Bug fixes for ENA ethernet driver
Subsystems (3)
v2 cover Bug fixes for ENA ethernet driver
This series adds the following:
- Fix undesired call to ena_restore after returning from suspend
- Fix condition inside a WARN_ON
- Fix overriding previous value when updating missed_tx statistic
v1->v2:
- fix bug when calling reset routine after device resources are freed (Jakub)
Shay Agroskin (3):
net: ena: Prevent reset after device destruction
net: ena: Change WARN_ON expression in ena_del_napi_in_range()
net: ena: Make missed_tx stat incremental
drivers/net/ethernet/amazon/ena/ena_netdev.c | 35 ++++++++++---------- 1 file changed, 18 insertions(+), 17 deletions(-)
Signature · 1 line
2.17.1
PATCH v2 1/3 net: ena: Prevent reset after device destruction
The reset work is scheduled by the timer routine whenever it detects that a device reset is required (e.g. when a keep_alive signal is missing).
When releasing device resources in ena_destroy_device() the driver cancels the scheduling of the timer routine without destroying the reset
work explicitly.
This creates the following bug:
The driver is suspended and the ena_suspend() function is called
-> This function calls ena_destroy_device() to free the net device
resources
-> The driver waits for the timer routine to finish its execution and then cancels it, thus preventing from it to be called again.
If, in its final execution, the timer routine schedules a reset, the reset routine might be called afterwards, and a redundant call to ena_restore_device() would be made.
By changing the reset routine we allow it to read the device's state accurately.
This is achieved by checking whether ENA_FLAG_TRIGGER_RESET flag is set before resetting the device and making both the destruction function and the flag check are under rtnl lock.
The ENA_FLAG_TRIGGER_RESET is cleared at the end of the destruction routine. Also surround the flag check with 'likely' because we expect that the reset routine would be called only when ENA_FLAG_TRIGGER_RESET flag is set.
The destruction of the timer and reset services in __ena_shutoff() have to stay, even though the timer routine is destroyed in ena_destroy_device(). This is to avoid a case in which the reset routine is scheduled after free_netdev() in __ena_shutoff(), which would create an access to freed memory in adapter->flags.
--- drivers/net/ethernet/amazon/ena/ena_netdev.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +10 -9
@@ -3601,16 +3601,14 @@ static void ena_fw_reset_device(struct work_struct *work) {struct ena_adapter *adapter = container_of(work, struct ena_adapter, reset_task); -struct pci_dev *pdev = adapter->pdev; -if (unlikely (!test_bit (ENA_FLAG_TRIGGER_RESET , &adapter->flags))) { -dev_err (&pdev->dev, -"device reset schedule while reset bit is off\n" ); -return ; - }rtnl_lock (); -ena_destroy_device (adapter, false); -ena_restore_device (adapter); + +if (likely (test_bit (ENA_FLAG_TRIGGER_RESET , &adapter->flags))) { +ena_destroy_device (adapter, false); +ena_restore_device (adapter); + } +rtnl_unlock (); }
@@ -4389,8 +4387,11 @@ static void __ena_shutoff(struct pci_dev *pdev, bool shutdown) netdev->rx_cpu_rmap = NULL; }#endif /* CONFIG_RFS_ACCEL */ -del_timer_sync (&adapter->timer_service); +/* Make sure timer and reset routine won't be called after +* freeing device resources. +*/ +del_timer_sync (&adapter->timer_service);cancel_work_sync (&adapter->reset_task);rtnl_lock ();/* lock released inside the below if-else block */
Signature · 1 line
2.17.1
Jakub Kicinski
maintainerRe: [PATCH V2 net 1/3] net: ena: Prevent reset after device destruction
Quoted 1 line · On Wed, 19 Aug 2020 12:04:41 +0300 Shay Agroskin wrote:
Fixes 84a629e ("[New feature] ena_netdev: Add hibernation support")
Fixes tag: Fixes: 84a629e ("[New feature] ena_netdev: Add hibernation support") Has these problem(s):
- Target SHA1 does not exist
Also hash needs to be 12 characters.
