KSZ8567S 7-Port 10/100 Ethernet AVB Switch with SGMII and RGMII/MII/RMII Interfaces Highlights * One port with 10/100/1000 Ethernet MAC and SGMII interface * One port with 10/100/1000 Ethernet MAC and configurable RGMII/MII/RMII interface * EtherSynch(R) with full support for IEEE 1588v2 Precision Time Protocol (PTP) * IEEE 802.1AS/Qav Audio Video Bridging (AVB) * IEEE 802.3az Energy Efficient Ethernet (EEE) * IEEE 802.1X access control support * Five ports with integrated 10/100BASE-T PHY transceivers with optional Quiet-WIRE(R) EMC filtering * Non-blocking wire-speed Ethernet switching fabric * Full-featured forwarding and filtering control, including Access Control List (ACL) filtering * Full VLAN and QoS support * EtherGreenTM power management features, including low power standby * Flexible management interface options: SPI, I2C, MIIM, and in-band management via any port * Extended temperature range support * 128-pin TQFP-EP (14 x 14mm) RoHS compliant pkg Target Applications * * * * * * Industrial Ethernet (Profinet, MODBUS, Ethernet/IP) Real-time Ethernet networks IEC 61850 networks w/ power substation automation Industrial control/automation switches Networked measurement and control systems Test and measurement equipment Features * Switch Management Capabilities - 10/100Mbps Ethernet switch basic functions: frame buffer management, address look-up table, queue management, MIB counters - Non-blocking store-and-forward switch fabric assures fast packet delivery by utilizing 4096 entry forwarding table with 256kByte frame buffer - Jumbo packet support up to 9000 bytes - Port mirroring/monitoring/sniffing: ingress and/or egress traffic to any port - Rapid spanning tree protocol (RSTP) support for topology management and ring/linear recovery - Multiple spanning tree protocol (MSTP) support * One External MAC Port with SGMII * One External MAC Port with RGMII/MII/RMII - RGMII v2.0, RMII v1.2 with 50MHz reference clock input/output option, MII in PHY/MAC mode 2017 Microchip Technology Inc. * Five Integrated PHY Ports - 100BASE-TX/10BASE-T/Te IEEE 802.3 Fast Link-up option significantly reduces link-up time Auto-negotiation and Auto-MDI/MDI-X support Energy-Efficient Ethernet (EEE) support with lowpower idle mode and clock stoppage - On-chip termination resistors and internal biasing for differential pairs to reduce power - LinkMD(R) cable diagnostic capabilities for determining cable opens, shorts, and length * Advanced Switch Capabilities - IEEE 802.1Q VLAN support for 128 active VLAN groups and the full range of 4096 VLAN IDs - IEEE 802.1p/Q tag insertion/removal on per port basis - VLAN ID on per port or VLAN basis - IEEE 802.3x full-duplex flow control and half-duplex back pressure collision control - IEEE 802.1X access control (Port and MAC address) - IGMP v1/v2/v3 snooping for multicast packet filtering - IPv6 multicast listener discovery (MLD) snooping - IPv4/IPv6 QoS support, QoS/CoS packet prioritization - 802.1p QoS packet classification with 4 priority queues - Programmable rate limiting at ingress/egress ports * IEEE 1588v2 PTP and Clock Synchronization - Transparent Clock (TC) with auto correction update Master and slave Ordinary Clock (OC) support End-to-end (E2E) or peer-to-peer (P2P) PTP multicast and unicast message support PTP message transport over IPv4/v6 and IEEE 802.3 IEEE 1588v2 PTP packet filtering Synchronous Ethernet support via recovered clock * Audio Video Bridging (AVB) - Compliant with IEEE 802.1BA/AS/Qat/Qav standards Priority queuing, Low latency cut-through mode gPTP time synchronization, credit-based traffic shaper Time aware traffic scheduler per port * Comprehensive Configuration Registers Access - High-speed 4-wire SPI (up to 50MHz), I2C interfaces provide access to all internal registers - MII Management (MIIM, MDC/MDIO 2-wire) Interface provides access to all PHY registers - In-band management via any of the data ports - I/O pin strapping facility to set certain register bits from I/O pins at reset time * Power Management - IEEE 802.3az Energy Efficient Ethernet (EEE) Energy detect power-down mode on cable disconnect Dynamic clock tree control Unused ports can be individually powered down Full-chip software power-down Wake-on-LAN (WoL) standby power mode with PME interrupt output for system wake upon triggered events DS00002391A-page 1 KSZ8567S TO OUR VALUED CUSTOMERS It is our intention to provide our valued customers with the best documentation possible to ensure successful use of your Microchip products. To this end, we will continue to improve our publications to better suit your needs. Our publications will be refined and enhanced as new volumes and updates are introduced. If you have any questions or comments regarding this publication, please contact the Marketing Communications Department via E-mail at docerrors@microchip.com. We welcome your feedback. Most Current Documentation To obtain the most up-to-date version of this documentation, please register at our Worldwide Web site at: http://www.microchip.com You can determine the version of a data sheet by examining its literature number found on the bottom outside corner of any page. The last character of the literature number is the version number, (e.g., DS30000000A is version A of document DS30000000). Errata An errata sheet, describing minor operational differences from the data sheet and recommended workarounds, may exist for current devices. As device/documentation issues become known to us, we will publish an errata sheet. The errata will specify the revision of silicon and revision of document to which it applies. To determine if an errata sheet exists for a particular device, please check with one of the following: * Microchip's Worldwide Web site; http://www.microchip.com * Your local Microchip sales office (see last page) When contacting a sales office, please specify which device, revision of silicon and data sheet (include -literature number) you are using. Customer Notification System Register on our web site at www.microchip.com to receive the most current information on all of our products. DS00002391A-page 2 2017 Microchip Technology Inc. KSZ8567S Table of Contents 1.0 Preface ............................................................................................................................................................................................ 4 2.0 Introduction ..................................................................................................................................................................................... 8 3.0 Pin Descriptions and Configuration ............................................................................................................................................... 10 4.0 Functional Description .................................................................................................................................................................. 19 5.0 Device Registers ........................................................................................................................................................................... 65 6.0 Operational Characteristics ......................................................................................................................................................... 205 7.0 Design Guidelines ....................................................................................................................................................................... 222 8.0 Package Information ................................................................................................................................................................... 224 Appendix A: Data Sheet Revision History ......................................................................................................................................... 226 The Microchip Web Site .................................................................................................................................................................... 227 Customer Change Notification Service ............................................................................................................................................. 227 Customer Support ............................................................................................................................................................................. 227 Product Identification System ........................................................................................................................................................... 228 2017 Microchip Technology Inc. DS00002391A-page 3 KSZ8567S 1.0 PREFACE 1.1 Glossary of Terms TABLE 1-1: GENERAL TERMS Term Description 10BASE-T 10 Mbps Ethernet, 3.3V signaling, IEEE 802.3 compliant 10BASE-Te 10 Mbps Ethernet, 2.5V signaling, IEEE 802.3 compliant 100BASE-TX 100 Mbps Fast Ethernet, IEEE 802.3u compliant ADC Analog-to-Digital Converter AN Auto-Negotiation AVB Audio Video Bridging (IEEE 802.1BA, 802.1AS, 802.1Qat, 802.1Qav) BLW Baseline Wander BPDU Bridge Protocol Data Unit. Messages which carry the Spanning Tree Protocol information. Byte 8 bits CRC Cyclic Redundancy Check. A common technique for detection data transmission errors. CRC for Ethernet is 32 bits long. CSR Control and Status Registers DA Destination Address DWORD 32 bits EEE Energy Efficient Ethernet FCS Frame Check Sequence. The extra checksum characters added to the end of an Ethernet frame, used for error detection and correction. FID Frame or Filter ID. Specifies the frame identifier. Alternately is the filter identifier. FIFO First In First Out buffer FSM Finite State Machine GPIO General Purpose I/O Host External system (Includes processor, application software, etc.) IGMP Internet Group Management Protocol. Defined by RFC 1112, RFC 2236, and RFC 4604 to establish multicast group membership in IPv4 networks. IPG Inter-Packet Gap. A time delay between successive data packets mandated by the network standard for protocol reasons. Jumbo Packet A packet larger than the standard Ethernet packet (1518 bytes). Large packet sizes allow for more efficient use of bandwidth, lower overhead, less processing, etc. lsb Least Significant Bit LSB Least Significant Byte MAC Media Access Controller. A functional block responsible for implementing the media access control layer, which is a sublayer of the data link layer. MDI Medium Dependent Interface. An Ethernet port connection that allows network hubs or switches to connect to other hubs or switches without a null-modem, or crossover, cable. MDIX Media Independent Interface with Crossover. An Ethernet port connection that allows networked end stations (i.e., PCs or workstations) to connect to each other using a null-modem, or crossover, cable. MIB Management Information Base. The MIB comprises the management portion of network devices. This can include monitoring traffic levels and faults (statistical), and can also change operating parameters in network nodes (static forwarding addresses). DS00002391A-page 4 2017 Microchip Technology Inc. KSZ8567S TABLE 1-1: GENERAL TERMS (CONTINUED) Term Description MII Media Independent Interface. The MII accesses PHY registers as defined in the IEEE 802.3 specification. MIIM Media Independent Interface Management MLD Multicast Listening Discovery. This protocol is defined by RFC 3810 and RFC 4604 to establish multicast group membership in IPv6 networks. MLT-3 Multi-Level Transmission Encoding (3-Levels). A tri-level encoding method where a change in the logic level represents a code bit "1" and the logic output remaining at the same level represents a code bit "0". msb Most Significant Bit MSB Most Significant Byte NRZ Non Return to Zero. A type of signal data encoding whereby the signal does not return to a zero state in between bits. NRZI Non Return to Zero Inverted. This encoding method inverts the signal for a "1" and leaves the signal unchanged for a "0" N/A Not Applicable NC No Connect OUI Organizationally Unique Identifier PHY A device or function block which performs the physical layer interface function in a network. PLL Phase Locked Loop. A electronic circuit that controls an oscillator so that it maintains a constant phase angle (i.e., lock) on the frequency of an input, or reference, signal. PTP Precision Time Protocol RESERVED Refers to a reserved bit field or address. Unless otherwise noted, reserved bits must always be zero for write operations. Unless otherwise noted, values are not guaranteed when reading reserved bits. Unless otherwise noted, do not read or write to reserved addresses. RTC Real-Time Clock SA Source Address SFD Start of Frame Delimiter. The 8-bit value indicating the end of the preamble of an Ethernet frame. SQE Signal Quality Error (also known as "heartbeat") SSD Start of Stream Delimiter TCP Transmission Control Protocol UDP User Datagram Protocol - A connectionless protocol run on top of IP networks UTP Unshielded Twisted Pair. Commonly a cable containing 4 twisted pairs of wire. UUID Universally Unique IDentifier VLAN Virtual Local Area Network WORD 16 bits 2017 Microchip Technology Inc. DS00002391A-page 5 KSZ8567S 1.2 Buffer Types TABLE 1-2: BUFFER TYPES Buffer Type I Description Input IPU IPU/O IPD IPD/O O8 Input with internal pull-up (58 k 30%) Input with internal pull-up (58 k 30%) during power-up/reset; output pin during normal operation Input with internal pull-down (58 k 30%) Input with internal pull-down (58 k 30%) during power-up/reset; output pin during normal operation Output with 8 mA sink and 8 mA source O24 Output with 24 mA sink and 24 mA source OPU Output (8mA) with internal pull-up (58 k 30%) OPD Output (8mA) with internal pull-down (58 k 30%) SGMII-I SGMII Input SGMII-O SGMII Output AIO Analog bidirectional ICLK Crystal oscillator input pin OCLK Crystal oscillator output pin P Power GND Ground Note: Refer to Section 6.3, "Electrical Characteristics," on page 206 for the electrical characteristics of the various buffers. DS00002391A-page 6 2017 Microchip Technology Inc. KSZ8567S 1.3 Register Nomenclature TABLE 1-3: REGISTER NOMENCLATURE Register Bit Type Notation R Read: A register or bit with this attribute can be read. W Write: A register or bit with this attribute can be written. RO Read only: Read only. Writes have no effect. RC Read to Clear: Contents is cleared after the read. Writes have no effect. WO Write only: If a register or bit is write-only, reads will return unspecified data. WC Write One to Clear: Writing a one clears the value. Writing a zero has no effect. LL Latch Low: Applies to certain RO status bits. If a status condition causes this bit to go low, it will maintain the low state until read, even if the status condition changes. A read clears the latch, allowing the bit to go high if dictated by the status condition. LH Latch High: Applies to certain RO status bits. If a status condition causes this bit to go high, it will maintain the high state until read, even if the status condition changes. A read clears the latch, allowing the bit to go low if dictated by the status condition. SC Self-Clearing: Contents are self-cleared after the being set. Writes of zero have no effect. Contents can be read. RESERVED 1.4 Register Bit Description Reserved Field: Reserved fields must be written with zeros, unless otherwise indicated, to ensure future compatibility. The value of reserved bits is not guaranteed on a read. References * NXP I2C-Bus Specification (UM10204, April 4, 2014): www.nxp.com/documents/user_manual/UM10204.pdf 2017 Microchip Technology Inc. DS00002391A-page 7 KSZ8567S 2.0 INTRODUCTION 2.1 General Description The KSZ8567S is a highly-integrated, IEEE 802.3 compliant networking device that incorporates a layer-2 managed high-performance Ethernet switch, five 10BASE-T/Te/100BASE-TX physical layer transceivers (PHYs) and associated MAC units, and two individually configurable MAC ports (one SGMII interface, one RGMII/MII/RMII interface) for direct connection to a host processor/controller, another Ethernet switch, or an Ethernet PHY transceiver. The SGMII port may be connected to a fiber optic transceiver. The KSZ8567S is built upon industry-leading Ethernet technology, with features designed to offload host processing and streamline the overall design: * * * * * * Non-blocking wire-speed Ethernet switch fabric Full-featured forwarding and filtering control, including port-based Access Control List (ACL) filtering Full VLAN and QoS support Traffic prioritization with per-port ingress/egress queues and by traffic classification Spanning Tree support IEEE 802.1X access control support As a member of the EtherSynch product family, the KSZ8567S incorporates full hardware support for the IEEE 1588v2 Precision Time Protocol (PTP), including hardware time-stamping at all PHY-MAC interfaces, and a high-resolution hardware "PTP clock". IEEE 1588 provides sub-microsecond synchronization for a range of industrial Ethernet applications. The KSZ8567S fully supports the IEEE family of Audio Video Bridging (AVB) standards, which provides high Quality of Service (QoS) for latency sensitive traffic streams over Ethernet. Time-stamping and time-keeping features support IEEE 802.1AS time synchronization. All ports feature credit based traffic shapers for IEEE 802.1Qav, and a time aware scheduler as proposed for IEEE 802.1Qbv. The 100Mbps PHYs feature Quiet-WIRE internal filtering to reduce line emissions and enhance immunity to environmental noise. It is ideal for automotive or industrial applications where stringent radiated emission limits must be met. A host processor can access all KSZ8567S registers for control over all PHY, MAC, and switch functions. Full register access is available via the integrated SPI or I2C interfaces, and by in-band management via any one of the data ports. PHY register access is provided by a MIIM interface. Flexible digital I/O voltage allows the MAC port to interface directly with a 1.8/2.5/3.3V host processor/controller/FPGA. Additionally, a robust assortment of power-management features including IEEE 802.3az Energy-Efficient Ethernet (EEE) for power savings with idle link, and Wake-on-LAN (WoL) for low power standby operation, have been designed to satisfy energy-efficient system requirements. The KSZ8567S is available in an extended (-40C to +105C) temperature range. An internal block diagram of the KSZ8567S is shown in Figure 2-1. DS00002391A-page 8 2017 Microchip Technology Inc. KSZ8567S Port 2 10/100 PHY 2 Port 3 10/100 PHY 3 Port 4 10/100 PHY 4 Port 5 10/100 PHY 5 Precision GPIO GPIO IEEE 1588 / 802.1AS Clock KSZ8567S 2017 Microchip Technology Inc. MAC 1 MAC 6 MAC 2 MAC 7 MAC 3 MAC 4 MAC 5 Control Registers Switch Engine 10/100 PHY 1 1588 & AVB Processing, Queue Management, QOS, Etc. Port 1 Address Lookup IEEE 1588 / 802.1AS Time Stamp INTERNAL BLOCK DIAGRAM IEEE 1588 / 802.1AS Time Stamp FIGURE 2-1: RGMII/MII/RMII SGMII MIB Counters Frame Buffers Queue Mgmt. 2 SPI/I C/MIIM DS00002391A-page 9 KSZ8567S 3.0 PIN DESCRIPTIONS AND CONFIGURATION 3.1 Pin Assignments The device pin diagram for the KSZ8567S can be seen in Figure 3-1. Table 3-1 provides a KSZ8567S pin assignment table. Pin descriptions are provided in Section 3.2, "Pin Descriptions". PIN ASSIGNMENTS (TOP VIEW) 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 RESET_N SYNCLKO INTRP_N PME_N LED2_1 LED2_0 GPIO_1 LED3_1 LED3_0 DVDDL LED4_1 LED4_0 VDDLS VDDHS S_OUT7M S_OUT7P GND S_IN7P S_IN7M GND S_REXT GND NC NC VDDHS VDDLS DVDDL GND VDDIO IBA DVDDL RXD6_0 FIGURE 3-1: 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 KSZ8567S 128-TQFP-EP ( Top View) G ND (Connect exposed pad to ground with a via field) RXD6_1 RXD6_2 RXD6_3 VDDIO CRS6 RX_ER6 RX_DV6/CRS_DV6/RX_CTL6 RX_CLK6/REFCLKO6 DVDDL TXD6_0 TXD6_1 TXD6_2 TXD6_3 COL6 TX_ER6 TX_EN6/TX_CTL6 TX_CLK6/REFCLKI6 GND GND DVDDL AVDDH NC NC AVDDL NC NC RX4M RX4P AVDDL TX4M TX4P AVDDH TX1P TX1M AVDDL RX1P RX1M NC NC NC NC AVDDH DVDDL TX2P TX2M AVDDL RX2P RX2M NC NC AVDDL NC NC AVDDH DVDDL TX3P TX3M RX3P RX3M NC NC AVDDL NC NC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 SDO SDI/SDA/MDIO VDDIO SCS_N SCL/MDC LED5_0 LED5_1 DVDDL LED1_0 LED1_1 GND NC GND DVDDL AVDDH TX5P TX5M AVDDL RX5P RX5M NC NC AVDDL NC NC AVDDH GND AVDDL XO XI ISET AVDDH Note: When an "_N" is used at the end of the signal name, it indicates that the signal is active low. For example, RESET_N indicates that the reset signal is active low. The buffer type for each signal is indicated in the "Buffer Type" column of the pin description tables in Section 3.2, "Pin Descriptions". A description of the buffer types is provided in Section 1.2, "Buffer Types". DS00002391A-page 10 2017 Microchip Technology Inc. KSZ8567S TABLE 3-1: PIN ASSIGNMENTS Pin Pin Name Pin Pin Name Pin Pin Name Pin Pin Name 1 TX1P 33 AVDDH 65 RXD6_0 (Note 3-1) 97 SDO 2 TX1M 34 TX4P 66 DVDDL 98 SDI/SDA/MDIO 3 AVDDL 35 TX4M 67 IBA (Note 3-1) 99 VDDIO 4 RX1P 36 AVDDL 68 VDDIO 100 SCS_N 5 RX1M 37 RX4P 69 GND 101 SCL/MDC 6 NC 38 RX4M 70 DVDDL 102 LED5_0 7 NC 39 NC 71 VDDLS 103 LED5_1 (Note 3-1) 8 NC 40 NC 72 VDDHS 104 DVDDL 9 NC 41 AVDDL 73 NC 105 LED1_0 (Note 3-1) 10 AVDDH 42 NC 74 NC 106 LED1_1 (Note 3-1) 11 DVDDL 43 NC 75 GND 107 GND 12 TX2P 44 AVDDH 76 S_REXT 108 NC 13 TX2M 45 DVDDL 77 GND 109 GND 14 AVDDL 46 GND 78 S_IN7M 110 DVDDL 15 RX2P 47 GND 79 S_IN7P 111 AVDDH 16 RX2M 48 TX_CLK6/REFCLKI6 80 GND 112 TX5P 17 NC 49 TX_EN6/TX_CTL6 81 S_OUT7P 113 TX5M 18 NC 50 TX_ER6 82 S_OUT7M 114 AVDDL 19 AVDDL 51 COL6 83 VDDHS 115 RX5P 20 NC 52 TXD6_3 84 VDDLS 116 RX5M 21 NC 53 TXD6_2 85 LED4_0 (Note 3-1) 117 NC 22 AVDDH 54 TXD6_1 86 LED4_1 (Note 3-1) 118 NC 23 DVDDL 55 TXD6_0 87 DVDDL 119 AVDDL 24 TX3P 56 DVDDL 88 LED3_0 120 NC 25 TX3M 57 RX_CLK6/REFCLKO6 89 LED3_1 (Note 3-1) 121 NC 26 RX3P 58 RX_DV6/CRS_DV6/ RX_CTL6 90 GPIO_1 122 AVDDH 27 RX3M 59 RX_ER6 91 LED2_0 (Note 3-1) 123 GND 28 NC 60 CRS6 92 LED2_1 (Note 3-1) 124 AVDDL 29 NC 61 VDDIO 93 PME_N 125 XO 30 AVDDL 62 RXD6_3 (Note 3-1) 94 INTRP_N 126 XI 31 NC 63 RXD6_2 (Note 3-1) 95 SYNCLKO 127 ISET 32 NC 64 RXD6_1 (Note 3-1) 96 RESET_N 128 AVDDH Exposed Pad Must be Connected to GND Note 3-1 This pin provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. 