RE: [PATCH v6 1/2] net/iavf: fix Rx/Tx burst in multi-process

["Zhang, Qi Z" <qi.z.zhang@intel.com>]

> -----Original Message-----
> From: Mingjin Ye <mingjinx.ye@intel.com>
> Sent: Friday, December 29, 2023 6:11 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming <qiming.yang@intel.com>; Ye, MingjinX
> <mingjinx.ye@intel.com>; stable@dpdk.org; Wu, Jingjing
> <jingjing.wu@intel.com>; Xing, Beilei <beilei.xing@intel.com>
> Subject: [PATCH v6 1/2] net/iavf: fix Rx/Tx burst in multi-process
> 
> In a multi-process environment, a secondary process operates on shared
> memory and changes the function pointer of the primary process, resulting
> in a crash when the primary process cannot find the function address during
> an Rx/Tx burst.
> 
> Fixes: 5b3124a0a6ef ("net/iavf: support no polling when link down")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Mingjin Ye <mingjinx.ye@intel.com>
> ---
> v2: Add fix for Rx burst.
> ---
>  drivers/net/iavf/iavf.h      |  42 +++++++-
>  drivers/net/iavf/iavf_rxtx.c | 184 ++++++++++++++++++++++++++++++-----
>  drivers/net/iavf/iavf_rxtx.h |   8 ++
>  3 files changed, 205 insertions(+), 29 deletions(-)
> 
> diff --git a/drivers/net/iavf/iavf.h b/drivers/net/iavf/iavf.h index
> 10868f2c30..8db9f3d7cd 100644
> --- a/drivers/net/iavf/iavf.h
> +++ b/drivers/net/iavf/iavf.h
> @@ -313,6 +313,44 @@ struct iavf_devargs {
> 
>  struct iavf_security_ctx;
> 
> +enum iavf_rx_burst_type {
> +	IAVF_RX_BURST_DEFAULT,
> +	IAVF_RX_BURST_FRXD,
> +	IAVF_RX_BURST_BULK_ALLOC,
> +	IAVF_RX_BURST_SCATTERED,
> +	IAVF_RX_BURST_SFRXD,
> +	IAVF_RX_BURST_VEC_SSE,
> +	IAVF_RX_BURST_VEC_AVX2,
> +	IAVF_RX_BURST_VEC_AVX2_OFFLOAD,
> +	IAVF_RX_BURST_VEC_SSE_FRXD,
> +	IAVF_RX_BURST_VEC_AVX2_FRXD,
> +	IAVF_RX_BURST_VEC_AVX2_FRXD_OFFLOAD,
> +	IAVF_RX_BURST_VEC_SSE_SCATTERED,
> +	IAVF_RX_BURST_VEC_AVX2_SCATTERED,
> +	IAVF_RX_BURST_VEC_AVX2_SCATTERED_OFFLOAD,
> +	IAVF_RX_BURST_VEC_SSE_SFLEX_RXD,
> +	IAVF_RX_BURST_VEC_AVX2_SFLEX_RXD,
> +	IAVF_RX_BURST_VEC_AVX2_SFRXD_OFFLOAD,
> +	IAVF_RX_BURST_VEC_AVX512,
> +	IAVF_RX_BURST_VEC_AVX512_OFFLOAD,
> +	IAVF_RX_BURST_VEC_AVX512_FRXD,
> +	IAVF_RX_BURST_VEC_AVX512_FRXD_OFFLOAD,
> +	IAVF_RX_BURST_VEC_AVX512_SCATTERED,
> +	IAVF_RX_BURST_VEC_AVX512_SCATTERED_OFFLOAD,
> +	IAVF_RX_BURST_VEC_AVX512_SFLEX_RXD,
> +	IAVF_RX_BURST_VEC_AVX512_SFRXD_OFFLOAD,

What is SFLEX, SFRXD, SFRXD, FRXD, please make it clear by following a consistent naming pattern.
Btw, you can consider removing BURST and VEC which didn't give any additional information if you are looking for a short name.

....

> @@ -3809,42 +3886,64 @@ iavf_set_rx_function(struct rte_eth_dev *dev)
>  			}
>  			if (use_flex) {
>  				dev->rx_pkt_burst =
> iavf_recv_scattered_pkts_vec_flex_rxd;
> +				rx_burst_type =
> IAVF_RX_BURST_VEC_SSE_SFLEX_RXD;
>  				if (use_avx2) {
> -					if (check_ret == IAVF_VECTOR_PATH)
> +					if (check_ret == IAVF_VECTOR_PATH)
> {
>  						dev->rx_pkt_burst =
> 
> 	iavf_recv_scattered_pkts_vec_avx2_flex_rxd;
> -					else
> +						rx_burst_type =
> +
> 	IAVF_RX_BURST_VEC_AVX2_SFLEX_RXD;

As you already introduce the burst_type, its not necessary to set the function pointer for each case.
Why not just 
dev->rx_pkt_burst = rx_burst_ops[rx_burst_type] at last?

....

> +struct iavf_rx_burst_ops {
> +	eth_rx_burst_t rx_pkt_burst;
> +};

Why create a wrapper here but not just use eth_rx_burst_t directly?