Re: [PATCH V2 net 1/3] net: ena: Prevent reset after device destruction
Quoted 9 lines · Jakub Kicinski
Quoted 1 line · On Wed, 19 Aug 2020 12:04:41 +0300 Shay Agroskin wrote:
Fixes 84a629e ("[New feature] ena_netdev: Add hibernation support")Fixes tag: Fixes: 84a629e ("[New feature]
ena_netdev: Add hibernation support") Has these problem(s):
- Target SHA1 does not existAlso hash needs to be 12 characters.
Yup, sorry for that. Should have been more careful with it.
I'll fix it in next patchset. Thanks
PATCH v2 2/3 net: ena: Change WARN_ON expression in ena_del_napi_in_range()
In reply to Shay AgroskinThe ena_del_napi_in_range() function unregisters the napi handler for rings in a given range.
This function had the following WARN_ON macro:
WARN_ON(ENA_IS_XDP_INDEX(adapter, i) && adapter->ena_napi[i].xdp_ring);
This macro prints the call stack if the expression inside of it is true [1], but the expression inside of it is the wanted situation. The expression checks whether the ring has an XDP queue and its index corresponds to a XDP one.
This patch changes the expression to
!ENA_IS_XDP_INDEX(adapter, i) && adapter->ena_napi[i].xdp_ring
which indicates an unwanted situation.
Also, change the structure of the function. The napi handler is unregistered for all rings, and so there's no need to check whether the index is an XDP index or not. By removing this check the code becomes much more readable.
--- drivers/net/ethernet/amazon/ena/ena_netdev.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +4 -7
@@ -2180,13 +2180,10 @@ static void ena_del_napi_in_range(struct ena_adapter *adapter,int i;for (i = first_index; i < first_index + count; i++) { -/* Check if napi was initialized before */ -if (!ENA_IS_XDP_INDEX (adapter, i) || - adapter->ena_napi[i].xdp_ring) -netif_napi_del (&adapter->ena_napi[i].napi); -else -WARN_ON (ENA_IS_XDP_INDEX (adapter, i) && - adapter->ena_napi[i].xdp_ring); +netif_napi_del (&adapter->ena_napi[i].napi); + +WARN_ON (!ENA_IS_XDP_INDEX (adapter, i) && + adapter->ena_napi[i].xdp_ring); } }
Signature · 1 line
2.17.1
PATCH v2 3/3 net: ena: Make missed_tx stat incremental
In reply to Shay AgroskinMost statistics in ena driver are incremented, meaning that a stat's value is a sum of all increases done to it since driver/queue initialization.
This patch makes all statistics this way, effectively making missed_tx statistic incremental.
Also added a comment regarding rx_drops and tx_drops to make it clearer how these counters are calculated.
--- drivers/net/ethernet/amazon/ena/ena_netdev.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +4 -1
@@ -3687,7 +3687,7 @@ static int check_missing_comp_in_tx_queue(struct ena_adapter *adapter, }u64_stats_update_begin (&tx_ring->syncp); - tx_ring->tx_stats.missed_tx = missed_tx; + tx_ring->tx_stats.missed_tx += missed_tx;u64_stats_update_end (&tx_ring->syncp);return rc;
@@ -4556,6 +4556,9 @@ static void ena_keep_alive_wd(void *adapter_data, tx_drops = ((u64 )desc->tx_drops_high <<32 ) | desc->tx_drops_low;u64_stats_update_begin (&adapter->syncp); +/* These stats are accumulated by the device, so the counters indicate +* all drops since last reset. +*/ adapter->dev_stats.rx_drops = rx_drops; adapter->dev_stats.tx_drops = tx_drops;u64_stats_update_end (&adapter->syncp);
Signature · 1 line
2.17.1