From: Christian Eggers <ceggers@arri.de>
To: Vladimir Oltean <olteanv@gmail.com>
Cc: Richard Cochran <richardcochran@gmail.com>,
Florian Fainelli <f.fainelli@gmail.com>,
Andrew Lunn <andrew@lunn.ch>,
Vivien Didelot <vivien.didelot@gmail.com>,
Jakub Kicinski <kuba@kernel.org>,
Rob Herring <robh+dt@kernel.org>,
Helmut Grohne <helmut.grohne@intenta.de>,
Paul Barker <pbarker@konsulko.com>,
Codrin Ciubotariu <codrin.ciubotariu@microchip.com>,
George McCollister <george.mccollister@gmail.com>,
Marek Vasut <marex@denx.de>,
Tristram Ha <Tristram.Ha@microchip.com>,
"David S . Miller" <davem@davemloft.net>,
Woojung Huh <woojung.huh@microchip.com>,
"Microchip Linux Driver Support" <UNGLinuxDriver@microchip.com>,
<netdev@vger.kernel.org>, <devicetree@vger.kernel.org>,
<linux-kernel@vger.kernel.org>
Subject: Re: [RFC PATCH net-next 7/9] net: dsa: microchip: ksz9477: add hardware time stamping support
Date: Thu, 5 Nov 2020 21:18:04 +0100 [thread overview]
Message-ID: <5844018.3araiXeC39@n95hx1g2> (raw)
In-Reply-To: <20201022113243.4shddtywgvpcqq6c@skbuf>
Hi Vladimir,
On Thursday, 22 October 2020, 13:32:43 CET, Vladimir Oltean wrote:
> On Thu, Oct 22, 2020 at 01:11:40PM +0200, Christian Eggers wrote:
> > On Thursday, 22 October 2020, 12:50:14 CEST, Vladimir Oltean wrote:
> > after applying the RX timestamp correctly to the correction field
> > (shifting
> > the nanoseconds by 16),
>
> That modification should have been done anyway, since the unit of
> measurement for correctionField is scaled ppb (48 bits nanoseconds, 16
> bits scaled nanoseconds), and not nanoseconds.
>
> > it seems that "moving" the timestamp back to the tail tag on TX is not
> > required anymore. Keeping the RX timestamp simply in the correction
> > field (negative value), works fine now. So this halves the effort in
> > the tag_ksz driver.
unfortunately I made a mistake when testing. Actually the timestamp *must* be
moved from the correction field (negative) to the egress tail tag.
> Ok, this makes sense.
> Depending on what Richard responds, it now looks like the cleanest
> approach would be to move your implementation that is currently in
> ksz9477_update_ptp_correction_field() into a generic function called
>
> static inline void ptp_onestep_p2p_move_t2_to_correction(struct sk_buff
> *skb, unsigned int ptp_type,
> struct ptp_header *ptp_header,
> ktime_t t2)
I have implemented this in ptp_classify.h. Passing t2 instead of the correction
field itself is fine for rx, but as this function is now still required for
transmit, it looks a little bit misused there (see below).
Shall I keep it as below, or revert it to passing value of the correction field
itself?
regards
Christian
static void ksz9477_xmit_timestamp(struct sk_buff *skb)
{
struct sk_buff *clone = DSA_SKB_CB(skb)->clone;
struct ptp_header *ptp_hdr;
u32 tstamp_raw = 0;
u64 correction;
if (!clone)
goto out_put_tag;
/* Use cached PTP header and type from ksz9477_ptp_should_tstamp(). Note
* that KSZ9477_SKB_CB(clone)->ptp_header != NULL implies that this is a
* Pdelay_resp message.
*/
ptp_hdr = KSZ9477_SKB_CB(clone)->ptp_header;
if (!ptp_hdr)
goto out_put_tag;
correction = get_unaligned_be64(&ptp_hdr->correction);
/* For PDelay_Resp messages we will likely have a negative value in the
* correction field (see ksz9477_rcv()). The switch hardware cannot
* correctly update such values, so it must be moved to the time stamp
* field in the tail tag.
*/
if ((s64)correction < 0) {
unsigned int ptp_type = KSZ9477_SKB_CB(clone)->ptp_type;
struct timespec64 ts;
u64 ns;
/* Move ingress time stamp from PTP header's correction field to
* tail tag. Format of the correction filed is 48 bit ns + 16
* bit fractional ns. Avoid shifting negative numbers.
*/
ns = -((s64)correction) >> 16;
ts = ns_to_timespec64(ns);
tstamp_raw = ((ts.tv_sec & 3) << 30) | ts.tv_nsec;
>>> /* Set correction field to 0 (by subtracting the negative value)
>>> * and update UDP checksum.
