On Mon, Feb 17, 2025 at 09:31:49AM +0200, Roger Quadros wrote:
> This allows us to re-use am65_cpsw_run_xdp() for zero copy
> case. Add AM65_CPSW_XDP_TX case for successful XDP_TX so we don't
> free the page while in flight.
>
> Signed-off-by: Roger Quadros <rogerq@xxxxxxxxxx>
Reviewed-by: Simon Horman <horms@xxxxxxxxxx>
> ---
> drivers/net/ethernet/ti/am65-cpsw-nuss.c | 13 ++++++++-----
> 1 file changed, 8 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
...
> @@ -1230,9 +1230,6 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_rx_flow *flow,
> ndev->stats.rx_dropped++;
> }
>
> - page = virt_to_head_page(xdp->data);
> - am65_cpsw_put_page(flow, page, true);
> -
> return ret;
It seems that before and after this patch ret is always initialised to
AM65_CPSW_XDP_CONSUMED and never changed. So it can be removed.
Given that with this patch the function only returns after the switch
statement, I think this would be a nice follow-up.
diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
index 20a4fc3e579f..4052c9153632 100644
--- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c
+++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
@@ -1172,7 +1172,6 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_rx_flow *flow,
{
struct am65_cpsw_common *common = flow->common;
struct net_device *ndev = port->ndev;
- int ret = AM65_CPSW_XDP_CONSUMED;
struct am65_cpsw_tx_chn *tx_chn;
struct netdev_queue *netif_txq;
int cpu = smp_processor_id();
@@ -1228,9 +1227,8 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_rx_flow *flow,
fallthrough;
case XDP_DROP:
ndev->stats.rx_dropped++;
+ return AM65_CPSW_XDP_CONSUMED;
}
-
- return ret;
}
/* RX psdata[2] word format - checksum information */
...