2017 Microchip Technology Inc. DS00002391A-page 11 KSZ8567S 3.2 Pin Descriptions This sections details the functions of the various device signals. TABLE 3-2: PIN DESCRIPTIONS Buffer Type Name Symbol Port 5-1 Ethernet TX + TX[5:1]P AIO Port 5-1 100BASE-TX/10BASE-T Differential Data (+) Transmit when in MDI mode, receive when in MDI-X mode. Port 5-1 Ethernet TX - TX[5:1]M AIO Port 5-1 100BASE-TX/10BASE-T Differential Data (-) Transmit when in MDI mode, receive when in MDI-X mode. Port 5-1 Ethernet RX + RX[5:1]P AIO Port 5-1 100BASE-TX/10BASE-T Differential Data (+) Receive when in MDI mode, transmit when in MDI-X mode. Port 5-1 Ethernet RX - RX[5:1]M AIO Port 5-1 100BASE-TX/10BASE-T Differential Data (-) Receive when in MDI mode, transmit when in MDI-X mode. Description Ports 5-1 10/100 Ethernet Pins Port 6 RGMII/MII/RMII Pins Port 6 Transmit/ Reference Clock TX_CLK6/ REFCLKI6 I/O8 MII Mode: TX_CLK6 is the Port 6 25/2.5MHz Transmit Clock. In PHY mode this pin is an output, in MAC mode it is an input. RMII Mode: REFCLKI6 is the Port 6 50MHz Reference Clock input when in RMII Normal mode. This pin is unused when in RMII Clock mode. RGMII Mode: TX_CLK6 is the Port 6 125/25/2.5MHz Transmit Clock input. Port 6 Transmit Enable/Control TX_EN6/ TX_CTL6 IPD MII/RMII Modes: TX_EN6 is the Port 6 Transmit Enable. Port 6 Transmit Error TX_ER6 IPD MII Mode: Port 6 Transmit Error input. RGMII Mode: TX_CTL6 is the Port 6 Transmit Control. RMII/RGMII Modes: Not used. Do not connect this pin in these modes of operation. Port 6 Collision Detect COL6 IPD/O8 MII Mode: Port 6 Collision Detect. In PHY mode this pin is an output, in MAC mode it is an input. RMII/RGMII Modes: Not used. Do not connect this pin in these modes of operation. Port 6 Transmit Data 3 TXD6_3 IPD MII/RGMII Modes: Port 6 Transmit Data bus bit 3. RMII Mode: Not used. Do not connect this pin in this mode of operation. Port 6 Transmit Data 2 TXD6_2 IPD MII/RGMII Modes: Port 6 Transmit Data bus bit 2. RMII Mode: Not used. Do not connect this pin in this mode of operation. Port 6 Transmit Data 1 TXD6_1 IPD MII/RMII/RGMII Modes: Port 6 Transmit Data bus bit 1. Port 6 Transmit Data 0 TXD6_0 IPD MII/RMII/RGMII Modes: Port 6 Transmit Data bus bit 0. DS00002391A-page 12 2017 Microchip Technology Inc. KSZ8567S TABLE 3-2: PIN DESCRIPTIONS (CONTINUED) Name Symbol Port 6 Receive/ Reference Clock RX_CLK6/ REFCLKO6 Buffer Type I/O24 Description MII Mode: RX_CLK6 is the Port 6 25/2.5MHz Receive Clock. In PHY mode this pin is an output, in MAC mode it is an input. RMII Mode: REFCLKO6 is the Port 6 50MHz Reference Clock output when in RMII Clock mode. This pin is unused when in RMII Normal mode. RGMII Mode: RX_CLK6 is the Port 6 125/25/2.5MHz Receive Clock output. Port 6 Receive Data Valid / Carrier Sense / Control RX_DV6/ CRS_DV6/ RX_CTL6 IPD/O24 MII Mode: RX_DV6 is the Port 6 Received Data Valid output. RMII Mode: CRS_DV6 is the Carrier Sense / Receive Data Valid output. RGMII Mode: RX_CTL6 is the Receive Control output. Port 6 Receive Error RX_ER6 IPD/O24 MII Mode: Port 6 Receive Error output. RMII/RGMII Modes: Not used. Do not connect this pin in these modes of operation. Port 6 Carrier Sense CRS6 IPD/O8 MII Mode: Port 6 Carrier Sense. In PHY mode this pin is an output, in MAC mode it is an input. RMII/RGMII Modes: Not used. Do not connect this pin in these modes of operation. Port 6 Receive Data 3 RXD6_3 IPD/O24 MII/RGMII Modes: Port 6 Receive Data bus bit 3. RMII Mode: Not used. Do not connect this pin in this mode of operation. Note: Port 6 Receive Data 2 RXD6_2 IPD/O24 This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. MII/RGMII Modes: Port 6 Receive Data bus bit 2. RMII Mode: Not used. Do not connect this pin in this mode of operation. Note: Port 6 Receive Data 1 RXD6_1 2017 Microchip Technology Inc. IPD/O24 This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. MII/RMII/RGMII Modes: Port 6 Receive Data bus bit 1. Note: This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. DS00002391A-page 13 KSZ8567S TABLE 3-2: PIN DESCRIPTIONS (CONTINUED) Name Symbol Buffer Type Port 6 Receive Data 0 RXD6_0 IPD/O24 Port 7 SGMII Differential Input Data + S_IN7P SGMII-I Port 7 SGMII Differential Input Data + Port 7 SGMII Differential Input Data - S_IN7M SGMII-I Port 7 SGMII Differential Input Data - Port 7 SGMII Differential Output Data + S_OUT7P SGMII-O Port 7 SGMII Differential Output Data + Port 7 SGMII Differential Output Data - S_OUT7M SGMII-O Port 7 SGMII Differential Output Data - Port 7 SGMII Reference Resistor S_REXT A Description MII/RMII/RGMII Modes: Port 6 Receive Data bus bit 0. Note: This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. Port 7 SGMII Pins SGMII reference resistor. Connect a 191 1% resistor between this pin and GND using a short trace to avoid noise coupling. SPI/I2C/MIIM Interface Pins SPI/I2C/MIIM Serial Clock SCL/MDC IPU SPI Data Out SDO O8 SPI/I2C Modes: SCL serial clock. MIIM Mode: MDC serial clock. SPI Mode: Data out (also known as MISO). I2C/MIIM Modes: Not used. SPI Data In / I2C/MIIM Data In/Out SDI/SDA/MDIO IPU/O8 SPI Mode: SDI Data In (also known as MOSI). I2C Mode: SDA Data In/Out. MIIM Mode: MDIO Data In/Out. SDI and MDIO are open-drain signals when in the output state. An external pull-up resistor to VDDIO (1.0k to 4.7k) is required. SPI Chip Select SCS_N IPU SPI Mode: Chip Select (active low). I2C/MIIM Modes: Not used. LED Pins Port 1 LED Indicator 0 LED1_0 IPU/O8 Port 1 LED Indicator 0. Active low output sinks current to light an external LED. Note: DS00002391A-page 14 This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. 2017 Microchip Technology Inc. KSZ8567S TABLE 3-2: PIN DESCRIPTIONS (CONTINUED) Name Symbol Buffer Type Port 1 LED Indicator 1 LED1_1 IPU/O8 Description Port 1 LED Indicator 1. Active low output sinks current to light an external LED. Note: Port 2 LED Indicator 0 LED2_0 IPU/O8 Port 2 LED Indicator 0. Active low output sinks current to light an external LED. Note: Port 2 LED Indicator 1 LED2_1 IPU/O8 This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. Port 2 LED Indicator 1. Active low output sinks current to light an external LED. Note: This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. Port 3 LED Indicator 0 LED3_0 IPU/O8 Port 3 LED Indicator 0. Active low output sinks current to light an external LED. Port 3 LED Indicator 1 LED3_1 IPU/O8 Port 3 LED Indicator 1. Active low output sinks current to light an external LED. Note: Port 4 LED Indicator 0 LED4_0 IPU/O8 Port 4 LED Indicator 0. Active low output sinks current to light an external LED. Note: Port 4 LED Indicator 1 LED4_1 IPU/O8 This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. Port 4 LED Indicator 1. Active low output sinks current to light an external LED. Note: This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. Port 5 LED Indicator 0 LED5_0 IPU/O8 Port 5 LED Indicator 0. Active low output sinks current to light an external LED. Port 5 LED Indicator 1 LED5_1 IPU/O8 Port 5 LED Indicator 1. Active low output sinks current to light an external LED. Note: 2017 Microchip Technology Inc. This pin also provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. DS00002391A-page 15 KSZ8567S TABLE 3-2: PIN DESCRIPTIONS (CONTINUED) Name Symbol Buffer Type Interrupt INTRP_N OPU Description Miscellaneous Pins Active low, open-drain interrupt. Note: Power Management Event PME_N O8 This pin requires an external pull-up resistor. Power Management Event. This output signal indicates that an energy detect event has occurred. It is intended to wake up the system from a low power mode. Note: The assertion polarity is programmable (default active low). An external pull-up resistor is required for active-low operation; an external pull-down resistor is required for active-high operation. System Reset RESET_N IPU Active low system reset. The device must be reset either during or after power-on. An RC circuit is suggested for power-on reset. Crystal Clock / Oscillator Input XI ICLK Crystal clock / oscillator input. When using a 25MHz crystal, this input is connected to one lead of the crystal. When using an oscillator, this pin is the input from the oscillator. The crystal oscillator should have a tolerance of 50ppm. Crystal Clock Output XO OCLK Crystal clock / oscillator output. When using a 25MHz crystal, this output is connected to one lead of the crystal. When using an oscillator, this pin is left unconnected. 25/125MHz Reference Clock Output SYNCLKO IPU/O24 25/125MHz reference clock output, derived from the crystal input or the recovered clock of any PHY. This signal may be used for Synchronous Ethernet. General Purpose Input/Output 1 GPIO_1 IPU/O8 This signal can be used as an input or output for use by the IEEE 1588 event trigger or timestamp capture units. It will be synchronized to the internal IEEE 1588 clock. This pin can also be controlled (as an output) or sampled (as an input) via device registers. Transmit Output Current Set Resistor ISET A Transmit output current set resistor. This pin configures the physical transmit output current. It must be connected to GND through a 6.04k 1% resistor. In-Band Management Configuration Strap IBA IPD In-Band Management Configuration strap. This pin provides configuration strap functions during hardware/software resets. Refer to Section 3.2.1, "Configuration Straps" for additional information. No Connect NC - DS00002391A-page 16 No Connect. For proper operation, this pin must be left unconnected. 2017 Microchip Technology Inc. KSZ8567S TABLE 3-2: PIN DESCRIPTIONS (CONTINUED) Buffer Type Name Symbol Description +3.3/2.5/1.8V I/O Power VDDIO P +3.3V / +2.5V / +1.8V I/O Power +3.3/2.5V Analog Power AVDDH P +3.3V / +2.5V Analog Power +1.2V Analog Power AVDDL P +1.2V Analog Power +1.2V Digital Power DVDDL P +1.2V Digital Power +1.2V SGMII Core Power VDDLS P +1.2V SGMII Core Power +2.5V SGMII I/O Power VDDHS P +2.5V SGMII I/O Power Ground GND GND Ground (pins and pad) Power/Ground Pins 3.2.1 CONFIGURATION STRAPS The KSZ8567S utilizes configuration strap pins to configure the device for different modes. While RESET_N is low, these pins are hi-Z. Pull-up/down resistors are used to create high or low states on these pins, which are internally sampled at the rising edge of RESET_N. All of these pins have a weak internal pull-up or pull-down resistor which provides a default level for strapping. To strap an LED pin low, use a 750 to 1k external pull-down resistor. To strap a non-LED pin high, use an external 1k to 10k pull-up resistor to VDDIO. Once RESET_N is high, all of these pins become driven outputs. Because the internal pull-up/down resistors are not strong, consideration must be given to any other pull-up/down resistors which may reside on the board or inside a device connected to these pins. When an LED pin is directly driving an LED, the effect of the LED and LED load resistor on the strapping level must be considered. This is the reason for using a small value resistor to pull an LED pin low. This is especially true when an LED is powered from a voltage that is higher than VDDIO. The configuration strap pins and their associated functions are detailed in Table 3-3. TABLE 3-3: CONFIGURATION STRAP DESCRIPTIONS Configuration Strap Pin Description LED1_0 Quiet-WIRE Filtering Enable 0: Quiet-WIRE filtering enabled 1: Quiet-WIRE filtering disabled (Default) LED1_1 Flow Control (All Ports) 0: Flow control disabled 1: Flow control enabled (Default) LED2_1 Link-up Mode (All PHYs) 0: Fast Link-up: Auto-negotiation and auto MDI/MDI-X are disabled 1: Normal Link-up: Auto-negotiation and auto MDI/MDI-X are enabled (Default) Note: 2017 Microchip Technology Inc. Since Fast Link-up disables auto-negotiation and auto-crossover, it is suitable only for specialized applications. DS00002391A-page 17 KSZ8567S TABLE 3-3: CONFIGURATION STRAP DESCRIPTIONS (CONTINUED) Configuration Strap Pin Description LED2_0, LED4_0 When LED2_1 = 1 at strap-in (Normal Link-up): [LED2_0, LED4_0]: Auto-Negotiation Enable (All PHYs) / NAND Tree Test Mode 00: Reserved 01: Auto-negotiation disabled 10: NAND Tree test mode 11: Auto-negotiation enabled (Default) When LED2_1 = 0 at strap-in (Fast Link-up; All PHYs Full-Duplex): LED2_0: MDI/MDI-X Mode (All PHYs) 0: MDI-X 1: MDI (Default) LED4_0: Not Used LED4_1, LED3_1 [LED4_1, LED3_1]: Management Interface Mode 00: MIIM (MDIO) 01: I2C 1x: SPI (Default) LED5_1 RXD6_3, RXD6_2 Switch Enable at Startup 0: Start Switch is disabled. The switch will not forward packets until the Start Switch bit is set in the Switch Operation Register. 1: Start Switch is enabled. The switch will forward packets immediately after reset. (Default) [RXD6_3, RXD6_2]: Port 6 Mode 00: RGMII (Default) 01: RMII 10: Reserved 11: MII RXD6_1 Port 6 MII/RMII Mode 0: MII: PHY Mode (Default) RMII: Clock Mode. RMII 50MHz reference clock is output on REFCLKO6. (Default) RGMII: No effect 1: MII: MAC Mode RMII: Normal Mode. RMII 50MHz reference clock is input on REFCLKI6. RGMII: No effect RXD6_0 Port 6 Speed Select 0: 1000Mbps Mode (Default) 1: 10/100Mbps Mode Note: IBA DS00002391A-page 18 If Port 6 is configured for MII or RMII, set the speed to 100Mbps. In-Band Management 0: Disable In-Band Management (Default) 1: Enable In-Band Management 2017 Microchip Technology Inc. KSZ8567S 4.0 FUNCTIONAL DESCRIPTION This section provides functional descriptions for the following: * * * * * * * * * * * * * Physical Layer Transceiver (PHY) LEDs Media Access Controller (MAC) Switch IEEE 1588 Precision Time Protocol Audio Video Bridging and Time Sensitive Networks NAND Tree Support Clocking Power Power Management Management Interface In-Band Management MAC Interface (Ports 6 and 7) 4.1 Physical Layer Transceiver (PHY) Ports 1 through 5 include completely integrated dual-speed (10BASE-T/Te, 100BASE-TX) Ethernet physical layer transceivers for transmission and reception of data over standard four-pair unshielded twisted pair (UTP), CAT-5 or better Ethernet cable. At 100Mbps, the optional Quiet-WIRE filtering feature reduced emissions while maintaining interoperability with standard 100BASE-TX devices. The device reduces board cost and simplifies board layout by using on-chip termination resistors for the differential pairs, eliminating the need for external termination resistors. The internal chip termination and biasing provides significant power savings when compared with using external biasing and termination resistors. 4.1.1 4.1.1.1 100BASE-TX TRANSCEIVER 100BASE-TX Transmit The 100BASE-TX transmit function performs parallel-to-serial conversion, 4B/5B coding, scrambling, NRZ-to-NRZI conversion, and MLT3 encoding and transmission. The circuitry starts with a parallel-to-serial conversion, which converts the MII data from the MAC into a 125MHz serial bit stream. The data and control stream is then converted into 4B/5B coding, followed by a scrambler. The serialized data is further converted from NRZ-to-NRZI format, and then transmitted in MLT3 current output. An external ISET resistor sets the output current for the 1:1 transformer ratio. The output signal has a typical rise/fall time of 4ns and complies with the ANSI TP-PMD standard regarding amplitude balance, overshoot, and timing jitter. The wave-shaped 10BASE-T/Te output driver is also incorporated into the 100BASE-TX driver. 4.1.1.2 100BASE-TX Receive The 100BASE-TX receiver function performs adaptive equalization, DC restoration, MLT3-to-NRZI conversion, data and clock recovery, NRZI-to-NRZ conversion, de-scrambling, 4B/5B decoding, and serial-to-parallel conversion. The receiving side starts with the equalization filter to compensate for inter-symbol interference (ISI) over the twisted pair cable. Since the amplitude loss and phase distortion is a function of the cable length, the equalizer has to adjust its characteristics to optimize performance. In this design, the variable equalizer makes an initial estimation based on comparisons of incoming signal strength against some known cable characteristics, and then tunes itself for optimization. This is an ongoing process and self-adjusts against environmental changes such as temperature variations. Next, the equalized signal goes through a DC restoration and data conversion block. The DC restoration circuit is used to compensate for the effect of baseline wander and to improve the dynamic range. The differential data conversion circuit converts the MLT3 format back to NRZI. The slicing threshold is also adaptive. The clock recovery circuit extracts the 125MHz clock from the edges of the NRZI signal. This recovered clock is then used to convert the NRZI signal into the NRZ format. This signal is sent through the de-scrambler followed by the 4B/ 5B decoder. Finally, the NRZ serial data is converted to an MII format and provided as the input data to the MAC. 2017 Microchip Technology Inc. DS00002391A-page 19 KSZ8567S 4.1.1.3 Scrambler/De-Scrambler The purpose of the scrambler is to spread the power spectrum of the signal to reduce electromagnetic interference (EMI) and baseline wander. The scrambler is used only for 100BASE-TX. Transmitted data is scrambled through the use of an 11-bit wide linear feedback shift register (LFSR). The scrambler generates a 2047-bit non-repetitive sequence. Then the receiver de-scrambles the incoming data stream using the same sequence as at the transmitter. 4.1.2 10BASE-T/Te TRANSCEIVER When the AVDDH supply is 3.3V, the 10Mbps interface is 10BASE-T. When AVDDH is 2.5V, the 10BASE-T signal has a reduced transmit signal amplitude and is known as 10BASE-Te. 10BASE-Te is interoperable to 100m with 10BASET when Cat5 cable is used. 4.1.2.1 10BASE-T/Te Transmit The 10BASE-T/Te driver is incorporated with the 100BASE-TX driver to allow for transmission using the same magnetics. They are internally wave-shaped and pre-emphasized into outputs with typical 2.5V amplitude for 10BASE-T, or 1.75V amplitude for 10BASE-Te. The harmonic contents are at least 27dB below the fundamental frequency when driven by an all-ones Manchester-encoded signal. 4.1.2.2 10BASE-T/Te Receive On the receive side, input buffers and level detecting squelch circuits are employed. A differential input receiver circuit and a phase-locked loop (PLL) perform the decoding function. The Manchester-encoded data stream is separated into clock signal and NRZ data. A squelch circuit rejects signals with levels less than 400mV or with short pulse widths to prevent noise at the RXP1 or RXM1 input from falsely triggering the decoder. When the input exceeds the squelch limit, the PLL locks onto the incoming signal and the device decodes a data frame. The receiver clock is maintained active during idle periods in between data reception. 4.1.3 AUTO MDI/MDI-X The automatic MDI/MDI-X feature, also known as auto crossover, eliminates the need to determine whether to use a straight cable or a crossover cable between the device and its link partner. The auto-sense function detects the MDI/ MDI-X pair mapping from the link partner, and assigns the MDI/MDI-X pair mapping of the device accordingly. Table 41 shows the device's 10/100 Mbps pin configuration assignments for MDI and MDI-X pin mapping. TABLE 4-1: MDI/MDI-X PIN DEFINITIONS Pin (RJ45 pair) MDI MDI-X 100BASE-TX 10BASE-T/Te 100BASE-TX 10BASE-T/Te TXP/M_A (1,2) TX+/- TX+/- RX+/- RX+/- RXP/M_A (3,6) RX+/- RX+/- TX+/- TX+/- Auto MDI/MDI-X is enabled by default. It can be disabled through the port control registers. If Auto MDI/MDI-X is disabled, the port control register can also be used to select between MDI and MDI-X settings. An isolation transformer with symmetrical transmit and receive data paths is recommended to support Auto MDI/MDI-X. 4.1.4 WAVE SHAPING, SLEW-RATE CONTROL, AND PARTIAL RESPONSE In communication systems, signal transmission encoding methods are used to provide the noise-shaping feature and to minimize distortion and error in the transmission channel. * For 100BASE-TX, a simple slew-rate control method is used to minimize EMI. * For 10BASE-T/Te, pre-emphasis is used to extend the signal quality through the cable. 