ACK/cmt: [SRU][B/E/F][PATCH] can: peak_canfd: provide hw timestamps in rx skbs

Khaled Elmously khalid.elmously at canonical.com
Wed Apr 22 07:32:05 UTC 2020



On 2020-04-21 11:38:24 , Sultan Alsawaf wrote:
> From: Stephane Grosjean <s.grosjean at peak-system.com>
> 
> BugLink: https://bugs.launchpad.net/bugs/1874124
> 
> PEAK-System's CAN FD interfaces based on an IP core provide a timestamp
> for each CAN and STATUS message received. This patch transfers these
> received timestamps (clocked in microseconds) to hardware timestamps
> (clocked in nanoseconds) in the corresponding skbs raised to the network
> layer.
> 
> Signed-off-by: Stephane Grosjean <s.grosjean at peak-system.com>
> Signed-off-by: Marc Kleine-Budde <mkl at pengutronix.de>
> (cherry picked from commit 2b1a4547c122dcb9999e1a876236b44408c7d01c)
> Signed-off-by: Sultan Alsawaf <sultan.alsawaf at canonical.com>
> ---
>  drivers/net/can/peak_canfd/peak_canfd.c | 18 ++++++++++++++++--
>  1 file changed, 16 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/can/peak_canfd/peak_canfd.c b/drivers/net/can/peak_canfd/peak_canfd.c
> index 6b0c6a99fc8d..e1876fbf7b14 100644
> --- a/drivers/net/can/peak_canfd/peak_canfd.c
> +++ b/drivers/net/can/peak_canfd/peak_canfd.c
> @@ -232,6 +232,20 @@ static int pucan_setup_rx_barrier(struct peak_canfd_priv *priv)
>  	return pucan_write_cmd(priv);
>  }
>  
> +static int pucan_netif_rx(struct sk_buff *skb, __le32 ts_low, __le32 ts_high)
> +{
> +	struct skb_shared_hwtstamps *hwts = skb_hwtstamps(skb);
> +	u64 ts_us;
> +
> +	ts_us = (u64)le32_to_cpu(ts_high) << 32;
> +	ts_us |= le32_to_cpu(ts_low);
> +
> +	/* IP core timestamps are ┬Ás. */
> +	hwts->hwtstamp = ns_to_ktime(ts_us * NSEC_PER_USEC);
> +
> +	return netif_rx(skb);
> +}
> +
>  /* handle the reception of one CAN frame */
>  static int pucan_handle_can_rx(struct peak_canfd_priv *priv,
>  			       struct pucan_rx_msg *msg)
> @@ -299,7 +313,7 @@ static int pucan_handle_can_rx(struct peak_canfd_priv *priv,
>  	stats->rx_bytes += cf->len;
>  	stats->rx_packets++;
>  
> -	netif_rx(skb);
> +	pucan_netif_rx(skb, msg->ts_low, msg->ts_high);
>  
>  	return 0;
>  }
> @@ -393,7 +407,7 @@ static int pucan_handle_status(struct peak_canfd_priv *priv,
>  
>  	stats->rx_packets++;
>  	stats->rx_bytes += cf->can_dlc;
> -	netif_rx(skb);
> +	pucan_netif_rx(skb, msg->ts_low, msg->ts_high);
>  
>  	return 0;
>  }

A link to the salesforce in the bug is fine.

Acked-by: Khalid Elmously <khalid.elmously at canonical.com>




More information about the kernel-team mailing list