]];
+
+		case symbol_search_kind::FUNCTION:
+		  if (type_regex.has_value ()
+		      && !treg_matches_sym_type_name (*type_regex, sym))
+		    continue;
+		  break;
+
+		case symbol_search_kind::MODULE:
+		  /* Skip module declarations, keep only definitions.  */
+		  if (sym->line () == 0)
+		    continue;
+		  break;
 		}
-	      if (sym->domain () == MODULE_DOMAIN && sym->line () == 0)
-		continue;
 
 	      /* Match, insert if not already in the results.  */
 	      if (result_set->size () < m_max_search_results)
-- 
2.55.0

[PATCH 5/9] gdb: make global_symbol_searcher::is_suitable_msymbol non-static

[Simon Marchi]

is_suitable_msymbol is a static method taking the search kind as a
parameter, but all callers pass m_kind.  Make it a const non-static
method that uses m_kind directly.

Change-Id: I803e7cddbb317a8af35a82ea9e528c64f6326f31
---
 gdb/symtab.c | 10 +++++-----
 gdb/symtab.h |  5 ++---
 2 files changed, 7 insertions(+), 8 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 2bb1cbf1b9ad..9ce6911666c6 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4722,7 +4722,7 @@ to_search_flags (symbol_search_kind kind)
 
 bool
 global_symbol_searcher::is_suitable_msymbol
-	(symbol_search_kind kind, const minimal_symbol *msymbol)
+	(const minimal_symbol *msymbol) const
 {
   switch (msymbol->type ())
     {
@@ -4730,12 +4730,12 @@ global_symbol_searcher::is_suitable_msymbol
     case mst_bss:
     case mst_file_data:
     case mst_file_bss:
-      return kind == symbol_search_kind::VARIABLE;
+      return m_kind == symbol_search_kind::VARIABLE;
     case mst_text:
     case mst_file_text:
     case mst_solib_trampoline:
     case mst_text_gnu_ifunc:
-      return kind == symbol_search_kind::FUNCTION;
+      return m_kind == symbol_search_kind::FUNCTION;
     default:
       return false;
     }
@@ -4793,7 +4793,7 @@ global_symbol_searcher::expand_symtabs
 	  if (msymbol->created_by_gdb)
 	    continue;
 
-	  if (is_suitable_msymbol (m_kind, msymbol))
+	  if (is_suitable_msymbol (msymbol))
 	    {
 	      if (!name_regex.has_value ()
 		  || name_regex->exec (msymbol->natural_name (), 0,
@@ -4911,7 +4911,7 @@ global_symbol_searcher::add_matching_msymbols
       if (msymbol->created_by_gdb)
 	continue;
 
-      if (is_suitable_msymbol (m_kind, msymbol))
+      if (is_suitable_msymbol (msymbol))
 	{
 	  if (!name_regex.has_value ()
 	      || name_regex->exec (msymbol->natural_name (), 0,
diff --git a/gdb/symtab.h b/gdb/symtab.h
index 1675b126b19d..50d7000681df 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2703,9 +2703,8 @@ class global_symbol_searcher
 			      const std::optional<compiled_regex> &name_regex,
 			      std::vector<symbol_search> *results) const;
 
-  /* Return true if MSYMBOL is of type KIND.  */
-  static bool is_suitable_msymbol (symbol_search_kind kind,
-				   const minimal_symbol *msymbol);
+  /* Return true if MSYMBOL is of type M_KIND.  */
+  bool is_suitable_msymbol (const minimal_symbol *msymbol) const;
 };
 
 /* When searching for Fortran symbols within modules (functions/variables)
-- 
2.55.0

[PATCH 6/9] gdb: update comment of global_symbol_searcher::expand_symtabs

[Simon Marchi]

Update the comment of expand_symtabs to describe more accurately what
the return value means.  It is true when a matching function msymbol for
which no debug info exists is found.

