From: Avinash Dayanand Date: Thu, 11 Jan 2018 14:35:51 -0800 Subject: [Intel-wired-lan] [PATCH next-queue 01/12] i40evf: Fix link up issue when queues are disabled In-Reply-To: <20180111223602.9082-1-avinash.dayanand@intel.com> References: <20180111223602.9082-1-avinash.dayanand@intel.com> Message-ID: <20180111223602.9082-2-avinash.dayanand@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: intel-wired-lan@osuosl.org List-ID: One of the previous patch fixes the link up issue by ignoring it if i40evf is not in __I40EVF_RUNNING state. However this doesn't fix the race condition when queues are disabled esp for ADq on VF. Hence check if all queues are enabled before starting all queues. Signed-off-by: Avinash Dayanand --- drivers/net/ethernet/intel/i40evf/i40evf.h | 1 + drivers/net/ethernet/intel/i40evf/i40evf_main.c | 1 + .../net/ethernet/intel/i40evf/i40evf_virtchnl.c | 28 +++++++++++++++------- 3 files changed, 22 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/intel/i40evf/i40evf.h b/drivers/net/ethernet/intel/i40evf/i40evf.h index b6991e8..89ce2f9 100644 --- a/drivers/net/ethernet/intel/i40evf/i40evf.h +++ b/drivers/net/ethernet/intel/i40evf/i40evf.h @@ -240,6 +240,7 @@ struct i40evf_adapter { #define I40EVF_FLAG_ALLMULTI_ON BIT(14) #define I40EVF_FLAG_LEGACY_RX BIT(15) #define I40EVF_FLAG_REINIT_ITR_NEEDED BIT(16) +#define I40EVF_FLAG_QUEUES_DISABLED BIT(17) /* duplicates for common code */ #define I40E_FLAG_DCB_ENABLED 0 #define I40E_FLAG_RX_CSUM_ENABLED I40EVF_FLAG_RX_CSUM_ENABLED diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c index 6fd0992..3014a5e 100644 --- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c +++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c @@ -1895,6 +1895,7 @@ static void i40evf_reset_task(struct work_struct *work) i40evf_free_all_rx_resources(adapter); i40evf_free_all_tx_resources(adapter); + adapter->flags |= I40EVF_FLAG_QUEUES_DISABLED; /* kill and reinit the admin queue */ i40evf_shutdown_adminq(hw); adapter->current_op = VIRTCHNL_OP_UNKNOWN; diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c b/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c index d57a672..794c671 100644 --- a/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c +++ b/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c @@ -1017,14 +1017,25 @@ void i40evf_virtchnl_completion(struct i40evf_adapter *adapter, if (adapter->link_up == link_up) break; - /* If we get link up message and start queues before - * our queues are configured it will trigger a TX hang. - * In that case, just ignore the link status message, - * we'll get another one after we enable queues and - * actually prepared to send traffic. - */ - if (link_up && adapter->state != __I40EVF_RUNNING) - break; + if (link_up) { + /* If we get link up message and start queues + * before our queues are configured it will + * trigger a TX hang. In that case, just ignore + * the link status message,we'll get another one + * after we enable queues and actually prepared + * to send traffic. + */ + if (adapter->state != __I40EVF_RUNNING) + break; + + /* For ADq enabled VF, we reconfigure VSIs and + * re-allocate queues. Hence wait till all + * queues are enabled. + */ + if (adapter->flags & + I40EVF_FLAG_QUEUES_DISABLED) + break; + } adapter->link_up = link_up; if (link_up) { @@ -1108,6 +1119,7 @@ void i40evf_virtchnl_completion(struct i40evf_adapter *adapter, case VIRTCHNL_OP_ENABLE_QUEUES: /* enable transmits */ i40evf_irq_enable(adapter, true); + adapter->flags &= ~I40EVF_FLAG_QUEUES_DISABLED; break; case VIRTCHNL_OP_DISABLE_QUEUES: i40evf_free_all_tx_resources(adapter); -- 2.9.4