Re: [PATCH v5 ethtool-next 1/1] add support for IEEE 802.3cg-2019 Clause 148

[Michal Kubecek <mkubecek@suse.cz>]

[-- Attachment #1: Type: text/plain, Size: 6963 bytes --]

On Thu, Feb 02, 2023 at 09:53:15AM +0100, Piergiorgio Beruto wrote:
> This patch adds support for the Physical Layer Collision Avoidance
> Reconciliation Sublayer which was introduced in the IEEE 802.3
> standard by the 802.3cg working group in 2019.
> 
> The ethtool interface has been extended as follows:
> - show if the device supports PLCA when ethtool is invoked without FLAGS
>    - additionally show what PLCA version is supported
>    - show the current PLCA status
> - add FLAGS for getting and setting the PLCA configuration
> 
> Signed-off-by: Piergiorgio Beruto <piergiorgio.beruto@gmail.com>

Sorry for the delay, I missed the patch in the mailing list and did not
get it directly into my inbox. Thankfully I noticed it in patchwork.

> ---
>  Makefile.am        |   1 +
>  ethtool.8.in       |  83 ++++++++++++-
>  ethtool.c          |  21 ++++
>  netlink/extapi.h   |   6 +
>  netlink/plca.c     | 296 +++++++++++++++++++++++++++++++++++++++++++++
>  netlink/settings.c |  82 ++++++++++++-
>  6 files changed, 486 insertions(+), 3 deletions(-)
>  create mode 100644 netlink/plca.c
> 
> diff --git a/Makefile.am b/Makefile.am
> index 999e7691e81c..cbc1f4f5fdf2 100644
> --- a/Makefile.am
> +++ b/Makefile.am
> @@ -42,6 +42,7 @@ ethtool_SOURCES += \
>  		  netlink/desc-ethtool.c netlink/desc-genlctrl.c \
>  		  netlink/module-eeprom.c netlink/module.c netlink/rss.c \
>  		  netlink/desc-rtnl.c netlink/cable_test.c netlink/tunnels.c \
> +		  netlink/plca.c \
>  		  uapi/linux/ethtool_netlink.h \
>  		  uapi/linux/netlink.h uapi/linux/genetlink.h \
>  		  uapi/linux/rtnetlink.h uapi/linux/if_link.h \
> diff --git a/ethtool.8.in b/ethtool.8.in
> index eaf6c55a84bf..c43c6d8b5263 100644
> --- a/ethtool.8.in
> +++ b/ethtool.8.in
[...]
> +.TP
> +.BI node\-cnt \ N
> +The node-cnt [1 .. 255] should be set after the maximum number of nodes that
> +can be plugged to the multi-drop network. This parameter regulates the minimum
> +length of the PLCA cycle. Therefore, it is only meaningful for the coordinator
> +node (\fBnod-id\fR = 0). Setting this parameter on a follower node has no
> +effect. The \fBnode\-cnt\fR parameter maps to IEEE 802.3cg-2019 clause 
> +30.16.1.1.3 (aPLCANodeCount).
> +.TP
> +.BI to\-tmr \ N
> +The TO timer parameter sets the value of the transmit opportunity timer in 
> +bit-times, and shall be set equal across all the nodes sharing the same
> +medium for PLCA to work. The default value of 32 is enough to cover a link of
> +roughly 50 mt. This parameter maps to  IEEE 802.3cg-2019 clause 30.16.1.1.5
> +(aPLCATransmitOpportunityTimer).
> +.TP
> +.BI burst\-cnt \ N
> +The \fBburst\-cnt\fR parameter [0 .. 255] indicates the extra number of packets
> +that the node is allowed to send during a single transmit opportunity.
> +By default, this attribute is 0, meaning that the node can send a sigle frame 
> +per TO. When greater than 0, the PLCA RS keeps the TO after any transmission,
> +waiting for the MAC to send a new frame for up to \fBburst\-tmr\fR BTs. This can
> +only happen a number of times per PLCA cycle up to the value of this parameter.
> +After that, the burst is over and the normal counting of TOs resumes.
> +This parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.6 (aPLCAMaxBurstCount).
> +.TP
> +.BI burst\-tmr \ N
> +The \fBburst\-tmr\fR parameter [0 .. 255] sets how many bit-times the PLCA RS
> +waits for the MAC to initiate a new transmission when \fBburst\-cnt\fR is 
> +greater than 0. If the MAC fails to send a new frame within this time, the burst
> +ends and the counting of TOs resumes. Otherwise, the new frame is sent as part
> +of the current burst. This parameter maps to IEEE 802.3cg-2019 clause
> +30.16.1.1.7 (aPLCABurstTimer). The value of \fBburst\-tmr\fR should be set
> +greater than the Inter-Frame-Gap (IFG) time of the MAC (plus some margin)
> +for PLCA burst mode to work as intended.