4.1.5 AUTO-NEGOTIATION The device conforms to the auto-negotiation protocol as described by IEEE 802.3. Auto-negotiation allows each port to operate at either 10BASE-T/Te or 100BASE-TX by allowing link partners to select the best common mode of operation. During auto-negotiation, the link partners advertise capabilities across the link to each other and then compare their own DS00002391A-page 20 2017 Microchip Technology Inc. KSZ8567S capabilities with those they received from their link partners. The highest speed and duplex setting that is common to the two link partners is selected as the mode of operation. Auto-negotiation is also used to negotiate support for Energy Efficient Ethernet (EEE) via the next page feature. The following list shows the speed and duplex operation mode from highest to lowest priority. * * * * Priority 1: 100BASE-TX, full-duplex Priority 2: 100BASE-TX, half-duplex Priority 3: 10BASE-T/Te, full-duplex Priority 4: 10BASE-T/Te, half-duplex If the KSZ8567S link partner doesn't support auto-negotiation or is forced to bypass auto-negotiation, the KSZ8567S port sets its operating mode by observing the signal at its receiver. This is known as parallel detection, and allows the KSZ8567S to establish a link by listening for a fixed signal protocol in the absence of the auto-negotiation advertisement protocol. The auto-negotiation link-up process is shown in Figure 4-1. FIGURE 4-1: AUTO-NEGOTIATION AND PARALLEL OPERATION Auto-negotiation is enabled by default after power-up or hardware reset. Afterwards, auto-negotiation can be enabled or disabled via bit 12 of the PHY Basic Control Register. If auto-negotiation is disabled, the speed is set by bit 13 of the PHY Basic Control Register, and the duplex is set by bit 8. If the speed is changed on the fly, the link goes down and either auto-negotiation or parallel detection initiate until a common speed between the KSZ8567S and its link partner is re-established for a link. If link is already established and there is no change of speed on the fly, the changes (for example, duplex and pause capabilities) will not take effect unless either auto-negotiation is restarted through bit 9 of the PHY Basic Control Register, or a link-down to link-up transition occurs (i.e. disconnecting and reconnecting the cable). 2017 Microchip Technology Inc. DS00002391A-page 21 KSZ8567S After auto-negotiation is completed, the link status is updated in the PHY Basic Status Register, and the link partner capabilities are updated in the PHY Auto-Negotiation Link Partner Ability Register and PHY Auto-Negotiation Expansion Status Register. 4.1.6 QUIET-WIRE FILTERING Quiet-WIRE is a feature to enhance 100BASE-TX EMC performance by reducing both conducted and radiated emissions from the TXP/M signal pair. It can be used either to reduce absolute emissions, or to enable replacement of shielded cable with unshielded cable, all while maintaining interoperability with standard 100BASE-TX devices. Quiet-WIRE filtering is implemented internally, with no additional external components required. It is enabled or disabled for all PHYs at power-up and reset by a strapping option on the LED1_0 pin. Once the device is powered up, QuietWIRE can be enabled or disabled by writing to the appropriate control register. The default setting for Quiet-WIRE reduces emissions primarily above 60MHz, with less reduction at lower frequencies. Several dB of reduction is possible. Signal attenuation is approximately equivalent to increasing the cable length by 10 to 20 meters, thus reducing cable reach by that amount. For applications needing more modest improvement in emissions, the level of filtering can be reduced by writing to certain registers. 4.1.7 FAST LINK-UP Link up time is normally determined by the time it takes to complete auto-negotiation. Additional time may be added by the auto MDI/MDI-X feature. The total link up time from power-up or cable connect is typically a second or more. Fast Link-up mode significantly reduces 100BASE-TX link-up time by disabling both auto-negotiation and auto MDI/ MDI-X, and fixing the TX and RX channels. This mode is enabled or disabled by the LED2_1 strapping option. It is not set by registers, so fast link-up is available immediately upon power-up. Fast Link-up is available at power-up only for 100BASE-TX link speed, which is selected by strapping the LED4_0 pin high. Fast Link-up is also available for 10BASET/Te, but this link speed must first be selected via a register write. Fast Link-up is intended for specialized applications where both link partners are known in advance. The link must also be known so that the fixed transmit channel of one device connects to the fixed receive channel of the other device, and vice versa. The TX and RX channel assignments are determined by the MDI/MDI-X strapping option on LED2_0. If a device in Fast Link-up mode is connected to a normal device (auto-negotiate and auto-MDI/MDI-X), there will be no problems linking, but the speed advantage of Fast Link-up will not be realized. For more information on configuration straps, refer to Section 3.2.1, "Configuration Straps," on page 17. 4.1.8 LinkMD(R) CABLE DIAGNOSTICS The LinkMD(R) function utilizes Time Domain Reflectometry (TDR) to analyze the cabling for common cabling problems, such as open circuits, short circuits and impedance mismatches. LinkMD(R) works by sending a pulse of known amplitude and duration down the MDI or MDI-X pair, and then analyzing the shape of the reflected signal to determine the type of fault. The time duration for the reflected signal to return provides the approximate distance to the cabling fault. The LinkMD(R) function processes this TDR information and presents it as a numerical value that can be translated to a cable distance. 4.1.9 LinkMD(R)+ ENHANCED DIAGNOSTICS: RECEIVE SIGNAL QUALITY INDICATOR The device provides a receive Signal Quality Indicator (SQI) feature, which indicates the relative quality of the 100BASE-TX receive signal. It approximates a signal-to-noise ratio, and is affected by cable length, cable quality, and coupled of environmental noise. The raw SQI value is available for reading at any time from the SQI register. A lower value indicates better signal quality, while a higher value indicates worse signal quality. Even in a stable configuration in a low-noise environment, the value read from this register may vary. The value should therefore be averaged by taking multiple readings. The update interval of the SQI register is 2s, so measurements taken more frequently than 2s will be redundant. In a quiet environment, 6 to 10 readings are suggested for averaging. In a noisy environment, individual readings are unreliable, so a minimum of 30 readings are suggested for averaging. The SQI circuit does not include any hysteresis. 4.1.10 REMOTE PHY LOOPBACK This loopback mode checks the line (differential pairs, transformer, RJ-45 connector, Ethernet cable) transmit and receive data paths between the KSZ8567S and its Ethernet PHY link partner, and is supported for 10/100 Mbps at fullduplex. DS00002391A-page 22 2017 Microchip Technology Inc. KSZ8567S The loopback data path is shown in Figure 4-2 and functions as follows: * The Ethernet PHY link partner transmits data to the KSZ8567S PHY port. * Data received at the external pins of the PHY port is looped back without passing through the MAC and internal switch fabric. * The same KSZ8567S PHY port transmits data back to the Ethernet PHY link partner. FIGURE 4-2: REMOTE PHY LOOPBACK Device PHY Port N (1-5) RJ-45 10/100 PHY MAC Switch Fabric CAT-5 (UTP) RJ-45 Ethernet PHY Link Partner The following programming steps and register settings are for remote PHY loopback mode for 100BASE-TX Mode, and 10BASE-T Mode. * 100BASE-TX Mode - Set Port N (1-5), PHY Auto-Negotiation Advertisement Register = 0x0181 - Set Port N (1-5), PHY Remote Loopback Register = 0x01F0 - Set Port N (1-5), PHY Basic Control Register = 0x3300 * 10BASE-T Mode - Set Port N (1-5), PHY Auto-Negotiation Advertisement Register = 0x0061 - Set Port N (1-5), PHY Remote Loopback Register = 0x01F0 - Set Port N (1-5), PHY Basic Control Register = 0x3300 2017 Microchip Technology Inc. DS00002391A-page 23 KSZ8567S 4.2 LEDs Each PHY port has two programmable LED output pins, LEDx_0 and LEDx_1, to indicate the PHY link and activity status. Two different LED modes are available. The LED mode can be changed individually for each PHY port by writing to the PHY Mode bit in the PHY indirect register: MMD 2, address 0, bit 4: * 1 = Single-LED Mode * 0 = Tri-Color Dual-LED Mode (Default) Each LED output pin can directly drive an LED with a series resistor (typically 220 to 470). LED outputs are activelow. 4.2.1 SINGLE-LED MODE In single-LED mode, the LEDx_1 pin indicates the link status while the LEDx_0 pin indicates the activity status, as shown in Figure 4-2. TABLE 4-2: SINGLE-LED MODE PIN DEFINITION LED Pin Pin State Pin LED Definition H OFF Link Off L ON Link On (any speed) LEDx_1 LEDx_0 4.2.2 Link/Activity H OFF No Activity Toggle Blinking Activity (RX,TX) TRI-COLOR DUAL-LED MODE In tri-color dual-LED mode, the link and activity status are indicated by the LEDx_1 pin for 1000BASE-T; by the LEDx_0 pin for 100BASE-TX; and by both LEDx_1 and LEDx_0 pins, working in conjunction, for 10BASE-T. This behavior is summarized in Figure 4-3. TABLE 4-3: TRI-COLOR DUAL-LED MODE PIN DEFINITION LED Pin (State) LED Pin (Definition) Link/Activity LEDx_1 LEDx_0 LEDx_1 LEDx_0 H H OFF OFF Link off L H ON OFF 1000Mbps Link / No Activity Toggle H Blinking OFF 1000Mbps Link / Activity (RX,TX) H L OFF ON 100Mbps Link / No Activity H Toggle OFF Blinking L L ON ON Toggle Toggle Blinking Blinking DS00002391A-page 24 100Mbps Link / Activity (RX,TX) 10Mbps Link / No Activity 10Mbps Link / Activity (RX,TX) 2017 Microchip Technology Inc. KSZ8567S 4.3 4.3.1 Media Access Controller (MAC) MAC OPERATION The device strictly abides by IEEE 802.3 standards to maximize compatibility. Additionally, there is an added MAC filtering function to filter unicast packets. The MAC filtering function is useful in applications, such as VoIP, where restricting certain packets reduces congestion and thus improves performance. The transmit MAC takes data from the egress buffer and creates full Ethernet frames by adding the preamble and the start-of-frame delimiter ahead of the data, and generates the FCS that is appended to the end of the frame. It also sends flow control packets as needed. The receive MAC accepts data via the integrated PHY or via the SGMII/MII/RMII/RGMII interface. It decodes the data bytes, strips off the preamble and SFD of each frame. The destination and source addresses and VLAN tag are extracted for use in filtering and address/ID lookup, and the MAC also calculates the CRC of the received frame, which is compared to the FCS field. The MAC can discard frames that are the wrong size, that have an FCS error, or when the source MAC address matches the Switch MAC address. The receive MAC also implements the Wake on LAN (WoL) feature. This system power saving feature is described in detail in the Section 4.10, "Power Management". MIB statistics are collected in both receive and transmit directions. 4.3.2 INTER-PACKET GAP (IPG) If a frame is successfully transmitted, then the minimum 96-bit time for IPG is specified as being between two consecutive packets. If the current packet is experiencing collisions, the minimum 96-bit time for IPG is specified as being from carrier sense (CRS) to the next transmit packet. 4.3.3 BACK-OFF ALGORITHM The device implements the IEEE standard 802.3 binary exponential back-off algorithm in half-duplex mode. After 16 collisions, the packet is dropped. 4.3.4 LATE COLLISION If a transmit packet experiences collisions after 512 bit times of the transmission, the packet is dropped. 4.3.5 LEGAL PACKET SIZE On all ports, the device discards received packets smaller than 64 bytes (excluding VLAN tag, including FCS) or larger than the maximum size. The default maximum size is the IEEE standard of 1518 bytes, but can be programmed to 2000 bytes. Ports operating at 1000Mbps may be programmed to accept jumbo packets up to 9000 bytes, but for performance reasons it is recommended that no more than two ports be enabled simultaneously for jumbo packets. 4.3.6 FLOW CONTROL The device supports standard MAC Control PAUSE (802.3x flow control) frames in both the transmit and receive directions for full-duplex connections. In the receive direction, if a PAUSE control frame is received on any port, the device will not transmit the next normal frame on that port until the timer, specified in the PAUSE control frame, expires. If another PAUSE frame is received before the current timer expires, the timer will then update with the new value in the second PAUSE frame. During this period (while it is flow controlled), only flow control packets from the device are transmitted. In the transmit direction, the device has intelligent and efficient ways to determine when to invoke flow control and send PAUSE frames. The flow control is based on availability of the system resources, including available buffers, available transmit queues and available receive queues. The device issues a PAUSE frame containing the maximum pause time defined in IEEE standard 802.3x. Once the resource is freed up, the device sends out another flow control frame with zero pause time to turn off the flow control (turn on transmission to the port). A hysteresis feature is provided to prevent the flow control mechanism from being constantly activated and deactivated. 4.3.7 HALF-DUPLEX BACK PRESSURE A half-duplex back pressure option (non-IEEE 802.3 standard) is also provided. The activation and deactivation conditions are the same as in full-duplex mode. If back pressure is required, the device sends preambles to defer the other stations' transmission (carrier sense deference). 2017 Microchip Technology Inc. DS00002391A-page 25 KSZ8567S To avoid jabber and excessive deference (as defined in the 802.3 standard), after a certain time, the device discontinues the carrier sense and then raises it again quickly. This short silent time (no carrier sense) prevents other stations from sending out packets thus keeping other stations in a carrier sense deferred state. If the port has packets to send during a back pressure situation, the carrier sense type back pressure is interrupted and those packets are transmitted instead. If there are no additional packets to send, carrier sense type back pressure is reactivated again until chip resources free up. If a collision occurs, the binary exponential back-off algorithm is skipped and carrier sense is generated immediately, thus reducing the chance of further collision and carrier sense is maintained to prevent packet reception. To ensure no packet loss in 10BASE-T/Te or 100BASE-TX half-duplex modes, the user must enable the following: * No excessive collision drop (Switch MAC Control 1 Register) * Back pressure (Port MAC Control 1 Register) 4.3.8 FLOW CONTROL AND BACK PRESSURE REGISTERS Table 4-4 provides a list of flow control and back pressure related registers. TABLE 4-4: FLOW CONTROL AND BACK PRESSURE REGISTERS Registers Description LED Configuration Strap Register LED configuration strap settings. (LED1_1 enables flow control and back pressure) Switch MAC Address 0 Register through Switch MAC Address 5 Register Switch's MAC address, used as source address of PAUSE control frames Switch MAC Control 0 Register "Aggressive back-off" enable Switch MAC Control 1 Register BP mode, "Fair mode" enable, "no excessive collision drop" enable Switch MAC Control 4 Register Pass PAUSE control frames Port Status Register Flow control enable (per port) PHY Auto-Negotiation Advertisement Register PHY - flow control advertisement (per port) Port MAC Control 1 Register Half-duplex back pressure enable (per port) Port Ingress Rate Limit Control Register Ingress rate limit flow control enable (per port) Port Control 0 Register Drop mode (per port) 4.3.9 BROADCAST STORM PROTECTION The device has an intelligent option to protect the switch system from receiving too many broadcast packets. As the broadcast packets are forwarded to all ports except the source port, an excessive number of switch resources (bandwidth and available space in transmit queues) may be utilized. The device has the option to include "multicast packets" for storm control. The broadcast storm rate parameters are programmed globally, and can be enabled or disabled on a per port basis. The rate is based on a 50ms interval for 100BASE-TX and a 500ms interval for 10BASE-T/Te. At the beginning of each interval, the counter is cleared to zero and the rate limit mechanism starts to count the number of bytes during the interval. The rate definition is described in control registers. The default setting equates to a rate of 1%. 4.3.10 SELF-ADDRESS FILTERING Received packets can be filtered (dropped) if their source address matches the device's MAC address. This feature is useful for automatically terminating packets once they have traversed a ring network and returned to their source. It can be enabled on a per-port basis via the Switch Lookup Engine Control 1 Register and Port Control 2 Register. DS00002391A-page 26 2017 Microchip Technology Inc. KSZ8567S 4.4 Switch 4.4.1 SWITCHING ENGINE A high-performance switching engine is used to move data to and from the MAC's packet buffers. It operates in store and forward mode, while an efficient switching mechanism reduces overall latency. The switching engine has a 256KByte internal frame buffer that is shared between all the ports. For the majority of switch functions, all of the data ports are treated equally. However, a few functions such as IGMP snooping, 802.1X, forwarding invalid VLAN packets, etc., give special recognition to the host port. Any port (but most commonly port 6 or port 7) may be assigned as the host port by enabling tail tagging mode for that port. Only one port may be a host port. When a switch receives a non-error packet, it checks the packet's destination MAC address. If the address is known, the packet is forwarded to the output port that is associated with the destination MAC address. The following paragraphs describe the key functions of destination address lookup and source address learning. These processes may be combined with VLAN support and other features, which are described in the subsequent sub-sections. 4.4.2 ADDRESS LOOKUP Destination address lookup is performed in three separate internal address tables in the device: 1. 2. 3. Address Lookup (ALU) Table: 4K dynamic + static entries Static Address Table: 16 static entries Reserved Multicast Address Table: 8 pre-configured static entries 4.4.2.1 Address Lookup (ALU) Table The Address Lookup (ALU) Table stores MAC addresses and their associated information. This table holds both dynamic and static entries. Dynamic entries are created automatically in hardware, as described in Section 4.4.2.4, "Learning". Static entries are created by management software. This table is a 4-way associative memory, with 1K buckets, for a total of 4K entries. A hash function translates the received packet's MAC address (and optionally the FID) into a 10-bit index for accessing the table. At each bucket are four fully-associative address entries. All four entries are simultaneously compared to the MAC address (plus optional FID) for a possible match. Three options are available for the hashing function, as described in Table 4-5. If VLAN is enabled (802.1Q VLAN Enable bit in the Switch Lookup Engine Control 0 Register), the VLAN group (FID) is included in the hashing function along with the MAC address. If VLAN is not enabled the hashing function is applied to MAC address and the FID in the default VLAN (VID=1) which is 0. TABLE 4-5: ADDRESS LOOKUP TABLE HASHING OPTIONS HASH_OPTION (Switch Lookup Engine Control 0 Register) Description 01b (Default) A hash algorithm based on the CRC of the MAC address plus FID. The hash algorithm uses the CRC-CCITT polynomial. The input to the hash is reduced to a 16-bit CRC hash value. Bits [9:0] of the hash value plus (binary addition) 7-bit FID (zero extended on the left) are used as an index to the table. The CRC-CCITT polynomial is: X16+X12+X5+1. 10b An XOR algorithm based on 16 bits of the XOR of the triple-folded MAC address. Bits [9:0] of the XOR value plus 7-bit FID (left-extended) are used to index the table. 00b or 11b A direct algorithm. The 10 least significant bits of the MAC address plus 7 bit FID are used to index the table. 4.4.2.2 Static Address Table The 16-entry Static Address Table is typically used to hold multicast addresses, but is not limited to this. As with static entries in the ALU table, entries in the Static Address Table are created by management software. It serves the same function as static entries that are created in the ALU table, so its use is optional. 2017 Microchip Technology Inc. DS00002391A-page 27 KSZ8567S 4.4.2.3 Reserved Multicast Address Table The Reserved Multicast Address Table holds 8 pre-configured address entries, as defined in Table 4-6. This table is an optional feature that is disabled at power-on. If desired, the forwarding ports may be modified. TABLE 4-6: Group RESERVED MULTICAST ADDRESS TABLE Address MAC Group Address Function Default PORT FORWARD Value Default Forwarding Action (defines forwarding port: P7...P1) 0 (01-80-C2-00)-00-00 Bridge Group Data 100_0000 Forward only to the highest numbered port (default host port) 1 (01-80-C2-00)-00-01 MAC Control Frame 000_0000 (typically flow control) Drop MAC flow control 2 (01-80-C2-00)-00-03 802.1X Access Control 100_0000 Forward to highest numbered port 3 (01-80-C2-00)-00-10 Bridge Management 111_1111 Flood to all ports 4 (01-80-C2-00)-00-20 GMRP 011_1111 Flood to all ports except highest numbered port 5 (01-80-C2-00)-00-21 GVRP 011_1111 Flood to all ports except highest numbered port 6 (01-80-C2-00)-00-02, (01-80-C2-00)-00-04 - (01-80-C2-00)-00-0F 100_0000 Forward to highest numbered port 7 (01-80-C2-00)-00-11 (01-80-C2-00)-00-1F, (01-80-C2-00)-00-22 (01-80-C2-00)-00-2F 011_1111 Flood to all ports except highest numbered port If a match is found in one of the tables, then the destination port is read from that table entry. If a match is found in more than one table, static entries will take priority over dynamic entries. 4.4.2.4 Learning The internal lookup engine updates the ALU table with a new dynamic entry if the following conditions are met: * * * * The received packet's source address (SA) does not exist in the lookup table. The received packet has no errors, and the packet size is of legal length. The received packet has a unicast SA. If VLAN is enabled, the received packet must belong to the indicated VLAN domain (FID). The lookup engine inserts the qualified SA into the table, along with the port number and time stamp. If all four table entries are valid, the oldest of the (up to four) dynamic entries may be deleted to make room for the new entry. Static entries are never deleted by the learning process. If all four entries are static entries, the address is not learned but an interrupt is generated and the table index number is made available to the interrupt service routine. 4.4.2.5 Migration The internal lookup engine also monitors whether a station has moved. If a station has moved, it updates the ALU table accordingly. Migration happens when the following conditions are met: * The received packet's SA is in the table but the associated source port information is different. * The received packet has no receiving errors, and the packet size is of legal length. The lookup engine updates the existing record in the table with the new source port information. DS00002391A-page 28 2017 Microchip Technology Inc. KSZ8567S 4.4.2.6 Aging The lookup engine updates the time stamp information of a dynamic record in the ALU table whenever the corresponding SA appears. The time stamp is used in the aging process. If a record is not updated for a period of time, the lookup engine removes the record from the table. The lookup engine constantly performs the aging process and continuously removes aging records. The aging period is about 300 seconds (75 seconds) and can be configured longer or shorter (1 second to 30 minutes). This feature can be enabled or disabled. Static entries are exempt from the aging process. 4.4.2.7 Forwarding The device forwards packets using the algorithm that is depicted in Figure 4-3. Figure 4-3 shows stage one of the forwarding algorithm where the search engine looks up the VLAN ID, static table, and dynamic table for the destination address, and comes up with "port to forward 1" (PTF1). PTF1 is then further modified by spanning tree, IGMP snooping, port mirroring, and port VLAN processes. The ACL process works in parallel with the flow outlined above. The authentication and ACL processes have the highest priority in the forwarding process, and the ACL result may override the result of the above flow. The output of the ACL process is the final "port-to-forward 2" (PTF2) destination port(s). The device will not forward the following packets: * Error packets: These include framing errors, frame check sequence (FCS) errors, alignment errors, and illegal size packet errors. * MAC Control PAUSE frames: The device intercepts these packets and performs full duplex flow control accordingly. * "Local" packets: Based on destination address (DA) lookup. If the destination port from the lookup table matches the port from which the packet originated, the packet is defined as "local". * In-Band Management packets. FIGURE 4-3: PACKET FORWARDING PROCESS FLOWCHART Start PTF1 no PTF1=NULL VLAN ID Valid? - Search VLAN table - Ingress VLAN filtering -Discard NPVID check Spanning Tree Process - Check receiving port's receive enable bit - Check destination port's transmit enable bit - Check whether packets are special (BPDU) yes Get PTF1 from Static Array found Search Static Array Search based on DA or DA+FID IGMP / MLD Process - IGMP / MLD packets are forwarded to Host port - Process does not apply to packets received at Host port Port Mirror Process - RX Mirror - TX Mirror - RX or TX Mirror - RX and TX Mirror not found Get PTF1 from Address Table found Search Address Look-up Table Search based on DA+FID not found Get PTF1 from VLAN Table Port Authentication & ACL PTF1 PTF2 2017 Microchip Technology Inc. DS00002391A-page 29 KSZ8567S 4.4.2.8 Lookup Engine Registers Table 4-7 provides a list of lookup engine related registers. TABLE 4-7: LOOKUP ENGINE REGISTERS Registers Description Global Interrupt Status Register, Global Interrupt Mask Register Top level LUE interrupt Switch Lookup Engine Control 0 Register, Switch Lookup Engine Control 1 Register, Switch Lookup Engine Control 2 Register, Switch Lookup Engine Control 3 Register Misc. Address Lookup Table Interrupt Register, Address Lookup Table Mask Register Low level LUE interrupts Address Lookup Table Entry Index 0 Register, Address Lookup Table Entry Index 1 Register Access failure address/index ALU Table Index 0 Register, ALU Table Index 1 Register, ALU Table Access Control Register, Static Address and Reserved Multicast Table Control Register, ALU / Static Address Table Entry 1 Register, ALU / Static Address / Reserved Multicast Table Entry 2 Register, ALU / Static Address Table Entry 3 Register, ALU / Static Address Table Entry 4 Register Address table access registers 4.4.3 IEEE 802.1Q VLAN Virtual LAN is a means of segregating a physical network into multiple virtual networks whereby traffic may be confined to specific subsets of the greater network. IEEE 802.1Q defines a VLAN protocol using a 4-byte tag that is added to the Ethernet frame header. The device supports port-based and tag-based VLANs, including tagging, un-tagging, forwarding and filtering. 4.4.3.1 Non-Tag Port-Based VLAN The simplest VLAN method establishes forwarding restrictions on a port-by-port basis without using VLAN tags. There is a register for each ingress port that is used to specify the allowed forwarding ports. An incoming packet is restricted from being forwarded to any egress port that is disallowed for that ingress port. The settings are made in the Port Control 1 Register. This function is always enabled; it is not enabled and disabled by the 802.1Q VLAN Enable bit in the Switch Lookup Engine Control 0 Register. The default setting is to allow all ingress-to-egress port paths. 4.4.3.2 Tag-Based VLAN When 802.1Q VLAN is enabled, an internal VLAN Table with 4k entries is used to a store port membership list, VLAN group ID (FID) and additional information relating to each VLAN. This table must be set up by an administrator prior to enabling 802.1Q VLAN. Enabling is done by setting the 802.1Q VLAN Enable bit in the Switch Lookup Engine Control 0 Register. In 802.1Q VLAN mode, the lookup process starts with VLAN Table lookup, using the tag's VID as the address. The first step is to determine whether the VID is valid. If the VID is not valid, the packet is dropped and its address is not learned. Alternatively, unknown VID packets may be forwarded to pre-defined ports or to the host port. If the VID is valid, the FID is retrieved for further lookup. The FID + Destination Address (hashed(DA) + FID) are used to determine the destination port. The FID + Source Address (hashed(SA) + FID) are used for address learning (see Table 4-9 and Table 4-10). DS00002391A-page 30 2017 Microchip Technology Inc. KSZ8567S The hashed(DA) + FID are hashed and used for forwarding lookup in the Address Lookup and Static Address Tables. For a successful address table lookup, the FID fields must also match. If the match fails, the packet is broadcast to all the VLAN port members defined in the VLAN Table entry. If there is a match and egress VLAN filtering is enabled, the packet is forwarded to those ports that are in both the address table port forwarding list and the VLAN table port membership list. A similar address table lookup is performed using the hashed(SA) + FID. If the lookup fails, the FID and SA are learned. If a non-tagged or null-VID-tagged packet is received, the ingress port default VID (Port Default Tag 0 Register and Port Default Tag 1 Register) is used for lookup. Table 4-8 details the forwarding and discarding actions that are taken for the various VLAN scenarios. The first entry in the table is explained by the fact that VLAN Table lookup is enabled even when 802.1Q VLAN is not enabled. Notice that in the Port Default Tag 0 Register and Port Default Tag 1 Register, the port default VID is 1 for each port. Correspondingly, the VLAN port membership list in the VLAN Table entry for VID=1 is pre-configured at power-on to all ones. This provides the standard Ethernet switch behavior of broadcasting all packets with unknown destination address. If the VLAN table entry # 1 is changed, or if the port default VID is changed, this may affect the forwarding action for "unknown packets" even when VLAN is not enabled. It should also be noted that the default values of the Egress VLAN Filtering bits are zero. These bits are zero only for backwards compatibility with previous "KSZ" switches. The resulting switch behavior, in the event of a successful VLAN and ALU lookups, is to forward the packet to the ports in the address table port forwarding list, without regard to the VLAN port membership list. It is suggested that the Egress VLAN Filtering bits be set to one so that the VLAN port membership list from the VLAN Table will be used to qualify the forwarding determined from the address lookup. TABLE 4-8: VLAN Enable (Note 4-1) VLAN FORWARDING VLAN Match/ Valid (Note 4-2) Forward Option (Note 4-3) Egress VLAN Filtering Unknown VID Forward Drop Invalid VID ALU Match/ Valid (Note 4-4) (Note 4-5) (Note 4-6) (Note 4-7) Action 0 X X X X X No Forward to port membership list of default VID in LAN table 0 X X X X X Yes Forward to Address Lookup port forwarding list 1 No X X 0 0 X Forward to host port 1 No X X 0 (def) 1 (def) X Discard 1 No X X 1 X X Forward to Unknown VID packet forward port list 1 Yes 0 X X X No Broadcast: Forward to VLAN table port membership list (PORT FORWARD) Multicast: Forward to Unknown Multicast ports if UM is enabled. Else, forward to VLAN table port membership list. Unicast: Forward to Unknown Unicast ports if UU is enabled. Else forward to VLAN table port membership list. 1 Yes 0 0 (def) X X Yes Forward to address table lookup port forwarding list 1 Yes 0 1 X X Yes Forward to address table lookup port forwarding list & VLAN table port membership list (bitwise AND) 1 Yes 1 X X X Yes Forward to VLAN table port membership list 2017 Microchip Technology Inc. DS00002391A-page 31 KSZ8567S Note: "(def)" indicates the default power-up value. Note 4-1 VLAN Enable is bit 7 in the Switch Lookup Engine Control 0 Register Note 4-2 VLAN Match/Valid indicates when the VLAN Table entry is valid Note 4-3 Forward Option is a bit in the VLAN Table Entry 0 Register Note 4-4 Egress VLAN Filtering are bits 5 and 4 in the Switch Lookup Engine Control 2 Register Note 4-5 Unknown VID Forwarding is in the Unknown VLAN ID Control Register Note 4-6 Drop Invalid VID is bit 6 in the Switch Lookup Engine Control 0 Register Note 4-7 ALU Match/Valid indicates when the Address Lookup is a success Table 4-9 describes in more detail the address lookup process that follows the VLAN Table lookup. Lookup occurs in both the Address Lookup Table and the Static Address Table simultaneously, and the resulting action depends on the results of the two lookups. TABLE 4-9: HASHED(DA) + FID LOOKUP IN VLAN MODE DA Found Use FID Flag? in Static (Static MAC Table) MAC Table? FID Match? DA+FID Found in ALU Table? Action No Don't Care Don't Care No Lookup has failed. Broadcast to the membership ports defined in the VLAN Table No Don't Care Don't Care Yes Send to the destination port defined in the Address Lookup (ALU) Table Yes 0 Don't Care Don't Care Yes 1 No No Lookup has failed. Broadcast to the membership ports defined in the VLAN Table. Yes 1 No Yes Send to the destination port defined in the Address Lookup (ALU) Table Yes 1 Yes Don't Care Send to the destination port(s) defined in the Static Address Table Send to the destination port(s) defined in the Static Address Table A source address (SA) lookup is also performed in the Address Lookup Table. SA lookup also performs SA filtering and MAC priority when the address is hit. Table 4-10 describes how learning is performed in the Address Lookup Table when a successful VLAN table lookup has been done and the no matching static entry is found in the Address Lookup Table or the Static Address Table. TABLE 4-10: HASHED(SA) + FID LOOKUP IN VLAN MODE FID + SA Found in Address Lookup (ALU) Table? Action No Learn and add FID + SA to the Address Lookup (ALU) Table Yes Update time stamp DS00002391A-page 32 2017 Microchip Technology Inc. KSZ8567S 4.4.3.2.1 Tag Insertion and Removal Tag insertion is enabled on all ports when the VLAN feature is enabled. At the ingress port, untagged packets are tagged with the ingress port's default tag. The default tag is separately programmable for each port. The switch does not add tags to already tagged packets unless double tagging is enabled. At the egress port, tagged packets will have their 802.1Q VLAN tags removed if un-tagging is enabled in the VLAN table entry. This feature is controlled on a per-port basis. Untagged packets will not be modified if 802.1Q is enabled. 4.4.3.2.2 Double Tagging The switch supports double tagging, also known as Q-in-Q or VLAN stacking. This feature can be used for service providers to append a second VLAN tag in addition to a first VLAN tag applied by the customer. VLAN support can be enabled either with or without double tagging. When double tagging is enabled, the outer tag is recognized and is used for VLAN and address lookup instead of the inner tag. The outer tag precedes the inner tag in the frame header: the outer tag is located immediately after the source address, and contains a different Tag Protocol Identifier (TPID) value than the inner tag. Additional controls are available for full control of the VLAN function. Some of these features can be enabled on a perport basis, while others are global: * * * * * * * * * Ingress VLAN Filtering: Discard packet if VID port membership in VLAN table does not include the ingress port. Discard non PVID Packet: Discard packet if VID does not match the ingress port default VID. Discard un-tagged Packet: Discard any received packet without a tag. Drop tag: Drops the packet if it is VLAN tagged. Unknown VID Forward: Forward to a fixed set of ports if VLAN lookup fails. Drop unknown VID: Additional options for unknown VID packets: discard or forward to the host port. Null VID Replacement: Replace a null VID with the ingress port default VID. PVID Replacement: Replace a non-null VID with the ingress port default VID. Double Tag Mcast Trap: In double tag mode, trap all reserved multicast packets and forward to the host port. 4.4.3.3 VLAN Registers Table 4-11 provides a list of VLAN related registers. TABLE 4-11: VLAN REGISTERS Registers Description Switch Operation Register Double tag enable Switch Lookup Engine Control 0 Register VLAN enable; Drop invalid VID frames Switch Lookup Engine Control 2 Register Trap double tagged MC frames; Dynamic & status egress VLAN filtering Unknown VLAN ID Control Register Forward unknown VID Switch MAC Control 2 Register Null VID replacement with PVID at egress VLAN Table Entry 0 Register, VLAN Table Entry 1 Register, VLAN Table Entry 2 Register, VLAN Table Index Register, VLAN Table Access Control Register Read/write access to the VLAN table Port Default Tag 0 Register, Port Default Tag 1 Register Port default tag Port Ingress MAC Control Register Drop non-VLAN frames; Tag drop Port Transmit Queue PVID Register PVID replacement at egress Port Control 2 Register VLAN table lookup for VID=0; Ingress VLAN filtering; PVID mismatch discard 2017 Microchip Technology Inc. DS00002391A-page 33 KSZ8567S 4.4.4 QUALITY-OF-SERVICE (QOS) PRIORITY SUPPORT The device provides quality-of-service (QoS) for applications such as VoIP. There are multiple methods for assigning priority to ingress packets. Depending on the packet prioritization method, the packet priority levels are mapped to the egress queues for each port. Each port can be configured for 1, 2, and 4 egress queues, which are prioritized. The default is 1 queue per port. When configured for 4 priority queues, Queue 3 is the highest priority queue and Queue 0 is the lowest priority. Likewise, for a 2-queue configuration, Queue 1 is the highest priority queue. If a port is not configured as 2 or 4 queues, then high priority and low priority packets have equal priority in the single transmit queue. There is an additional option for every port to select either to always deliver packets from the highest priority queue first, or use weighted round robin queuing amongst the multiple queues. This is described later in Section 4.4.13, "Scheduling and Rate Limiting". 4.4.4.1 Port-Based Priority With port-based priority, each ingress port is individually classified as a specific priority level. All packets received at the high-priority receiving port are marked as high priority and are sent to the high-priority transmit queue if the corresponding transmit queue is split into 2 or 4 queues. 4.4.4.2 IEEE 802.1p-Based Priority For IEEE 802.1p-based priority, the device examines the ingress packets to determine whether they are tagged. If tagged, the 3-bit PCP priority field in the VLAN tag is retrieved and used to look up the "priority mapping" value. The "priority mapping" value is programmable. Figure 4-4 illustrates how the 802.1p priority field is embedded in the 802.1Q VLAN tag. FIGURE 4-4: 4.4.4.3 802.P PRIORITY FIELD FORMAT IEEE 802.1p Priority Field Re-Mapping This is a QoS feature that allows the device to set the "User Priority Ceiling" at any ingress port. If the ingress packet's priority field has a higher priority value than the default tag's priority field of the ingress port, the packet's priority field is replaced with the default tag's priority field. 4.4.4.4 DiffServ (DSCP) Priority (IP) DiffServ-based priority from the DSCP field in the IP header can be used to determine packet priority. The 6-bit DSCP value is used as an index to a set of registers which translate the 6-bit DSCP value to a 2-bit value that specifies one of the 4 (or 2) queues. These registers are fully programmable. 4.4.4.5 ACL Priority The Access Control List (ACL) Filtering feature can also be used to assign priority to received packets. This is discussed in Section 4.4.18, "Access Control List (ACL) Filtering". DS00002391A-page 34 2017 Microchip Technology Inc. KSZ8567S 4.4.5 4.4.5.1 TRAFFIC CONDITIONING & POLICING Two Rate Three Color Marker The Two Rate Three Color Marker meters an IP packet stream and marks its packets green, yellow, or red. A packet is marked red if it exceeds the Peak Information Rate (PIR). Otherwise, it is marked either yellow or green depending on whether it exceeds or doesn't exceed the Committed Information Rate (CIR). The Meter operates in one of two modes. In the Color-Blind mode, the Meter assumes that the packet stream is uncolored. In the Color-Aware mode, the Meter assumes that some preceding entity has pre-colored the incoming packet stream so that each packet is green, yellow, or red. The Marker (re)colors an IP packet according to the results of the Meter. 