public inbox for dev@dpdk.org
 help / color / mirror / Atom feed
From: spinler@cesnet.cz
To: dev@dpdk.org
Cc: Martin Spinler <spinler@cesnet.cz>, stable@dpdk.org
Subject: [PATCH v5 2/6] net/nfb: fix bad pointer access in queue stats
Date: Mon,  2 Feb 2026 20:33:26 +0100	[thread overview]
Message-ID: <20260202193330.3324681-3-spinler@cesnet.cz> (raw)
In-Reply-To: <20260202193330.3324681-1-spinler@cesnet.cz>

From: Martin Spinler <spinler@cesnet.cz>

The driver code has dereferenced the dev->data->rx_queues pointer
without checking for its validity.
Pointer invalidation can occur when the eth_dev_rx_queue_config
is called with set to 0, for example.

Moreover, an array of pointers (to a structure) was used like array
of structures (which worked with early dereference just for one queue).

Fixes: 6435f9a0ac22 ("net/nfb: add new netcope driver")
Cc: stable@dpdk.org

Signed-off-by: Martin Spinler <spinler@cesnet.cz>
---
 drivers/net/nfb/nfb_stats.c | 46 ++++++++++++++++++-------------------
 1 file changed, 23 insertions(+), 23 deletions(-)

diff --git a/drivers/net/nfb/nfb_stats.c b/drivers/net/nfb/nfb_stats.c
index 4ea6b7be21..27a01c3160 100644
--- a/drivers/net/nfb/nfb_stats.c
+++ b/drivers/net/nfb/nfb_stats.c
@@ -20,28 +20,28 @@ nfb_eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats,
 	uint64_t rx_total_bytes = 0;
 	uint64_t tx_total_bytes = 0;
 
-	struct ndp_rx_queue *rx_queue = *((struct ndp_rx_queue **)
-		dev->data->rx_queues);
-	struct ndp_tx_queue *tx_queue = *((struct ndp_tx_queue **)
-		dev->data->tx_queues);
+	struct ndp_rx_queue *rx_queue;
+	struct ndp_tx_queue *tx_queue;
 
 	for (i = 0; i < nb_rx; i++) {
+		rx_queue = dev->data->rx_queues[i];
 		if (qstats && i < RTE_ETHDEV_QUEUE_STAT_CNTRS) {
-			qstats->q_ipackets[i] = rx_queue[i].rx_pkts;
-			qstats->q_ibytes[i] = rx_queue[i].rx_bytes;
+			qstats->q_ipackets[i] = rx_queue->rx_pkts;
+			qstats->q_ibytes[i] = rx_queue->rx_bytes;
 		}
-		rx_total += rx_queue[i].rx_pkts;
-		rx_total_bytes += rx_queue[i].rx_bytes;
+		rx_total += rx_queue->rx_pkts;
+		rx_total_bytes += rx_queue->rx_bytes;
 	}
 
 	for (i = 0; i < nb_tx; i++) {
+		tx_queue = dev->data->tx_queues[i];
 		if (qstats && i < RTE_ETHDEV_QUEUE_STAT_CNTRS) {
-			qstats->q_opackets[i] = tx_queue[i].tx_pkts;
-			qstats->q_obytes[i] = tx_queue[i].tx_bytes;
+			qstats->q_opackets[i] = tx_queue->tx_pkts;
+			qstats->q_obytes[i] = tx_queue->tx_bytes;
 		}
-		tx_total += tx_queue[i].tx_pkts;
-		tx_total_bytes += tx_queue[i].tx_bytes;
-		tx_err_total += tx_queue[i].err_pkts;
+		tx_total += tx_queue->tx_pkts;
+		tx_total_bytes += tx_queue->tx_bytes;
+		tx_err_total += tx_queue->err_pkts;
 	}
 
 	stats->ipackets = rx_total;
@@ -59,20 +59,20 @@ nfb_eth_stats_reset(struct rte_eth_dev *dev)
 	uint16_t nb_rx = dev->data->nb_rx_queues;
 	uint16_t nb_tx = dev->data->nb_tx_queues;
 
-	struct ndp_rx_queue *rx_queue = *((struct ndp_rx_queue **)
-		dev->data->rx_queues);
-	struct ndp_tx_queue *tx_queue = *((struct ndp_tx_queue **)
-		dev->data->tx_queues);
+	struct ndp_rx_queue *rx_queue;
+	struct ndp_tx_queue *tx_queue;
 
 	for (i = 0; i < nb_rx; i++) {
-		rx_queue[i].rx_pkts = 0;
-		rx_queue[i].rx_bytes = 0;
-		rx_queue[i].err_pkts = 0;
+		rx_queue = dev->data->rx_queues[i];
+		rx_queue->rx_pkts = 0;
+		rx_queue->rx_bytes = 0;
+		rx_queue->err_pkts = 0;
 	}
 	for (i = 0; i < nb_tx; i++) {
-		tx_queue[i].tx_pkts = 0;
-		tx_queue[i].tx_bytes = 0;
-		tx_queue[i].err_pkts = 0;
+		tx_queue = dev->data->tx_queues[i];
+		tx_queue->tx_pkts = 0;
+		tx_queue->tx_bytes = 0;
+		tx_queue->err_pkts = 0;
 	}
 
 	return 0;
-- 
2.52.0


  parent reply	other threads:[~2026-02-02 19:33 UTC|newest]

