[PATCH 2/2] net/mana: optimize completion queue polling by processing a batch at a time

[Ferruh Yigit]

On 3/17/2023 11:32 PM, longli at linuxonhyperv.com wrote:
> -	while (pkt_received < pkts_n &&
> -	       gdma_poll_completion_queue(&rxq->gdma_cq, &comp) == 1) {
> -		struct mana_rxq_desc *desc;
> -		struct mana_rx_comp_oob *oob =
> -			(struct mana_rx_comp_oob *)&comp.completion_data[0];
> -
> -		if (comp.work_queue_number != rxq->gdma_rq.id) {
> -			DRV_LOG(ERR, "rxq comp id mismatch wqid=0x%x rcid=0x%x",
> -				comp.work_queue_number, rxq->gdma_rq.id);
> -			rxq->stats.errors++;
> -			break;
> -		}
> +	num_pkts = gdma_poll_completion_queue(&rxq->gdma_cq, rxq->gdma_comp_buf, pkts_n);
> +	for (uint32_t i = 0; i < num_pkts; i++) {
> +		struct mana_rx_comp_oob *oob = (struct mana_rx_comp_oob *)
> +			rxq->gdma_comp_buf[i].cqe_data;
> +		struct mana_rxq_desc *desc =
> +			&rxq->desc_ring[rxq->desc_ring_tail];


Checkpatch complains about defining variable inside for() [1], but as we
are switching to C99 standard, I don't see any reason to prevent this usage.

I can see relevant checkpatch commit mentions about compiler warning [2]
but not sure if that is still valid with C99 switch, I will send a patch
to discuss this on.



[1]
Warning in drivers/net/mana/rx.c:
Declaring a variable inside for()

[2]
43e73483a4b8 ("devtools: forbid variable declaration inside for")