| Message ID | 20230921121946.3025771-1-yong.liang.choong@linux.intel.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | TSN auto negotiation between 1G and 2.5G | expand |
> Auto-negotiation between 10, 100, 1000Mbps will use > in-band auto negotiation. Auto-negotiation between 10/100/1000Mbps and > 2.5Gbps will work as the following proposed flow, the stmmac driver reads > the PHY link status registers then identifies the negotiated speed. I don't think you replied to my comment. in-band is just an optimisation. It in theory allows you to avoid a software path, the PHY driver talking to the MAC driver about the PHY status. As an optimisation, it is optional. Linux has the software path and the MAC driver you are using basically has it implemented. Why use this odd mix of in-band and out of band? It seems the change will be simpler if you just use the out of band method all the time and ignore in-band. Andrew
On Thu, Sep 21, 2023 at 03:14:59PM +0200, Andrew Lunn wrote: > > Auto-negotiation between 10, 100, 1000Mbps will use > > in-band auto negotiation. Auto-negotiation between 10/100/1000Mbps and > > 2.5Gbps will work as the following proposed flow, the stmmac driver reads > > the PHY link status registers then identifies the negotiated speed. > > I don't think you replied to my comment. > > in-band is just an optimisation. It in theory allows you to avoid a > software path, the PHY driver talking to the MAC driver about the PHY > status. As an optimisation, it is optional. Linux has the software > path and the MAC driver you are using basically has it implemented. Sorry Andrew, I have to disagree. It isn't always optional - there are PHYs out there where they won't pass data until the in-band exchange has completed. If you try to operate out-of-band without the PHY being told that is the case, and program the MAC/PCS end not to respond to the in-band frames from the PHY, the PHY will report link up as normal (since it reports the media side), but no data will flow because the MAC facing side of the PHY hasn't completed. The only exception are PHYs that default to in-band but have an inband bypass mode also enabled to cover the case where the MAC/PCS doesn't respond to the inband messages.
On 21/9/2023 10:09 pm, Russell King (Oracle) wrote: > On Thu, Sep 21, 2023 at 03:14:59PM +0200, Andrew Lunn wrote: >>> Auto-negotiation between 10, 100, 1000Mbps will use >>> in-band auto negotiation. Auto-negotiation between 10/100/1000Mbps and >>> 2.5Gbps will work as the following proposed flow, the stmmac driver reads >>> the PHY link status registers then identifies the negotiated speed. >> >> I don't think you replied to my comment. >> >> in-band is just an optimisation. It in theory allows you to avoid a >> software path, the PHY driver talking to the MAC driver about the PHY >> status. As an optimisation, it is optional. Linux has the software >> path and the MAC driver you are using basically has it implemented. > > Sorry Andrew, I have to disagree. It isn't always optional - there are > PHYs out there where they won't pass data until the in-band exchange > has completed. If you try to operate out-of-band without the PHY being > told that is the case, and program the MAC/PCS end not to respond to > the in-band frames from the PHY, the PHY will report link up as normal > (since it reports the media side), but no data will flow because the > MAC facing side of the PHY hasn't completed. > > The only exception are PHYs that default to in-band but have an inband > bypass mode also enabled to cover the case where the MAC/PCS doesn't > respond to the inband messages. > Russell is correct, we did set out-of-band for PCS and configured MAC. Due to the PHY not being completed, there will be no data flow through.
Intel platforms’ integrated Gigabit Ethernet controllers support 2.5Gbps mode statically using BIOS programming. In the current implementation, the BIOS menu provides an option to select between 10/100/1000Mbps and 2.5Gbps modes. Based on the selection, the BIOS programs the Phase Lock Loop (PLL) registers. The BIOS also read the TSN lane registers from Flexible I/O Adapter (FIA) block and provided 10/100/1000Mbps/2.5Gbps information to the stmmac driver. But auto-negotiation between 10/100/1000Mbps and 2.5Gbps is not allowed. The new proposal is to support auto-negotiation between 10/100/1000Mbps and 2.5Gbps . Auto-negotiation between 10, 100, 1000Mbps will use in-band auto negotiation. Auto-negotiation between 10/100/1000Mbps and 2.5Gbps will work as the following proposed flow, the stmmac driver reads the PHY link status registers then identifies the negotiated speed. Based on the speed stmmac driver will identify TSN lane registers from FIA then send IPC command to the Power Management controller (PMC) through PMC driver/API. PMC will act as a proxy to programs the PLL registers. changelog: v1 -> v2: - Add static to pmc_lpm_modes declaration - Add cur_link_an_mode to the kernel doc - Combine 2 commits i.e. "stmmac: intel: Separate driver_data of ADL-N from TGL" and "net: stmmac: Add 1G/2.5G auto-negotiation support for ADL-N" into 1 commit. v2 -> v3: - Create `pmc_ipc.c` file for `intel_pmc_ipc()` function and allocate the file in `arch/x86/platform/intel/` directory. - Update phylink's AN mode during phy interface change and not exposing phylink's AN mode into phylib. --- Choong Yong Liang (2): net: phy: update in-band AN mode when changing interface by PHY driver stmmac: intel: Add 1G/2.5G auto-negotiation support for ADL-N David E. Box (1): arch: x86: Add IPC mailbox accessor function and add SoC register access Tan, Tee Min (2): net: pcs: xpcs: combine C37 SGMII AN and 2500BASEX for Intel mGbE controller net: stmmac: enable Intel mGbE 1G/2.5G auto-negotiation support MAINTAINERS | 2 + arch/x86/Kconfig | 9 + arch/x86/platform/intel/Makefile | 1 + arch/x86/platform/intel/pmc_ipc.c | 75 +++++++ drivers/net/ethernet/stmicro/stmmac/Kconfig | 1 + .../net/ethernet/stmicro/stmmac/dwmac-intel.c | 183 +++++++++++++++++- .../net/ethernet/stmicro/stmmac/dwmac-intel.h | 81 ++++++++ .../net/ethernet/stmicro/stmmac/stmmac_main.c | 20 ++ drivers/net/pcs/pcs-xpcs.c | 72 +++++-- drivers/net/phy/phylink.c | 30 ++- include/linux/pcs/pcs-xpcs.h | 1 + .../linux/platform_data/x86/intel_pmc_ipc.h | 34 ++++ include/linux/stmmac.h | 1 + 13 files changed, 493 insertions(+), 17 deletions(-) create mode 100644 arch/x86/platform/intel/pmc_ipc.c create mode 100644 include/linux/platform_data/x86/intel_pmc_ipc.h