Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 10 Jun 2010 16:02:23 -0400
From:      Alexander Sack <pisymbol@gmail.com>
To:        Juli Mallett <jmallett@freebsd.org>
Cc:        freebsd-net@freebsd.org, Jack Vogel <jfvogel@gmail.com>
Subject:   Re: Dual-rate transceivers with ixgbe?
Message-ID:  <AANLkTil3ojaY_WsG4mgwlBGFHEyyEERBJ7eI6mZzQipw@mail.gmail.com>
In-Reply-To: <AANLkTinOnzvjrmmLb3w5jOYEk02CY2tWCwnKxTv8sHS8@mail.gmail.com>
References:  <AANLkTinO9NZ8F9TeS68I2ULQgdlMGzlXkinCsywWosAM@mail.gmail.com> <AANLkTinS607kd3wc3F2WWmA6Zk9KL4GhscxEHPtcvxA5@mail.gmail.com> <AANLkTimkxOn9h6SAkTPDqfUM9kl2CZiFrZC_BuNDfRyB@mail.gmail.com> <AANLkTikcQMXk8UebmaynOeeInGiwx8yr0NMGE1yJfm8u@mail.gmail.com> <AANLkTil_YRvU54qHtIMO7mP4yYjojeHVrCHaRcl2K2Ug@mail.gmail.com> <AANLkTim5Ao9nSh6T6HF7NztLgvbTzxuVyr8lSXAJ7bMo@mail.gmail.com> <AANLkTim9-Za5mzLTw7MDAHY_TuIQsQ0SF0_1xpxyGY7v@mail.gmail.com> <AANLkTik-V2frmirwBLtg4RemdEVvPhUmVsOP7CqEkvUi@mail.gmail.com> <AANLkTikqjQic_M3mX7OTx-V0OJxbk4vzxajPmHmIUAKa@mail.gmail.com> <AANLkTikjs42mE5QHnSvZ9x9DI1xfYowvIES-DRORz6hH@mail.gmail.com> <AANLkTilqb4y0zEmom0jyg-HEi4yp8D3nkkbUWGHUjWPt@mail.gmail.com> <AANLkTinOnzvjrmmLb3w5jOYEk02CY2tWCwnKxTv8sHS8@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jun 10, 2010 at 3:59 PM, Alexander Sack <pisymbol@gmail.com> wrote:
> On Thu, Jun 10, 2010 at 3:20 PM, Juli Mallett <jmallett@freebsd.org> wrot=
e:
>> On Thu, Jun 10, 2010 at 12:04, Alexander Sack <pisymbol@gmail.com> wrote=
:
>>> However, would it be possible to please make this a kenv tunable in
>>> the driver? =A0Its kinda stupid I have to recompile to add a SFP. =A0It
>>> can certainly be an unsupported feature by you.
>>
>> This is a good idea, but asking Jack to commit something that goes
>> against Intel's wishes and which he's not going to support is probably
>> a losing strategy. =A0Whether he minds someone else committing might be
>> another story :)
>
> Sorry Jack! =A0:) =A0I am not trying to cause you grief but....
>
> It would SEEM BENEFICIAL to the community to allow unsupported SFPs
> (again use at your own risk, don't ask Jack why doesn't it work and
> expect a lot of attention). =A0That's all I am saying. =A0I don't know th=
e
> history here, so if this topic has been heavily debated on the lists,
> please forgive.
>
> I am also not trying to upset Jack or Intel or anyone here. =A0Good
> intentions an all that!?
>
>> One thing that the base driver probably ought to do is not fail in
>> attach if there's an unrecognized SFP+ module. =A0Since we get
>> interrupts on module change (although this doesn't seem to always work
>> *entirely* right in the stock sources, mostly wrt stored values of
>> AUTOC and the like) it should be possible to bring the interface up
>> with the unsupported (and disabled) SFP+ module and do the SFP+ module
>> probing we already do on hot-swap.
>
> Alright, let me see if I can test that. =A0Let me rephrase so I validate
> what you are saying:
>
> The driver can come up with an unsupported module but disable the
> interface (ifconfig shows the interface, etc.).
>
> If you then hot-swap a supported SFP, it should come up then with a
> ifconfig down/up cycle. =A0Right?
>
> As it stand now, if you load the driver with an unsupported module, it
> will not attach at all causing you to reload the entire driver OR
> reboot the box to have it reattach to the other SFP.
>
> Do I understand you correctly?
>
>>> Patch attached. =A0Tested with CURRENT driver on a 7.2-amd64-release
>>> machine. =A0If you set the tunable to 1, ixgbe loads without issue. =A0=
If
>>> you leave it to zero (default), it will not attach to unsupported
>>> SFPs.
>>
>> If you're going to do this, you ought to support SFP as well as SFP+
>> modules which requires a few more changes. =A0But not very many :)
>
> LOL....right....let me adjust...let me look at the SFP path.
>
>> Although some SFP modules seem to return something for the
>> comp_codes_10g read, which makes deciding which SFI mode to configure
>> non-trivial, short of providing an actual vendor/model switch.
>
> I want to avoid this completely. =A0There is already a vendor_oui switch
> which I now understand are the specific SFPs Intel tests and certifies
> with! =A0(right?)
>
> I think the policy should be (ideally):
>
> - Attach to all supported tested Intel stamp-of-approval SFPs (normal
> driver path)
> - Disable any unsupported SFPs but still complete driver attachment
> - Tunable to control attaching to an unsupported SFP (YOU AT YOUR OWN RIS=
K)
>
> Is that right?
>

Sorry for the mixed conversation in that last email....I think you
guys can figure out who is who!  :-)!

-aps



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTil3ojaY_WsG4mgwlBGFHEyyEERBJ7eI6mZzQipw>