Skip to contentSkip to conversation

ENA driver bug fixes

PATCH In review [email protected] 3 participants 7 messages Mar 17, 2020 – Mar 18, 2020

Subsystems (3)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

v2 cover ENA driver bug fixes

From: Arthur Kiyanovski <[email protected]>

ENA driver bug fixes

Arthur Kiyanovski (4):
net: ena: fix incorrect setting of the number of msix vectors
net: ena: fix request of incorrect number of IRQ vectors
net: ena: avoid memory access violation by validating req_id properly
net: ena: fix continuous keep-alive resets

 drivers/net/ethernet/amazon/ena/ena_netdev.c | 27 ++++++++++++++------
 1 file changed, 19 insertions(+), 8 deletions(-)
Signature · 1 line
2.17.1
author maintainer
#1 lore raw

PATCH v2 1/4 net: ena: fix incorrect setting of the number of msix vectors

From: Arthur Kiyanovski <[email protected]>

Overview:
We don't frequently change the msix vectors throughout the life cycle of the driver. We do so in two functions: ena_probe() and ena_restore(). ena_probe() is only called when the driver is loaded. ena_restore() on the other hand is called during device reset / resume operations.

We use num_io_queues for calculating and allocating the number of msix vectors. At ena_probe() this value is equal to max_num_io_queues and thus this is not an issue, however ena_restore() might be called after the number of io queues has changed.

A possible bug scenario is as follows:

* Change number of queues from 8 to 4.
(num_io_queues = 4, max_num_io_queues = 8, msix_vecs = 9,)
* Trigger reset occurs -> ena_restore is called. (num_io_queues = 4, max_num_io_queues =8 , msix_vecs = 5)
* Change number of queues from 4 to 6.
(num_io_queues = 6, max_num_io_queues = 8, msix_vecs = 5)
* The driver will reset due to failure of check_for_rx_interrupt_queue()

Fix:
This can be easily fixed by always using max_num_io_queues to init the msix_vecs, since this number won't change as opposed to num_io_queues.

Fixes 4d19266022ec ("net: ena: multiple queue creation related cleanups")
---
 drivers/net/ethernet/amazon/ena/ena_netdev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +1 -1
index 0b2fd96b93d7..39f290f4458f 100644
@@ -1968,7 +1968,7 @@ static int ena_enable_msix(struct ena_adapter *adapter)
 	}
 
 	/* Reserved the max msix vectors we might need */
-	msix_vecs = ENA_MAX_MSIX_VEC(adapter->num_io_queues);
+	msix_vecs = ENA_MAX_MSIX_VEC(adapter->max_num_io_queues);
 	netif_dbg(adapter, probe, adapter->netdev,
 		  "trying to enable MSI-X, vectors %d\n", msix_vecs);
 
Signature · 1 line
2.17.1
author maintainer
#2 lore raw

PATCH v2 2/4 net: ena: fix request of incorrect number of IRQ vectors

In reply to [email protected]

From: Arthur Kiyanovski <[email protected]>

Bug:
In short the main issue is caused by the fact that the number of queues is changed using ethtool after ena_probe() has been called and before ena_up() was executed. Here is the full scenario in detail:

* ena_probe() is called when the driver is loaded, the driver is not up yet at the end of ena_probe().
* The number of queues is changed -> io_queue_count is changed as well - ena_up() is not called since the "dev_was_up" boolean in ena_update_queue_count() is false.
* ena_up() is called by the kernel (it's called asynchronously some time after ena_probe()). ena_setup_io_intr() is called by ena_up() and it uses io_queue_count to get the suitable irq lines for each msix vector. The function ena_request_io_irq() is called right after that and it uses msix_vecs - This value only changes during ena_probe() and ena_restore() - to request the irq vectors. This results in "Failed to request I/O IRQ" error for i > io_queue_count.

Numeric example:
* After ena_probe() io_queue_count = 8, msix_vecs = 9.
* The number of queues changes to 4 -> io_queue_count = 4, msix_vecs = 9.
* ena_up() is executed for the first time:
** ena_setup_io_intr() inits the vectors only up to io_queue_count. ** ena_request_io_irq() calls request_irq() and fails for i = 5.

How to reproduce:
simply run the following commands:

    sudo rmmod ena && sudo insmod ena.ko;
    sudo ethtool -L eth1 combined 3;

Fix:
Use ENA_MAX_MSIX_VEC(adapter->num_io_queues + adapter->xdp_num_queues) instead of adapter->msix_vecs. We need to take XDP queues into consideration as they need to have msix vectors assigned to them as well. Note that the XDP cannot be attached before the driver is up and running but in XDP mode the issue might occur when the number of queues changes right after a reset trigger.
The ENA_MAX_MSIX_VEC simply adds one to the argument since the first msix vector is reserved for management queue.

