Re: [PATCH 04/40] Fix TAB-completion + .gdb_index slowness (generalize filename_seen_cache)

[Pedro Alves <palves@redhat.com>]

On 07/13/2017 09:41 PM, Keith Seitz wrote:


>> +
>> +     NOTE: We don't manage space for FILE, we assume FILE lives as
>> +     long as the caller needs.  */
>> +  bool filename_seen (const char *file, bool add);
> 
> I'm not really a fan of this style of interface. [I realize we've done this for many years.] When reading through code that uses this, e.g.,

Yeah...  I'm not a fan either.  It's known as "the boolean trap".
This one was preexisting, and I should have probably tried
to fix it, because then I'd have realized then that _all_ callers
pass "true".  :-)  So we can just remove the parameter.
I did that now.  I also shortened the method name from "filename_seen" 
to just "seen", since the prefix wasn't really adding
anything.

> 
>    bool seen = cache.filename_seen (file, true);
> 
> that "true" doesn't really tell me what that means, and in the context of the method itself, it isn't obvious (to me, at least) what "true" might mean when asking if we've seen a filename. A reader could think that it somehow altered how the search was performed, e.g., case-insensitive or whatnot.
> 
> Now if this read
> 
>   bool seen = cache.filename_seen (file, ADD);
> 
> That reveals much more about what is going to happen.
> 
> But that's nitpicky style comment (aka there's no reason to change anything). I didn't otherwise see any issues with the patch.

I was about to push it when I realized that I missed releasing the
new heap-allocated dwarf2_per_objfile::filenames_cache...

To address that, I made it possible for dwarf2_per_objfile to be
a non-POD:

  https://sourceware.org/ml/gdb-patches/2017-07/msg00202.html

and made dwarf2_per_objfile::filenames_cache an optional, so the
cache object (though not its elements) is on the obstack too.

Here's the updated patch.

From b83f66dac43ec3458e0bd4b3883437aefe28d1b6 Mon Sep 17 00:00:00 2001
From: Pedro Alves <palves@redhat.com>
Date: Fri, 14 Jul 2017 20:24:02 +0100
Subject: [PATCH] Fix TAB-completion + .gdb_index slowness (generalize
 filename_seen_cache)

Tab completion when debugging a program binary that uses GDB index is
surprisingly much slower than when GDB uses psymtabs instead.  Around
1.5x/3x slower.  That's surprising, because the whole point of GDB
index is to speed things up...

For example, with:

 set pagination off
 set $count = 0
 while $count < 400
   complete b string_prin         # matches gdb's string_printf
   printf "count = %d\n", $count
   set $count = $count + 1
 end

 $ time ./gdb --batch -q  ./gdb-with-index -ex "source script.cmd"
 real    0m11.042s
 user    0m10.920s
 sys     0m0.042s

 $ time ./gdb --batch -q  ./gdb-without-index -ex "source script.cmd"
 real    0m4.635s
 user    0m4.590s
 sys     0m0.037s

Same but with:
 -   complete b string_prin
 +   complete b zzzzzz
to exercise the no-matches worst case, master currently gets you
something like:

 with index           without index
 real    0m11.971s    0m8.413s
 user    0m11.912s    0m8.355s
 sys     0m0.035s     0m0.035s

Running gdb under perf shows 80% spent inside
maybe_add_partial_symtab_filename, and 20% spent in the lbasename
inside that.

The problem that tab completion walks over all compunit symtabs, and
for each, walks the contained file symtabs.  And there a huge number
of file symtabs (each included system header, etc.) that appear in
each compunit symtab's file symtab list.  As in, when debugging GDB, I
have 367381 symtabs iterated, when of those only 5371 filenames are
unique...

This was a regression from the earlier (nice) split of symtabs in
compunit symtabs + file symtabs.

The fix here is to add a cache of unique filenames per objfile so that
the walk / uniquing is only done once.  There's already a abstraction
for this in symtab.c; this patch moves that code out to a separate
file and C++ifies it bit.

