OPTIM: connection: fold the error handling with handshake handling
Both of them are rare and are detected from the same flags source, so
let's detect errors in the handshake loop and remove two tests in the
fast path. This seems to improve overall performance by less than 0.5%
on connection-bound workloads.
diff --git a/src/connection.c b/src/connection.c
index ad523b9..b477b46 100644
--- a/src/connection.c
+++ b/src/connection.c
@@ -49,17 +49,16 @@
conn_refresh_polling_flags(conn);
flags = conn->flags & ~CO_FL_ERROR; /* ensure to call the wake handler upon error */
- if (unlikely(conn->flags & CO_FL_ERROR))
- goto leave;
-
process_handshake:
/* The handshake callbacks are called in sequence. If either of them is
* missing something, it must enable the required polling at the socket
* layer of the connection. Polling state is not guaranteed when entering
* these handlers, so any handshake handler which does not complete its
- * work must explicitly disable events it's not interested in.
+ * work must explicitly disable events it's not interested in. Error
+ * handling is also performed here in order to reduce the number of tests
+ * around.
*/
- while (unlikely(conn->flags & CO_FL_HANDSHAKE)) {
+ while (unlikely(conn->flags & (CO_FL_HANDSHAKE | CO_FL_ERROR))) {
if (unlikely(conn->flags & (CO_FL_ERROR|CO_FL_WAIT_RD|CO_FL_WAIT_WR)))
goto leave;
@@ -117,13 +116,10 @@
conn->data->send(conn);
}
- if (unlikely(conn->flags & CO_FL_ERROR))
- goto leave;
-
/* It may happen during the data phase that a handshake is
* enabled again (eg: SSL)
*/
- if (unlikely(conn->flags & CO_FL_HANDSHAKE))
+ if (unlikely(conn->flags & (CO_FL_HANDSHAKE | CO_FL_ERROR)))
goto process_handshake;
if (unlikely(conn->flags & CO_FL_WAIT_L4_CONN) && !(conn->flags & CO_FL_WAIT_WR)) {