Commit 9cbbc451 authored by Edward Cree's avatar Edward Cree Committed by David S. Miller
Browse files

sfc: take correct lock in ef100_reset() 



When downing and upping the ef100 filter table, we need to take a write
 lock on efx->filter_sem, not just a read lock, because we may kfree()
 the table pointers.
Without this, resets cause a WARN_ON from efx_rwsem_assert_write_locked().

Fixes: a9dc3d56 ("sfc_ef100: RX filter table management and related gubbins")
Signed-off-by: default avatarEdward Cree <ecree@solarflare.com>
Reviewed-by: default avatarJesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent db06ea34

drivers/net/ethernet/sfc/ef100_nic.c

+4 −4
Original line number Diff line number Diff line
@@ -431,18 +431,18 @@ static int ef100_reset(struct efx_nic *efx, enum reset_type reset_type) 
		/* A RESET_TYPE_ALL will cause filters to be removed, so we remove filters

		 * and reprobe after reset to avoid removing filters twice

		 */

		down_read(&efx->filter_sem);

		down_write(&efx->filter_sem);

		ef100_filter_table_down(efx);

		up_read(&efx->filter_sem);

		up_write(&efx->filter_sem);

		rc = efx_mcdi_reset(efx, reset_type);

		if (rc)

			return rc;



		netif_device_attach(efx->net_dev);



		down_read(&efx->filter_sem);

		down_write(&efx->filter_sem);

		rc = ef100_filter_table_up(efx);

		up_read(&efx->filter_sem);

		up_write(&efx->filter_sem);

		if (rc)

			return rc;