Hard Processor System Technical Reference Manual: Agilex™ 5 SoCs

ID 814346
Date 4/01/2024
Public
Document Table of Contents

12.3.1.3.1. HPS GMII to RGMII Adapter Intel FPGA IP

You can implement the HPS GMII to RGMII Adapter Intel FPGA IP into FPGA fabric to convert the GMII interface of the Ethernet controller in the hard processor system (HPS) to RGMII interface via FPGA IO connectivity. By default, the HPS Ethernet controller was configured as RGMII interface routed to HPS I/O but you have the option to configure EMAC interface to expose as GMII interface in FPGA fabric, and connect to FPGA I/O.

Intel offers HPS GMII-to-RGMII Adapter Intel FPGA IP, as the RTL adaptation solution for you to interface EMAC to an external RGMII PHY through the FPGA logic.

Figure 299. HPS GMII-to-RGMII Adapter Intel FPGA IP Block Diagram

Features

The following features are supported by the HPS GMII to RGMII Adapter Intel FPGA IP.

  • Performs GMII interface to RGMII interface conversion
  • Supports tri-speed (10/100/1000 Mbps) operation
  • Supports dynamic speed switching
  • Supports generation time option to enable pipeline registers for the transmit and receive paths
Note: The HPS GMII to RGMII Adapter Intel FPGA IP does not support an internal delay of the TX/RX clock. However, you may utilize the FPGA programmable IOE delay features to meet the 2 ns delay for center-aligned data transmission/reception through the FPGA I/O buffer. This delay feature is also commonly supported by external PHY device and can be handled with trace delay at board level.

You can refer to the RGMII Design Guidelines on guidelines to meet desired RGMII clock skew relationship.

Data Path

Figure 300. Transmit and Receive Data Path

For transmit path, the GMII data goes through the transmit pipeline register stage before going into the RGMII Output Standard Function (RGMII_O SF) Converter Block. The pipeline logic is optional and you can enabled or disabled during IP generation time.

For receive path, the GMII data right after the RGMII Input Standard Function (RGMII_I SF) Converter Block goes through the receive pipeline register stage then directly to the HPS EMAC GMII interface. Similarly, you can enable or disable this pipeline logic during IP generation time.

The RGMII_I/O SF converter block manages single data rate to double data rate conversion and vice-versa. The Agilex™ 5 FPGA HVIO component is used to perform this task. This block also decodes collision and carrier sense condition through In-Band status detection.

Signal Interfaces

The following table provides the signal interfaces information.

Table 364.  Signal Interfaces
Signal Name Direction Type Description
clk Input Clock Peripheral clock source
rst_n Input Reset

Active low peripheral asynchronous reset source.

This signal is expected to be asynchronously asserted and synchronously de-asserted associated to the peri_clock.

Synchronous de-assertion must be provided external to

this core.

pll_25m_clock Input Clock 25 MHz input clock from FPGA PLL
pll_2_5m_clock Input Clock 2.5 MHz input clock from FPGA PLL
pll_125m_tx_clock Input Clock 125 MHz input clock from FPGA PLL
locked_pll_250m_tx Input Clock Plase-locked indicator for 250 MHz input clock from FPGA PLL
pll_250m_tx_clock Input Clock 250 MHz input clock from FPGA PLL
HPS GMII Interface
mac_tx_clk_o Input Conduit GMII transmit clock from HPS
mac_tx_clk_i Output Conduit GMII transmit clock to HPS
mac_rx_clk Output Conduit GMII receive clock to HPS
mac_rst_tx_n Input Conduit GMII transmit reset source from HPS. Active low reset
mac_rst_rx_n Input Conduit GMII receive reset source from HPS. Active low reset
mac_txd[7:0] Input Conduit GMII transmit data from HPS
mac_txen Input Conduit GMII transmit enable from HPS
mac_txer Input Conduit GMII transmit error from HPS
mac_rxdv Output Conduit GMII receive data valid to HPS
mac_rxer Output Conduit GMII receive data error to HPS
mac_rxd[7:0] Output Conduit GMII receive data to HPS
mac_col Output Conduit GMII collision detect to HPS
mac_crs Output Conduit GMII carrier sense to HPS
mac_speed[2:0] Input Conduit

MAC speed indication from HPS

3'b011: 1G

3'b100: 100M

3'b111: 10M

PHY RGMII Interface
rgmii_tx_clk Output Conduit RGMII transmit clock to PHY
rgmii_rx_clk Input Conduit RGMII receive clock from PHY
rgmii_txd[3:0] Output Conduit RGMII transmit data to PHY
rgmii_tx_ctl Output Conduit RGMII transmit control to PHY
rgmii_rxd[3:0] Input Conduit RGMII receive data from PHY
rgmii_rx_ctl Input Conduit RGMII receive control from PHY

Clocks

All transmit sequential logic in the HPS GMII to RGMII Adapter Intel FPGA IP is clocked by the HPS PLL for 1000 Mbps speed mode, or by the FPGA PLL during 10/100 Mbps speed mode as shown in the figure below.

Figure 301. TX Clocking Scheme

All receive sequential logic in the HPS GMII to RGMII Adapter Intel FPGA IP is clocked and driven by external PHY rgmii_rx_clk.