[PATCH net V2 2/2] net: xilinx: axienet: Fix BQL accounting for multi-BD TX packets

[Sean Anderson]

On 3/27/26 03:32, Suraj Gupta wrote:
> When a TX packet spans multiple buffer descriptors (scatter-gather),
> axienet_free_tx_chain sums the per-BD actual length from descriptor
> status into a caller-provided accumulator. That sum is reset on each
> NAPI poll. If the BDs for a single packet complete across different
> polls, the earlier bytes are lost and never credited to BQL. This
> causes BQL to think bytes are permanently in-flight, eventually
> stalling the TX queue.
> 
> The SKB pointer is stored only on the last BD of a packet. When that
> BD completes, use skb->len for the byte count instead of summing
> per-BD status lengths. This matches netdev_sent_queue(), which debits
> skb->len, and naturally survives across polls because no partial
> packet contributes to the accumulator.
> 
> Fixes: c900e49d58eb ("net: xilinx: axienet: Implement BQL")
> Signed-off-by: Suraj Gupta <suraj.gupta2 at amd.com>
> ---
>  drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 9 ++++-----
>  1 file changed, 4 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
> index b06e4c37ff61..263c4b67fd5a 100644
> --- a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
> +++ b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
> @@ -770,8 +770,8 @@ static int axienet_device_reset(struct net_device *ndev)
>   * @first_bd:	Index of first descriptor to clean up
>   * @nr_bds:	Max number of descriptors to clean up
>   * @force:	Whether to clean descriptors even if not complete
> - * @sizep:	Pointer to a u32 filled with the total sum of all bytes
> - *		in all cleaned-up descriptors. Ignored if NULL.
> + * @sizep:	Pointer to a u32 accumulating the total byte count of
> + *		completed packets (using skb->len). Ignored if NULL.
>   * @budget:	NAPI budget (use 0 when not called from NAPI poll)
>   *
>   * Would either be called after a successful transmit operation, or after
> @@ -805,6 +805,8 @@ static int axienet_free_tx_chain(struct axienet_local *lp, u32 first_bd,
>  				 DMA_TO_DEVICE);
>  
>  		if (cur_p->skb && (status & XAXIDMA_BD_STS_COMPLETE_MASK)) {
> +			if (sizep)
> +				*sizep += cur_p->skb->len;
>  			napi_consume_skb(cur_p->skb, budget);
>  			packets++;
>  		}
> @@ -818,9 +820,6 @@ static int axienet_free_tx_chain(struct axienet_local *lp, u32 first_bd,
>  		wmb();
>  		cur_p->cntrl = 0;
>  		cur_p->status = 0;
> -
> -		if (sizep)
> -			*sizep += status & XAXIDMA_BD_STS_ACTUAL_LEN_MASK;
>  	}
>  
>  	if (!force) {

Reviewed-by: Sean Anderson <sean.anderson at linux.dev>

Although FWIW this may result in slightly-different statistics. Not sure
if we care since this will only affect packets that could not be sent
for whatever reason (collisions, loss of carrier, etc.)