[Intel-wired-lan] [PATCH iwl-next v2] ice: Disable Cage Max Power override
Drewek, Wojciech
wojciech.drewek at intel.com
Tue Aug 29 09:12:22 UTC 2023
> -----Original Message-----
> From: Ido Schimmel <idosch at idosch.org>
> Sent: niedziela, 27 sierpnia 2023 10:47
> To: Drewek, Wojciech <wojciech.drewek at intel.com>
> Cc: Jakub Kicinski <kuba at kernel.org>; intel-wired-lan at lists.osuosl.org; netdev at vger.kernel.org; Kitszel, Przemyslaw
> <przemyslaw.kitszel at intel.com>; idosch at nvidia.com
> Subject: Re: [PATCH iwl-next v2] ice: Disable Cage Max Power override
>
> On Fri, Aug 25, 2023 at 11:01:07AM +0000, Drewek, Wojciech wrote:
> > CC: Ido
> >
> > > -----Original Message-----
> > > From: Jakub Kicinski <kuba at kernel.org>
> > > Sent: czwartek, 24 sierpnia 2023 17:32
> > > To: Drewek, Wojciech <wojciech.drewek at intel.com>
> > > Cc: intel-wired-lan at lists.osuosl.org; netdev at vger.kernel.org; Kitszel, Przemyslaw <przemyslaw.kitszel at intel.com>
> > > Subject: Re: [PATCH iwl-next v2] ice: Disable Cage Max Power override
> > >
> > > On Thu, 24 Aug 2023 10:54:59 +0200 Wojciech Drewek wrote:
> > > > NVM module called "Cage Max Power override" allows to
> > > > change max power in the cage. This can be achieved
> > > > using external tools. The responsibility of the ice driver is to
> > > > go back to the default settings whenever port split is done.
> > > > This is achieved by clearing Override Enable bit in the
> > > > NVM module. Override of the max power is disabled so the
> > > > default value will be used.
> > >
> > > Can you say more? We have ETHTOOL_MSG_MODULE_GET / SET, sounds like
> > > something we could quite easily get ethtool to support?
> >
> > So you're suggesting that ethtool could support setting the maximum power in the cage?
> > Something like:
> > - new "--set-module" parameter called "power-max"
> > - new "--get-module" parameters: "power-max-allowed", "power-min-allowed" indicating limitations reported by the HW.
> >
> > About the patch itself, it's only about restoration of the default settings upon port split. Those might be overwritten by
> > Intel's external tools.
>
> Can you please explain why this setting needs to be changed in the first
> place and why it needs to be restored to the default on port split?
In some cases users are trying to use media with power exceeding max allowed value.
Port split require system reboot so it feels natural to me to restore default settings.
More information about the Intel-wired-lan
mailing list