There are some trailing spaces in these paragraphs.

> +.RE
> +.TP
> +.B \-\-get\-plca\-status
> +Show the current PLCA status for the given interface. If \fBon\fR, the PHY is
> +successfully receiving or generating the BEACON signal. If \fBoff\fR, the PLCA
> +function is temporarily disabled and the PHY is operating in plain CSMA/CD mode.
>  .SH BUGS
>  Not supported (in part or whole) on all network drivers.
>  .SH AUTHOR
> @@ -1532,7 +1610,8 @@ Alexander Duyck,
>  Sucheta Chakraborty,
>  Jesse Brandeburg,
>  Ben Hutchings,
> -Scott Branden.
> +Scott Branden,
> +Piergiorgio Beruto.
>  .SH AVAILABILITY
>  .B ethtool
>  is available from

Do you have a strong reason to add your name here? In general, I rather
see lists like these as a historical relic. In this case, the list has
not been updated since 2017 and even before that, I'm pretty sure it is
only a small fraction of contributors. IMHO the git log serves the
purpose much better today.

[...]
> diff --git a/netlink/settings.c b/netlink/settings.c
> index 1107082167d4..a349e270dff9 100644
> --- a/netlink/settings.c
> +++ b/netlink/settings.c
[...]
> @@ -923,7 +987,10 @@ int nl_gset(struct cmd_context *ctx)
>  	    netlink_cmd_check(ctx, ETHTOOL_MSG_LINKINFO_GET, true) ||
>  	    netlink_cmd_check(ctx, ETHTOOL_MSG_WOL_GET, true) ||
>  	    netlink_cmd_check(ctx, ETHTOOL_MSG_DEBUG_GET, true) ||
> -	    netlink_cmd_check(ctx, ETHTOOL_MSG_LINKSTATE_GET, true))
> +	    netlink_cmd_check(ctx, ETHTOOL_MSG_LINKSTATE_GET, true) ||
> +	    netlink_cmd_check(ctx, ETHTOOL_MSG_LINKSTATE_GET, true) ||

You accidentally duplicated the line here.

> +	    netlink_cmd_check(ctx, ETHTOOL_MSG_PLCA_GET_CFG, true) ||
> +	    netlink_cmd_check(ctx, ETHTOOL_MSG_PLCA_GET_STATUS, true))
>  		return -EOPNOTSUPP;
>  
>  	nlctx->suppress_nlerr = 1;
> @@ -943,6 +1010,12 @@ int nl_gset(struct cmd_context *ctx)
>  	if (ret == -ENODEV)
>  		return ret;
>  
> +	ret = gset_request(nlctx, ETHTOOL_MSG_PLCA_GET_CFG,
> +			   ETHTOOL_A_PLCA_HEADER, plca_cfg_reply_cb);
> +
> +	if (ret == -ENODEV)
> +		return ret;
> +
>  	ret = gset_request(nlctx, ETHTOOL_MSG_DEBUG_GET, ETHTOOL_A_DEBUG_HEADER,
>  			   debug_reply_cb);
>  	if (ret == -ENODEV)
> @@ -950,6 +1023,13 @@ int nl_gset(struct cmd_context *ctx)
>  
>  	ret = gset_request(nlctx, ETHTOOL_MSG_LINKSTATE_GET,
>  			   ETHTOOL_A_LINKSTATE_HEADER, linkstate_reply_cb);
> +
> +	if (ret == -ENODEV)
> +		return ret;
> +
> +
> +	ret = gset_request(nlctx, ETHTOOL_MSG_PLCA_GET_STATUS,
> +			   ETHTOOL_A_PLCA_HEADER, plca_status_reply_cb);
>  	if (ret == -ENODEV)
>  		return ret;

Please make the whitespace consistent with existing code, i.e. no empty
line between gset_request() call and the test of ret and no double empty
line.

As I have no actual objections, I can adjust the whitespace issues here
and in ethtool.8.in if you would prefer to avoid v6 before moving on to
the MAC merge series.

Michal

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]