Change-Id: I7c26c065b31575066ed2398563cdc1dac8869f8b
---
 gdb/symtab.h | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/gdb/symtab.h b/gdb/symtab.h
index 50d7000681df..acbe41d131b7 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2676,8 +2676,9 @@ class global_symbol_searcher
   std::optional<compiled_regex> compile_type_regex () const;
 
   /* Expand symtabs in OBJFILE that match NAME_REGEX, are of type M_KIND.
-     Return true if any msymbols were seen that we should later consider
-     adding to the results list.  */
+     Return true if a matching function msymbol without debug info was
+     seen, meaning that msymbols should later be considered for the results
+     list.  */
   bool expand_symtabs (objfile *objfile,
 		       const std::optional<compiled_regex> &name_regex) const;
 
-- 
2.55.0

[PATCH 7/9] gdb: add matching helpers to global_symbol_searcher

[Simon Marchi]

The methods of global_symbol_searcher repeat the same checks in several
places.  Add some helpers to make the code easier to follow:

 - name_matches: check a name against the (optional) name regexp

 - symtab_matches_filenames: check a symtab against the searched file
   names

 - symbol_matches: all the filtering done on full symbols in
   add_matching_symbols

 - msymbol_matches: the filtering done on minimal symbols, shared by
   expand_symtabs and add_matching_msymbols

Switch expand_symtabs, add_matching_symbols and add_matching_msymbols to
use early continues to avoid deeply nested ifs.

Change-Id: I553f70fb8ea9c7e55b89bb1204a402938c116c3b
---
 gdb/symtab.c | 205 ++++++++++++++++++++++++++++-----------------------
 gdb/symtab.h |  13 ++++
 2 files changed, 127 insertions(+), 91 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 9ce6911666c6..79e43eafde18 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4741,6 +4741,93 @@ global_symbol_searcher::is_suitable_msymbol
     }
 }
 
+/* Return true if NAME matches NAME_REGEX, or if NAME_REGEX is empty.  */
+
+static bool
+name_matches (const std::optional<compiled_regex> &name_regex,
+	      const char *name)
+{
+  return (!name_regex.has_value ()
+	  || name_regex->exec (name, 0, nullptr, 0) == 0);
+}
+
+/* See symtab.h.  */
+
+bool
+global_symbol_searcher::symtab_matches_filenames (symtab *symtab) const
+{
+  /* Check first sole SYMTAB->FILENAME.  It does not need to be a substring
+     of symtab_to_fullname as it may contain "./" etc.  */
+  if (file_matches (symtab->filename (), m_filenames, false))
+    return true;
+
+  if (!basenames_may_differ
+      && !file_matches (lbasename (symtab->filename ()), m_filenames, true))
+    return false;
+
+  return file_matches (symtab_to_fullname (symtab), m_filenames, false);
+}
+
+/* See symtab.h.  */
+
+bool
+global_symbol_searcher::symbol_matches
+	(const symbol *sym,
+	 const std::optional<compiled_regex> &name_regex,
+	 const std::optional<compiled_regex> &type_regex) const
+{
+  if (!symtab_matches_filenames (sym->symtab ()))
+    return false;
+
+  if (!sym->matches (to_search_flags (m_kind)))
+    return false;
+
+  if (!name_matches (name_regex, sym->natural_name ()))
+    return false;
+
+  switch (m_kind)
+    {
+    case symbol_search_kind::VARIABLE:
+      /* LOC_CONST can be used for more than just enums, e.g., C++ static
+	 const members.  We only want to skip enums here.  */
+      if (sym->loc_class () == LOC_UNRESOLVED
+	  || (sym->loc_class () == LOC_CONST
+	      && sym->type ()->code () == TYPE_CODE_ENUM))
+	return false;
+      [[fallthrough]];
+
+    case symbol_search_kind::FUNCTION:
+      if (type_regex.has_value ()
+	  && !treg_matches_sym_type_name (*type_regex, sym))
+	return false;
+      break;
+
+    case symbol_search_kind::MODULE:
+      /* Skip module declarations, keep only definitions.  */
+      if (sym->line () == 0)
+	return false;
+      break;
+    }
+
+  return true;
+}
+
+/* See symtab.h.  */
+
+bool
+global_symbol_searcher::msymbol_matches
+	(const minimal_symbol *msymbol,
+	 const std::optional<compiled_regex> &name_regex) const
+{
+  if (msymbol->created_by_gdb)
+    return false;
+
+  if (!is_suitable_msymbol (msymbol))
+    return false;
+
+  return name_matches (name_regex, msymbol->natural_name ());
+}
+
 /* See symtab.h.  */
 
 bool
@@ -4763,8 +4850,7 @@ global_symbol_searcher::expand_symtabs
      &lookup_name_info::match_any (),
      [&] (const char *symname)
      {
-       return (!name_regex.has_value ()
-	       || name_regex->exec (symname, 0, NULL, 0) == 0);
+       return name_matches (name_regex, symname);
      },
      NULL,
      SEARCH_GLOBAL_BLOCK | SEARCH_STATIC_BLOCK,
@@ -4790,25 +4876,16 @@ global_symbol_searcher::expand_symtabs
 	{
 	  QUIT;
 
