On Wed, 2023-02-08 at 16:34 +0900, Yoshihiro Shimoda wrote:
> The gptp flag is completely related to the !dir_tx in struct
> rswitch_gwca_queue. In the future, a new queue handling for
> timestamp will be implemented and this gptp flag is confusable.
> So, remove the gptp flag.
>
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@xxxxxxxxxxx>
Based on these changes I am assuming that gptp == !dir_tx? Am I
understanding it correctly? It would be useful if you called that out
in the patch description.
> ---
> drivers/net/ethernet/renesas/rswitch.c | 26 +++++++++++---------------
> drivers/net/ethernet/renesas/rswitch.h | 1 -
> 2 files changed, 11 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/net/ethernet/renesas/rswitch.c b/drivers/net/ethernet/renesas/rswitch.c
> index b256dadada1d..e408d10184e8 100644
> --- a/drivers/net/ethernet/renesas/rswitch.c
> +++ b/drivers/net/ethernet/renesas/rswitch.c
> @@ -280,11 +280,14 @@ static void rswitch_gwca_queue_free(struct net_device *ndev,
> {
> int i;
>
> - if (gq->gptp) {
> + if (!gq->dir_tx) {
> dma_free_coherent(ndev->dev.parent,
> sizeof(struct rswitch_ext_ts_desc) *
> (gq->ring_size + 1), gq->rx_ring, gq->ring_dma);
> gq->rx_ring = NULL;
> +
> + for (i = 0; i < gq->ring_size; i++)
> + dev_kfree_skb(gq->skbs[i]);
> } else {
> dma_free_coherent(ndev->dev.parent,
> sizeof(struct rswitch_ext_desc) *
> @@ -292,11 +295,6 @@ static void rswitch_gwca_queue_free(struct net_device *ndev,
> gq->tx_ring = NULL;
> }
>
> - if (!gq->dir_tx) {
> - for (i = 0; i < gq->ring_size; i++)
> - dev_kfree_skb(gq->skbs[i]);
> - }
> -
> kfree(gq->skbs);
> gq->skbs = NULL;
> }
One piece I don't understand is why freeing of the skbs stored in the
array here was removed. Is this cleaned up somewhere else before we
call this function?
