Use execute_batch instead of executemany in places (#9181)

`execute_batch` does fewer round trips in postgres than `executemany`, but does not give a correct `txn.rowcount` result after.

synapse/storage/database.py

@@ -267,8 +267,7 @@ class LoggingTransaction:
 
             self._do_execute(lambda *x: execute_batch(self.txn, *x), sql, args)
         else:
-            for val in args:
-                self.execute(sql, val)
+            self.executemany(sql, args)
 
     def execute_values(self, sql: str, *args: Any) -> List[Tuple]:
         """Corresponds to psycopg2.extras.execute_values. Only available when
@@ -888,7 +887,7 @@ class DatabasePool:
             ", ".join("?" for _ in keys[0]),
         )
 
-        txn.executemany(sql, vals)
+        txn.execute_batch(sql, vals)
 
     async def simple_upsert(
         self,