Thread overview: 58+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-15 14:01 [PATCH 0/6] spinler
2026-01-15 14:01 ` [PATCH 1/6] net/nfb: use constant values for max Rx/Tx queues count spinler
2026-01-15 14:01 ` [PATCH 2/6] net/nfb: fix bad pointer access in queue stats spinler
2026-01-15 14:01 ` [PATCH 3/6] net/nfb: update timestamp calculation to meaningful value spinler
2026-01-15 14:01 ` [PATCH 4/6] net/nfb: use process private variable for internal data spinler
2026-01-15 14:01 ` [PATCH 5/6] net/nfb: release allocated resources correctly spinler
2026-01-15 14:01 ` [PATCH 6/6] net/nfb: stop only started queues in fail path spinler
2026-01-15 14:40 ` [PATCH v2 0/6] net/nfb: code cleanup spinler
2026-01-15 14:40   ` [PATCH v2 1/6] net/nfb: use constant values for max Rx/Tx queues count spinler
2026-01-15 14:40   ` [PATCH v2 2/6] net/nfb: fix bad pointer access in queue stats spinler
2026-01-15 14:40   ` [PATCH v2 3/6] net/nfb: update timestamp calculation to meaningful value spinler
2026-01-15 14:40   ` [PATCH v2 4/6] net/nfb: use process private variable for internal data spinler
2026-01-15 14:40   ` [PATCH v2 5/6] net/nfb: release allocated resources correctly spinler
2026-01-15 14:40   ` [PATCH v2 6/6] net/nfb: stop only started queues in fail path spinler
2026-01-16  5:48   ` [PATCH v2 0/6] net/nfb: code cleanup Stephen Hemminger
2026-01-16  9:42     ` Martin Spinler
2026-01-16 17:39       ` Stephen Hemminger
2026-01-16 15:20 ` spinler
2026-01-16 15:20   ` [PATCH v3 1/6] net/nfb: use constant values for max Rx/Tx queues count spinler
2026-02-02 17:47     ` Stephen Hemminger
2026-02-02 18:58       ` Martin Špinler
2026-01-16 15:20   ` [PATCH v3 2/6] net/nfb: fix bad pointer access in queue stats spinler
2026-01-16 15:20   ` [PATCH v3 3/6] net/nfb: update timestamp calculation to meaningful value spinler
2026-01-16 15:20   ` [PATCH v3 4/6] net/nfb: use process private variable for internal data spinler
2026-01-20  0:13     ` Stephen Hemminger
2026-01-20 14:13       ` Martin Spinler
2026-01-20 16:11         ` Stephen Hemminger
2026-01-16 15:20   ` [PATCH v3 5/6] net/nfb: release allocated resources correctly spinler
2026-01-20  0:10     ` Stephen Hemminger
2026-01-20 14:14       ` Martin Spinler
2026-01-16 15:20   ` [PATCH v3 6/6] net/nfb: stop only started queues in fail path spinler
2026-01-20  0:09     ` Stephen Hemminger
2026-01-20 14:14       ` Martin Spinler
2026-01-16 15:22 ` [PATCH v3 0/6] net/nfb: code cleanup spinler
2026-01-21  4:57   ` Stephen Hemminger
2026-01-21 17:01 ` [PATCH v4 " spinler
2026-01-21 17:01   ` [PATCH v4 1/6] net/nfb: use constant values for max Rx/Tx queues count spinler
2026-01-21 17:01   ` [PATCH v4 2/6] net/nfb: fix bad pointer access in queue stats spinler
2026-01-21 17:01   ` [PATCH v4 3/6] net/nfb: update timestamp calculation to meaningful value spinler
2026-01-21 17:33     ` Stephen Hemminger
2026-01-27  8:12       ` Martin Spinler
2026-01-27  0:34     ` Stephen Hemminger
2026-01-27  8:16       ` Martin Spinler
2026-01-21 17:01   ` [PATCH v4 4/6] net/nfb: use process private variable for internal data spinler
2026-01-21 17:01   ` [PATCH v4 5/6] net/nfb: release allocated resources correctly spinler
2026-01-21 17:01   ` [PATCH v4 6/6] net/nfb: stop only started queues in fail path spinler
2026-01-21 17:35   ` [PATCH v4 0/6] net/nfb: code cleanup Stephen Hemminger
2026-02-02 19:33 ` [PATCH v5 " spinler
2026-02-02 19:33   ` [PATCH v5 1/6] net/nfb: use constant values for max Rx/Tx queues count spinler
2026-02-02 19:33   ` spinler [this message]
2026-02-10  0:51     ` [PATCH v5 2/6] net/nfb: fix bad pointer access in queue stats Stephen Hemminger
2026-02-02 19:33   ` [PATCH v5 3/6] net/nfb: update timestamp calculation to meaningful value spinler
2026-02-02 19:33   ` [PATCH v5 4/6] net/nfb: use process private variable for internal data spinler
2026-02-02 19:33   ` [PATCH v5 5/6] net/nfb: release allocated resources correctly spinler
2026-02-10  0:52     ` Stephen Hemminger
2026-02-02 19:33   ` [PATCH v5 6/6] net/nfb: stop only started queues in fail path spinler
2026-02-03  1:50   ` [PATCH v5 0/6] net/nfb: code cleanup Stephen Hemminger
2026-02-03  6:36     ` Martin Spinler

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260202193330.3324681-3-spinler@cesnet.cz \
    --to=spinler@cesnet.cz \
    --cc=dev@dpdk.org \
    --cc=stable@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox