On Fri, Dec 10, 2021 at 02:45:46PM +0100, Maciej Machnikowski wrote:
Synchronous Ethernet networks use a physical layer clock to syntonize the frequency across different network elements.
Basic SyncE node defined in the ITU-T G.8264 consist of an Ethernet Equipment Clock (EEC) and have the ability to synchronize to reference frequency sources.
This patch series is a prerequisite for EEC object and adds ability to enable recovered clocks in the physical layer of the netdev object. Recovered clocks can be used as one of the reference signal by the EEC.
The dependency is the other way around. It doesn't make sense to add APIs to configure the inputs of an object that doesn't exist. First add the EEC object, then we can talk about APIs to configure its inputs from netdevs.
With these four patches alone, user space doesn't know how many EECs there are in the system, it doesn't know the mapping from netdev to EEC, it doesn't know the state of the EEC, it doesn't know which source is chosen in case more than one source is enabled. Patch #3 tries to work around it by having ice print to kernel log, when the information should really be exposed via the EEC object.
+ dev_warn(ice_pf_to_dev(pf), + "<DPLL%i> state changed to: %d, pin %d", + ICE_CGU_DPLL_SYNCE, + pf->synce_dpll_state, + pin);
Further work is required to add the DPLL subsystem, link it to the netdev object and create API to read the EEC DPLL state.
When the EEC object materializes, we might find out that this API needs to be changed / reworked / removed, but we won't be able to do that given it's uAPI. I don't know where the confidence that it won't happen stems from when there are so many question marks around this new object.
v5:
- rewritten the documentation
- fixed doxygen headers
v4:
- Dropped EEC_STATE reporting (TBD: DPLL subsystem)
- moved recovered clock configuration to ethtool netlink
v3:
- remove RTM_GETRCLKRANGE
- return state of all possible pins in the RTM_GETRCLKSTATE
- clarify documentation
v2:
- improved documentation
- fixed kdoc warning
RFC history: v2:
- removed whitespace changes
- fix issues reported by test robot
v3:
- Changed naming from SyncE to EEC
- Clarify cover letter and commit message for patch 1
v4:
- Removed sync_source and pin_idx info
- Changed one structure to attributes
- Added EEC_SRC_PORT flag to indicate that the EEC is synchronized to the recovered clock of a port that returns the state
v5:
- add EEC source as an optiona attribute
- implement support for recovered clocks
- align states returned by EEC to ITU-T G.781
v6:
- fix EEC clock state reporting
- add documentation
- fix descriptions in code comments
Maciej Machnikowski (4): ice: add support detecting features based on netlist ethtool: Add ability to configure recovered clock for SyncE feature ice: add support for monitoring SyncE DPLL state ice: add support for recovered clocks
Documentation/networking/ethtool-netlink.rst | 62 ++++ drivers/net/ethernet/intel/ice/ice.h | 7 + .../net/ethernet/intel/ice/ice_adminq_cmd.h | 70 ++++- drivers/net/ethernet/intel/ice/ice_common.c | 224 +++++++++++++++ drivers/net/ethernet/intel/ice/ice_common.h | 20 +- drivers/net/ethernet/intel/ice/ice_devids.h | 3 + drivers/net/ethernet/intel/ice/ice_ethtool.c | 96 +++++++ drivers/net/ethernet/intel/ice/ice_lib.c | 6 +- drivers/net/ethernet/intel/ice/ice_ptp.c | 35 +++ drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 49 ++++ drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 36 +++ drivers/net/ethernet/intel/ice/ice_type.h | 1 + include/linux/ethtool.h | 9 + include/uapi/linux/ethtool_netlink.h | 21 ++ net/ethtool/Makefile | 3 +- net/ethtool/netlink.c | 20 ++ net/ethtool/netlink.h | 4 + net/ethtool/synce.c | 267 ++++++++++++++++++ 18 files changed, 929 insertions(+), 4 deletions(-) create mode 100644 net/ethtool/synce.c
-- 2.26.3