From: "Arınç ÜNAL" <arinc.unal@arinc9.com>
To: Vladimir Oltean <olteanv@gmail.com>
Cc: Sean Wang <sean.wang@mediatek.com>,
Landen Chao <Landen.Chao@mediatek.com>,
DENG Qingfang <dqfext@gmail.com>,
Daniel Golle <daniel@makrotopia.org>,
Andrew Lunn <andrew@lunn.ch>,
Florian Fainelli <f.fainelli@gmail.com>,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Matthias Brugger <matthias.bgg@gmail.com>,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@collabora.com>,
Russell King <linux@armlinux.org.uk>,
Richard van Schagen <richard@routerhints.com>,
Richard van Schagen <vschagen@cs.com>,
Frank Wunderlich <frank-w@public-files.de>,
Bartel Eerdekens <bartel.eerdekens@constell8.be>,
erkin.bozoglu@xeront.com, mithat.guner@xeront.com,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linux-mediatek@lists.infradead.org
Subject: Re: [PATCH net-next 29/30] net: dsa: introduce preferred_default_local_cpu_port and use on MT7530
Date: Sun, 4 Jun 2023 13:02:20 +0300 [thread overview]
Message-ID: <886ae203-1aca-0cb3-cf32-416984a7c37c@arinc9.com> (raw)
In-Reply-To: <20230526171755.nk643aphoojvhjpg@skbuf>
On 26.05.2023 20:17, Vladimir Oltean wrote:
> On Mon, May 22, 2023 at 03:15:31PM +0300, arinc9.unal@gmail.com wrote:
>> From: Vladimir Oltean <olteanv@gmail.com>
>>
>> When multiple CPU ports are being used, the numerically smallest CPU port
>> becomes the port all user ports become affine to. This may not be the best
>> choice for all switches as there may be a numerically greater CPU port with
>> more bandwidth than the numerically smallest one.
>>
>> Such switches are MT7530 and MT7531BE, which the MT7530 DSA subdriver
>> controls. Port 5 of these switches has got RGMII whilst port 6 has got
>> either TRGMII or SGMII.
>>
>> Therefore, introduce the preferred_default_local_cpu_port operation to the
>> DSA subsystem and use it on the MT7530 DSA subdriver to prefer port 6 as
>> the default CPU port.
>>
>> To prove the benefit of this operation, I (Arınç) have done a bidirectional
>> speed test between two DSA user ports on the MT7531BE switch using iperf3.
>> The user ports are 1 Gbps full duplex and on different networks so the SoC
>> MAC would have to do 2 Gbps TX and 2 Gbps RX to deliver full speed.
>
> I think the real argument would sound like this:
>
> Since the introduction of the OF bindings, DSA has always had a policy
> that in case multiple CPU ports are present in the device tree, the
> numerically first one is always chosen.
>
> The MT7530 switch family has 2 CPU ports, 5 and 6, where port 6 is
> preferable because it has higher bandwidth.
>
> The MT7530 driver developers had 3 options:
> - to modify DSA when the driver was introduced, such as to prefer the
> better port
> - to declare both CPU ports in device trees as CPU ports, and live with
> the sub-optimal performance resulting from not preferring the better
> port
> - to declare just port 6 in the device tree as a CPU port
>
> Of course they chose the path of least resistance (3rd option), kicking
> the can down the road. The hardware description in the device tree is
> supposed to be stable - developers are not supposed to adopt the
> strategy of piecemeal hardware description, where the device tree is
> updated in lockstep with the features that the kernel currently supports.
>
> Now, as a result of the fact that they did that, any attempts to modify
> the device tree and describe both CPU ports as CPU ports would make DSA
> change its default selection from port 6 to 5, effectively resulting in
> a performance degradation visible to users as can be seen below vvvvv
>
>>
>> Without preferring port 6:
>>
>> [ ID][Role] Interval Transfer Bitrate Retr
>> [ 5][TX-C] 0.00-20.00 sec 374 MBytes 157 Mbits/sec 734 sender
>> [ 5][TX-C] 0.00-20.00 sec 373 MBytes 156 Mbits/sec receiver
>> [ 7][RX-C] 0.00-20.00 sec 1.81 GBytes 778 Mbits/sec 0 sender
>> [ 7][RX-C] 0.00-20.00 sec 1.81 GBytes 777 Mbits/sec receiver
>>
>> With preferring port 6:
>>
>> [ ID][Role] Interval Transfer Bitrate Retr
>> [ 5][TX-C] 0.00-20.00 sec 1.99 GBytes 856 Mbits/sec 273 sender
>> [ 5][TX-C] 0.00-20.00 sec 1.99 GBytes 855 Mbits/sec receiver
>> [ 7][RX-C] 0.00-20.00 sec 1.72 GBytes 737 Mbits/sec 15 sender
>> [ 7][RX-C] 0.00-20.00 sec 1.71 GBytes 736 Mbits/sec receiver
>>
>> Using one port for WAN and the other ports for LAN is a very popular use
>> case which is what this test emulates.