This makes the worst-case scenario above consistently drop to ~2.5s
(1.5s for the "string_prin" hit case), making it over 3.3x times
faster than psymtabs in this use case (7x in the "string_prin" hit
case).

gdb/ChangeLog:
2017-07-14  Pedro Alves  <palves@redhat.com>

	* Makefile.in (COMMON_OBS): Add filename-seen-cache.o.
	* dwarf2read.c: Include "filename-seen-cache.h".
	* dwarf2read.c (dwarf2_per_objfile) <filenames_cache>: New field.
	(dw2_map_symbol_filenames): Build and use a filenames_seen_cache.
	* filename-seen-cache.c: New file.
	* filename-seen-cache.h: New file.
	* symtab.c: Include "filename-seen-cache.h".
	(struct filename_seen_cache, INITIAL_FILENAME_SEEN_CACHE_SIZE)
	(create_filename_seen_cache, clear_filename_seen_cache)
	(delete_filename_seen_cache, filename_seen): Delete, parts moved
	to filename-seen-cache.h/filename-seen-cache.c.
	(output_source_filename, sources_info)
	(maybe_add_partial_symtab_filename)
	(make_source_files_completion_list): Adjust to use
	filename_seen_cache.
---
 gdb/ChangeLog             |  18 ++++++++
 gdb/Makefile.in           |   1 +
 gdb/dwarf2read.c          | 104 ++++++++++++++++++++++++++--------------------
 gdb/filename-seen-cache.c |  66 +++++++++++++++++++++++++++++
 gdb/filename-seen-cache.h |  63 ++++++++++++++++++++++++++++
 gdb/symtab.c              | 100 ++++++--------------------------------------
 6 files changed, 220 insertions(+), 132 deletions(-)
 create mode 100644 gdb/filename-seen-cache.c
 create mode 100644 gdb/filename-seen-cache.h

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index 8c6a4f4..c9eae0a 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,5 +1,23 @@
 2017-07-14  Pedro Alves  <palves@redhat.com>
 
+	* Makefile.in (COMMON_OBS): Add filename-seen-cache.o.
+	* dwarf2read.c: Include "filename-seen-cache.h".
+	* dwarf2read.c (dwarf2_per_objfile) <filenames_cache>: New field.
+	(dw2_map_symbol_filenames): Build and use a filenames_seen_cache.
+	* filename-seen-cache.c: New file.
+	* filename-seen-cache.h: New file.
+	* symtab.c: Include "filename-seen-cache.h".
+	(struct filename_seen_cache, INITIAL_FILENAME_SEEN_CACHE_SIZE)
+	(create_filename_seen_cache, clear_filename_seen_cache)
+	(delete_filename_seen_cache, filename_seen): Delete, parts moved
+	to filename-seen-cache.h/filename-seen-cache.c.
+	(output_source_filename, sources_info)
+	(maybe_add_partial_symtab_filename)
+	(make_source_files_completion_list): Adjust to use
+	filename_seen_cache.
+
+2017-07-14  Pedro Alves  <palves@redhat.com>
+
 	* symtab.c (make_file_symbol_completion_list_1): Iterate over
 	symtabs matching all symtabs with SRCFILE as file name instead of
 	only considering the first hit, with lookup_symtab.
diff --git a/gdb/Makefile.in b/gdb/Makefile.in
index b27f698..c6e618a 100644
--- a/gdb/Makefile.in
+++ b/gdb/Makefile.in
@@ -1714,6 +1714,7 @@ COMMON_OBS = $(DEPFILES) $(CONFIG_OBS) $(YYOBJ) \
 	f-typeprint.o \
 	f-valprint.o \
 	fileio.o \
+	filename-seen-cache.o \
 	filestuff.o \
 	filesystem.o \
 	findcmd.o \
diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index 334adaf..beb3401 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -74,7 +74,7 @@
 #include "common/gdb_optional.h"
 #include "common/underlying.h"
 #include "common/byte-vector.h"
-
+#include "filename-seen-cache.h"
 #include <fcntl.h>
 #include <sys/types.h>
 #include <algorithm>
@@ -345,6 +345,9 @@ public:
 
