From: Vladimir Oltean <vladimir.oltean@nxp.com>
To: "Y.b. Lu" <yangbo.lu@nxp.com>
Cc: Tobias Waldekranz <tobias@waldekranz.com>,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
Richard Cochran <richardcochran@gmail.com>,
"David S . Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>,
Jonathan Corbet <corbet@lwn.net>,
Kurt Kanzenbach <kurt@linutronix.de>,
Andrew Lunn <andrew@lunn.ch>,
Vivien Didelot <vivien.didelot@gmail.com>,
Florian Fainelli <f.fainelli@gmail.com>,
Claudiu Manoil <claudiu.manoil@nxp.com>,
Alexandre Belloni <alexandre.belloni@bootlin.com>,
"UNGLinuxDriver@microchip.com" <UNGLinuxDriver@microchip.com>,
"linux-doc@vger.kernel.org" <linux-doc@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [net-next, v2, 3/7] net: dsa: free skb->cb usage in core driver
Date: Fri, 7 May 2021 11:48:37 +0000 [thread overview]
Message-ID: <20210507114837.aybpln6twhstelqd@skbuf> (raw)
In-Reply-To: <DB7PR04MB5017DD1C470A06A66ADBB5DEF8579@DB7PR04MB5017.eurprd04.prod.outlook.com>
On Fri, May 07, 2021 at 11:26:32AM +0000, Y.b. Lu wrote:
> > From: Tobias Waldekranz <tobias@waldekranz.com>
> > On Mon, Apr 26, 2021 at 17:37, Yangbo Lu <yangbo.lu@nxp.com> wrote:
> > > Free skb->cb usage in core driver and let device drivers decide to use
> > > or not. The reason having a DSA_SKB_CB(skb)->clone was because
> > > dsa_skb_tx_timestamp() which may set the clone pointer was called
> > > before p->xmit() which would use the clone if any, and the device
> > > driver has no way to initialize the clone pointer.
> > >
> > > Although for now putting memset(skb->cb, 0, 48) at beginning of
> > > dsa_slave_xmit() by this patch is not very good, there is still way to
> > > improve this. Otherwise, some other new features, like one-step
> >
> > Could you please expand on this improvement?
> >
> > This memset makes it impossible to carry control buffer information from
> > driver callbacks that run before .ndo_start_xmit, for
> > example .ndo_select_queue, to a tagger's .xmit.
> >
> > It seems to me that if the drivers are to handle the CB internally from now on,
> > that should go for both setting and clearing of the required fields.
>
> For the timestamping case, dsa_skb_tx_timestamp may not touch
> .port_txtstamp callback, so we had to put skb->cb initialization at
> beginning of dsa_slave_xmit.
> To avoid breaking future skb->cb usage you mentioned, do you think we
> can back to Vladimir's idea initializing only field required, or even
> just add a callback for cb initialization for timestamping?
I would like to avoid overengineering, which a callback for skb->cb
initialization would introduce, given the needs we have now.
FWIW, we may not even be able to touch skb->cb in .ndo_select_queue for
Tobias's use case, that discussion is here:
https://patchwork.kernel.org/project/netdevbpf/patch/20210426170411.1789186-7-tobias@waldekranz.com/
next prev parent reply other threads:[~2021-05-07 11:48 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-26 9:37 [net-next, v2, 0/7] Support Ocelot PTP Sync one-step timestamping Yangbo Lu
2021-04-26 9:37 ` [net-next, v2, 1/7] net: dsa: check tx timestamp request in core driver Yangbo Lu
2021-04-26 18:36 ` Florian Fainelli
2021-04-26 9:37 ` [net-next, v2, 2/7] net: dsa: no longer identify PTP packet " Yangbo Lu
2021-04-26 18:39 ` Florian Fainelli
2021-04-26 9:37 ` [net-next, v2, 3/7] net: dsa: free skb->cb usage " Yangbo Lu
2021-04-26 13:38 ` Richard Cochran
2021-04-26 18:39 ` Vladimir Oltean
2021-04-27 4:25 ` Y.b. Lu
2021-04-27 15:56 ` Richard Cochran
2021-04-27 4:26 ` Y.b. Lu
2021-04-26 18:16 ` Vladimir Oltean
2021-04-27 4:13 ` Y.b. Lu
2021-04-28 20:29 ` Tobias Waldekranz
2021-05-07 11:26 ` Y.b. Lu
2021-05-07 11:48 ` Vladimir Oltean [this message]
2021-04-26 9:37 ` [net-next, v2, 4/7] net: dsa: no longer clone skb " Yangbo Lu
2021-04-26 9:38 ` [net-next, v2, 5/7] docs: networking: timestamping: update for DSA switches Yangbo Lu
2021-04-26 9:38 ` [net-next, v2, 6/7] net: mscc: ocelot: convert to ocelot_port_txtstamp_request() Yangbo Lu
2021-04-26 9:38 ` [net-next, v2, 7/7] net: mscc: ocelot: support PTP Sync one-step timestamping Yangbo Lu
2021-04-26 13:47 ` [net-next, v2, 0/7] Support Ocelot " Richard Cochran
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=20210507114837.aybpln6twhstelqd@skbuf \
--to=vladimir.oltean@nxp.com \
--cc=UNGLinuxDriver@microchip.com \
--cc=alexandre.belloni@bootlin.com \
--cc=andrew@lunn.ch \
--cc=claudiu.manoil@nxp.com \
--cc=corbet@lwn.net \
--cc=davem@davemloft.net \
--cc=f.fainelli@gmail.com \
--cc=kuba@kernel.org \
--cc=kurt@linutronix.de \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=richardcochran@gmail.com \
--cc=tobias@waldekranz.com \
--cc=vivien.didelot@gmail.com \
--cc=yangbo.lu@nxp.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).