net/mlx5: Wait for inactive autogroups

Currently, if one thread tries to add an entry to an autogrouped table
with no free matching group, while another thread is in the process of
creating a new matching autogroup, it doesn't wait for the new group
creation, and creates an unnecessary new autogroup.

Instead of skipping inactive, wait on the write lock of those groups.

Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 2da45e9..52af602 100644 (file)
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -1755,11 +1755,13 @@ skip_search:
        list_for_each_entry(iter, match_head, list) {
                g = iter->g;
 
-               if (!g->node.active)
-                       continue;
-
                nested_down_write_ref_node(&g->node, FS_LOCK_PARENT);
 
+               if (!g->node.active) {
+                       up_write_ref_node(&g->node, false);
+                       continue;
+               }
+
                err = insert_fte(g, fte);
                if (err) {
                        up_write_ref_node(&g->node, false);