mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-16 03:52:35 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions

mhnsw: don't create many empty layers

Browse source
This commit is contained in:
Sergei Golubchik 2024-06-05 14:59:25 +02:00
parent 058976533f
commit c91cab3b6f
1 changed files with 20 additions and 22 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

42
sql/vector_mhnsw.cc
View file

@ -546,8 +546,6 @@ int mhnsw_insert(TABLE *table, KEY *keyinfo)
return write_neighbors(&ctx, 0, target);
}
longlong max_layer= graph->field[0]->val_int();
List<FVectorNode> candidates;
List<FVectorNode> start_nodes;
String ref_str, *ref_ptr;
@ -572,36 +570,37 @@ int mhnsw_insert(TABLE *table, KEY *keyinfo)
double new_num= my_rnd(&thd->rand);
double log= -std::log(new_num) * NORMALIZATION_FACTOR;
longlong new_node_layer= static_cast<longlong>(std::floor(log));
longlong max_layer= graph->field[0]->val_int();
// XXX what is that?
for (longlong cur_layer= new_node_layer; cur_layer >= max_layer + 1;
cur_layer--)
if (new_node_layer > max_layer)
{
if (int err= write_neighbors(&ctx, cur_layer, target))
if (int err= write_neighbors(&ctx, max_layer + 1, target))
return err;
new_node_layer= max_layer;
}
else
{
for (longlong cur_layer= max_layer; cur_layer > new_node_layer; cur_layer--)
{
if (int err= search_layer(&ctx, start_nodes,
thd->variables.hnsw_ef_constructor, cur_layer,
&candidates))
return err;
start_nodes.empty();
start_nodes.push_back(candidates.head(), &ctx.root); // XXX ef=1
candidates.empty();
}
}
for (longlong cur_layer= max_layer; cur_layer > new_node_layer; cur_layer--)
{
if (int err= search_layer(&ctx, start_nodes,
thd->variables.hnsw_ef_constructor, cur_layer,
&candidates))
return err;
start_nodes.empty();
start_nodes.push_back(candidates.head(), &ctx.root); // XXX ef=1
candidates.empty();
}
for (longlong cur_layer= std::min(max_layer, new_node_layer);
cur_layer >= 0; cur_layer--)
for (longlong cur_layer= new_node_layer; cur_layer >= 0; cur_layer--)
{
if (int err= search_layer(&ctx, start_nodes,
thd->variables.hnsw_ef_constructor, cur_layer,
&candidates))
return err;
uint max_neighbors= (cur_layer == 0) ? // heuristics from the paper
thd->variables.hnsw_max_connection_per_layer * 2
uint max_neighbors= (cur_layer == 0) // heuristics from the paper
? thd->variables.hnsw_max_connection_per_layer * 2
: thd->variables.hnsw_max_connection_per_layer;
if (int err= select_neighbors(&ctx, cur_layer, target, candidates,
@ -611,7 +610,6 @@ int mhnsw_insert(TABLE *table, KEY *keyinfo)
return err;
start_nodes= candidates;
}
start_nodes.empty();
dbug_tmp_restore_column_map(&table->read_set, old_map);