[tracker/wip/carlosg/update-perf: 6/18] core: Maintain array for refcount changes

[Carlos Garnacho <carlosg src gnome org>]

commit d8bbd4d5ce53c481104756b4c0b75ba6e2b4c310
Author: Carlos Garnacho <carlosg gnome org>
Date:   Fri Aug 19 03:21:05 2022 +0200

    core: Maintain array for refcount changes
    
    Since refcount changes are very frequent, and each of them requires
    a hashtable lookup+update to change (the refcount is the value of
    the hashtable, stored with GINT_TO_POINTER), plus the replacement
    goes with a new copy of the TrackerRowid for the hashtable key. Overall,
    this maintenance is somewhat expensive to perform.
    
    As there are small buffers for triple updates (64 items), a hash table
    does not bring a lot of benefit in lookups, nor updates. Switching to
    a simple unordered array to keep this accounting looks faster in
    profiling.

 src/libtracker-sparql/core/tracker-data-update.c | 51 ++++++++++++++----------
 1 file changed, 31 insertions(+), 20 deletions(-)
---
diff --git a/src/libtracker-sparql/core/tracker-data-update.c 
b/src/libtracker-sparql/core/tracker-data-update.c
index 1b3c119f4..930ade036 100644
--- a/src/libtracker-sparql/core/tracker-data-update.c
+++ b/src/libtracker-sparql/core/tracker-data-update.c
@@ -106,13 +106,18 @@ struct _TrackerDataUpdateBuffer {
        TrackerDBStatementMru stmt_mru;
 };
 
+typedef struct {
+       TrackerRowid rowid;
+       gint refcount_change;
+} RefcountEntry;
+
 struct _TrackerDataUpdateBufferGraph {
        gchar *graph;
 
        /* id -> TrackerDataUpdateBufferResource */
        GHashTable *resources;
        /* id -> integer */
-       GHashTable *refcounts;
+       GArray *refcounts;
 
        TrackerDBStatement *insert_ref;
        TrackerDBStatement *update_ref;
@@ -1214,15 +1219,25 @@ tracker_data_update_refcount (TrackerData  *data,
                               gint          refcount)
 {
        const TrackerDataUpdateBufferGraph *graph;
-       gint old_refcount;
+       RefcountEntry entry;
+       guint i;
 
        g_assert (data->resource_buffer != NULL);
        graph = data->resource_buffer->graph;
 
-       old_refcount = GPOINTER_TO_INT (g_hash_table_lookup (graph->refcounts, &id));
-       g_hash_table_insert (graph->refcounts,
-                            tracker_rowid_copy (&id),
-                            GINT_TO_POINTER (old_refcount + refcount));
+       for (i = 0; i < graph->refcounts->len; i++) {
+               RefcountEntry *ptr;
+
+               ptr = &g_array_index (graph->refcounts, RefcountEntry, i);
+               if (ptr->rowid == id) {
+                       ptr->refcount_change += refcount;
+                       return;
+               }
+       }
+
+       entry.rowid = id;
+       entry.refcount_change = refcount;
+       g_array_append_val (graph->refcounts, entry);
 }
 
 static void
@@ -1279,10 +1294,9 @@ tracker_data_flush_graph_refcounts (TrackerData                   *data,
                                     GError                       **error)
 {
        TrackerDBInterface *iface;
-       GHashTableIter iter;
-       gpointer key, value;
        TrackerRowid id;
        gint refcount;
+       guint i;
        GError *inner_error = NULL;
        const gchar *database;
        gchar *query;
@@ -1290,11 +1304,12 @@ tracker_data_flush_graph_refcounts (TrackerData                   *data,
        iface = tracker_data_manager_get_writable_db_interface (data->manager);
        database = graph->graph ? graph->graph : "main";
 
-       g_hash_table_iter_init (&iter, graph->refcounts);
+       for (i = 0; i < graph->refcounts->len; i++) {
+               RefcountEntry *entry;
 
-       while (g_hash_table_iter_next (&iter, &key, &value)) {
-               id = *(TrackerRowid *) key;
-               refcount = GPOINTER_TO_INT (value);
+               entry = &g_array_index (graph->refcounts, RefcountEntry, i);
+               id = entry->rowid;
+               refcount = entry->refcount_change;
 
                if (refcount > 0) {
                        if (!graph->insert_ref) {
@@ -1370,8 +1385,6 @@ tracker_data_flush_graph_refcounts (TrackerData                   *data,
                                break;
                        }
                }
-
-               g_hash_table_iter_remove (&iter);
        }
 }
 
@@ -1382,7 +1395,7 @@ graph_buffer_free (TrackerDataUpdateBufferGraph *graph)
        g_clear_object (&graph->update_ref);
        g_clear_object (&graph->delete_ref);
        g_hash_table_unref (graph->resources);
-       g_hash_table_unref (graph->refcounts);
+       g_array_unref (graph->refcounts);
        g_free (graph->graph);
        g_slice_free (TrackerDataUpdateBufferGraph, graph);
 }
@@ -1505,7 +1518,7 @@ tracker_data_update_buffer_flush (TrackerData  *data,
                }
 
                g_hash_table_remove_all (graph->resources);
-               g_hash_table_remove_all (graph->refcounts);
+               g_array_set_size (graph->refcounts, 0);
        }
 
 out:
@@ -1533,7 +1546,7 @@ tracker_data_update_buffer_clear (TrackerData *data)
        for (i = 0; i < data->update_buffer.graphs->len; i++) {
                graph = g_ptr_array_index (data->update_buffer.graphs, i);
                g_hash_table_remove_all (graph->resources);
-               g_hash_table_remove_all (graph->refcounts);
+               g_array_set_size (graph->refcounts, 0);
        }
 
        g_hash_table_remove_all (data->update_buffer.new_resources);
@@ -2402,9 +2415,7 @@ ensure_graph_buffer (TrackerDataUpdateBuffer  *buffer,
        }
 
        graph_buffer = g_slice_new0 (TrackerDataUpdateBufferGraph);
-       graph_buffer->refcounts =
-               g_hash_table_new_full (tracker_rowid_hash, tracker_rowid_equal,
-                                      (GDestroyNotify) tracker_rowid_free, NULL);
+       graph_buffer->refcounts = g_array_sized_new (FALSE, FALSE, sizeof (RefcountEntry), UPDATE_LOG_SIZE);
        graph_buffer->graph = g_strdup (name);
 
        graph_buffer->resources =