When IB device is undergoing unregistration, the GID cache is always
cleaned up after all clients are unregistered with the below flow.
__ib_unregister_device()
disable_device()
ib_cache_cleanup_one()
gid_table_cleanup_one()
cleanup_gid_table_port()
There is no use in generating a GID change event at this stage, where
there is no active client of the device and device is nearly unregistered.
Link: https://lore.kernel.org/r/20191020065427.8772-4-leon@kernel.org
Signed-off-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
struct ib_gid_table *table)
{
int i;
- bool deleted = false;
if (!table)
return;
mutex_lock(&table->lock);
for (i = 0; i < table->sz; ++i) {
- if (is_gid_entry_valid(table->data_vec[i])) {
+ if (is_gid_entry_valid(table->data_vec[i]))
del_gid(ib_dev, port, table, i);
- deleted = true;
- }
}
mutex_unlock(&table->lock);
-
- if (deleted)
- dispatch_gid_change_event(ib_dev, port);
}
void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port,