[dpdk-dev] [PATCH 1/5] net/fm10k: fix dereference after null check coverity

Ye Xiaolong xiaolong.ye at intel.com
Mon Aug 5 05:35:09 CEST 2019

Hi, Xiao

It's better to have a cover letter for a patch series that contains more than
1 patches, where you can have an overall description of the patchset.

On 08/05, Xiao Zhang wrote:
>The address of receive queue start segment was not updated when found by
>iterated checking, update the address to fix coverity issue.

I suggest to describe more details (dereference after null check) about the
issue other than just say general coverity issue.

>Coverity issue: 343416
>Fixes: fe65e1e1 ("fm10k: add vector scatter Rx")
>Cc: stable at dpdk.org
>Signed-off-by: Xiao Zhang <xiao.zhang at intel.com>
> drivers/net/fm10k/fm10k_rxtx_vec.c | 1 +
> 1 file changed, 1 insertion(+)
>diff --git a/drivers/net/fm10k/fm10k_rxtx_vec.c b/drivers/net/fm10k/fm10k_rxtx_vec.c
>index 788e248..9f2c4ac 100644
>--- a/drivers/net/fm10k/fm10k_rxtx_vec.c
>+++ b/drivers/net/fm10k/fm10k_rxtx_vec.c
>@@ -676,6 +676,7 @@ fm10k_recv_scattered_pkts_vec(void *rx_queue,
> 		/* find the first split flag, and only reassemble then*/
> 		while (i < nb_bufs && !split_flags[i])
> 			i++;
>+		rxq->pkt_first_seg = rx_pkts[i];

Minor nit, if i == nb_bufs, above assignment is  meaningless, suggest to move
it after i == nb_bufs check although it doesn't affect the functionality. 
I think it applies for the rest of patches too.


> 		if (i == nb_bufs)
> 			return nb_bufs;
> 	}

More information about the dev mailing list