[PATCH v3 2/3] drm/xe: Expose PCIe Gen4 downspeed attributes

[Vivi, Rodrigo]

On Wed, 2025-04-23 at 19:41 +0300, Raag Jadav wrote:
> On Wed, Apr 23, 2025 at 09:25:27PM +0530, Vivi, Rodrigo wrote:
> > On Wed, 2025-04-23 at 18:01 +0300, Raag Jadav wrote:
> 
> ...
> 
> > > And all of this will take place when the user is doing a
> > > "firmware
> > > upgrade",
> > > and this can potentially open the door for "link speed downgrade"
> > > to
> > > be
> > > confused with "firmware rollback (downgrade) done for the link
> > > speed",
> > > which is not the case.
> > 
> > well, gen4 is not a firmware, but also
> > 
> > is gen4_downspeed == gen3 ?!
> 
> I understand it as "downgrade to" than "downgrade from".
> 
> Would you prefer gen5_downgrade instead?

again
"yes, naming is hard and this is why we stick with the spec names"

> 
> > yes, naming is hard and this is why we stick with the spec names
> > without re-inventing it, which only increases the confusion.
> > 
> > > 
> > > > > > > 2. This information is for the end user and has to be
> > > > > > > translatable enough
> > > > > > >     regardless of what spec says about it and the
> > > > > > > distinction
> > > > > > > reduces the
> > > > > > >     chances of misinterpretation.
> > > > 
> > > > when you introduce a new term that is not known, I'd say the
> > > > effect
> > > > is
> > > > pretty much the opposite and it can be even worse if a future
> > > > pcie
> > > > spec
> > > > starts to use that term.
> > > 
> > > Agree, unless we're expecting users to be informed about the spec
> > > to
> > > be
> > > able to flash their firmwares. But if you insist, sure will
> > > _upgrade_
> > > the
> > > document ;)
> > 
> > I believe you meant 'update' the document! ;)
> 
> Well, I meant to use 'downgrade' but we'll be going up a revision,
> so.. :D
> 
> Raag