   /* Table containing line_header indexed by offset and offset_in_dwz.  */
   htab_t line_header_hash {};
+
+  /* Table containing all filenames.  */
+  gdb::optional<filename_seen_cache> filenames_cache;
 };
 
 static struct dwarf2_per_objfile *dwarf2_per_objfile;
@@ -4311,64 +4314,75 @@ static void
 dw2_map_symbol_filenames (struct objfile *objfile, symbol_filename_ftype *fun,
 			  void *data, int need_fullname)
 {
-  int i;
-  htab_up visited (htab_create_alloc (10, htab_hash_pointer, htab_eq_pointer,
-				      NULL, xcalloc, xfree));
-
   dw2_setup (objfile);
 
-  /* The rule is CUs specify all the files, including those used by
-     any TU, so there's no need to scan TUs here.
-     We can ignore file names coming from already-expanded CUs.  */
-
-  for (i = 0; i < dwarf2_per_objfile->n_comp_units; ++i)
+  if (!dwarf2_per_objfile->filenames_cache)
     {
-      struct dwarf2_per_cu_data *per_cu = dw2_get_cutu (i);
+      dwarf2_per_objfile->filenames_cache.emplace ();
 
-      if (per_cu->v.quick->compunit_symtab)
-	{
-	  void **slot = htab_find_slot (visited.get (),
-					per_cu->v.quick->file_names,
-					INSERT);
+      htab_up visited (htab_create_alloc (10,
+					  htab_hash_pointer, htab_eq_pointer,
+					  NULL, xcalloc, xfree));
 
-	  *slot = per_cu->v.quick->file_names;
-	}
-    }
-
-  for (i = 0; i < dwarf2_per_objfile->n_comp_units; ++i)
-    {
-      int j;
-      struct dwarf2_per_cu_data *per_cu = dw2_get_cu (i);
-      struct quick_file_names *file_data;
-      void **slot;
+      /* The rule is CUs specify all the files, including those used
+	 by any TU, so there's no need to scan TUs here.  We can
+	 ignore file names coming from already-expanded CUs.  */
 
-      /* We only need to look at symtabs not already expanded.  */
-      if (per_cu->v.quick->compunit_symtab)
-	continue;
+      for (int i = 0; i < dwarf2_per_objfile->n_comp_units; ++i)
+	{
+	  struct dwarf2_per_cu_data *per_cu = dw2_get_cutu (i);
 
-      file_data = dw2_get_file_names (per_cu);
-      if (file_data == NULL)
-	continue;
+	  if (per_cu->v.quick->compunit_symtab)
+	    {
+	      void **slot = htab_find_slot (visited.get (),
+					    per_cu->v.quick->file_names,
+					    INSERT);
 
-      slot = htab_find_slot (visited.get (), file_data, INSERT);
-      if (*slot)
-	{
-	  /* Already visited.  */
-	  continue;
+	      *slot = per_cu->v.quick->file_names;
+	    }
 	}
-      *slot = file_data;
 
-      for (j = 0; j < file_data->num_file_names; ++j)
+      for (int i = 0; i < dwarf2_per_objfile->n_comp_units; ++i)
 	{
-	  const char *this_real_name;
+	  int j;
+	  struct dwarf2_per_cu_data *per_cu = dw2_get_cu (i);
+	  struct quick_file_names *file_data;
+	  void **slot;
 
