iavf: fix reset_task for early reset event
Subsystems (3)
PATCH 1/1 iavf: fix reset_task for early reset event
From: Ahmed Zaki <[email protected]>
If a reset event is received from the PF early in the init cycle, the state machine hangs for about 25 seconds.
Reproducer:
echo 1 > /sys/class/net/$PF0/device/sriov_numvfs ip link set dev $PF0 vf 0 mac $NEW_MAC
The log shows:
[792.620416] ice 0000:5e:00.0: Enabling 1 VFs [792.738812] iavf 0000:5e:01.0: enabling device (0000 -> 0002) [792.744182] ice 0000:5e:00.0: Enabling 1 VFs with 17 vectors and 16 queues per VF [792.839964] ice 0000:5e:00.0: Setting MAC 52:54:00:00:00:11 on VF 0. VF driver will be reinitialized [813.389684] iavf 0000:5e:01.0: Failed to communicate with PF; waiting before retry [818.635918] iavf 0000:5e:01.0: Hardware came out of reset. Attempting reinit. [818.766273] iavf 0000:5e:01.0: Multiqueue Enabled: Queue pair count = 16
Fix it by scheduling the reset task and making the reset task capable of resetting early in the init cycle.
---
This should be applied after "iavf: get rid of the crit lock"
--- drivers/net/ethernet/intel/iavf/iavf_main.c | 11 +++++++++++ drivers/net/ethernet/intel/iavf/iavf_virtchnl.c | 17 +++++++++++++++++ 2 files changed, 28 insertions(+)
drivers/net/ethernet/intel/iavf/iavf_main.c +11 -0
@@ -3209,6 +3209,17 @@ static void iavf_reset_task(struct work_struct *work) } continue_reset: +/* If we are still early in the state machine, just restart. */ +if (adapter->state <= __IAVF_INIT_FAILED) { +iavf_shutdown_adminq (hw); +iavf_change_state (adapter, __IAVF_STARTUP); +iavf_startup (adapter); +queue_delayed_work (adapter->wq, &adapter->watchdog_task, +msecs_to_jiffies (30 )); +netdev_unlock (netdev); +return ; + } +/* We don't use netif_running() because it may be true prior to * ndo_open() returning, so we can't assume it means all our open * tasks have finished, since we're not holding the rtnl_lock here.
drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +17 -0
@@ -79,6 +79,23 @@ iavf_poll_virtchnl_msg(struct iavf_hw *hw, struct iavf_arq_event_info *event,return iavf_status_to_errno (status); received_op = (enum virtchnl_ops )le32_to_cpu (event->desc.cookie_high); + +if (received_op ==VIRTCHNL_OP_EVENT ) { +struct iavf_adapter *adapter = hw->back; +struct virtchnl_pf_event *vpe = + (struct virtchnl_pf_event *)event->msg_buf; + +if (vpe->event !=VIRTCHNL_EVENT_RESET_IMPENDING ) +continue ; + +dev_info (&adapter->pdev->dev,"Reset indication received from the PF\n" ); +if (!(adapter->flags &IAVF_FLAG_RESET_PENDING )) +iavf_schedule_reset (adapter, +IAVF_FLAG_RESET_PENDING ); + +return -EIO ; + } +if (op_to_poll == received_op)break ; }
Signature · 1 line
2.49.0
Quoted 21 lines · Marcin Szycik
From: Ahmed Zaki <[email protected]>
If a reset event is received from the PF early in the init cycle, the state machine hangs for about 25 seconds.
Reproducer:
echo 1 >/sys/class/net/$PF0/device/sriov_numvfsip link set dev $PF0 vf 0 mac $NEW_MACThe log shows:
[792.620416] ice 0000:5e:00.0: Enabling 1 VFs [792.738812] iavf 0000:5e:01.0: enabling device (0000 -> 0002) [792.744182] ice 0000:5e:00.0: Enabling 1 VFs with 17 vectors and 16 queues per VF [792.839964] ice 0000:5e:00.0: Setting MAC 52:54:00:00:00:11 on VF 0. VF driver will be reinitialized [813.389684] iavf 0000:5e:01.0: Failed to communicate with PF; waiting before retry [818.635918] iavf 0000:5e:01.0: Hardware came out of reset. Attempting reinit. [818.766273] iavf 0000:5e:01.0: Multiqueue Enabled: Queue pair count = 16Fix it by scheduling the reset task and making the reset task capable of resetting early in the init cycle.
Quoted 7 lines
Fixes ef8693eb90ae3 ("i40evf: refactor reset handling")Signed-off-by Ahmed Zaki <[email protected]>Tested-by Przemek Kitszel <[email protected]>Reviewed-by Przemek Kitszel <[email protected]>Signed-off-by Marcin Szycik <[email protected]>---
This should be applied after "iavf: get rid of the crit lock"
Romanowski, Rafal
Tested-byQuoted 45 lines
-----Original Message-----
From: Intel-wired-lan <[email protected]> On Behalf Of Simon Horman
Sent: Monday, April 28, 2025 2:44 PM
To: Marcin Szycik <[email protected]>
Cc: [email protected]; [email protected]; Zaki, Ahmed <[email protected]>; Kitszel, Przemyslaw <[email protected]>
Subject: Re: [Intel-wired-lan] [PATCH iwl-net] iavf: fixreset_taskfor early reset eventQuoted 5 lines · Marcin Szycik
From: Ahmed Zaki <[email protected]>
If a reset event is received from the PF early in the init cycle, the state machine hangs for about 25 seconds.
Quoted 8 lines
Reproducer:
echo 1 >/sys/class/net/$PF0/device/sriov_numvfsip link set dev $PF0 vf 0 mac $NEW_MACThe log shows:
[792.620416] ice 0000:5e:00.0: Enabling 1 VFs [792.738812] iavf 0000:5e:01.0: enabling device (0000 -> 0002) [792.744182] ice 0000:5e:00.0: Enabling 1 VFs with 17 vectors and 16 queuesper VF
Quoted 1 line
[792.839964] ice 0000:5e:00.0: Setting MAC 52:54:00:00:00:11 on VF 0. VFdriver will be reinitialized
Quoted 1 line
[813.389684] iavf 0000:5e:01.0: Failed to communicate with PF; waitingbefore retry
Quoted 1 line
[818.635918] iavf 0000:5e:01.0: Hardware came out of reset. Attemptingreinit.
Quoted 6 lines
[818.766273] iavf 0000:5e:01.0: Multiqueue Enabled: Queue pair count= 16
Fix it by scheduling the reset task and making the reset task capable of resetting early in the init cycle.
Quoted 7 lines
Fixes ef8693eb90ae3 ("i40evf: refactor reset handling")Signed-off-by Ahmed Zaki <[email protected]>Tested-by Przemek Kitszel <[email protected]>Reviewed-by Przemek Kitszel <[email protected]>Signed-off-by Marcin Szycik <[email protected]>---
This should be applied after "iavf: get rid of the crit lock"
Quoted 1 line
Reviewed-by Simon Horman <[email protected]>