4.4.5.2 Weighted Random Early Detection (WRED) The WRED feature monitors the average queue size of packet memory and ingress queue size of each traffic class, and drops packets based on memory and queue utilization. If the buffers are almost empty, all incoming traffic is accepted. As the buffer utilization increases, the probability for dropping an incoming packet also increases. WRED is intended to avoid the problem of global synchronization. Global synchronization can occur when a switch becomes congested and begins dropping incoming packets all at once. For TCP streams, packet drops invoke the TCP congestion control mechanism, which reduce the transmission rate until there are no more packet drops. If there are many TCP streams and their congestion control mechanisms act in unison, this can cause an undesirable oscillation in traffic rates. By selectively dropping some packets early rather than waiting until the buffer is full, WRED avoids dropping large numbers of packets at once and minimizes the chances of global synchronization. The packet drop probability is based on the minimum threshold, maximum threshold, and a probability multiplier. When the average queue depth is above the minimum threshold, packets start getting dropped. The rate of packet drop increases linearly as the average queue size increases until the average queue size reaches the maximum threshold. The probability multiplier is the fraction of packets dropped when the average queue depth is at the maximum threshold. When the average queue size is above the maximum threshold, all packets are dropped. AVB traffic streams (SR streams) can be exempted from WRED policing. 4.4.6 SPANNING TREE SUPPORT To support spanning tree, one port is the designated port for the host processor, which is defined as the port for which tail tagging is enabled. Each of the other ports can be configured in one of the five spanning tree states via "transmit enable", "receive enable" and "learning disable" register bits. Table 4-12 shows the setting and software actions taken for each of the five spanning tree states. TABLE 4-12: SPANNING TREE STATES Disable State Port Setting Software Action The port should not forward or transmit enable = 0 receive any packets. receive enable = 0 Learning is disabled. learning disable = 1 The processor should not send any packets to the port. The switch may still send specific packets to the processor (packets that match some entries in the "Static MAC Table" with "overriding bit" set) and the processor should discard those packets. Address learning is disabled on the port in this state. Blocking State Software Action Port Setting Only packets to the processor transmit enable = 0 are forwarded. receive enable = 0 Learning is disabled. learning disable = 1 2017 Microchip Technology Inc. The processor should not send any packets to the port(s) in this state. The processor should program the "Static MAC Table" with the entries that it needs to receive (for example, BPDU packets). The "overriding" bit should also be set so that the switch will forward those specific packets to the processor. Address learning is disabled on the port in this state. DS00002391A-page 35 KSZ8567S TABLE 4-12: SPANNING TREE STATES (CONTINUED) Listening State Port Setting Software Action Only packets to and from the processor are forwarded. Learning is disabled. transmit enable = 0 receive enable = 0 learning disable = 1 The processor should program the "Static MAC Table" with the entries that it needs to receive (for example, BPDU packets). The "overriding" bit should be set so that the switch will forward those specific packets to the processor. The processor may send packets to the port(s) in this state. Address learning is disabled on the port in this state. Learning State Port Setting Software Action Only packets to and from the processor are forwarded. Learning is enabled. transmit enable = 0 receive enable = 0 learning disable = 0 The processor should program the "Static MAC Table" with the entries that it needs to receive (for example, BPDU packets). The "overriding" bit should be set so that the switch will forward those specific packets to the processor. The processor may send packets to the port(s) in this state. Address learning is enabled on the port in this state. Forwarding State Port Setting Software Action Packets are forwarded and received normally. Learning is enabled. transmit enable = 1 receive enable = 1 learning disable = 0 The processor programs the "Static MAC Table" with the entries that it needs to receive (for example, BPDU packets). The "overriding" bit is set so that the switch forwards those specific packets to the processor. The processor can send packets to the port(s) in this state. Address learning is enabled on the port in this state. 4.4.7 RAPID SPANNING TREE SUPPORT There are three operational states assigned to each port for the Rapid Spanning Tree Protocol (RSTP): 1. 2. 3. Discarding State Learning State Forwarding State 4.4.7.1 Discarding State Discarding ports do not participate in the active topology and do not learn MAC addresses. * Discarding state: the state includes three states of the disable, blocking and listening of STP. * Port setting: transmit enable = "0", receive enable = "0", learning disable = "1". * Software action: The host processor should not send any packets to the port. The switch may still send specific packets to the processor (packets that match some entries in the static table with "overriding bit" set) and the processor should discard those packets. When the port's learning capability (learning disable = '1') is disabled, port related entries in the ALU table and static MAC table can be rapidly flushed. 4.4.7.2 Learning State Ports in "learning state" learn MAC addresses, but do not forward user traffic. * Learning State: Only packets to and from the host processor are forwarded. Learning is enabled. * Port setting for Learning State: transmit enable = "0", receive enable = "0", learning disable = "0". * Software action: The processor should program the Static Address Table with the entries that it needs to receive (e.g., BPDU packets). The "overriding" bit should be set so that the switch will forward those specific packets to the processor. The processor may send packets to the port(s) in this state (see Section 4.4.9, "Tail Tagging Mode" for details). Address learning is enabled on the port in this state. 4.4.7.3 Forwarding State Ports in "forwarding states" fully participate in both data forwarding and MAC learning. * Forwarding state: Packets are forwarded and received normally. Learning is enabled. * Port setting: transmit enable = "1", receive enable = "1", learning disable = "0". DS00002391A-page 36 2017 Microchip Technology Inc. KSZ8567S * Software action: The host processor should program the Static Address Table with the entries that it needs to receive (e.g., BPDU packets). The "overriding" bit should be set so that the switch will forward those specific packets to the processor. The processor may send packets to the port(s) in this state (see Section 4.4.9, "Tail Tagging Mode" for details). Address learning is enabled on the port in this state. RSTP uses only one type of BPDU called RSTP BPDUs. They are similar to STP configuration BPDUs with the exception of a type field set to "version 2" for RSTP and "version 0" for STP, and a flag field carrying additional information. 4.4.8 MULTIPLE SPANNING TREE SUPPORT Multiple Spanning Tree Protocol (MSTP) is an extension of RSTP that allows different VLANs to have different spanning tree configurations. The VLAN Table, Address Lookup Table, Static Address Table and Reserved Multicast Address Table all contain a 3-bit field which can be used to specify one of eight spanning trees. Each port contains state registers for specifying unique states for each of the spanning trees. 4.4.9 TAIL TAGGING MODE Tail tagging is a method to communicate ingress and egress port information between the host processor and the switch. It is useful for spanning tree protocol, IGMP/MLD snooping, IEEE 1588, and other applications. When the switch forwards a frame to the host port, two tail tagging bytes are added to the frame by the switch to indicate to the host processor the port that the frame was received on. In the other direction, the host processor adds the tail tagging bytes to a frame to indicate the intended egress destination port to the switch. When multiple priority queues are enabled, the tail tag is also used to indicate the priority queue. The tail tagging bytes are removed before the frame egresses the switch. Tail tagging may be enabled on any one port, and this defines the host port. tail tagging must not be enabled on multiple ports. Tail tagging is implemented by inserting two additional bytes at the end of the packet, between the data field and the 4byte CRC / FCS, as shown in Figure 4-5. FIGURE 4-5: BYTES 4.4.10 TAIL TAG FRAME FORMAT 6 6 (4) 2 46 (42) - 1500 2 4 DEST ADDRESS SOURCE ADDRESS 802.1Q TAG ETYPE or LENGTH PAYLOAD TAIL TAG FCS IGMP SUPPORT For Internet Group Management Protocol (IGMP) support in Layer 2, the device provides two components: * "IGMP" Snooping * "Multicast Address Insertion" in the Static MAC Table 4.4.10.1 "IGMP" Snooping The device traps IGMP packets and forwards them only to the processor (host port). The IGMP packets are identified as IP packets (either Ethernet IP packets, or IEEE 802.3 SNAP IP packets) with IP version = 0x4 and protocol version number = 0x2. Note: 4.4.10.2 The port for which Tail Tagging Mode is enabled is the host port. "Multicast Address Insertion" in the Static MAC Table Once the multicast address is programmed in the Static Address Table or Address Lookup Table, the multicast session is trimmed to the subscribed ports, instead of broadcasting to all ports. So that the host processor knows which port the IGMP packet was received on, Tail Tagging Mode must be enabled. 4.4.11 IPV6 MLD SNOOPING The device traps IPv6 Multicast Listener Discovery (MLD) packets and forwards them only to the processor (host port). 2017 Microchip Technology Inc. DS00002391A-page 37 KSZ8567S 4.4.12 PORT MIRRORING the device supports "port mirroring" comprehensively as: * "Receive Only" Mirror-on-a-Port * "Transmit Only" Mirror-on-a-Port * "Receive and Transmit" Mirror-on-a-Port 4.4.12.1 "Receive Only" Mirror-on-a-Port All the packets received on the port are mirrored on the sniffer port. For example, 1 is programmed to be "receive sniff" and the host port is programmed to be the "sniffer". A packet received on port 1 is destined to port 2 after the internal lookup. The packet is forwarded to both port 2 and the host port. The device can optionally even forward "bad" received packets to the "sniffer port". 4.4.12.2 "Transmit Only" Mirror-on-a-Port All the packets transmitted on the port are mirrored on the sniffer port. For example, port 1 is programmed to be "transmit sniff" and the host port is programmed to be the "sniffer port". A packet received on port 2 is destined to port 1 after the internal lookup. The device forwards the packet to both port 1 and the host port. 4.4.12.3 "Receive and Transmit" Mirror-on-a-Port All the packets received on port A and transmitted on port B are mirrored on the sniffer port. For example, port 1 is programmed to be "receive sniff", port 2 is programmed to be "transmit sniff", and the host port is programmed to be the "sniffer port". A packet received on port 1 is destined to port 2 after the internal lookup. The device forwards the packet to both port 2 and the host port. Multiple ports can be selected as "receive sniff" or "transmit sniff". In addition, any port can be selected as the "sniffer port". 4.4.13 SCHEDULING AND RATE LIMITING Each device port has two egress packet scheduling options, which can be applied when the port is configured for two or four queues. Additionally, each port has ingress and egress rate limiter features. 4.4.13.1 Strict Priority Scheduling When an egress port is configured as two or four queues, and strict priority scheduling is selected, each queue will take absolute priority over all lower priority queues. If a packet is available to transmit from queue 3 (the highest priority queue), then it will take priority for transmission over any packet that will also be available in any of the other queues. A packet in queue 2 will be transmitted only if no packet is available in queue 3. Weighted round robin is an alternative to strict priority scheduling. 4.4.13.2 Weighted Round Robin (WRR) Scheduling WRR scheduling is an alternative to strict priority scheduling for egress queues. It is referred to as fair queuing because it gives proportionally higher priority to the highest priority queue, but not absolute priority. 4.4.13.3 Rate Limiting The device supports independent ingress and egress hardware rate limiting on each port. Normally these two features are considered mutually exclusive, and users are discouraged from using both on the same port. For 10BASE-T/Te, a rate setting above 10Mbps means the rate is not limited. Likewise, for 100BASE-TX, a rate setting above 100Mbps means the rate is not limited. On the receive side, the data receive rate for each priority at each port can be limited by setting up ingress rate control registers. On the transmit side, the data transmit rate for each priority queue at each port can be limited by setting up egress rate control registers. The size of each frame has options to include minimum inter-frame gap (IFG) or preamble byte, in addition to the data field (from packet DA to FCS). For ingress rate limiting, the device provides options to selectively choose frames from all types, multicast, broadcast, and flooded unicast frames. The data rate from those selected type of frames is counted. Packets are dropped at the ingress port when the data rate exceeds the specified rate limit. For egress rate limiting, the leaky bucket algorithm is applied to each output priority queue for shaping output traffic. Inter-frame gap is stretched on a per frame base to generate smooth, non-burst egress traffic. The throughput of each output priority queue is limited by the egress rate specified. DS00002391A-page 38 2017 Microchip Technology Inc. KSZ8567S If any egress queue receives more traffic than the specified egress rate throughput, packets may be accumulated in the output queue and packet memory. After the memory of the queue or the port is used up, packet dropping or flow control will be triggered. As a result of congestion, the actual egress rate may be dominated by flow control/dropping at the ingress end, and may be therefore slightly less than the specified egress rate. To reduce congestion, it is a good practice to ensure that the egress bandwidth exceeds the ingress bandwidth. 4.4.14 EGRESS TRAFFIC SHAPING Two options are available for shaping the transmission of time-sensitive traffic: the IEEE 802.1Qav Credit-Based Traffic Shaper and the Time Aware Traffic Scheduler (TAS). The Credit-Based Shaper is defined in IEEE 802.1Qav for Audio Video Bridging (AVB). It attempts to minimize the jitter of traffic in the egress queue to which it applies, but the transmission of large packets from the "best effort" queue(s) can negatively impact its performance. The Time Aware Shaper reserves periodic time windows for the transmission of data from its queue. This ensures that low priority traffic will not block the periodic transmission of time-sensitive traffic, but it can also reduce the overall bandwidth of the port. When a shaper is used, the corresponding egress ports are normally configured for either two or four queues. One or more queues may be used for time-sensitive (i.e. scheduled) traffic, and the remaining queue(s) are used for lower priority unscheduled (best effort) traffic. The shapers are applied on a per queue basis at each port. Shaping is individually configurable for each port and each queue. The Credit-Based Shaper may be applied to either one or two queues on the same port. However, the Time Aware Scheduler may only be applied to one queue per port. It is possible to use both the Time Aware Scheduler (TAS) and Credit based Shaper (CBS) options on the same port at the same time. The TAS and CBS queue/priority assignments are configured via software. Section 4.4.4, "Quality-of-Service (QoS) Priority Support" discusses how ingress packets are mapped to egress queues based on priority. The most common method is by the PCP field of the VLAN tag. 4.4.14.1 IEEE 802.1Qav Credit-Based Traffic Shaper The traffic shaper is used to meter high priority (AVB) egress traffic as determined by the reserved bandwidth of the SR class traffic. A separate traffic shaper is available at the egress of each priority queue at each port. If a port is configured for four queues, then the two highest priority queues may be used for SR traffic classes. If a port is configured for two queues, then the highest priority queue (1) may be used for the SR traffic class. The lower priority queue(s) are used for "best effort" class traffic and are configured for strict priority scheduling, with the traffic shaper disabled for these queues. The traffic shaper is more sophisticated than the traditional "leaky bucket" rate limiting feature described earlier. A certain percentage of a port's bandwidth may be reserved for an AVB stream reservation (SR) traffic class. The traffic shaper may be configured for the reserved bandwidth, allowing the queue to egress packets at a rate up to but not exceeding this bandwidth. The queue may accumulate "credits" if an egress packet is delayed by a packet egressing from another queue. Accumulated credits may be used to permit catching up, in order to maintain an average rate. If a traffic class supported by the credit-based shaper uses less than the bandwidth allocated to it, then the unused bandwidth can be used by other traffic classes, in accordance with the relative priorities of the traffic classes and the transmission queuing algorithms associated with them. 4.4.14.2 Time Aware Traffic Scheduler (TAS) Like the Credit-Based Shaper, the Time Aware Scheduler is used to regulate the egress flow of high priority traffic. The TAS defines a periodic window of time during which only scheduled (high priority) traffic may egress a port. Additionally, transmission of low priority traffic will not be initiated during a guard band period prior to the high priority transmit window. This ensures that no blocking can occur during the transmit window. The TAS is configured individually for each egress port. Typically the egress port shall be configured for multiple queues - either 2 or 4 - with at least one queue reserved for unscheduled "best effort" traffic. For each port, only one queue may be set up for TAS. With a 4-queue configuration, it is possible to have both TAS queue and one or two Credit-Based Shaper queues at the same time. Once the guard band period begins, no packet transmission will be initiated unless the non-TAS queue packet length is smaller than the remaining guard band time. Any packets already being transmitted will be allowed to complete transmission. The guard band period shall be set up to match the transmit time for the largest possible non-scheduled packet. This ensures that transmission is always complete before the end of the guard band time. The guard band time is immediately followed by the transmit window. During the transmit window, transmission is allowed only for the TAS queue. All 2017 Microchip Technology Inc. DS00002391A-page 39 KSZ8567S other queues remain blocked. At the end of the transmit window, the block is lifted from all queues, and they are allowed to transmit per the scheduling scheme - either strict priority or weighted round robin (WRR). The cycle then repeats with start of the next guard band period. A 128-entry table is used to define the time intervals that relate the guard band, transmit window and cycle time. The size of the table makes it possible to define two or more concurrent cycles which correspond to two or more individual streams. Each cycle can have a unique period, guard band and transmit window time. 4.4.15 LOW LATENCY CUT-THROUGH MODE Cut-Through Mode allows for the reduction of packet latency through the switch by forwarding packets directly to the egress port without first waiting for receipt of the entire packet. This feature works in conjunction with the Time Aware Traffic Scheduler (TAS), and applies only if the incoming packet is being sent to a destination port with a corresponding TAS queue. Cut-through mode is enabled or disabled individually for each egress port. Cut-through mode does not guarantee that all packets in the scheduled traffic class can be expedited. The following conditions must be met in order for cut-through to occur for a given packet. If the conditions are not met, the packet will be stored and queued in the usual manner. * * * * * * * The Time Aware Scheduler must be applied to a queue at the egress port. The packet must belong to the scheduled traffic class that uses the Time Aware Scheduler. The TAS transmission window must be open. No other packet is currently being transmitted at that port. The TAS queue for that port is empty. The speed of the egress port must be equal to the speed of the ingress port. The packet must be unicast. The only traffic going to a cut-through egress port should be a single unicast cut-through stream. The egress port may be configured as a single queue, with the Time Aware Scheduler applied to that one queue. Set the transmit window to 100%. When cut-through occurs, the packet is forwarded after the first 64 bytes have been received. Latency through the switch is approximately 900ns for 1Gb/s traffic, regardless of the packet size. For non-cut-through packets, the minimum latency is proportional to the size of the packet. 