Fixes 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
---
 drivers/net/ethernet/amazon/ena/ena_netdev.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +6 -3
index 39f290f4458f..836fda585391 100644
@@ -2068,6 +2068,7 @@ static int ena_request_mgmnt_irq(struct ena_adapter *adapter)
 
 static int ena_request_io_irq(struct ena_adapter *adapter)
 {
+	u32 io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues;
 	unsigned long flags = 0;
 	struct ena_irq *irq;
 	int rc = 0, i, k;
@@ -2078,7 +2079,7 @@ static int ena_request_io_irq(struct ena_adapter *adapter)
 		return -EINVAL;
 	}
 
-	for (i = ENA_IO_IRQ_FIRST_IDX; i < adapter->msix_vecs; i++) {
+	for (i = ENA_IO_IRQ_FIRST_IDX; i < ENA_MAX_MSIX_VEC(io_queue_count); i++) {
 		irq = &adapter->irq_tbl[i];
 		rc = request_irq(irq->vector, irq->handler, flags, irq->name,
 				 irq->data);
@@ -2119,6 +2120,7 @@ static void ena_free_mgmnt_irq(struct ena_adapter *adapter)
 
 static void ena_free_io_irq(struct ena_adapter *adapter)
 {
+	u32 io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues;
 	struct ena_irq *irq;
 	int i;
 
@@ -2129,7 +2131,7 @@ static void ena_free_io_irq(struct ena_adapter *adapter)
 	}
 #endif /* CONFIG_RFS_ACCEL */
 
-	for (i = ENA_IO_IRQ_FIRST_IDX; i < adapter->msix_vecs; i++) {
+	for (i = ENA_IO_IRQ_FIRST_IDX; i < ENA_MAX_MSIX_VEC(io_queue_count); i++) {
 		irq = &adapter->irq_tbl[i];
 		irq_set_affinity_hint(irq->vector, NULL);
 		free_irq(irq->vector, irq->data);
@@ -2144,12 +2146,13 @@ static void ena_disable_msix(struct ena_adapter *adapter)
 
 static void ena_disable_io_intr_sync(struct ena_adapter *adapter)
 {
+	u32 io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues;
 	int i;
 
 	if (!netif_running(adapter->netdev))
 		return;
 
-	for (i = ENA_IO_IRQ_FIRST_IDX; i < adapter->msix_vecs; i++)
+	for (i = ENA_IO_IRQ_FIRST_IDX; i < ENA_MAX_MSIX_VEC(io_queue_count); i++)
 		synchronize_irq(adapter->irq_tbl[i].vector);
 }
 
Signature · 1 line
2.17.1
author maintainer
#3 lore raw

PATCH v2 3/4 net: ena: avoid memory access violation by validating req_id properly

In reply to [email protected]

From: Arthur Kiyanovski <[email protected]>

Rx req_id is an index in struct ena_eth_io_rx_cdesc_base. The driver should validate that the Rx req_id it received from the device is in range [0, ring_size -1]. Failure to do so could yield to potential memory access violoation.
The validation was mistakenly done when refilling
the Rx submission queue and not in Rx completion queue.

Fixes ad974baef2a1 ("net: ena: add support for out of order rx buffers refill")
---
 drivers/net/ethernet/amazon/ena/ena_netdev.c | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +11 -4
index 836fda585391..51333a05c14d 100644
@@ -1018,13 +1018,9 @@ static int ena_refill_rx_bufs(struct ena_ring *rx_ring, u32 num)
 		struct ena_rx_buffer *rx_info;
 
 		req_id = rx_ring->free_ids[next_to_use];
-		rc = validate_rx_req_id(rx_ring, req_id);
-		if (unlikely(rc < 0))
-			break;
 
 		rx_info = &rx_ring->rx_buffer_info[req_id];
 
-
 		rc = ena_alloc_rx_page(rx_ring, rx_info,
 				       GFP_ATOMIC | __GFP_COMP);
 		if (unlikely(rc < 0)) {
@@ -1379,9 +1375,15 @@ static struct sk_buff *ena_rx_skb(struct ena_ring *rx_ring,
 	struct ena_rx_buffer *rx_info;
 	u16 len, req_id, buf = 0;
 	void *va;
+	int rc;
 
 	len = ena_bufs[buf].len;
 	req_id = ena_bufs[buf].req_id;
+
+	rc = validate_rx_req_id(rx_ring, req_id);
+	if (unlikely(rc < 0))
+		return NULL;
+
 	rx_info = &rx_ring->rx_buffer_info[req_id];
 
 	if (unlikely(!rx_info->page)) {
@@ -1454,6 +1456,11 @@ static struct sk_buff *ena_rx_skb(struct ena_ring *rx_ring,
 		buf++;
 		len = ena_bufs[buf].len;
 		req_id = ena_bufs[buf].req_id;
+
+		rc = validate_rx_req_id(rx_ring, req_id);
+		if (unlikely(rc < 0))
+			return NULL;
+
 		rx_info = &rx_ring->rx_buffer_info[req_id];
 	} while (1);
 
Signature · 1 line
2.17.1
author maintainer
#4 lore raw

PATCH v2 4/4 net: ena: fix continuous keep-alive resets

In reply to [email protected]

From: Arthur Kiyanovski <[email protected]>

last_keep_alive_jiffies is updated in probe and when a keep-alive event is received. In case the driver times-out on a keep-alive event, it has high chances of continuously timing-out on keep-alive events. This is because when the driver recovers from the keep-alive-timeout reset the value of last_keep_alive_jiffies is very old, and if a keep-alive event is not received before the next timer expires, the value of last_keep_alive_jiffies will cause another keep-alive-timeout reset and so forth in a loop.

Solution:
Update last_keep_alive_jiffies whenever the device is restored after reset.

Fixes 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
---
 drivers/net/ethernet/amazon/ena/ena_netdev.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/amazon/ena/ena_netdev.c +1 -0
index 51333a05c14d..4647d7656761 100644
@@ -3486,6 +3486,7 @@ static int ena_restore_device(struct ena_adapter *adapter)
 		netif_carrier_on(adapter->netdev);
 
 	mod_timer(&adapter->timer_service, round_jiffies(jiffies + HZ));
+	adapter->last_keep_alive_jiffies = jiffies;
 	dev_err(&pdev->dev,
 		"Device reset completed successfully, Driver info: %s\n",
 		version);
Signature · 1 line
2.17.1