-	  if (msymbol->created_by_gdb)
+	  if (!msymbol_matches (msymbol, name_regex))
 	    continue;
 
-	  if (is_suitable_msymbol (msymbol))
-	    {
-	      if (!name_regex.has_value ()
-		  || name_regex->exec (msymbol->natural_name (), 0,
-				       NULL, 0) == 0)
-		{
-		  /* An important side-effect of this lookup function is
-		     to expand the symbol table if msymbol is found, later
-		     in the process we will add matching symbols or
-		     msymbols to the results list, and that requires that
-		     the symbols tables are expanded.  */
-		  if (find_compunit_symtab_for_pc
-			(msymbol->value_address (objfile)) == nullptr)
-		    found_func_msymbol_without_debug_info = true;
-		}
-	    }
+	  /* An important side-effect of this lookup function is to expand
+	     the symbol table if msymbol is found, later in the process we
+	     will add matching symbols or msymbols to the results list, and
+	     that requires that the symbols tables are expanded.  */
+	  if (find_compunit_symtab_for_pc
+		(msymbol->value_address (objfile)) == nullptr)
+	    found_func_msymbol_without_debug_info = true;
 	}
     }
 
@@ -4824,8 +4901,6 @@ global_symbol_searcher::add_matching_symbols
 	 const std::optional<compiled_regex> &type_regex,
 	 std::set<symbol_search> *result_set) const
 {
-  domain_search_flags domain = to_search_flags (m_kind);
-
   /* Add matching symbols (if not already present).  */
   for (compunit_symtab &cust : objfile->compunits ())
     {
@@ -4837,54 +4912,11 @@ global_symbol_searcher::add_matching_symbols
 
 	  for (struct symbol *sym : block_iterator_range (b))
 	    {
-	      struct symtab *real_symtab = sym->symtab ();
-
 	      QUIT;
 
-	      /* Check first sole REAL_SYMTAB->FILENAME.  It does
-		 not need to be a substring of symtab_to_fullname as
-		 it may contain "./" etc.  */
-	      if (!(file_matches (real_symtab->filename (), m_filenames, false)
-		    || ((basenames_may_differ
-			 || file_matches (lbasename (real_symtab->filename ()),
-					  m_filenames, true))
-			&& file_matches (symtab_to_fullname (real_symtab),
-					 m_filenames, false))))
+	      if (!symbol_matches (sym, name_regex, type_regex))
 		continue;
 
-	      if (!sym->matches (domain))
-		continue;
-
-	      if (name_regex.has_value ()
-		  && name_regex->exec (sym->natural_name (), 0,
-				       nullptr, 0) != 0)
-		continue;
-
-	      switch (m_kind)
-		{
-		case symbol_search_kind::VARIABLE:
-		  /* LOC_CONST can be used for more than just enums, e.g.,
-		     C++ static const members.  We only want to skip enums
-		     here.  */
-		  if (sym->loc_class () == LOC_UNRESOLVED
-		      || (sym->loc_class () == LOC_CONST
-			  && sym->type ()->code () == TYPE_CODE_ENUM))
-		    continue;
-		  [[fallthrough]];
-
-		case symbol_search_kind::FUNCTION:
-		  if (type_regex.has_value ()
-		      && !treg_matches_sym_type_name (*type_regex, sym))
-		    continue;
-		  break;
-
-		case symbol_search_kind::MODULE:
-		  /* Skip module declarations, keep only definitions.  */
-		  if (sym->line () == 0)
-		    continue;
-		  break;
-		}
-
 	      /* Match, insert if not already in the results.  */
 	      if (result_set->size () < m_max_search_results)
 		result_set->emplace (block, sym);
@@ -4908,34 +4940,25 @@ global_symbol_searcher::add_matching_msymbols
     {
       QUIT;
 
-      if (msymbol->created_by_gdb)
+      if (!msymbol_matches (msymbol, name_regex))
 	continue;
 
