Smatch complains that "local_rx_pd->priority" can't be trusted because
it comes from skb->data and it can go up to 255 instead of being capped
in the 0-7 range.  A few lines earlier, on the other side of the if
statement, we cap priority so it seems harmless to add a bounds check
here as well.

Signed-off-by: Dan Carpenter <[email protected]>
---
 drivers/net/wireless/marvell/mwifiex/sta_rx.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/wireless/marvell/mwifiex/sta_rx.c 
b/drivers/net/wireless/marvell/mwifiex/sta_rx.c
index fb28a5c7f441..52a2ce2e78b0 100644
--- a/drivers/net/wireless/marvell/mwifiex/sta_rx.c
+++ b/drivers/net/wireless/marvell/mwifiex/sta_rx.c
@@ -250,7 +250,8 @@ int mwifiex_process_sta_rx_packet(struct mwifiex_private 
*priv,
                                                             local_rx_pd->nf);
                }
        } else {
-               if (rx_pkt_type != PKT_TYPE_BAR)
+               if (rx_pkt_type != PKT_TYPE_BAR &&
+                   local_rx_pd->priority < MAX_NUM_TID)
                        priv->rx_seq[local_rx_pd->priority] = seq_num;
                memcpy(ta, priv->curr_bss_params.bss_descriptor.mac_address,
                       ETH_ALEN);
-- 
2.17.1

Reply via email to