>
> As such, this change proposes that we retroactively modify stable
> kernels to keep the mt7530 driver preferring port 6 even with device
> trees where the hardware is more fully described.
>
> Fixes: b8f126a8d543 ("net-next: dsa: add dsa support for Mediatek MT7530 switch")
>
>>
>> This doesn't affect the remaining switches, MT7531AE and the switch on the
>> MT7988 SoC. Both CPU ports of the MT7531AE switch have got SGMII and there
>> is only one CPU port on the switch on the MT7988 SoC.
>>
>> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
>> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
>> ---
>
> See the difference in intent?
Yeah, nicely put.
Arınç
next prev parent reply other threads:[~2023-06-04 10:03 UTC|newest]
Thread overview: 121+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-22 12:15 [PATCH net-next 00/30] net: dsa: mt7530: improve, trap BPDU & LLDP, and prefer CPU port arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 01/30] net: dsa: mt7530: add missing @p5_interface to mt7530_priv description arinc9.unal
2023-05-23 23:29 ` Andrew Lunn
2023-05-22 12:15 ` [PATCH net-next 02/30] net: dsa: mt7530: use p5_interface_select as data type for p5_intf_sel arinc9.unal
2023-05-23 23:31 ` Andrew Lunn
2023-05-24 7:41 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 03/30] net: dsa: mt7530: properly support MT7531AE and MT7531BE arinc9.unal
2023-05-24 14:48 ` Vladimir Oltean
2023-05-25 6:00 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 04/30] net: dsa: mt7530: improve comments regarding port 5 and 6 arinc9.unal
2023-05-23 23:35 ` Andrew Lunn
2023-05-24 14:49 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 05/30] net: dsa: mt7530: read XTAL value from correct register arinc9.unal
2023-05-23 23:37 ` Andrew Lunn
2023-05-24 16:57 ` Vladimir Oltean
2023-05-25 6:20 ` Arınç ÜNAL
2023-05-25 13:31 ` Vladimir Oltean
2023-06-04 6:34 ` Arınç ÜNAL
2023-06-04 7:11 ` Vladimir Oltean
2023-06-04 16:22 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 06/30] net: dsa: mt7530: improve code path for setting up port 5 arinc9.unal
2023-05-24 17:35 ` Vladimir Oltean
2023-05-25 6:42 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 07/30] net: dsa: mt7530: do not run mt7530_setup_port5() if port 5 is disabled arinc9.unal
2023-05-24 17:45 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 08/30] net: dsa: mt7530: change p{5,6}_interface to p{5,6}_configured arinc9.unal
2023-05-24 17:51 ` Vladimir Oltean
2023-05-25 6:49 ` Arınç ÜNAL
2023-05-26 13:01 ` Vladimir Oltean
2023-06-03 12:15 ` Arınç ÜNAL
2023-06-03 12:26 ` Russell King (Oracle)
2023-06-04 10:46 ` Arınç ÜNAL
2023-06-04 12:18 ` Russell King (Oracle)
2023-06-04 12:55 ` Vladimir Oltean
2023-06-04 13:07 ` Russell King (Oracle)
2023-06-04 13:14 ` Arınç ÜNAL
2023-06-04 13:25 ` Vladimir Oltean
2023-06-04 15:13 ` Russell King (Oracle)
2023-06-04 16:00 ` Russell King (Oracle)
2023-06-04 16:06 ` Russell King (Oracle)
2023-06-04 16:14 ` Arınç ÜNAL
2023-06-10 10:57 ` Arınç ÜNAL
2023-06-10 17:55 ` Vladimir Oltean
2023-06-11 7:23 ` Arınç ÜNAL
2024-01-10 11:15 ` Arınç ÜNAL
2024-01-10 14:27 ` Vladimir Oltean
2024-01-10 17:15 ` Arınç ÜNAL
2024-01-10 18:05 ` Vladimir Oltean
2024-01-10 18:31 ` Russell King (Oracle)
2024-01-11 9:19 ` Vladimir Oltean
2023-06-05 14:36 ` Arınç ÜNAL
2023-06-03 12:27 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 09/30] net: dsa: mt7530: empty default case on mt7530_setup_port5() arinc9.unal
2023-05-24 18:04 ` Vladimir Oltean
2023-05-24 18:05 ` Vladimir Oltean
2023-05-25 6:51 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 10/30] net: dsa: mt7530: call port 6 setup from mt7530_mac_config() arinc9.unal
2023-05-24 18:08 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 11/30] net: dsa: mt7530: remove pad_setup function pointer arinc9.unal
2023-05-24 18:09 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 12/30] net: dsa: mt7530: move XTAL check to mt7530_setup() arinc9.unal
2023-05-23 23:40 ` Andrew Lunn
2023-05-24 18:15 ` Vladimir Oltean
2023-05-25 7:04 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 13/30] net: dsa: mt7530: move enabling port 6 to mt7530_setup_port6() arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 14/30] net: dsa: mt7530: switch to if/else statements on mt7530_setup_port6() arinc9.unal
