Re: [PATCH net-next 0/2] net: phy: sfp: Add single-byte SMBus SFP access

["Bjørn Mork" <bjorn@mork.no>]

Andrew Lunn <andrew@lunn.ch> writes:

>> So, not only do I think that hwmon should be disabled if using SMBus,
>> but I also think that the kernel should print a warning that SMBus is
>> being used and therefore e.g. copper modules will be unreliable. We
>> don't know how the various firmwares in various microprocessors that
>> convert I2C to MDIO will behave when faced with SMBus transfers.
>
> I agree, hwmon should be disabled, and that the kernel should printing
> a warning that the hardware is broken and that networking is not
> guaranteed to be reliable.

What do you think will be the effect of such a warning?  Who is the
target audience?

You can obviously add it, and I don't really care.  But I believe the
result will be an endless stream of end users worrying about this scary
warning and wanting to know what they can do about it.  What will be
your answer?

No SoC/phy designer will ever see the warning.  They finished designing
these chips decades ago.  The switch designers might see it. But
probably not. They're not worried about running mainline Linux at all.
They have they're vendor SDK.  Linux based of course, but it's never
going to have that warning no matter what you do.  Firmware developers?
Same as switch designers really.

The hardware exists. It's not perfect. We agree so far.  But I do not
understand your way of dealing with that.

If your intention is detecting this hardware problem in bug reports etc,
then it makes more sense to me.  But I believe a more subtle method will
be more effeicient than a standalone and scary warning. Like embedding
"smbus" or similar in existing debug/warning/error messages, e.g by
making it part of the mdio bus name.


Bjørn