4.4.16 INGRESS MAC ADDRESS FILTERING FUNCTION When a packet is received, the destination MAC address is looked up in both the static and dynamic MAC address tables. If the address is not found in either of these tables, then the destination MAC address is "unknown". By default, an unknown packet is forwarded to all ports except the port at which it was received. An optional feature makes it possible to specify the port or ports to which to forward unknown packets. It is also possible to specify no ports, meaning that unknown packets will be discarded. This feature is implemented separately for unknown unicast, unknown multicast and unknown VID packets. 4.4.17 802.1X ACCESS CONTROL IEEE 802.1X is a Port-based authentication protocol. EAPOL is the protocol normally used by the authentication process as uncontrolled Port. By receiving and extracting special EAPOL frames, the host processor can control whether the ingress and egress ports should forward packets or not. If a user port wants service from another port (authenticator), it must get approved by the authenticator. The device detects EAPOL frames by checking the destination address of the frame. The destination addresses should be either a multicast address as defined in IEEE 802.1x (01-80-C2-0000-03) or an address used in the programmable reserved multicast address domain with offset -00-03. Once EAPOL frames are detected, the frames are forwarded to the host port so it can send the frames to the authenticator server. Eventually, the CPU determines whether the requester is qualified or not based on its source MAC address, and frames are either accepted or dropped. When the device is configured as an authenticator, the ports of the switch must then be configured for authorization. In an authenticator-initiated port authorization, a client is powered up or plugs into the port, and the authenticator port sends an Extensible Authentication Protocol (EAP) PDU to the supplicant requesting the identification of the supplicant. At this point in the process, the port on the switch is connected from a physical standpoint; however, the 802.1X process has not authorized the port and no frames are passed from the port on the supplicant into the switching fabric. If the supplicant attached to the switch (KSZ8567S) did not understand the EAP PDU that it was receiving from the switch, it DS00002391A-page 40 2017 Microchip Technology Inc. KSZ8567S would not be able to send an ID and the port would remain unauthorized. In this state, the port would be blocked from passing any user traffic. If the supplicant is running the 802.1X EAP, it would respond to the request with its configured ID. (This could be a user name/password combination or a certificate.) After the device receives the ID from the supplicant, it passes the ID information to an authentication server (RADIUS server) that can verify the identification information. The RADIUS server responds to the switch with either a success or failure message. If the response is a success, the port will be authorized and user traffic will be allowed to pass through the port like any switch port connected to an access device. If the response is a failure, the port will remain unauthorized and, therefore, unused. If there is no response from the server, the port will also remain unauthorized and will not pass any traffic. Port control can be performed via the Access Control List (ACL) Filtering feature. 4.4.18 ACCESS CONTROL LIST (ACL) FILTERING An Access Control List (ACL) can be created for each port to perform filtering on incoming layer 2 MAC, layer 3 IP or layer 4 TCP/UDP packets. Multicast filtering is handled in the Static Address Table and the Reserved Multicast Address Table, but the ACL provides additional capabilities for filtering routed network protocols. As shown in Figure 4-3, ACL filtering may take precedence over other forwarding functions. The ACL allows the switch to filter ingress traffic based on the following header fields: * * * * * * Source or destination MAC address and/or EtherType Source or destination IPv4 address with programmable mask IPv4 protocol Source or destination UDP port Source or destination TCP port TCP Flag with programmable mask The ACL is implemented as an ordered list of up to 16 access control rules which are programmed into the ACL Table. Each entry specifies certain rules (a set of matching conditions and action rules) to control the forwarding and priority of packets. When a packet is received on an interface, the switch compares the fields in the packet against any applied ACLs to verify that the packet has the permissions required to be forwarded, based on the conditions specified in the lists. Multiple match conditions can be either AND'ed or OR'ed together. The ACL can also implement a count function that generates an interrupt rather than a forwarding action. The counter can be either a watchdog timer or an event counter. As a watchdog timer, an interrupt is generated if a packet with a specific MAC address and EtherType is not received within a specified time interval. As an event counter, an interrupt is generated once a specified number of packets with a specific MAC address and EtherType have been received. The ACL consists of three parts: matching rules, action rules, and processing entries. A matching rule specifies what comparison test shall be performed on the incoming packet. It can also enable a counter function. An action rule specifies the forwarding action to be taken if the matching test succeeds. Alternatively, when a count function is enabled in a matching rule, the 11-bit count value is stored in the corresponding action rule field and there is no forwarding action. In general, the 16 matching rules are not directly linked to the 16 action rules. For example, matching entry #0 is not necessarily related to action entry #0. The exception is when the counter function is enabled in a matching rule, whereby the matching rule and action rule fields at the same ACL table entry will function together and are no longer independent. Each of the 16 processing entries is used to link any number of matching rules (specified in RuleSet) to any one action rule (specified in FRN). When there are multiple matching rules in a RuleSet, those rules are AND'ed together. Only if all of those matching results are true will the FRN action be taken. It is also possible to configure the ACL table so that multiple processing entries specify the same action rule. In this way, the final matching result is the OR of the matching results from each of the multiple RuleSets. The 16 ACL rules represent an ordered list, with entry #0 having the highest priority and entry #15 having the lowest priority. All matching rules are evaluated. If there are multiple true match results and multiple corresponding actions, the highest priority (lowest numbered) of those actions will be the one taken. 2017 Microchip Technology Inc. DS00002391A-page 41 KSZ8567S 4.4.18.1 Processing Entry Description The Processing Entry consists of two parameters as described in Table 4-13. TABLE 4-13: ACL PROCESSING ENTRY PARAMETERS Parameter Description FRN[3:0] First Rule Number Pointer to an Action rule entry. Possible values are 0 to 15. If all Matching rules specified in the RuleSet are evaluated true, then this is the resulting Action rule. RuleSet[15:0] Specifies a set of one or more Matching rule entries. RuleSet has one bit for each of the 16 Matching rule entries. If multiple Matching rules are selected, then all conditions will be AND'ed to produce a final match result. 0 = Matching rule not selected 1 = Matching rule selected FIGURE 4-6: ACL STRUCTURE AND EXAMPLE RULE VALUES Processing Field Entry Number RuleSet Matching Rule Entry #0 0 #0 0 Entry #0 Entry #1 0 #1 1 Entry #1 Entry #2 1 #2 2 Entry #2 Action Field #3 none Entry #3 Entry #3 #4 4, 5, 6 Entry #4 #5 none #6 6 Entry #6 #7 none Entry #7 #8 7, 11 Entry #8 Action Rule Entry #4 OR FRN 4 Entry #5 Entry #6 5 Entry #7 Entry #8 9 AND When counter function is enabled in Matching Rule. Entry #5 Entry #9 #9 none Entry #9 Entry #10 #10 none Entry #10 Entry #11 #11 none Entry #11 Entry #12 #12 none Entry #12 Entry #13 #13 none Entry #13 Entry #14 #14 none Entry #14 Entry #15 #15 none Entry #15 The examples in Figure 4-6 are interpreted as follows: * * * * Rule #0: Test the matching rule entry #0. If true, apply action rule entry #0. Rule #1: Test the matching rule entry #1. If true, apply action rule entry #0. Rule #2: Test the matching rule entry #2. If true, apply action rule entry #1. Matching rule entry #3 is configured for the counter function. Action entry #3 is used to hold the corresponding count value. * Rule #4: Test the matching rule entries #4, 5 and 6. If all are true, apply action rule entry #4. * Rule #6: Test the matching rule entry #6. If true, apply action rule entry #5. * Rule #8: Test the matching rule entries #7 and 11. If both are true, apply action rule entry #9. No more than one action can be taken for any packet. If the matching conditions are true for multiple RuleSets, then the corresponding FRN field with the lowest value (highest priority) determines the action to be taken. DS00002391A-page 42 2017 Microchip Technology Inc. KSZ8567S Note that processing entries #0 and 1 produce an OR function: action #0 is taken if RuleSet #0 or RuleSet #1 is true. Notice that processing entries #4 and 6 have overlapping RuleSets, but different FRNs. This can be summarized as: If match #4, 5 and 6 are all true, then apply action #4, Else if match #6 is true, then apply action #5. Table 4-14 summarizes the available matching options. The MD and ENB fields are used to select the desired matching option. More configuration details are given in the following section. TABLE 4-14: MATCHING RULE OPTIONS MD[1:0] ENB[1:0] 00 XX Matching rule disabled 01 (Layer 2 matching: MAC address, EtherType) 00 Action field is used as count value for packets matching MAC address and EtherType 01 Compare MAC address only 10 Compare EtherType only 10 (Layer 3 matching: IP address) 11 (Layer 4 matching: TCP, UDP, IP protocol) 4.4.18.2 Matching Rule 11 Compare both MAC address and EtherType 00 Reserved 01 Compare IPv4 source and destination address (with mask) 10 Compare both source and destination IPv4 addresses (without mask) 11 Reserved 00 Compare IPv4 protocol 01 Compare TCP source port or destination port 10 Compare UDP source port or destination port 11 Compare TCP sequence number Matching Rule Description The Matching Rule consists of several parameters. The first two parameters, MD[1:0] and ENB[1:0], determine the organization of the remainder of each Matching Rule. When MD = 00, the Matching Rule is disabled. TABLE 4-15: ACL MATCHING RULE PARAMETERS FOR MD = 01 Parameter Description MD[1:0] MODE 00 = Matching rule is disabled 01 = Layer 2 MAC header or counter filtering 10 = Layer 3 IP header filtering 11 = Layer 4 TCP header (and IP protocol) filtering ENB[1:0] 00 = Count Mode. Both the MAC Address and TYPE are tested. A count value (either time or packet count) is also incorporated. Details are given below this table. 01 = Comparison is performed only on the MAC Address value 10 = Comparison is performed only on the TYPE value 11 = Both the MAC Address and TYPE are tested S/D Source / Destination 0 = Destination address 1 = Source address EQ Equal / Not Equal 0 = Not Equal produces true result 1 = Equal produces true result MAC ADDRESS[47:0] 48-bit MAC address TYPE[15:0] EtherType Details for MD = 01, ENB = 00: 2017 Microchip Technology Inc. DS00002391A-page 43 KSZ8567S The 11 bits of the aggregated bit fields from PM, P, RPE, RP and MM in the Action rule entry specify a count value for packets matching MAC Address and TYPE in the Matching Field. The count unit is determined by the TU bit (located in the Action rule). * When 0, the unit is microsecond. * When 1, the unit is millisecond. The CA bit (located in the Action rule) determines the algorithm used to generate an interrupt when the count terminates. * When 0, an 11-bit counter will be loaded with the count value from the list and start counting down every unit time. An interrupt will be generated when the timer expires, i.e. the next qualified packet has not been received within the period specified by the value. * When 1, the counter is incremented with every matched packet received. An interrupt is generated when the terminal count is reached. The count resets thereafter. Time units are not used in this mode. TABLE 4-16: ACL MATCHING RULE PARAMETERS FOR MD = 10 Parameter Description MD[1:0] MODE 00 = Matching rule is disabled 01 = Layer 2 MAC header or counter filtering 10 = Layer 3 IP header filtering 11 = Layer 4 TCP header (and IP protocol) filtering ENB[1:0] 00 = Reserved 01 = IPv4 source or destination address (with mask) 10 = IPv4 source and destination address (without mask) 11 = Reserved S/D Source / Destination 0 = Destination address 1 = Source address EQ Equal / Not Equal 0 = Not Equal produces true result 1 = Equal produces true result IP ADDRESS[31:0] IPv4 address Source or destination address (determined by S/D) when ENB = 01, Source address when ENB = 10 IP MASK[31:0] Mask bits for the IPv4 address when ENB = 01: 0 = This bit of the address is compared 1 = This bit of the address is not compared Destination IPv4 address when ENB = 10 TABLE 4-17: ACL MATCHING RULE PARAMETERS FOR MD = 11 Parameter Description MD[1:0] MODE 00 = Matching rule is disabled 01 = Layer 2 MAC header or counter filtering 10 = Layer 3 IP header filtering 11 = Layer 4 TCP header (and IP protocol) filtering ENB[1:0] 00 = IP Protocol comparison is enabled 01 = TCP source/destination port comparison is enabled 10 = UDP source/destination port comparison is enabled 11 = TCP sequence number is compared S/D Source / Destination 0 = Destination address 1 = Source address DS00002391A-page 44 2017 Microchip Technology Inc. KSZ8567S TABLE 4-17: ACL MATCHING RULE PARAMETERS FOR MD = 11 (CONTINUED) Parameter Description EQ Equal / Not Equal 0 = Not Equal produces true result 1 = Equal produces true result MAX PORT[15:0] MIN PORT[15:0] Max and Min Ports for TCP/UDP or TCP Sequence Number[31:0] PC[1:0] Port Comparison 00 = Port comparison is disabled 01 = Port matches either one of MAX or MIN 10 = Match if port number is in the range of MIN to MIN 11 = Match if port number is out of the range PRO[7:0] IPv4 protocol to be matched FME TCP Flag Match Enable 0 = TCP FLAG matching disabled 1 = TCP FLAG matching enabled FMASK[7:0] TCP FLAG Mask 0 = This bit of the Flag field is compared 1 = This bit of the Flag field is not compared FLAG[7:0] TCP Flag to be matched 4.4.18.3 Action Rule Description TABLE 4-18: ACL ACTION RULE PARAMETERS FOR NON-COUNT MODES (MD 01 OR ENB 00) Parameter Description PM[1:0] Priority Mode 00 = ACL does not specify the packet priority. Priority is determined by standard QoS functions. 01 = Change packet priority to P[2:0] if it is greater than QoS result. 10 = Change packet priority to P[2:0] if it is smaller than the QoS result. 11 = Always change packet priority to P[2:0]. P[2:0] Priority value RPE Remark Priority Enable 0 = Disable priority remarking 1 = Enable priority remarking. VLAN tag priority (PCP) bits are replaced by RP[2:0]. RP[2:0] Remarked Priority value MM[1:0] Map Mode 00 = No forwarding remapping 01 = The forwarding map in FORWARD is OR'ed with the forwarding map from the Address Lookup Table. 10 = The forwarding map in FORWARD is AND'ed with the forwarding map from the Address Lookup Table. 11 = The forwarding map in FORWARD replaces the forwarding map from the Address Lookup Table. FORWARD[N-1:0] Forwarding Ports Bit 0 corresponds to port 1 Bit 1 corresponds to port 2, etc. 0 = Do not forward to this port 1 = Forward to this port 2017 Microchip Technology Inc. DS00002391A-page 45 KSZ8567S TABLE 4-19: ACL ACTION RULE PARAMETERS FOR COUNT MODE (MD = 01 OR ENB = 00) Parameter Description COUNT[10:0] Count value TU Time unit for counter. 0 = Microseconds 1 = Milliseconds CA Counter Algorithm. 0 = An 11-bit counter will be loaded with the count value from the list and start counting down every unit time. An interrupt will be generated when the timer expires, i.e. the next qualified packet has not been received within the period specified by the value. 1 = The counter is incremented with every matched packet received. An interrupt is generated when the terminal count is reached. The count resets thereafter. Time units are not used in this mode. Figure 4-7 shows basic organization of the ACL Table. The table has 16 entries, and each entry includes a matching field, action field and process field. Although these fields are stored together in one table, it is important to note that for a given table entry, the Matching, Action and Process fields generally do not form an associated group. The one exception is when the Matching Rule is in Count Mode (MD = 01 and ENB = 00). In that case, the Matching and Action fields are used in tandem. FIGURE 4-7: ACL TABLE FORMAT MD = 01 ENB = 00 F R N MD [1:0] ENB [1:0] S E / Q D MAC ADDRESS [47:0] TYPE [15:0] MD = 01 ENB != 00 F R N MD [1:0] ENB [1:0] S E / Q D MAC ADDRESS [47:0] TYPE [15:0] MD = 10 F R N MD [1:0] ENB [1:0] S E / Q D MD = 11 F R N MD [1:0] ENB [1:0] S E / Q D PROCESS Field 4.4.18.4 IP ADDRESS [31:0] MAX PORT [15:0] MIN PORT [15:0] Resvd (5) MATCHING Rule IP MASK [31:0] PC [1:0] PRO [7:0] F FMSK FLAG M [7:0] [7:0] E COUNT[10:0] TU CA unu sed RULESET [15:0] PM [1:0] P [2:0] R P E RP [2:0] MM [1:0] FORWARD [# ports] RULESET [15:0] PM [1:0] P [2:0] R P E RP [2:0] MM [1:0] FORWARD [# ports] RULESET [15:0] PM [1:0] P [2:0] R P E RP [2:0] MM [1:0] FORWARD [# ports] RULESET [15:0] ACTION Rule PROCESS Field ACL Interrupts The ACL filtering functions do not generate interrupts. Interrupts apply only for the Count Mode (MD = 01, ENB = 00). The Matching Rule can be configured either to timeout if the interval between packets of a specific type (MAC address and EtherType), or when a set number of these packets are received. There is a separate interrupt for each port. Port specific interrupt status and masks are located in the Port Interrupt Status Register and Port Interrupt Mask Register. The top level interrupt registers for each port are in the Global Port Interrupt Status Register and Global Port Interrupt Mask Register. DS00002391A-page 46 2017 Microchip Technology Inc. KSZ8567S 4.4.18.5 ACL Registers Table 4-20 provides a list of ACL related registers. TABLE 4-20: ACL REGISTERS Registers Description Port Interrupt Status Register, Port Interrupt Mask Register ACL interrupt Port ACL Access 0 Register through Port ACL Access F Register, Port ACL Byte Enable MSB Register, Port ACL Byte Enable LSB Register, Port ACL Access Control 0 Register ACL Table access Port Priority Control Register Priority classification Port Authentication Control Register ACL enable 4.5 IEEE 1588 Precision Time Protocol The IEEE 1588 precision time protocol (PTP) provides a method for establishing time synchronization across nodes in a network. The device supports V2 (2008) of the IEEE 1588 PTP specification and can be programmed as either an end-to-end (E2E) or peer-to-peer (P2P) transparent clock (TC) between ports. In addition, the host port can be programmed as either a slave or master ordinary clock (OC) port. Ingress timestamp capture, egress timestamp recording, correction field update with residence time and link delay, delay turn-around time insertion, egress timestamp insertion, and checksum update are supported. PTP frame filtering is implemented to enhance overall system performance. Delay adjustments are implemented to fine tune the synchronization. Versatile event trigger outputs and timestamp capture inputs are implemented to meet various real time application requirements through the GPIO_1 pin. The key features of the IEEE 1588 implementation are: * Either one-step or two-step event message format * Implementation of precision time clock per specification - Upper 16 bits of the second clock not implemented * Both E2E and P2P TC * Both slave and master OC * PTP multicast and unicast addressing * Transport of PTP over IPv4/IPv6 UDP and IEEE 802.3/Ethernet * Precision time stamping of input signal on the GPIO_1 pin * Creation and delivery of clock, pulses, or other unique serial bit stream on the GPIO_1 pin with respect to the precision time clock time. * IEEE 802.1AS gPTP supported IEEE 1588 defines two essential functions: The measurement of link and residence (switching) delays by using the Delay_Req/Resp or Pdelay_Req/Resp messages, and the distribution of time information by using the Sync/Follow_Up messages. The 1588 PTP event messages are periodically sent from the grandmaster in the network to all slave clock devices. Link delays are measured by each slave node to all its link partners to compensate for the delay of PTP messages sent through the network. 