2023-05-26 13:08 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 15/30] net: dsa: mt7530: set TRGMII RD TAP if trgmii is being used arinc9.unal
2023-05-26 13:09 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 16/30] net: dsa: mt7530: move lowering port 5 RGMII driving to mt7530_setup() arinc9.unal
2023-05-26 13:17 ` Vladimir Oltean
2023-06-04 7:05 ` Arınç ÜNAL
2023-06-04 7:14 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 17/30] net: dsa: mt7530: fix port capabilities for MT7988 arinc9.unal
2023-05-26 13:20 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 18/30] net: dsa: mt7530: remove .mac_port_config for MT7988 and make it optional arinc9.unal
2023-05-26 13:21 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 19/30] net: dsa: mt7530: set interrupt register only for MT7530 arinc9.unal
2023-05-26 13:25 ` Vladimir Oltean
2023-06-04 7:18 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 20/30] net: dsa: mt7530: properly reset MT7531 switch arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 21/30] net: dsa: mt7530: get rid of useless error returns on phylink code path arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 22/30] net: dsa: mt7530: rename p5_intf_sel and use only for MT7530 switch arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 23/30] net: dsa: mt7530: run mt7530_pll_setup() only with 40 MHz XTAL arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 24/30] net: dsa: mt7530: rename MT7530_MFC to MT753X_MFC arinc9.unal
2023-05-26 15:42 ` Vladimir Oltean
2023-05-26 15:50 ` Russell King (Oracle)
2023-06-04 8:06 ` Arınç ÜNAL
2023-06-04 13:17 ` Vladimir Oltean
2023-06-04 13:25 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 25/30] net: dsa: mt7530: properly set MT7531_CPU_PMAP arinc9.unal
2023-05-26 15:51 ` Vladimir Oltean
2023-06-04 8:21 ` Arınç ÜNAL
2023-06-04 13:08 ` Vladimir Oltean
2023-06-04 13:33 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 26/30] net: dsa: mt7530: properly set MT7530_CPU_PORT arinc9.unal
2023-05-26 16:55 ` Vladimir Oltean
2023-06-04 8:33 ` Arınç ÜNAL
2023-05-22 12:15 ` [PATCH net-next 27/30] net: dsa: mt7530: introduce BPDU trapping for MT7530 switch arinc9.unal
2023-05-26 17:02 ` Vladimir Oltean
2023-06-04 8:51 ` Arınç ÜNAL
2023-06-04 9:23 ` Vladimir Oltean
2023-06-04 9:39 ` Arınç ÜNAL
2023-06-04 12:47 ` Vladimir Oltean
2023-06-10 8:32 ` Arınç ÜNAL
2023-06-10 17:57 ` Vladimir Oltean
2023-05-22 12:15 ` [PATCH net-next 28/30] net: dsa: mt7530: introduce LLDP frame trapping arinc9.unal
2023-05-22 12:15 ` [PATCH net-next 29/30] net: dsa: introduce preferred_default_local_cpu_port and use on MT7530 arinc9.unal
2023-05-26 17:17 ` Vladimir Oltean
2023-06-04 10:02 ` Arınç ÜNAL [this message]
2023-05-22 12:15 ` [PATCH net-next 30/30] MAINTAINERS: add me as maintainer of MEDIATEK SWITCH DRIVER arinc9.unal
2023-05-24 14:37 ` Vladimir Oltean
2023-05-26 17:38 ` Vladimir Oltean
2023-05-22 12:25 ` [PATCH net-next 00/30] net: dsa: mt7530: improve, trap BPDU & LLDP, and prefer CPU port Andrew Lunn
2023-05-22 13:37 ` Arınç ÜNAL
2023-05-22 15:43 ` Vladimir Oltean
2023-05-22 14:09 ` Horatiu Vultur
2023-05-22 14:35 ` Arınç ÜNAL
2023-05-22 18:13 ` Russell King (Oracle)
2023-05-23 1:54 ` Jakub Kicinski
2023-05-26 17:47 ` Vladimir Oltean
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=886ae203-1aca-0cb3-cf32-416984a7c37c@arinc9.com \
--to=arinc.unal@arinc9.com \
--cc=Landen.Chao@mediatek.com \
--cc=andrew@lunn.ch \
--cc=angelogioacchino.delregno@collabora.com \
--cc=bartel.eerdekens@constell8.be \
--cc=daniel@makrotopia.org \
--cc=davem@davemloft.net \
--cc=dqfext@gmail.com \
--cc=edumazet@google.com \
--cc=erkin.bozoglu@xeront.com \
--cc=f.fainelli@gmail.com \
--cc=frank-w@public-files.de \
--cc=kuba@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux@armlinux.org.uk \
--cc=matthias.bgg@gmail.com \
--cc=mithat.guner@xeront.com \
--cc=netdev@vger.kernel.org \
--cc=olteanv@gmail.com \
--cc=pabeni@redhat.com \
--cc=richard@routerhints.com \
--cc=sean.wang@mediatek.com \
--cc=vschagen@cs.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
all inboxes | Powered by JetHome®