|SUCCESS| net/sxe2: fix 32-bit SSE build

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: SUCCESS
https://dpdk.org/patch/164527

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-05-28

git sha: message id 20260528084717.1754057-1-thomas at monjalon.net
Author: Thomas Monjalon <thomas at monjalon.net>
Subject: [PATCH] net/sxe2: fix 32-bit SSE build

This patch addresses a compiler error on i586 Debian builds with
32-bit compilation when using SSE intrinsics. The patch replaces
rte_memcpy with standard memcpy for copying an array of pointers.

> diff --git a/drivers/net/sxe2/sxe2_txrx_vec_common.h b/drivers/net/sxe2/sxe2_txrx_vec_common.h
> index 99e1663f03..6b1649c390 100644
> --- a/drivers/net/sxe2/sxe2_txrx_vec_common.h
> +++ b/drivers/net/sxe2/sxe2_txrx_vec_common.h

[ ... ]

> @@ -230,7 +230,8 @@ sxe2_rx_pkts_refactor(struct sxe2_rx_queue *rxq,
>  	}
>  	rxq->pkt_first_seg = first_seg;
>  	rxq->pkt_last_seg  = last_seg;
> -	rte_memcpy(mbuf_bufs, done_pkts, done_num * (sizeof(struct rte_mbuf *)));
> +	memcpy(mbuf_bufs, done_pkts, done_num * sizeof(*done_pkts));

While the change from rte_memcpy to memcpy addresses the immediate
build issue, does the size calculation remain correct? The original code
uses sizeof(struct rte_mbuf *), while the new code uses
sizeof(*done_pkts). Can you verify that these are equivalent for the
done_pkts array declaration?

>  	return done_num;
>  }
> +
>  #endif /* SXE2_TXRX_VEC_COMMON_H */

The added blank line at the end of the file appears to be an unrelated
style change. Is this intentional or was it added inadvertently?