-      if (is_suitable_msymbol (msymbol))
-	{
-	  if (!name_regex.has_value ()
-	      || name_regex->exec (msymbol->natural_name (), 0,
-				   NULL, 0) == 0)
-	    {
-	      /* For functions we can do a quick check of whether the
-		 symbol might be found via find_pc_symtab.  */
-	      if (m_kind != symbol_search_kind::FUNCTION
-		  || (find_compunit_symtab_for_pc
-		      (msymbol->value_address (objfile)) == NULL))
-		{
-		  if (lookup_symbol_in_objfile_from_linkage_name
-		      (objfile, msymbol->linkage_name (),
-		       SEARCH_VFT).symbol == NULL)
-		    {
-		      /* Matching msymbol, add it to the results list.  */
-		      if (results->size () < m_max_search_results)
-			results->emplace_back (GLOBAL_BLOCK, msymbol, objfile);
-		      else
-			return false;
-		    }
-		}
-	    }
-	}
+      /* For functions we can do a quick check of whether the
+	 symbol might be found via find_pc_symtab.  */
+      if (m_kind == symbol_search_kind::FUNCTION
+	  && (find_compunit_symtab_for_pc
+	      (msymbol->value_address (objfile)) != NULL))
+	continue;
+
+      if (lookup_symbol_in_objfile_from_linkage_name
+	    (objfile, msymbol->linkage_name (), SEARCH_VFT).symbol != NULL)
+	continue;
+
+      /* Matching msymbol, add it to the results list.  */
+      if (results->size () < m_max_search_results)
+	results->emplace_back (GLOBAL_BLOCK, msymbol, objfile);
+      else
+	return false;
     }
 
   return true;
diff --git a/gdb/symtab.h b/gdb/symtab.h
index acbe41d131b7..5dd0475e4491 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2675,6 +2675,19 @@ class global_symbol_searcher
   /* Compile M_SYMBOL_TYPE_REGEXP, if set.  */
   std::optional<compiled_regex> compile_type_regex () const;
 
+  /* Return true if SYMTAB matches one of M_FILENAMES.  */
+  bool symtab_matches_filenames (symtab *symtab) const;
+
+  /* Return true if SYM is of type M_KIND, is in a file matching
+     M_FILENAMES, and matches NAME_REGEX and TYPE_REGEX.  */
+  bool symbol_matches (const symbol *sym,
+		       const std::optional<compiled_regex> &name_regex,
+		       const std::optional<compiled_regex> &type_regex) const;
+
+  /* Return true if MSYMBOL is of type M_KIND and matches NAME_REGEX.  */
+  bool msymbol_matches (const minimal_symbol *msymbol,
+			const std::optional<compiled_regex> &name_regex) const;
+
   /* Expand symtabs in OBJFILE that match NAME_REGEX, are of type M_KIND.
      Return true if a matching function msymbol without debug info was
      seen, meaning that msymbols should later be considered for the results
-- 
2.55.0

[PATCH 8/9] gdb: pass result containers by reference in global_symbol_searcher

[Simon Marchi]

add_matching_symbols and add_matching_msymbols take the container to
add results to as a pointer, but it can never be nullptr.  Pass it by
reference instead.

Change-Id: I65d8e14f2f3152d83afbe3b348fc91a8817b87fa
---
 gdb/symtab.c | 17 ++++++++---------
 gdb/symtab.h |  4 ++--
 2 files changed, 10 insertions(+), 11 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 79e43eafde18..10a794004d33 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4899,7 +4899,7 @@ global_symbol_searcher::add_matching_symbols
 	(objfile *objfile,
 	 const std::optional<compiled_regex> &name_regex,
 	 const std::optional<compiled_regex> &type_regex,
-	 std::set<symbol_search> *result_set) const
+	 std::set<symbol_search> &result_set) const
 {
   /* Add matching symbols (if not already present).  */
   for (compunit_symtab &cust : objfile->compunits ())
@@ -4918,8 +4918,8 @@ global_symbol_searcher::add_matching_symbols
 		continue;
 
 	      /* Match, insert if not already in the results.  */
-	      if (result_set->size () < m_max_search_results)
-		result_set->emplace (block, sym);
+	      if (result_set.size () < m_max_search_results)
+		result_set.emplace (block, sym);
 	      else
 		return false;
 	    }
@@ -4934,7 +4934,7 @@ global_symbol_searcher::add_matching_symbols
 bool
 global_symbol_searcher::add_matching_msymbols
 	(objfile *objfile, const std::optional<compiled_regex> &name_regex,
-	 std::vector<symbol_search> *results) const
+	 std::vector<symbol_search> &results) const
 {
   for (minimal_symbol *msymbol : objfile->msymbols ())
     {
@@ -4955,8 +4955,8 @@ global_symbol_searcher::add_matching_msymbols
 	continue;
 
