Comment on memory handle performance.

evoskuil · Jun 27, 2024 · 489fabf · 489fabf
1 parent 9a3027a
commit 489fabf
Showing 1 changed file with 3 additions and 0 deletions.
diff --git a/include/bitcoin/database/impl/query/translate.ipp b/include/bitcoin/database/impl/query/translate.ipp
@@ -480,7 +480,10 @@ spend_set CLASS::to_spend_set(const tx_link& link) const NOEXCEPT
 
     // This is not concurrent because to_spend_sets is (by tx).
     table::spend::get_prevout_sequence get{};
+
+    // This reduced a no-bypass 840k sync/confirmable/confirm run by 8.3%.
     const auto ptr = store_.spend.get_memory();
+
     for (const auto& spend_fk: puts.spend_fks)
     {
         if (!store_.spend.get(ptr, spend_fk, get))