Re: [PATCH for net-next] if_ether.h: Clarify ethertype validity for gsw1xx dsa

Re: [PATCH for net-next] if_ether.h: Clarify ethertype validity for gsw1xx dsa

[Peter Enderborg]

(resend due to html bounce)
On 11/15/25 21:41, Andrew Lunn wrote:
> On Fri, Nov 14, 2025 at 02:59:36PM +0100, Peter Enderborg wrote:
>> Ref https://standards-oui.ieee.org/ethertype/eth.txt
>>
>>
> Is this actually registered with IANA?

No.

> https://www.iana.org/assignments/ieee-802-numbers/ieee-802-numbers.xhtml
>
> Does not list it. Please keep the "NOT AN OFFICIALLY REGISTERED ID" if
> it is not.

Let med quote  text from your link.

Note

    This page has assignments under the control of the IEEE
    Registration Authority that are of primarily historic interest
    that and have traditionally been on the IANA web pages.
    For allocations under the IANA OUI [RFC9542 <https://www.iana.org/go/rfc9542>], see the "Ethernet
    Numbers" IANA web page.  Contact information for the IEEE
    Registration Authority is as follows:http://standards.ieee.org/develop/regauth

Ethertypes:

Registration Procedure(s)

    Not assigned by IANA. Per [RFC9542 <https://www.iana.org/go/rfc9542>], updates to this registry
    are coordinated with the expert.

Note

    The following list of Ethertypes is contributed unverified
    information from various sources.  See the IEEE Registration
    Authority web pages at [http://standards.ieee.org/develop/regauth]
    for a public list of Ethertypes.

    Another list of Ethertypes is maintained by Michael A. Patton
    and is accessible at:http://www.cavebear.com/archive/cavebear/Ethernet/index.html
         

> 	Andrew

IEEE is the official source to use for ethertype number assignment.

/Peter

Re: [PATCH for net-next] if_ether.h: Clarify ethertype validity for gsw1xx dsa

[Andrew Lunn]

On Mon, Nov 17, 2025 at 10:02:05AM +0100, Peter Enderborg wrote:
> (resend due to html bounce)
> On 11/15/25 21:41, Andrew Lunn wrote:
> > On Fri, Nov 14, 2025 at 02:59:36PM +0100, Peter Enderborg wrote:
> > > Ref https://standards-oui.ieee.org/ethertype/eth.txt
> > > 
> > > 
> > Is this actually registered with IANA?
> 
> No.
> 
> > https://www.iana.org/assignments/ieee-802-numbers/ieee-802-numbers.xhtml
> > 
> > Does not list it. Please keep the "NOT AN OFFICIALLY REGISTERED ID" if
> > it is not.

> IEEE is the official source to use for ethertype number assignment.

What i want to make clear is, if IEEE officially allocate this to
something else, this is an unofficial allocation, and the official use
can replace it. We have a few ID like this, and never had a collision
yet, but it could happen.

	Andrew

Re: [PATCH for net-next] if_ether.h: Clarify ethertype validity for gsw1xx dsa

[Peter Enderborg]

You have to pay us$4000 to get a ethertype.

On 11/17/25 14:28, Andrew Lunn wrote:
> On Mon, Nov 17, 2025 at 10:02:05AM +0100, Peter Enderborg wrote:
>> (resend due to html bounce)
>> On 11/15/25 21:41, Andrew Lunn wrote:
>>> On Fri, Nov 14, 2025 at 02:59:36PM +0100, Peter Enderborg wrote:
>>>> Ref https://standards-oui.ieee.org/ethertype/eth.txt
>>>>
>>>>
>>> Is this actually registered with IANA?
>> No.
>>
>>> https://www.iana.org/assignments/ieee-802-numbers/ieee-802-numbers.xhtml
>>>
>>> Does not list it. Please keep the "NOT AN OFFICIALLY REGISTERED ID" if
>>> it is not.
>> IEEE is the official source to use for ethertype number assignment.
> What i want to make clear is, if IEEE officially allocate this to
> something else, this is an unofficial allocation, and the official use
> can replace it. We have a few ID like this, and never had a collision
> yet, but it could happen.

This is the public list from IEEE. It wont get more official. Yes, official
allocation can be changed in non backward compatible way. It happen
in 1983 so it can happen again. If the standard changes we need to adopt.

"NOT AN OFFICIALLY REGISTERED ID" is wrong, it is "Infineon Technologies Corporate Research ST".

> 	Andrew

Re: [PATCH for net-next] if_ether.h: Clarify ethertype validity for gsw1xx dsa

[Andrew Lunn]

On Mon, Nov 17, 2025 at 03:59:13PM +0100, Peter Enderborg wrote:
> You have to pay us$4000 to get a ethertype.
> 
> On 11/17/25 14:28, Andrew Lunn wrote:
> > On Mon, Nov 17, 2025 at 10:02:05AM +0100, Peter Enderborg wrote:
> > > (resend due to html bounce)
> > > On 11/15/25 21:41, Andrew Lunn wrote:
> > > > On Fri, Nov 14, 2025 at 02:59:36PM +0100, Peter Enderborg wrote:
> > > > > Ref https://standards-oui.ieee.org/ethertype/eth.txt
> > > > > 
> > > > > 
> > > > Is this actually registered with IANA?
> > > No.
> > > 
> > > > https://www.iana.org/assignments/ieee-802-numbers/ieee-802-numbers.xhtml
> > > > 
> > > > Does not list it. Please keep the "NOT AN OFFICIALLY REGISTERED ID" if
> > > > it is not.
> > > IEEE is the official source to use for ethertype number assignment.
> > What i want to make clear is, if IEEE officially allocate this to
> > something else, this is an unofficial allocation, and the official use
> > can replace it. We have a few ID like this, and never had a collision
> > yet, but it could happen.
> 
> This is the public list from IEEE. It wont get more official. Yes, official
> allocation can be changed in non backward compatible way. It happen
> in 1983 so it can happen again. If the standard changes we need to adopt.
> 
> "NOT AN OFFICIALLY REGISTERED ID" is wrong, it is "Infineon Technologies Corporate Research ST".

Ah. Let me see if i have this correct.....

Infineon made a spin off called Lantiq.

Lantiq was acquired by Intel

MaxLinear acquired Intels Connected Home division.

Assuming the rights to this ID transferred along the way, it is
O.K. for MaxLinear to use this ID. But, IANAL...

Please expand the commit message with this history, which is then a
justification for removing the NOT AN OFFICIALLY REGISTERED ID, and
why it is O.K. for MaxLinear to use it.

	Andrew