>>> */
>>> ptp_onestep_p2p_move_t2_to_correction(skb, ptp_type, ptp_hdr, ns_to_ktime(-ns));
}
out_put_tag:
put_unaligned_be32(tstamp_raw, skb_put(skb, KSZ9477_PTP_TAG_LEN));
}
Addtionally ptp_onestep_p2p_move_t2_to_correction() must be able to handle negative values:
static inline
void ptp_onestep_p2p_move_t2_to_correction(struct sk_buff *skb,
unsigned int type,
struct ptp_header *hdr,
ktime_t t2)
{
u8 *ptr = skb_mac_header(skb);
struct udphdr *uhdr = NULL;
s64 ns = ktime_to_ns(t2);
__be64 correction_old;
s64 correction;
/* previous correction value is required for checksum update. */
memcpy(&correction_old, &hdr->correction, sizeof(correction_old));
correction = (s64)be64_to_cpu(correction_old);
/* PTP correction field consists of 32 bit nanoseconds and 16 bit
* fractional nanoseconds. Avoid shifting negative numbers.
*/
>>> if (ns >= 0)
>>> correction -= ns << 16;
>>> else
>>> correction += -ns << 16;
/* write new correction value */
put_unaligned_be64((u64)correction, &hdr->correction);
...
}
next prev parent reply other threads:[~2020-11-05 20:19 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-19 17:24 [RFC PATCH 0/9] net: dsa: microchip: PTP support for KSZ956x Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 1/9] dt-bindings: net: dsa: convert ksz bindings document to yaml Christian Eggers
2020-10-21 6:52 ` Kurt Kanzenbach
2020-10-21 8:46 ` Christian Eggers
2020-10-22 0:16 ` Vladimir Oltean
2020-10-22 0:40 ` Florian Fainelli
2020-10-22 10:54 ` Kurt Kanzenbach
2020-10-22 12:37 ` Vladimir Oltean
2020-10-22 19:17 ` Florian Fainelli
2020-10-26 13:54 ` Rob Herring
2020-10-19 17:24 ` [RFC PATCH net-next 2/9] net: dsa: microchip: split ksz_common.h Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 3/9] net: dsa: microchip: rename ksz9477.c to ksz9477_main.o Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 4/9] dt-bindings: net: dsa: microchip,ksz: add interrupt property Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 5/9] net: dsa: microchip: ksz9477: basic interrupt support Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 6/9] net: dsa: microchip: ksz9477: add Posix clock support for chip PTP clock Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 7/9] net: dsa: microchip: ksz9477: add hardware time stamping support Christian Eggers
2020-10-20 0:10 ` Vladimir Oltean
2020-10-20 8:39 ` Christian Eggers
2020-10-21 23:39 ` Vladimir Oltean
2020-10-22 2:42 ` Richard Cochran
2020-10-22 7:30 ` Christian Eggers
2020-10-22 10:17 ` Christian Eggers
2020-10-30 18:24 ` Vladimir Oltean
2020-11-01 9:35 ` Christian Eggers
2020-11-01 11:10 ` Vladimir Oltean
2020-11-01 22:14 ` Christian Eggers
2020-11-01 23:41 ` Vladimir Oltean
2020-11-01 23:55 ` Vladimir Oltean
2020-11-02 10:35 ` Christian Eggers
2020-11-02 12:28 ` Vladimir Oltean
2020-10-22 3:02 ` Richard Cochran
2020-10-22 9:01 ` Vladimir Oltean
2020-10-22 10:50 ` Vladimir Oltean
2020-10-22 11:11 ` Christian Eggers
2020-10-22 11:32 ` Vladimir Oltean
2020-10-22 14:34 ` Richard Cochran
2020-11-05 20:18 ` Christian Eggers [this message]
2020-11-10 1:42 ` Vladimir Oltean
2020-11-10 14:36 ` Christian Eggers
2020-11-10 16:40 ` Vladimir Oltean
2020-11-10 19:32 ` Vladimir Oltean
2020-11-11 21:49 ` Christian Eggers
2020-11-11 21:50 ` Christian Eggers
2020-11-12 15:28 ` Christian Eggers
2020-11-12 15:38 ` Vladimir Oltean
2020-11-17 11:27 ` Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 8/9] net: dsa: microchip: ksz9477: add Pulse Per Second (PPS) support Christian Eggers
2020-10-19 17:46 ` Vladimir Oltean
2020-10-20 8:38 ` Christian Eggers
2020-10-19 17:24 ` [RFC PATCH net-next 9/9] net: dsa: microchip: ksz9477: add periodic output support Christian Eggers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5844018.3araiXeC39@n95hx1g2 \
--to=ceggers@arri.de \
--cc=Tristram.Ha@microchip.com \
--cc=UNGLinuxDriver@microchip.com \
--cc=andrew@lunn.ch \
--cc=codrin.ciubotariu@microchip.com \
--cc=davem@davemloft.net \
--cc=devicetree@vger.kernel.org \
--cc=f.fainelli@gmail.com \
--cc=george.mccollister@gmail.com \
--cc=helmut.grohne@intenta.de \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=marex@denx.de \
--cc=netdev@vger.kernel.org \
--cc=olteanv@gmail.com \
--cc=pbarker@konsulko.com \
--cc=richardcochran@gmail.com \
--cc=robh+dt@kernel.org \
--cc=vivien.didelot@gmail.com \
--cc=woojung.huh@microchip.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).