netdev - Re: [PATCH v6 net-next 06/13] dt-bindings: net: ti: add new cpsw switch driver bindings

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 12 Nov 2019 08:22:30 -0800
From:   Tony Lindgren <tony@...mide.com>
To:     Grygorii Strashko <grygorii.strashko@...com>
Cc:     Florian Fainelli <f.fainelli@...il.com>, netdev@...r.kernel.org,
        Ilias Apalodimas <ilias.apalodimas@...aro.org>,
        Andrew Lunn <andrew@...n.ch>,
        "David S . Miller" <davem@...emloft.net>,
        Ivan Khoronzhuk <ivan.khoronzhuk@...aro.org>,
        Jiri Pirko <jiri@...nulli.us>, Sekhar Nori <nsekhar@...com>,
        linux-kernel@...r.kernel.org, linux-omap@...r.kernel.org,
        Murali Karicheri <m-karicheri2@...com>,
        Ivan Vecera <ivecera@...hat.com>,
        Rob Herring <robh+dt@...nel.org>, devicetree@...r.kernel.org
Subject: Re: [PATCH v6 net-next 06/13] dt-bindings: net: ti: add new cpsw
 switch driver bindings

Hi,

* Grygorii Strashko <grygorii.strashko@...com> [191112 09:54]:
> No, sorry I do not agree. The MDIO is inseparable part of CPSW and it's enabled when CPSW is enabled
> (on interconnect level), more over I want to get rid of platform device in MDIO for most of the cases
> as it only introduces boot/probing complexity.

Well the fact that mdio is enabled at the interconnect level is why
I think the cpsw child modules are independent components :)

So I did the following quick test on pocketbeagle with Linux next,
it has no Ethernet wired up, and by default we have ethernet@0
set to status = "disabled".

Manually enable the target module at 0x4a100000:
# echo on > /sys/devices/platform/ocp/4a000000.interconnect/\
4a000000.interconnect:segment@...a101200.target-module/power/control

Dump out mdio registers at offset 0x1000:
# rwmem 0x4a101000+0x100
0x4a101000 = 0x40070106
0x4a101004 = 0x810000ff
0x4a101008 = 0000000000
0x4a10100c = 0000000000
0x4a101010 = 0000000000
0x4a101014 = 0000000000
0x4a101018 = 0000000000
...

So yup, it seems quite independent of the other child devices
on the same interconnect target mdoule. I'm quessing it's the
same story for other modules like cppi_dma and so on, this
should be easy to check.

Hmm and isn't the some version of mdio also used stuffed into
davinci_emac and netcp too?

Anyways, up to you. But my experience is that having separate
driver modules is the way to go than trying to treat any TI
"subsystem" as a single device. This is because the child modules
tend to get updated and changed and moved around over time.

Regards,

Tony