i40e/i40evf: check for stopped admin queue
Subsystems (2)
PATCH 1/1 i40e/i40evf: check for stopped admin queue
It's possible that while we are waiting for the spinlock, another entity (that owns the spinlock) has shut down the admin queue. If we then attempt to use the queue, we will panic.
Add a check for this condition on the receive side. This matches an existing check on the send queue side.
Signed-off-by Mitch Williams <[email protected]>
--- drivers/net/ethernet/intel/i40e/i40e_adminq.c | 9 +++++++++ drivers/net/ethernet/intel/i40evf/i40e_adminq.c | 9 +++++++++ 2 files changed, 18 insertions(+)
drivers/net/ethernet/intel/i40e/i40e_adminq.c +9 -0
@@ -946,6 +946,13 @@ i40e_status i40e_clean_arq_element(struct i40e_hw *hw,/* take the lock before we start messing with the ring */ mutex_lock (&hw->aq.arq_mutex); +if (hw->aq.arq.count ==0 ) { +i40e_debug (hw,I40E_DEBUG_AQ_MESSAGE , +"AQRX: Admin queue not initialized.\n" ); + ret_code =I40E_ERR_QUEUE_EMPTY ; + goto clean_arq_element_err; + } +/* set next_to_use to head */ ntu = (rd32 (hw, hw->aq.arq.head) &I40E_PF_ARQH_ARQH_MASK );if (ntu == ntc) {
@@ -1007,6 +1014,8 @@ clean_arq_element_out:/* Set pending if needed, unlock and return */ if (pending !=NULL ) *pending = (ntc > ntu ? hw->aq.arq.count :0 ) + (ntu - ntc); + +clean_arq_element_err:mutex_unlock (&hw->aq.arq_mutex);if (i40e_is_nvm_update_op (&e->desc)) {
drivers/net/ethernet/intel/i40evf/i40e_adminq.c +9 -0
@@ -898,6 +898,13 @@ i40e_status i40evf_clean_arq_element(struct i40e_hw *hw,/* take the lock before we start messing with the ring */ mutex_lock (&hw->aq.arq_mutex); +if (hw->aq.arq.count ==0 ) { +i40e_debug (hw,I40E_DEBUG_AQ_MESSAGE , +"AQRX: Admin queue not initialized.\n" ); + ret_code =I40E_ERR_QUEUE_EMPTY ; + goto clean_arq_element_err; + } +/* set next_to_use to head */ ntu = (rd32 (hw, hw->aq.arq.head) &I40E_PF_ARQH_ARQH_MASK );if (ntu == ntc) {
@@ -959,6 +966,8 @@ clean_arq_element_out:/* Set pending if needed, unlock and return */ if (pending !=NULL ) *pending = (ntc > ntu ? hw->aq.arq.count :0 ) + (ntu - ntc); + +clean_arq_element_err:mutex_unlock (&hw->aq.arq_mutex);return ret_code;
Signature · 1 line
2.4.3
PATCH 1/1 i40e/i40evf: check for stopped admin queue
Quoted 8 lines · On Mon, 28 Sep 2015 15:10:01 -0700 Mitch Williams <[email protected]>…
It's possible that while we are waiting for the spinlock, another entity (that owns the spinlock) has shut down the admin queue. If we then attempt to use the queue, we will panic.
Add a check for this condition on the receive side. This matches an existing check on the send queue side.
Signed-off-by Mitch Williams <[email protected]>
Acked-by Jesse Brandeburg <[email protected]>