[dpdk-stable] [PATCH] net/thunderx: fix access an array out of bounds
Jerin Jacob
jerin.jacob at caviumnetworks.com
Tue Apr 11 12:00:44 CEST 2017
-----Original Message-----
> Date: Fri, 7 Apr 2017 16:38:41 +0200
> From: marcin.wilk at caviumnetworks.com
> To: jerin.jacob at caviumnetworks.com, maciej.czekaj at caviumnetworks.com
> Cc: dev at dpdk.org, stable at dpdk.org, Marcin Wilk
> <marcin.wilk at caviumnetworks.com>
> Subject: [PATCH] net/thunderx: fix access an array out of bounds
> X-Mailer: git-send-email 2.7.4
>
> From: Marcin Wilk <marcin.wilk at caviumnetworks.com>
>
> Trying to assign more queues to stats struct break only from one loop
> when the maximum size is reached. Outside loop interation is continued.
> This leads to access an array out of bounds.
>
> Fixes: 21e3fb0050b9 ("net/thunderx: add final bits for secondary queue support")
>
> Signed-off-by: Marcin Wilk <marcin.wilk at caviumnetworks.com>
> ---
> drivers/net/thunderx/nicvf_ethdev.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/thunderx/nicvf_ethdev.c b/drivers/net/thunderx/nicvf_ethdev.c
> index 6c3670a..40d6671 100644
> --- a/drivers/net/thunderx/nicvf_ethdev.c
> +++ b/drivers/net/thunderx/nicvf_ethdev.c
> @@ -251,6 +251,7 @@ nicvf_dev_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
> uint16_t rx_start, rx_end;
> uint16_t tx_start, tx_end;
> size_t i;
> + bool breakout = false;
>
> /* RX queue indices for the first VF */
> nicvf_rx_range(dev, nic, &rx_start, &rx_end);
> @@ -289,8 +290,10 @@ nicvf_dev_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
>
> /* Reading per RX ring stats */
> for (qidx = rx_start; qidx <= rx_end; qidx++) {
> - if (qidx == RTE_ETHDEV_QUEUE_STAT_CNTRS)
IMO, adding the qidx >= RTE_ETHDEV_QUEUE_STAT_CNTRS is much simpler than
breakout logic.
With that change:
Acked-by: Jerin Jacob <jerin.jacob at caviumnetworks.com>
> + if (qidx == RTE_ETHDEV_QUEUE_STAT_CNTRS) {
> + breakout = true;
> break;
> + }
>
> nicvf_hw_get_rx_qstats(snic, &rx_qstats,
> qidx % MAX_RCV_QUEUES_PER_QS);
> @@ -302,14 +305,18 @@ nicvf_dev_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
> nicvf_tx_range(dev, snic, &tx_start, &tx_end);
> /* Reading per TX ring stats */
> for (qidx = tx_start; qidx <= tx_end; qidx++) {
> - if (qidx == RTE_ETHDEV_QUEUE_STAT_CNTRS)
> + if (qidx == RTE_ETHDEV_QUEUE_STAT_CNTRS) {
> + breakout = true;
> break;
> + }
>
> nicvf_hw_get_tx_qstats(snic, &tx_qstats,
> qidx % MAX_SND_QUEUES_PER_QS);
> stats->q_obytes[qidx] = tx_qstats.q_tx_bytes;
> stats->q_opackets[qidx] = tx_qstats.q_tx_packets;
> }
> + if (breakout)
> + break;
> }
>
> nicvf_hw_get_stats(nic, &port_stats);
> --
> 2.7.4
>
More information about the stable
mailing list