4.5.1 IEEE 1588 PTP SYSTEM TIME CLOCK The device's system time clock (STC) is a readable and writable high-precision counter that is used to keep the PTP time. The counter resolution is 2-32ns. Figure 4-8 details the PTP Clock. 2017 Microchip Technology Inc. DS00002391A-page 47 KSZ8567S . FIGURE 4-8: 4.5.2 PTP SYSTEM CLOCK OVERVIEW IEEE 1588 PTP MESSAGING PROCESSING The device supports IEEE 1588 PTP time synchronization when 1588 PTP mode and message detection are enabled. Different operations will be applied to PTP packet processing based on the setting of P2P or E2E in transparent clock mode, master or slave in ordinary clock mode, one-step or two-step clock mode, and if the domain checking is enabled. For the IPv4/UDP egress packet, the checksum can be updated by either re-calculating the two-bytes or by setting it to zero. For the IPv6/UDP egress packet, the checksum is always updated. 4.5.3 IEEE 1588 PTP EVENT TRIGGERING AND TIMESTAMPING An event trigger output signal can be generated when the target and activation time matches the IEEE1588 PTP system clock time. Likewise, an event timestamp input can be captured from an external event input signal and the corresponding time on the IEEE1588 PTP system clock will be captured. Both the output event block and the input timestamp block can be programmed to generate interrupts. 4.6 Audio Video Bridging and Time Sensitive Networks AVB defines a set of features and protocols for ensuring high QoS for time-sensitive traffic such as audio and video streams. The device provides the necessary hardware features for implementation of AVB: gPTP timing synchronization features (802.1AS, based on 1588 PTP), prioritized queuing, and credit-based traffic shapers (802.1Qav). Ports supporting AVB must operate at 100 Mbps at full-duplex. 10 Mbps and/or half-duplex are not compatible with AVB. Non-AVB devices are not allowed to participate in an AVB network, but non-AVB traffic is allowed to occupy available bandwidth unused by the AVB traffic. Protocols are defined for network discovery, path setup, and bandwidth reservation across the network. AVB defines Stream Reservation (SR) traffic classes, which are given a high priority for queuing and egress scheduling. With up to four egress queues per port, the device can accommodate up to two SR traffic classes. Ingress AVB traffic packets are typically VLAN tagged. If not, the device can assign them a customizable VLAN tag on a port-by-port basis. The information in this tag, including the 3-bit PCP priority field, is used to map the packets to the appropriate high priority queue. At least one queue must be reserved for non-SR traffic classes, in order to accommodate regular "best effort" network traffic. DS00002391A-page 48 2017 Microchip Technology Inc. KSZ8567S The egress of SR class traffic from the high priority queue(s) is regulated at each queue by a credit-based traffic shaper as discussed below. If the SR traffic streams do not exceed their reserved bandwidth, then the traffic shaper will ensure that the packets egress in a relatively uniform manner. The 802.1AS standard, which ensures node synchronization in an AVB network, relies heavily on the IEEE 1588 standard for PTP. It defines a specific profile for 1588, and adds certain other requirements in the context of 802.1Q. The resulting entity is referred to as generalized PTP (gPTP). Section 4.5, "IEEE 1588 Precision Time Protocol" discusses IEEE 1588 and 802.1AS together. The device provides two egress traffic shaping options. The IEEE 802.1Qav Credit-Based Traffic Shaper provides traffic shaping as defined in IEEE 802.1Qav. This is configurable on a per-queue basis. Each port also has a Time Aware Traffic Scheduler (TAS) that provides periodic timing windows for transmission of time sensitive traffic class data. The shaper and scheduler options and details are individually configurable for each egress port. 4.7 NAND Tree Support The KSZ8567S provides parametric NAND tree support for fault detection between chip I/Os and board. The NAND tree is a chain of nested NAND gates in which each KSZ8567S digital I/O (NAND tree input) pin is an input to one NAND gate along the chain. At the end of the chain, the INTRP_N pin provides the output for the last NAND gate. The NAND tree test process includes: * Enabling NAND tree mode * Pulling all NAND tree input pins high * Driving low each NAND tree input pin sequentially per the NAND tree pin order, starting with the first row of Table 4-21. * Checking the NAND tree output to ensure there is a toggle high-to-low or low-to-high for each NAND tree input driven low. TABLE 4-21: NAND TREE TEST PIN ORDER NAND Tree Sequence Pin Number Pin Name NAND Tree Description 1 85 LED4_0 Input 2 86 LED4_1 Input 3 88 LED3_0 Input 4 89 LED3_1 Input 5 90 GPIO_1 Input 6 91 LED2_0 Input 7 92 LED2_1 Input 8 93 PME_N Input 9 95 SYNCLKO Input 10 97 SDO Input 11 98 SDI/SDA/MDIO Input 12 100 SCS_N Input 13 101 SCL/MDC Input 14 102 LED5_0 Input 15 103 LED5_1 Input 16 105 LED1_0 Input 17 106 LED1_1 Input 18 94 INTRP_N Output 2017 Microchip Technology Inc. DS00002391A-page 49 KSZ8567S The following procedure can be used to check for faults on the KSZ8567S digital I/O pin connections to the board: 1. 2. 3. Enable NAND tree mode via the LED2_1, LED2_0, and LED4_0 configuration strap pins option. Use board logic to drive all KSZ8567S NAND tree input pins high and verify that the INTRP_N pin output is high. Use board logic to drive each NAND tree input pin, per the NAND Tree pin order, as follows: a) Toggle the first pin in the NAND tree sequence (LED4_0) from high to low, and verify the INTRP_N pin switches from high to low to indicate that the first pin is connected properly. b) Leave the first pin (LED4_0) low. c) Toggle the second pin in the NAND tree sequence (LED4_1) from high to low, and verify the INTRP_N pin switches from low to high to indicate that the second pin is connected properly. d) Leave the first pin (LED4_0) and the second pin (LED4_1) low. e) Toggle the third pin in the NAND tree sequence (LED3_0) from high to low, and verify the INTRP_N pin switches from high to low to indicate that the third pin is connected properly. f) Continue with this sequence until all KSZ8567S NAND tree input pins have been toggled. Each KSZ8567S NAND tree input pin must cause the INTRP_N output pin to toggle high-to-low or low-to-high to indicate a good connection. If the INTRP_N pin fails to toggle when the KSZ8567S input pin toggles from high to low, the input pin has a fault. 4.8 4.8.1 Clocking PRIMARY CLOCK The device requires a 25MHz reference clock input at the XI pin. This clock is internally multiplied up and used to clock all of the internal logic and switching functions. It is also normally used as to clock the PHY transmit paths. This clock may be supplied by connecting a crystal between the XI and XO pins (and appropriate load capacitors to ground). Alternatively, an external CMOS clock signal may drive XI, while XO is left unconnected. The XI/XO block is powered from AVDDH. 4.8.2 PORT 6 RGMII/MII/RMII CLOCKS The MII interface is clocked asymmetrically, with the PHY device driving both the RX_CLK6 receive clock and the TX_CLK6 transmit clock to the MAC device. The MII port may be configured at reset by a strapping option to take the role of either the PHY or the MAC. RX_CLK6 and TX_CLK6 are therefore either both inputs or both outputs, depending on the MII mode. The RMII interface uses a single 50MHz clock. This REFCLK may be sourced either from the KSZ8567S or from the connected device. A strapping option is used to select the mode. "Normal Mode" is the mode where the other device supplies the clock, and the clock is an input to the REFCLKI6 pin of the device. "Clock Mode" is the mode where the KSZ8567S generates the 50MHz clock on the REFCLKO6 pin. The RGMII interface employs source synchronous clocking, so it is symmetrical and does not require a mode selection. An output clock is generated on the RX_CLK6 pin, while an input clock is received on the TX_CLK6 pin. The clock speed scales with the interface data rate - either 10, 100 or 1000 Mbps. A strapping option is used to select between the 100 and 1000 Mbps speeds. If the 10 Mbps rate is required, then a register setting is used to set that speed. The Port 6 MAC interface is powered from VDDIO. Note: 4.8.3 Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information on using configuration straps. PORT 7 SGMII CLOCK The SGMII interface consists of two differential serial data signals: one transmit pair and one receive pair. No clock is transmitted or received. Each receiver must perform clock and data recovery from the received data signal. The internal SGMII transmit clock is derived from the 25MHz reference clock at XI. Note: Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information on using configuration straps. DS00002391A-page 50 2017 Microchip Technology Inc. KSZ8567S 4.8.4 SERIAL MANAGEMENT INTERFACE CLOCK Whether configured to be SPI, I2C or MIIM, the KSZ8567S is always a slave and receives the clock as an input. The serial management interface is powered from VDDIO. 4.8.5 SYNCHRONOUS ETHERNET AND SYNCLKO An output clock is provided on the SYNCLKO pin. By default it is derived from the 25MHz input reference clock on XI, but the source can be selected to be the recovered clock from any of the PHY ports. This recovered clock may then feed an external device with a low bandwidth PLL and hold-over feature for use in Synchronous Ethernet applications. A 25MHz clock derived from SYNCLKO may then be used as the input to XI. The output frequency choices are 25MHz (default) and 125MHz. If not needed, this output clock can also be disabled. SYNCLKO is controlled via the Output Clock Control Register, and is powered from VDDIO. Note: In order to utilize synchronous Ethernet functions, additional hardware design rules must be taken into consideration. Contact your local Microchip sales representative for additional information. 2017 Microchip Technology Inc. DS00002391A-page 51 KSZ8567S 4.9 Power The KSZ8567S requires two to three supply voltages. The device core operates from a 1.2V supply (DVDDL and AVDDL). The PHY transceivers and XI/XO crystal/clock interface operate from a 2.5V or 3.3V supply (AVDDH). The digital I/O's can be operated from 1.8V, 2.5V or 3.3V (VDDIO). The digital I/Os powered from VDDIO include RGMII, RMII, MII, SPI, I2C, MIIM, LED, RESET_N, PME_N, INTRP_N and SYNCLKO. The SGMII interface requires 2.5V (VDDHS) and 1.2V (VDDLS) supplies. An example power connection diagram can be seen in Figure 4-9. FIGURE 4-9: POWER CONNECTION DIAGRAM +2.5V 22F +1.2V AVDDH AVDDH 22F 0.1F 0.1F AVDDH VDDLS AVDDH VDDLS AVDDH AVDDH AVDDH 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 22F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F DVDDL 22F DVDDL DVDDL VDDHS DVDDL VDDHS DVDDL +1.8V, 2.5V or 3.3V DVDDL DVDDL DVDDL 10F VDDIO VDDIO 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F 0.1F DS00002391A-page 52 0.1F 0.1F 0.1F GND 22F 0.1F 0.1F DVDDL VDDIO 0.1F 0.1F AVDDL AVDDL AVDDL AVDDL AVDDL AVDDL AVDDL AVDDL AVDDL GND GND GND GND GND GND GND GND GND (exposed pad) 2017 Microchip Technology Inc. KSZ8567S 4.10 Power Management The device supports enhanced power management features in a low-power state with energy detection to ensure lowpower dissipation during device idle periods. There are three operation modes under the power management function which are implemented globally (i.e., applying to all ports): * Normal Operation Mode * Energy Detect Mode * Global Soft Power Down Mode Table 4-22 summarizes all internal function blocks status under the three power-management operation modes. TABLE 4-22: MDI/MDI-X PIN DEFINITIONS Functional Blocks Power Management Operation Modes Normal Mode Energy Detect Mode Soft Power Down Mode Internal PLL Clock Enabled Disabled Disabled TX/RX PHYs Enabled Energy Detect at RX Disabled MACs Enabled Disabled Disabled Host Interface Enabled Disabled Disabled There are two additional power saving modes that may be implemented on a per-port basis: * Port-Based Power Down * Energy Efficient Ethernet (EEE) - Active only if enabled in the device and auto-negotiated (per-port) with a link partner. EEE can be enabled on a per-port basis, but is not available on the non-PHY ports 6 and 7. The first three global power modes are mutually exclusive; only one mode may be selected at a time. Port-based power down may be enabled independent of the global power mode. 4.10.1 NORMAL OPERATION MODE At power-up, the device enters into Normal operation mode. It is also selected via bits [4:3] = 00 in the Power Down Control 0 Register. When the device is in normal operation mode, all PLL clocks are running, PHYs and MACs are on, and the CPU is ready to read or write the device registers through the serial interface (SPI, I2C or MIIM). During normal operation mode, the host processor can change the power management mode bits in the Power Down Control 0 Register to transition to any of the other power management modes. 4.10.2 ENERGY-DETECT MODE Energy-detect mode, also known as energy-detect power down (EDPD) mode, is enabled by setting bits [4:3] to 01 in the Power Down Control 0 Register. Energy-detect mode provides a mechanism to save power when the device is not connected to an active link partner. Auto-negotiation must be enabled when in energy-detect mode. Energy-detect mode consists of two states, normal-power state and low-power state. When the device is in this mode, it will monitor the cable energy. If there is no energy on the cable for a time longer than a pre-configured value, the device will go into the low-power state. While in low-power state, the device reduces power consumption by disabling all circuitry except the energy-detect circuitry of the receiver, which consumes minimal power. When the device is in the lowpower state, it will transmit link pulses at long intervals, with a very low duty cycle. At the same time, it continuously monitors for energy on the cable. Once energy is detected from the cable and is present for a time longer than 100ns, the device will enter the normal-power state. 4.10.3 GLOBAL SOFT POWER-DOWN MODE Soft power-down mode is used to power down the device when it is not in use after power-up. This mode disables all internal functions except for the serial (SPI or I2C) management interface. When soft power-down mode is exited, all registers are reset to their default values, and all configuration strap pins are sampled to set the device settings. 4.10.4 PORT-BASED POWER DOWN Unused ports may be powered down individually to save power. 2017 Microchip Technology Inc. DS00002391A-page 53 KSZ8567S 4.10.5 ENERGY EFFICIENT ETHERNET (EEE) Energy-efficient Ethernet (EEE) is implemented in the device as described in the IEEE 802.3AZ specification for ports 1 through 5. EEE is not available on ports 6 and 7. EEE saves power by keeping the voltage on the Ethernet cable at approximately 0V for as often as possible during periods of no traffic activity. This is called the low-power idle (LPI) state. However, the link will respond automatically when traffic resumes and do so in such a way as to not cause blocking or dropping of any packets (the wake-up time for 100BASE-TX is specified to be less than 30s). The transmit and receive directions are independently controlled. The EEE feature is disabled by default. EEE must be auto-negotiated, and is enabled only if both nodes on a link support it. The time during which LPI mode is active is called quiet time. This is shown in Figure 4-10. FIGURE 4-10: 4.10.5.1 TRAFFIC ACTIVITY AND EEE Transmit Direction Control for MII Mode For EEE ports, low-power idle (LPI) state for the transmit direction will be entered when the internal EEE MAC signals to its PHY to do so. The PHY will stay in the transmit LPI state as long as indicated by the MAC. The TX_CLK is not stopped. Even though the PHY is in LPI state, it will periodically leave the LPI state to transmit a refresh signal using specific transmit code bits. This allows the link partner to keep track of the long-term variation of channel characteristics and clock drift between the two partners. Approximately every 20ms - 22ms, the PHY will transmit a bit pattern to its link partner of duration 200s - 220s. The refresh times are shown in Figure 4-10. 4.10.5.2 Receive Direction Control for MII Mode If enabled for LPI mode, upon receiving a P Code bit pattern (refresh), the PHY will enter the LPI state and signal to the internal MAC. If the PHY receives some non-P Code bit pattern, it will signal to the MAC to return to "normal frame" mode. The PHY can turn off the RX_CLK after nine or more clocks have occurred in the LPI state. In the EEE-compliant environment, the internal PHYs will be monitoring and expecting the P Code (refresh) bit pattern from its link partner that is generated approximately every 20ms - 22ms, with a duration of about 200s - 220s. This allows the link partner to keep track of the long term variation of channel characteristics and clock drift between the two partners. 4.10.6 WAKE ON LAN (WOL) Wake on LAN allows a computer to be turned on or woken up by a network message. The message is usually sent by a program executed on another computer on the same local area network. Wake-up frame events are used to awaken the system whenever meaningful data is presented to the system over the network. Examples of meaningful data include the reception of a Magic Packet, a management request from a remote administrator, or simply network traffic directly targeted to the local system. The device can be programmed to notify the host of the Wake-Up frame detection with the assertion of the power management event signal (PME_N). The device's MACs support the detection of the following Wake-Up events: * Detection of energy signal over a pre-configured value * Detection of a linkup in the network link state * Receipt of a Magic Packet DS00002391A-page 54 2017 Microchip Technology Inc. KSZ8567S There are also other types of Wake-Up events that are not listed here as manufacturers may choose to implement these in their own way. 4.10.6.1 Direction of Energy The energy is detected from the cable and is continuously presented for a time longer than pre-configured value, especially when this energy change may impact the level at which the system should re-enter to the normal power state. 4.10.6.2 Direction of Link-up Link status wake events are useful to indicate a linkup in the network's connectivity status. 4.10.6.3 Magic PacketTM The Magic Packet is a broadcast frame containing anywhere within its payload 6 bytes of all 1s (FF FF FF FF FF FF) followed by sixteen repetitions of the target computer's 48-bit DA MAC address. Since the magic packet is only scanned for the above string, and not actually parsed by a full protocol stack, it may be sent as any network- and transport-layer protocol. Magic Packet technology is used to remotely wake up a sleeping or powered-off PC on a LAN. This is accomplished by sending a specific packet of information, called a Magic Packet frame, to a node on the network. When a PC capable of receiving the specific frame goes to sleep, it enables the Magic Packet RX mode in the LAN controller, and when the LAN controller receives a Magic Packet frame, it will alert the system to wake up. Once the device has been enabled for Magic Packet Detection, it scans all incoming frames addressed to the node for a specific data sequence, which indicates to the controller this is a Magic Packet frame. A Magic Packet frame must also meet the basic requirements for the LAN technology chosen, such as Source Address (SA), Destination Address (DA), which may be the receiving station's IEEE MAC address, or a multicast or broadcast address and CRC. The specific sequence consists of 16 duplications of the MAC address of this node, with no breaks or interruptions. This sequence can be located anywhere within the packet, but must be preceded by a synchronization stream. The synchronization stream is defined as 6 bytes of 0xFF. The device will also accept a broadcast frame, as long as the 16 duplications of the IEEE address match the address of the machine to be awakened. 