       /* Matching msymbol, add it to the results list.  */
-      if (results->size () < m_max_search_results)
-	results->emplace_back (GLOBAL_BLOCK, msymbol, objfile);
+      if (results.size () < m_max_search_results)
+	results.emplace_back (GLOBAL_BLOCK, msymbol, objfile);
       else
 	return false;
     }
@@ -5056,8 +5056,7 @@ global_symbol_searcher::search () const
 	 RESULT_SET set.  Use a set here so that we can easily detect
 	 duplicates as we go, and can therefore track how many unique
 	 matches we have found so far.  */
-      if (!add_matching_symbols (&objfile, name_regex, type_regex,
-				 &result_set))
+      if (!add_matching_symbols (&objfile, name_regex, type_regex, result_set))
 	break;
     }
 
@@ -5077,7 +5076,7 @@ global_symbol_searcher::search () const
       gdb_assert (m_kind == symbol_search_kind::VARIABLE
 		  || m_kind == symbol_search_kind::FUNCTION);
       for (objfile &objfile : current_program_space->objfiles ())
-	if (!add_matching_msymbols (&objfile, name_regex, &result))
+	if (!add_matching_msymbols (&objfile, name_regex, result))
 	  break;
     }
 
diff --git a/gdb/symtab.h b/gdb/symtab.h
index 5dd0475e4491..e382b64b427c 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2705,7 +2705,7 @@ class global_symbol_searcher
   bool add_matching_symbols (objfile *objfile,
 			     const std::optional<compiled_regex> &name_regex,
 			     const std::optional<compiled_regex> &type_regex,
-			     std::set<symbol_search> *result_set) const;
+			     std::set<symbol_search> &result_set) const;
 
   /* Add msymbols from OBJFILE that match NAME_REGEX and M_KIND, to the
      results vector RESULTS.  Return false if we stop adding results early
@@ -2715,7 +2715,7 @@ class global_symbol_searcher
      didn't _not_ add a result due to reaching MAX_SEARCH_RESULTS.  */
   bool add_matching_msymbols (objfile *objfile,
 			      const std::optional<compiled_regex> &name_regex,
-			      std::vector<symbol_search> *results) const;
+			      std::vector<symbol_search> &results) const;
 
   /* Return true if MSYMBOL is of type M_KIND.  */
   bool is_suitable_msymbol (const minimal_symbol *msymbol) const;
-- 
2.55.0

[PATCH 9/9] gdb: fix some comments in global_symbol_searcher

[Simon Marchi]

Fix a few comments in and around global_symbol_searcher:

 - The comment of global_symbol_searcher::m_exclude_minsyms refers to
   M_SYMBOL_REGEXP, which does not (or no longer) exist.  It means
   M_SYMBOL_NAME_REGEXP.

 - The comment in global_symbol_searcher::add_matching_msymbols refers
   to find_pc_symtab, which does not exist anymore.  The code calls
   find_compunit_symtab_for_pc.

 - The comment of global_symbol_searcher calls it a structure, but it
   is a class.

Change-Id: I0ec18651688406bb525752a53b9558096a798aef
---
 gdb/symtab.c | 2 +-
 gdb/symtab.h | 8 ++++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/gdb/symtab.c b/gdb/symtab.c
index 10a794004d33..611815a568d6 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -4944,7 +4944,7 @@ global_symbol_searcher::add_matching_msymbols
 	continue;
 
       /* For functions we can do a quick check of whether the
-	 symbol might be found via find_pc_symtab.  */
+	 symbol might be found via find_compunit_symtab_for_pc.  */
       if (m_kind == symbol_search_kind::FUNCTION
 	  && (find_compunit_symtab_for_pc
 	      (msymbol->value_address (objfile)) != NULL))
diff --git a/gdb/symtab.h b/gdb/symtab.h
index e382b64b427c..1d92907fce2c 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -2604,8 +2604,8 @@ enum class symbol_search_kind
 };
 
 /* In order to search for global symbols of a particular kind matching
-   particular regular expressions, create an instance of this structure and
-   call the SEARCH member function.  */
+   particular regular expressions, create an instance of this class and call
+   the SEARCH member function.  */
 class global_symbol_searcher
 {
 public:
@@ -2661,8 +2661,8 @@ class global_symbol_searcher
   /* Regular expression to match against the symbol type.  */
   const char *m_symbol_type_regexp = nullptr;
 
-  /* When this flag is false then minsyms that match M_SYMBOL_REGEXP will
-     be included in the results, otherwise they are excluded.  */
+  /* When this flag is false then minsyms that match M_SYMBOL_NAME_REGEXP
+     will be included in the results, otherwise they are excluded.  */
   bool m_exclude_minsyms = false;
 
   /* Maximum number of search results.  We currently impose a hard limit
-- 
2.55.0