-	  if (need_fullname)
-	    this_real_name = dw2_get_real_path (objfile, file_data, j);
-	  else
-	    this_real_name = NULL;
-	  (*fun) (file_data->file_names[j], this_real_name, data);
+	  /* We only need to look at symtabs not already expanded.  */
+	  if (per_cu->v.quick->compunit_symtab)
+	    continue;
+
+	  file_data = dw2_get_file_names (per_cu);
+	  if (file_data == NULL)
+	    continue;
+
+	  slot = htab_find_slot (visited.get (), file_data, INSERT);
+	  if (*slot)
+	    {
+	      /* Already visited.  */
+	      continue;
+	    }
+	  *slot = file_data;
+
+	  for (int j = 0; j < file_data->num_file_names; ++j)
+	    {
+	      const char *filename = file_data->file_names[j];
+	      dwarf2_per_objfile->filenames_cache->seen (filename);
+	    }
 	}
     }
+
+  dwarf2_per_objfile->filenames_cache->traverse ([&] (const char *filename)
+    {
+      const char *this_real_name;
+
+      if (need_fullname)
+	this_real_name = gdb_realpath (filename);
+      else
+	this_real_name = NULL;
+      (*fun) (filename, this_real_name, data);
+    });
 }
 
 static int
diff --git a/gdb/filename-seen-cache.c b/gdb/filename-seen-cache.c
new file mode 100644
index 0000000..6282105
--- /dev/null
+++ b/gdb/filename-seen-cache.c
@@ -0,0 +1,66 @@
+/* Filename-seen cache for the GNU debugger, GDB.
+
+   Copyright (C) 1986-2017 Free Software Foundation, Inc.
+
+   This file is part of GDB.
+
+   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 "defs.h"
+#include "filename-seen-cache.h"
+#include "filenames.h"
+
+  /* Initial size of the table.  It automagically grows from here.  */
+#define INITIAL_FILENAME_SEEN_CACHE_SIZE 100
+
+/* filename_seen_cache constructor.  */
+
+filename_seen_cache::filename_seen_cache ()
+{
+  m_tab = htab_create_alloc (INITIAL_FILENAME_SEEN_CACHE_SIZE,
+			     filename_hash, filename_eq,
+			     NULL, xcalloc, xfree);
+}
+
+/* See filename-seen-cache.h.  */
+
+void
+filename_seen_cache::clear ()
+{
+  htab_empty (m_tab);
+}
+
+/* See filename-seen-cache.h.  */
+
+filename_seen_cache::~filename_seen_cache ()
+{
+  htab_delete (m_tab);
+}
+
+/* See filename-seen-cache.h.  */
+
+bool
+filename_seen_cache::seen (const char *file)
+{
+  void **slot;
+
+  /* Is FILE in tab?  */
+  slot = htab_find_slot (m_tab, file, INSERT);
+  if (*slot != NULL)
+    return true;
+
+  /* No; add it to tab.  */
+  *slot = (char *) file;
+  return false;
+}
diff --git a/gdb/filename-seen-cache.h b/gdb/filename-seen-cache.h
new file mode 100644
index 0000000..175234f
--- /dev/null
+++ b/gdb/filename-seen-cache.h
@@ -0,0 +1,63 @@
+/* Filename-seen cache for the GNU debugger, GDB.
+
+   Copyright (C) 1986-2017 Free Software Foundation, Inc.
+
+   This file is part of GDB.
+
+   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 "defs.h"
+#include "common/function-view.h"
+
+/* Cache to watch for file names already seen.  */
+
+class filename_seen_cache
+{
+public:
+  filename_seen_cache ();
+  ~filename_seen_cache ();
+
+  /* Disable copy.  */
+  filename_seen_cache (const filename_seen_cache &) = delete;
+  void operator= (const filename_seen_cache &) = delete;
+
+  /* Empty the cache, but do not delete it.  */
+  void clear ();
+
+  /* If FILE is not already in the table of files in CACHE, add it and
+     return false; otherwise return true.
+
+     NOTE: We don't manage space for FILE, we assume FILE lives as
+     long as the caller needs.  */
+  bool seen (const char *file);
+
+  /* Traverse all cache entries, calling CALLBACK on each.  The
+     filename is passed as argument to CALLBACK.  */
+  void traverse (gdb::function_view<void (const char *)> callback)
+  {
+    htab_traverse_noresize (m_tab,
+			    [] (void **slot, void *info) -> int
+			    {
+			      auto filename = (const char *) *slot;
+			      auto cb = (gdb::function_view<void (const char *)> *) info;
+			      (*cb) (filename);
+			      return 1;
+			    },
+			    &callback);
+  }
+
+private:
+  /* Table of files seen so far.  */
+  htab_t m_tab;
+};
diff --git a/gdb/symtab.c b/gdb/symtab.c
index c7f1311..c80f637 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -63,6 +63,7 @@
 #include "completer.h"
 #include "progspace-and-thread.h"
 #include "common/gdb_optional.h"
+#include "filename-seen-cache.h"
 
