[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220410072930.GC212299@hoboy.vegasvil.org>
Date: Sun, 10 Apr 2022 00:29:30 -0700
From: Richard Cochran <richardcochran@...il.com>
To: Gerhard Engleder <gerhard@...leder-embedded.com>
Cc: vinicius.gomes@...el.com, yangbo.lu@....com, davem@...emloft.net,
kuba@...nel.org, mlichvar@...hat.com, netdev@...r.kernel.org
Subject: Re: [PATCH net-next v2 4/5] ptp: Support late timestamp determination
On Sun, Apr 03, 2022 at 07:55:43PM +0200, Gerhard Engleder wrote:
> @@ -887,18 +885,28 @@ void __sock_recv_timestamp(struct msghdr *msg, struct sock *sk,
> if (shhwtstamps &&
> (sk->sk_tsflags & SOF_TIMESTAMPING_RAW_HARDWARE) &&
> !skb_is_swtx_tstamp(skb, false_tstamp)) {
> + rcu_read_lock();
> + orig_dev = dev_get_by_napi_id(skb_napi_id(skb));
__sock_recv_timestamp() is hot path.
No need to call dev_get_by_napi_id() for the vast majority of cases
using plain old MAC time stamping.
Make this conditional on (sk->sk_tsflags & SOF_TIMESTAMPING_BIND_PHC).
Thanks,
Richard
> + if (orig_dev) {
> + if_index = orig_dev->ifindex;
> + hwtstamp = netdev_get_tstamp(orig_dev, shhwtstamps,
> + sk->sk_tsflags & SOF_TIMESTAMPING_BIND_PHC);
> + } else {
> + if_index = 0;
> + hwtstamp = shhwtstamps->hwtstamp;
> + }
> + rcu_read_unlock();
> +
> if (sk->sk_tsflags & SOF_TIMESTAMPING_BIND_PHC)
> - hwtstamp = ptp_convert_timestamp(&shhwtstamps->hwtstamp,
> + hwtstamp = ptp_convert_timestamp(&hwtstamp,
> sk->sk_bind_phc);
> - else
> - hwtstamp = shhwtstamps->hwtstamp;
>
> if (ktime_to_timespec64_cond(hwtstamp, tss.ts + 2)) {
> empty = 0;
>
> if ((sk->sk_tsflags & SOF_TIMESTAMPING_OPT_PKTINFO) &&
> !skb_is_err_queue(skb))
> - put_ts_pktinfo(msg, skb);
> + put_ts_pktinfo(msg, skb, if_index);
> }
> }
> if (!empty) {
> --
> 2.20.1
>
Powered by blists - more mailing lists