> -----Original Message-----
> From: Su, Simei <simei...@intel.com>
> Sent: Thursday, January 14, 2021 1:23 PM
> To: Zhang, Qi Z <qi.z.zh...@intel.com>
> Cc: dev@dpdk.org; Guo, Junfeng <junfeng....@intel.com>; Guo, Jia
> <jia....@intel.com>; Cao, Yahui <yahui....@intel.com>; Su, Simei
> <simei...@intel.com>
> Subject: [PATCH] net/iavf: fix null pointer dereference
>
> A pointer has already been dereferenced before checking if it is NULL.
> It doesn't make any sense, so correct to avoid it.
>
> Fixes: 4f3cfcbc3df3 ("net/iavf: support eCPRI msg type 0 for RSS") Coverity
> issue:
> 365290
>
> Signed-off-by: Simei Su <simei...@intel.com>
> ---
> drivers/net/iavf/iavf_hash.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/iavf/iavf_hash.c b/drivers/net/iavf/iavf_hash.c index
> ebaac58..9901f4b 100644
> --- a/drivers/net/iavf/iavf_hash.c
> +++ b/drivers/net/iavf/iavf_hash.c
> @@ -597,11 +597,13 @@ iavf_hash_parse_pattern(const struct rte_flow_item
> pattern[], uint64_t *phint,
> break;
> case RTE_FLOW_ITEM_TYPE_ECPRI:
> ecpri = item->spec;
> - ecpri_common.u32 = rte_be_to_cpu_32(
> - ecpri->hdr.common.u32);
> if (!ecpri)
> break;
> - else if (ecpri_common.type !=
> +
> + ecpri_common.u32 = rte_be_to_cpu_32(
> + ecpri->hdr.common.u32);
patchwork warning CHECK:OPEN_ENDED_LINE for above code
As we have 100 char line limit now, the wrap is not necessary.
Acked-by: Qi Zhang <qi.z.zh...@intel.com>
Applied to dpdk-next-net-intel after capture above fix.
Thanks
Qi
