> -----Original Message----- > From: Intel-wired-lan <[email protected]> On Behalf Of > Tony Nguyen > Sent: Friday, May 9, 2025 10:53 PM > To: Korba, Przemyslaw <[email protected]>; intel-wired- > [email protected] > Cc: [email protected]; Kitszel, Przemyslaw > <[email protected]>; Olech, Milena <[email protected]> > Subject: Re: [Intel-wired-lan] [PATCH iwl-next] ice: add recovery clock and > clock 1588 control for E825c > > > > On 5/6/2025 2:35 AM, Przemyslaw Korba wrote: > > Add control for E825 input pins: phy clock recovery and clock 1588. > > E825 does not provide control over platform level DPLL but it > > provides control over PHY clock recovery, and PTP/timestamp driven > > inputs for platform level DPLL. > > > > Introduce a software controlled layer of abstraction to: > > - create a DPLL of type EEC for E825c, > > - create recovered clock pin for each PF, and control them through > > writing to registers, > > - create pin to control clock 1588 for PF0, and control it through > > writing to registers. > > > > Reviewed-by: Milena Olech <[email protected]> > > Signed-off-by: Przemyslaw Korba <[email protected]> > > --- > > drivers/net/ethernet/intel/ice/ice_dpll.c | 856 ++++++++++++++++++-- > > drivers/net/ethernet/intel/ice/ice_dpll.h | 24 +- > > drivers/net/ethernet/intel/ice/ice_main.c | 3 +- > > drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 40 +- > > drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 2 + > > drivers/net/ethernet/intel/ice/ice_tspll.h | 7 + > > drivers/net/ethernet/intel/ice/ice_type.h | 6 + > > 7 files changed, 865 insertions(+), 73 deletions(-) > >
Apologies for late feedback and not answering directly to original message (I lost it somehow) This patch requires two more changes: 1) functional, we can't use pf_id for rclk pins initialization in case of E825C devices. It applies to 2XNC mode variants specifically. There is a mapping needed to achieve unique pin indexes for PFs from both NAC (there is a function ice_get_phy_lane_number which covers that and is suitable for other E800 products). 2) after one more analysis of the patch, we think that CLK_OUT pin definition is redundant (not used) and it would be good to remove it from the patch. Thanks Grzegorz