 /* Forward declarations for local functions.  */
 
@@ -3967,74 +3968,6 @@ operator_chars (const char *p, const char **end)
 }
 \f
 
-/* Cache to watch for file names already seen by filename_seen.  */
-
-struct filename_seen_cache
-{
-  /* Table of files seen so far.  */
-  htab_t tab;
-  /* Initial size of the table.  It automagically grows from here.  */
-#define INITIAL_FILENAME_SEEN_CACHE_SIZE 100
-};
-
-/* filename_seen_cache constructor.  */
-
-static struct filename_seen_cache *
-create_filename_seen_cache (void)
-{
-  struct filename_seen_cache *cache = XNEW (struct filename_seen_cache);
-
-  cache->tab = htab_create_alloc (INITIAL_FILENAME_SEEN_CACHE_SIZE,
-				  filename_hash, filename_eq,
-				  NULL, xcalloc, xfree);
-
-  return cache;
-}
-
-/* Empty the cache, but do not delete it.  */
-
-static void
-clear_filename_seen_cache (struct filename_seen_cache *cache)
-{
-  htab_empty (cache->tab);
-}
-
-/* filename_seen_cache destructor.
-   This takes a void * argument as it is generally used as a cleanup.  */
-
-static void
-delete_filename_seen_cache (void *ptr)
-{
-  struct filename_seen_cache *cache = (struct filename_seen_cache *) ptr;
-
-  htab_delete (cache->tab);
-  xfree (cache);
-}
-
-/* If FILE is not already in the table of files in CACHE, return zero;
-   otherwise return non-zero.  Optionally add FILE to the table if ADD
-   is non-zero.
-
-   NOTE: We don't manage space for FILE, we assume FILE lives as long
-   as the caller needs.  */
-
-static int
-filename_seen (struct filename_seen_cache *cache, const char *file, int add)
-{
-  void **slot;
-
-  /* Is FILE in tab?  */
-  slot = htab_find_slot (cache->tab, file, add ? INSERT : NO_INSERT);
-  if (*slot != NULL)
-    return 1;
-
-  /* No; maybe add it to tab.  */
-  if (add)
-    *slot = (char *) file;
-
-  return 0;
-}
-
 /* Data structure to maintain printing state for output_source_filename.  */
 
 struct output_source_filename_data
@@ -4064,7 +3997,7 @@ output_source_filename (const char *name,
      symtabs; it doesn't hurt to check.  */
 
   /* Was NAME already seen?  */
-  if (filename_seen (data->filename_seen_cache, name, 1))
+  if (data->filename_seen_cache->seen (name))
     {
       /* Yes; don't print it again.  */
       return;
@@ -4096,16 +4029,15 @@ sources_info (char *ignore, int from_tty)
   struct symtab *s;
   struct objfile *objfile;
   struct output_source_filename_data data;
-  struct cleanup *cleanups;
 
   if (!have_full_symbols () && !have_partial_symbols ())
     {
       error (_("No symbol table is loaded.  Use the \"file\" command."));
     }
 
