ixgbevf: fix AER error handling
Subsystems (2)
PATCH 1/1 ixgbevf: fix AER error handling
Make sure that we free the IRQs in ixgbevf_io_error_detected() when responding to an PCIe AER error and also restore them when the interface recovers from it.
Previously it was possible to trigger BUG_ON() check in free_msix_irqs() in the case where we call ixgbevf_remove() after a failed recovery from AER error because the inerrupts were not freed.
Also moved the down and free functions into ixgbevf_close_suspend() same as with ixgbe.
--- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 43 ++++++++++++--------- 1 file changed, 25 insertions(+), 18 deletions(-)
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +25 -18
@@ -3228,6 +3228,21 @@ int ixgbevf_open(struct net_device *netdev) }/** +* ixgbevf_close_suspend - actions necessary to both suspend and close flows +* @adapter: the private adapter struct +* +* This function should contain the necessary work common to both suspending +* and closing of the device. +*/ +static void ixgbevf_close_suspend (struct ixgbevf_adapter *adapter) +{ +ixgbevf_down (adapter); +ixgbevf_free_irq (adapter); +ixgbevf_free_all_tx_resources (adapter); +ixgbevf_free_all_rx_resources (adapter); +} + +/** * ixgbevf_close - Disables a network interface * @netdev: network interface device structure *
@@ -3242,14 +3257,8 @@ int ixgbevf_close(struct net_device *netdev) {struct ixgbevf_adapter *adapter =netdev_priv (netdev); -if (!netif_device_present (netdev)) -return 0 ; - -ixgbevf_down (adapter); -ixgbevf_free_irq (adapter); - -ixgbevf_free_all_tx_resources (adapter); -ixgbevf_free_all_rx_resources (adapter); +if (netif_device_present (netdev)) +ixgbevf_close_suspend (adapter);return 0 ; }
@@ -3802,13 +3811,10 @@ static int ixgbevf_suspend(struct pci_dev *pdev, pm_message_t state)rtnl_lock ();netif_device_detach (netdev); -if (netif_running (netdev)) { -ixgbevf_down (adapter); -ixgbevf_free_irq (adapter); -ixgbevf_free_all_tx_resources (adapter); -ixgbevf_free_all_rx_resources (adapter); -ixgbevf_clear_interrupt_scheme (adapter); - } +if (netif_running (netdev)) +ixgbevf_close_suspend (adapter); + +ixgbevf_clear_interrupt_scheme (adapter);rtnl_unlock ();#ifdef CONFIG_PM
@@ -4247,7 +4253,7 @@ static pci_ers_result_t ixgbevf_io_error_detected(struct pci_dev *pdev, }if (netif_running (netdev)) -ixgbevf_down (adapter); +ixgbevf_close_suspend (adapter);if (!test_and_set_bit (__IXGBEVF_DISABLED, &adapter->state))pci_disable_device (pdev);
@@ -4295,12 +4301,13 @@ static pci_ers_result_t ixgbevf_io_slot_reset(struct pci_dev *pdev)static void ixgbevf_io_resume (struct pci_dev *pdev) {struct net_device *netdev =pci_get_drvdata (pdev); -struct ixgbevf_adapter *adapter =netdev_priv (netdev); +rtnl_lock ();if (netif_running (netdev)) -ixgbevf_up (adapter); +ixgbevf_open (netdev);netif_device_attach (netdev); +rtnl_unlock (); }/* PCI Error Recovery (ERS) */
Bowers
PATCH 1/1 ixgbevf: fix AER error handling
Quoted 15 lines
-----Original Message-----
From: Intel-wired-lan [mailto:intel-wired-lan-bounces at lists.osuosl.org] On Behalf Of Emil Tantilov
Sent: Wednesday, November 16, 2016 11:26 AM
To: intel-wired-lan at lists.osuosl.org
Subject: [Intel-wired-lan] [PATCH] ixgbevf: fix AER error handlingMake sure that we free the IRQs in
ixgbevf_io_error_detected()when responding to an PCIe AER error and also restore them when the interface recovers from it.Previously it was possible to trigger
BUG_ON()check infree_msix_irqs()in the case where we callixgbevf_remove()after a failed recovery from AER error because the inerrupts were not freed.
Quoted 8 lines
Also moved the down and free functions into
ixgbevf_close_suspend()same as with ixgbe.Signed-off-by Emil Tantilov <[email protected]>--- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 43 ++++++++++++---- ----- 1 file changed, 25 insertions(+), 18 deletions(-)