netdev - Re: [PATCH net-next 00/11] mlxsw: extend line card model by devices and info

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]

Message-ID: <YmfoXsw+o9LE9dF3@nanopsycho>
Date:   Tue, 26 Apr 2022 14:41:02 +0200
From:   Jiri Pirko <jiri@...nulli.us>
To:     Andrew Lunn <andrew@...n.ch>
Cc:     Ido Schimmel <idosch@...sch.org>, Jakub Kicinski <kuba@...nel.org>,
        Ido Schimmel <idosch@...dia.com>, netdev@...r.kernel.org,
        davem@...emloft.net, pabeni@...hat.com, jiri@...dia.com,
        petrm@...dia.com, dsahern@...il.com, mlxsw@...dia.com
Subject: Re: [PATCH net-next 00/11] mlxsw: extend line card model by devices
 and info

Tue, Apr 26, 2022 at 02:27:54PM CEST, andrew@...n.ch wrote:
>> It is not a separate devlink device, not removetely. A devlink device is
>> attached to some bus on the host, it contains entities like netdevices,
>> etc.
>> 
>> Line card devices, on contrary, are accessible over ASIC FW interface,
>> they reside on line cards. ASIC FW is using build-in SDK to communicate
>> with them. There is really nothing to expose, except for the face they
>> are there, with some FW version and later on (follow-up patchset) to be
>> able to flash FW on them.
>
>But isn't this just an implementation detail?
>
>Say the flash was directly accessible to the host? It is just another
>mtd devices? The gearbox is just another bunch of MMIO registers. You

Not really, the ASIC FW is also not mtd device. ASIC FW exposes set of
registers to work with the ASIC FW flash. The linecard gearbox
implements the same set of registers, tunneled over MDDT register.

Also, the gearbox is not accessable from the host. The ASIC FW is the
only one able to access it.


>can access the SFP socket via a host i2c bus, etc. More of a SoC like
>implementation, which the enterprise routers are like.
>
>This is a completely different set of implementation details, but i
>still have the same basic building blocks. Should it look the same,
>and the implementation details are hidden, or do you want to expose
>your implementation details?

Well, I got your point. If the HW would be designed in the way the
building blocks are exposed to the host, that would work. However, that
is not the case here, unfortunatelly.


>
>	Andrew