-  data.filename_seen_cache = create_filename_seen_cache ();
-  cleanups = make_cleanup (delete_filename_seen_cache,
-			   data.filename_seen_cache);
+  filename_seen_cache filenames_seen;
+
+  data.filename_seen_cache = &filenames_seen;
 
   printf_filtered ("Source files for which symbols have been read in:\n\n");
 
@@ -4121,13 +4053,11 @@ sources_info (char *ignore, int from_tty)
   printf_filtered ("Source files for which symbols "
 		   "will be read in on demand:\n\n");
 
-  clear_filename_seen_cache (data.filename_seen_cache);
+  filenames_seen.clear ();
   data.first = 1;
   map_symbol_filenames (output_partial_symbol_filename, &data,
 			1 /*need_fullname*/);
   printf_filtered ("\n");
-
-  do_cleanups (cleanups);
 }
 
 /* Compare FILE against all the NFILES entries of FILES.  If BASENAMES is
@@ -5551,7 +5481,7 @@ maybe_add_partial_symtab_filename (const char *filename, const char *fullname,
 
   if (not_interesting_fname (filename))
     return;
-  if (!filename_seen (data->filename_seen_cache, filename, 1)
+  if (!data->filename_seen_cache->seen (filename)
       && filename_ncmp (filename, data->text, data->text_len) == 0)
     {
       /* This file matches for a completion; add it to the
@@ -5563,7 +5493,7 @@ maybe_add_partial_symtab_filename (const char *filename, const char *fullname,
       const char *base_name = lbasename (filename);
 
       if (base_name != filename
-	  && !filename_seen (data->filename_seen_cache, base_name, 1)
+	  && !data->filename_seen_cache->seen (base_name)
 	  && filename_ncmp (base_name, data->text, data->text_len) == 0)
 	add_filename_to_list (base_name, data->text, data->word, data->list);
     }
@@ -5584,23 +5514,20 @@ make_source_files_completion_list (const char *text, const char *word)
   VEC (char_ptr) *list = NULL;
   const char *base_name;
   struct add_partial_filename_data datum;
-  struct filename_seen_cache *filename_seen_cache;
-  struct cleanup *back_to, *cache_cleanup;
+  struct cleanup *back_to;
 
   if (!have_full_symbols () && !have_partial_symbols ())
     return list;
 
   back_to = make_cleanup (do_free_completion_list, &list);
 
-  filename_seen_cache = create_filename_seen_cache ();
-  cache_cleanup = make_cleanup (delete_filename_seen_cache,
-				filename_seen_cache);
+  filename_seen_cache filenames_seen;
 
   ALL_FILETABS (objfile, cu, s)
     {
       if (not_interesting_fname (s->filename))
 	continue;
-      if (!filename_seen (filename_seen_cache, s->filename, 1)
+      if (!filenames_seen.seen (s->filename)
 	  && filename_ncmp (s->filename, text, text_len) == 0)
 	{
 	  /* This file matches for a completion; add it to the current
@@ -5615,13 +5542,13 @@ make_source_files_completion_list (const char *text, const char *word)
 	     command do when they parse file names.  */
 	  base_name = lbasename (s->filename);
 	  if (base_name != s->filename
-	      && !filename_seen (filename_seen_cache, base_name, 1)
+	      && !filenames_seen.seen (base_name)
 	      && filename_ncmp (base_name, text, text_len) == 0)
 	    add_filename_to_list (base_name, text, word, &list);
 	}
     }
 
-  datum.filename_seen_cache = filename_seen_cache;
+  datum.filename_seen_cache = &filenames_seen;
   datum.text = text;
   datum.word = word;
   datum.text_len = text_len;
@@ -5629,7 +5556,6 @@ make_source_files_completion_list (const char *text, const char *word)
   map_symbol_filenames (maybe_add_partial_symtab_filename, &datum,
 			0 /*need_fullname*/);
 
-  do_cleanups (cache_cleanup);
   discard_cleanups (back_to);
 
   return list;
-- 
2.5.5