4.10.6.4 Interrupt Generation on Power Management Related Events There are two ways an interrupt can be generated to the host whenever a power management related event takes place. The resulting interrupts are via the PME_N signal pin or via the INTRP_N signal pin. 4.11 Management Interface The management interface may be used by an external host processor to read and write the device's registers. This interface has three available modes of operation: SPI, I2C or MIIM. The interface mode is selected at the deassertion of reset by a strapping option (refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information). Of the three interface options, SPI provides the highest performance, while MIIM performance is the lowest. Most importantly, MIIM provides access to the PHY control and status registers, but not to any of the switch registers. The vast majority of applications therefore can use SPI or I2C, but not MIIM. Register access is also available through the high-performance in-band management interface as described in Section 4.12, "In-Band Management," on page 59. 4.11.1 SPI SLAVE BUS The KSZ8567S supports a slave mode SPI interface that provides complete access to all device registers via an SPI master device. The SPI master device supplies the clock (SCL), select (SCS_N), and serial input data (SDI). Serial output data (SDO) is driven by the KSZ8567S. SCL is expected to stay low when SPI operation is idle. SPI operations start with the falling edge of SCS_N and end with the rising edge of SCS_N. A single read or write access consists of a 27-bit command/address phase, then a 5-bit turnaround (TA) phase, then an 8-bit data phase. For burst read or write access, SCS_N is held low while SCL continues to toggle. For every 8 cycles of SCL, the device will increment the address counter, and the corresponding data byte will be transferred on SDI or SDO in succession. All commands, addresses and data are transferred most significant bit first. Input data on SDI is latched on the rising edge of clock SCL. Output data on SDO is clocked on the falling edge of SCL. As shown in Figure 4-23, there are two commands: register read and register write. Figure 4-11 and Figure 4-12 show the timing for these two operations. 2017 Microchip Technology Inc. DS00002391A-page 55 KSZ8567S TABLE 4-23: REGISTER ACCESS USING THE SPI INTERFACE SPI Operation Command/Address Phase (SDI pin) Command Data Phase (SDO or SDI pins) TA bits (Note 4-8) Register Address Register Read 011 A23 A22 A21 A20 ... A7 A6 A5 A4 A3 A2 A1 A0 XXXXX D7 D6 D5 D4 D3 D2 D1 D0 Register Write 010 A23 A22 A21 A20 ... A7 A6 A5 A4 A3 A2 A1 A0 XXXXX D7 D6 D5 D4 D3 D2 D1 D0 Note 4-8 Note: TA bits are turn-around bits. They are "don't care" bits. The actual device address space is 16 bits (A15 - A0), so the values of address bits A23 - A16 in the SPI command/address phase are "don't care". FIGURE 4-11: SPI REGISTER READ OPERATION SCS_N 1 2 3 0 1 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 SCL SDI (MOSI) A23 A22 A21 A20 A19 A18 A17 A16 A15 A14 A13 A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0 TA TA TA TA TA SDO D7 D6 D5 D4 D3 D2 D1 D0 (MISO) Read Command Read Address FIGURE 4-12: TurnAround Read Data SPI REGISTER WRITE OPERATION SCS_N 1 2 3 0 1 0 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 SCL SDI (MOSI) A23 A22 A21 A20 A19 A18 A17 A16 A15 A14 A13 A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0 TA TA TA TA TA D7 D6 D5 D4 D3 D2 D1 D0 SDO (MISO) Write Command DS00002391A-page 56 Write Address TurnAround Write Data 2017 Microchip Technology Inc. KSZ8567S 4.11.2 I2C BUS The management interface may be configured to be an I2C slave. In this mode, an I2C master has complete programming access to the device's internal control and status registers, including all MIB counters, address lookup tables, VLAN table and ACL table. The 7-bit device address is fixed as 1011_111. Because of the fixed address, only one KSZ8567S may be on the I2C bus at a time. The R/W control bit is then appended as the least significant bit to form these 8-bit address/control words: 1011_1110 <write> 1011_1111 <read> The internal registers and tables of the device are accessed using 16-bit addressing and 8-bit data. The access formats are as follows: FIGURE 4-13: SINGLE BYTE REGISTER WRITE FIGURE 4-14: SINGLE BYTE REGISTER READ FIGURE 4-15: BURST REGISTER WRITE FIGURE 4-16: BURST REGISTER READ 2017 Microchip Technology Inc. DS00002391A-page 57 KSZ8567S 4.11.3 MII MANAGEMENT (MIIM) INTERFACE The device supports the IEEE 802.3 MII management interface, also known as the management data input/output (MDIO) interface. This interface allows upper-layer devices to monitor and control the states of the KSZ8567S PHY blocks, but it does not provide access to the switch registers. An external device with MDC/MDIO capability can read the PHY status or configure the PHY settings. Details on the MIIM interface can be found in Clauses 22 and 45 of the IEEE 802.3 Specification. Use of MIIM conflicts with use of the In-Band Management interface. These interfaces cannot be used simultaneously. The MIIM interface consists of the following: * A physical connection that uses a data signal (MDIO) and a clock signal (MDC) for communication between an external controller and the KSZ8567S. Note that the MDIO signal is open-drain. * A specific protocol that operates across the two signal physical connection that allows an external controller to communicate with the internal PHY devices. * Access to a set of standard, vendor-specific and extended (MMD) 16-bit registers. These registers are also directly accessible via the SPI and I2C interface options. The MIIM Interface can operate up to a maximum clock speed of 5MHz. Access is limited to only the registers in the PHY blocks of ports 1 through 5. Table 4-24 summarizes the MII management interface frame format. TABLE 4-24: MII MANAGEMENT INTERFACE FRAME FORMAT Operation Mode Preamble (32-bit) Start of Frame (2-bit) Operation Code (2-bit) PHY Address (5-bit) Register Address (5-bit) Turn Around (2-bit) Register Data (16-bit) Idle Read All 1s 01 10 A[4:0] Reg[4:0] Z0 D[15:0] Z Write All 1s 01 01 A[4:0] Reg[4:0] 10 D[15:0] Z The MIIM PHY address to PHY port mapping is as follows: * * * * * PHY Address 1h to PHY port 1 PHY Address 2h to PHY port 2 PHY Address 3h to PHY port 3 PHY Address 4h to PHY port 4 PHY Address 5h to PHY port 5 The MIIM register address space consists of two distinct areas. * Standard MIIM Registers (Direct) * MDIO Manageable Device (MMD) Registers (Indirect) 4.11.3.1 Standard MIIM Registers (Direct) Standard registers provide direct read/write access to a 32-register address space, as defined in Clause 22 of the IEEE 802.3 Specification. Within this address space, the first 16 registers (Registers 0h to Fh) are defined according to the IEEE specification, while the remaining 16 registers (Registers 10h to 1Fh) are defined specific to the PHY vendor. The KSZ8567S supports the standard registers listed in Table 4-25 for each PHY port. Each 16-bit MIIM Standard Register Address maps to two corresponding 8-bit Port N Register Addresses. The register bit map and description are located at the 8-bit Port N Register Addresses. TABLE 4-25: STANDARD MIIM REGISTERS MIIM Standard Register Address (hex) Port N Register Address (hex) Description 0h 0xN100 - 0xN101 PHY Basic Control Register 1h 0xN102 - 0xN103 PHY Basic Status Register 2h 0xN104 - 0xN105 PHY ID High Register IEEE-Defined Registers DS00002391A-page 58 2017 Microchip Technology Inc. KSZ8567S TABLE 4-25: STANDARD MIIM REGISTERS (CONTINUED) MIIM Standard Register Address (hex) Port N Register Address (hex) 3h 0xN106 - 0xN107 PHY ID Low Register 4h 0xN108 - 0xN109 PHY Auto-Negotiation Advertisement Register 5h 0xN10A - 0xN10B PHY Auto-Negotiation Link Partner Ability Register 6h 0xN10C - 0xN10D PHY Auto-Negotiation Expansion Status Register 7h 0xN10E - 0xN10F PHY Auto-Negotiation Next Page Register 8h 0xN110 - 0xN111 PHY Auto-Negotiation Link Partner Next Page Ability Register Description 9h-Ch - Dh 0xN11A - 0xN11B PHY MMD Setup Register RESERVED Eh 0xN11C - 0xN11D PHY MMD Data Register Fh - RESERVED Vendor-Specific Registers 10h - 11h 0xN122 - 0xN123 12h 0xN124 - 0xN125 PHY LinkMD Register 13h 0xN126 - 0xN127 PHY Digital PMA/PCS Status Register 14h - 15h 0xN12A - 0xN12B RESERVED PHY Remote Loopback Register RESERVED Port RXER Count Register 16h-1Ah - 1Bh 0xN136 - 0xN137 Port Interrupt Control / Status Register PHY Auto MDI / MDI-X Register 1Ch 0xN138 - 0xN139 1Dh-1Eh - 1Fh 0xN13E - 0xN13F 4.11.3.2 RESERVED RESERVED PHY Control Register MDIO Manageable Device (MMD) Registers (Indirect) The MIIM interface provides indirect access to a set of MMD registers as defined in Section 5.4, "MDIO Manageable Device (MMD) Registers (Indirect)," on page 199. 4.12 In-Band Management The in-band management access (IBA) is a feature that provides full register read and write access via any one of the seven data ports. Port 7 is the default IBA port. The in-band feature is enabled or disabled by a strapping option at power-up and reset. To use a different port instead of port 7 for IBA, the SPI or I2C interface or IBA must be used to write to a control register. IBA may not be used on more than one port at a time. In-band management frames are processed differently from normal network frames. They are recognized as special frames, so address and VID lookup, VLAN tagging, source address filtering, un-tag discard, tagged frame drop, etc. are not applied to them. Received in-band management frames are never forwarded to the switch fabric or to any other port. The In-Band Management (IBA) Control Register is used to enable and control the IBA feature and to specify one of the seven ports as the IBA port. The IBA frame format is shown in Figure 4-17. The layer 2 portion of the IBA frame contains normal destination address (DA) and source address (SA) fields. The DA of the frames are defined to be the switch MAC address (default 00-10A1-FF-FF-FF), and the SA is the MAC address of the source device. The DA and SA will be swapped in the response frame. A special 4-byte IBA tag follows the SA. This is then followed by the 2-byte EtherType/Length field that serves to identify this as an IBA frame. 2017 Microchip Technology Inc. DS00002391A-page 59 KSZ8567S Only one IBA frame can be processed at a time. Any subsequent IBA frames received by the device will be dropped unless the most recent response frame has been fully transmitted. There are six types of read/write commands: READ, WRITE, WAIT on 0, WAIT on 1, MODIFY to 0 and MODIFY to 1. The burst commands offer fast and bundled data return, up to the capacity of the IBA frame buffer. There are two types of operations in burst command: READ burst and Write burst. FIGURE 4-17: IN-BAND MANAGEMENT FRAME FORMAT LSB 0 1 2 3 4 7 Bytes PREAMBLE 1 Byte SFD 6 Bytes DES. ADDRESS 6 Bytes SRC. ADDRESS 5 6 7 MSB i/g u/l 46 bit address i/g=0, Individual Address; i/g=1, Group Address u/l=0, Globally Admin. Addr. u/l=1, Locally Admin. Addr. 010xxxxx 11111110 p TPID Sequence # 3 bits of Priority, 1 bit of Canonical Format Indicator = 0 4 bits of mode, 8 bits of sequence # 10011000 00000000 0x9800 for 98xx Family Reserved cfi mode 2 Bytes IBAF TAG TYPE 2 Bytes IBAF TAG CONTROL INFO. 2 Bytes MAC LENGTH/TYPE 2 Bytes ACCESS FORMAT 00000000 00000000 2 Bytes Reserved 0x0001 = Rd/Wr, 0x0002 = Burst IBAF TAG 2 Bytes IBAF Layer 3 Data 38/42/46 - 1500 MAC DATA ACCESS CODE 4 Bytes ACCESS COMMAND 4 Bytes DATA 4 Bytes ACCESS COMMAND/DATA COMMAND = 0x0001 (Rd/Wr) OP. 0eeeeAAAAAAAA Code 4 Bytes DATA AAAAAAAAAAAAAAAA 4 Bytes ACCESS COMMAND/DATA 4 Bytes DATA 4 Bytes ACCESS COMMAND/DATA 4 Bytes DATA 4 Bytes ACCESS COMMAND/DATA 4 Bytes DATA COMMAND = 0x0002 (Burst) Dir 000000AAAAAAAA AAAAAAAAAAAAAAAA 3 bits of OP code: 001 = READ, 010 = WRITE, 100 = WAIT on 0, 101 = WAIT on 1, 110 = MODIFY to 0, 111 = MODIFY to 1, COMMAND = 0x0002000 (Dump) = end of command list OP. eeee = Byte enable 0000000000000 Code AAAA..AAAA[23:0] = Register Address 00000000000aaaaa 2 bits of Direction: 01 = READ, 10 = WRITE, AAAA..AAAA[23:0] = Starting Register Address COMMAND = 0x0001 (Rd/Wr) 0xdddddddd dddddddd = data written/read COMMAND = 0x0001 (WAIT/ MODIFY) 0xmaskmask Variable PAD 4 Bytes FRAME CHECK SEQUENCE Variable EXTENSION Mask (1) = bit in reg. to be tested for OP code 100/101 OR set (1) = bit to be set to 0/1 based on OP code 110/111 COMMAND = 0x0002 (Burst) 26 0's and 6 bits of burst count Bits Transmitted from Left to Right 4.13 MAC Interface (Ports 6 and 7) Strapping options are used to individually select any of these MAC interface options for port 6: * Media Independent Interface (MII) (Port 6): Supports 100 and 10 Mbps data rates * Reduced Media Independent Interface (RMII) (Port 6): Supports 100 and 10 Mbps data rates * Reduced Gigabit Media Independent Interface (RGMII) (Port 6): Supports 1000, 100 and 10 Mbps data rates Port 7 is permanently configured as an SGMII port: DS00002391A-page 60 2017 Microchip Technology Inc. KSZ8567S * Serial Gigabit Media Independent Interface (SGMII) (Port 7): Supports 1000, 100 and 10 Mbps data rates Note that the signals on the KSZ8567S MAC interfaces are named as they would be for a PHY: the TX direction is into the KSZ8567S, while the RX direction is out of the KSZ8567S, as if to a host processor with integrated MAC. Signal connection to such a "MAC" device is TX-to-TX, and RX-to-RX. An external PHY (such as the Microchip KSZ9031RNX) may be connected to port 6, but in that case the signal connection will be RX-to-TX, and TX-to-RX. The Port 6 MII/RMII/RGMII interface is powered by the VDDIO power supply. 4.13.1 MEDIA INDEPENDENT INTERFACE (MII) (PORT 6) The media independent interface (MII) is specified in Clause 22 of the IEEE 802.3 standard. It provides a common interface between PHY layer and MAC layer devices. The data interface is 4-bits wide and runs at one quarter the network bit rate; either 2.5MHz in 10BASE-T/Te or 25MHz in 100BASE-TX (not encoded). Additional signals on the transmit side indicate when data is valid or when an error occurs during transmission. Similarly, the receive side provides signals that convey when the data is valid and without physical layer errors. For half duplex operation, the COL signal indicates if a collision has occurred during transmission. The MII interface operates in either PHY Mode or MAC Mode. Select PHY Mode when the port is connected to a processor or other device with a MAC function; select MAC Mode when connecting to an external PHY. Note that the direction of the TX_CLK6, RX_CLK6, COL6 and CRS6 signals is affected by the PHY mode or MAC mode setting, while other MII signals do not change direction. MII mode is selected at reset by a configuration strap option on pins RXD6_3 and RXD6_2 for port 6. The Speed strapping option (on pin RXD6_0 for port 6) should be set for 100/10 Mbps Mode. PHY Mode or MAC Mode is selected by a configuration strap option on pin RXD6_1 (port 6). Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. The interface contains two distinct groups of signals, one for transmission and the other for reception. Table 4-26 and Table 4-27 describe the signals used by the MII interface to connect to an external MAC or to an external PHY, respectively. TABLE 4-26: MII (PHY MODE) CONNECTION TO EXTERNAL MAC MII Interface Signals Description KSZ8567S Signals in PHY Mode External MAC Device Signals Transmit Enable TX_EN6 (input) TX_EN (output) Transit Error TX_ER6 (input) TX_ER (output) Transmit Data Bits [3:0] TXD6_[3:0] (input) TXD[3:0] (output) Transmit Clock TX_CLK6 (output) TX_CLK (input) Collision Detection COL6 (output) COL (input) Carrier Sense CRS6 (output) CRS (input) Received Data Valid RX_DV6 (output) RX_DV (input) Receive Error RX_ER6 (output) RX_ER (input) Receive Data Bits [3:0] RXD6_[3:0] (output) RXD[3:0] (input) Receive Clock RX_CLK6 (output) RX_CLK (input) TABLE 4-27: MII (MAC MODE) CONNECTION TO EXTERNAL PHY MII Interface Signals Description KSZ8567S Signals in MAC Mode External PHY Device Signals Transmit Enable RX_DV6 (output) TX_EN (input) Transit Error RX_ER6 (output) TX_ER (input) Transmit Data Bits [3:0] RXD6_[3:0] (output) TXD[3:0] (input) Transmit Clock RX_CLK6 (input) TX_CLK (output) Collision Detection COL6 (input) COL (output) Carrier Sense CRS6 (input) CRS (output) Received Data Valid TX_EN6 (input) RX_DV (output) 2017 Microchip Technology Inc. DS00002391A-page 61 KSZ8567S TABLE 4-27: MII (MAC MODE) CONNECTION TO EXTERNAL PHY (CONTINUED) MII Interface Signals Description KSZ8567S Signals in MAC Mode External PHY Device Signals Receive Error TX_ER6 (input) RX_ER (output) Receive Data Bits [3:0] TXD6_[3:0] (input) RXD[3:0] (output) Receive Clock TX_CLK6 (input) RX_CLK (output) 4.13.2 REDUCED MEDIA INDEPENDENT INTERFACE (RMII) (PORT 6) The reduced media independent interface (RMII) specifies a low pin count interface, which is based on MII, that provides communication with a MAC attached to the port. As with MII, RMII provides a common interface between physical layer and MAC layer devices, or between two MAC layer devices, and has the following key characteristics: * * * * Supports network data rates of either 10Mbps or 100Mbps. Uses a single 50MHz clock reference (provided internally or externally) for both transmit and receive data. Uses independent 2-bit wide transmit and receive data paths. Contains two distinct groups of signals: one for transmission and the other for reception. The user selects one of the two RMII clocking modes by setting the appropriate strapping option. While in RMII Normal Mode, the port will require an external 50MHz signal to be input to TX_CLK6/REFCLKI6 from an external source. This mode is selected by strapping the RXD6_1 pin high during reset. While in RMII Clock Mode, the port will output a 50MHz clock on RX_CLK6/REFCLKO6, which is derived from the 25MHz crystal or oscillator attached to the XI clock input. The TX_CLK6/REFCLKI6 input is unused in this mode. This mode is selected by strapping the RXD6_1 pin low during reset. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional configuration strap information. Table 4-28 describes the signals used by the RMII interface. Refer to the RMII specification for full details on the signal descriptions. TABLE 4-28: RMII SIGNAL DESCRIPTIONS RMII Signal Name (per spec) RMII Signal (per KSZ8567S) REF_CLK REFCLKI6 Input Input or Output Synchronous 50MHz reference clock, when port is in RMII Normal Mode n/a REFCLKO6 Output Input Synchronous 50MHz reference clock, when port is in RMII Clock Mode TX_EN TX_EN6 Input Output Transmit Enable Transmit Data Bit [1:0] Pin Direction Pin Direction (with Respect to (with Respect to PHY, KSZ8567S) MAC) TXD[1:0] TXD6_[1:0] Input Output CRS_DV RX_DV6 Output Input RX_ER RX_ER6 Output Input or not required RXD[1:0] RXD6_[1:0] Output Input RMII Signal Description Carrier Sense / Receive Data Valid Receive Error Receive Data Bit [1:0] A device port in RMII mode may connect to either an external MAC device (such as a host processor) or to an external PHY; but unlike MII, RMII does not provide separate PHY and MAC modes of operation. However, it is necessary to connect the pins properly. TABLE 4-29: RMII CONNECTION TO EXTERNAL MAC RMII Interface Signals Description KSZ8567S Signals External MAC Device Signals Transmit Enable TX_EN6 (input) TX_EN (output) Transmit Data Bits [1:0] TXD6_[1:0] (input) TXD[1:0] (output) Reference Clock REFCLKI6 (input) or REFCLKO6 (output) REF_CLK (input or output) DS00002391A-page 62 2017 Microchip Technology Inc. KSZ8567S TABLE 4-29: RMII CONNECTION TO EXTERNAL MAC (CONTINUED) RMII Interface Signals Description KSZ8567S Signals External MAC Device Signals Carrier Sense Data Valid RX_DV6 (output) CRS_DV (input) Receive Error RX_ER6 (output) RX_ER (input) Receive Data Bits [1:0] RXD6_[1:0] (output) RXD[1:0] (input) TABLE 4-30: RMII CONNECTION TO EXTERNAL PHY RMII Interface Signals Description KSZ8567S Signals External PHY Device Signals Transmit Enable RX_DV6 (output) TX_EN (input) Transmit Data Bits [1:0] RXD6_[1:0] (output) TXD[1:0] (input) Reference Clock REFCLKI6 (input) or REFCLKO6 (output) REF_CLK (input or output) Carrier Sense Data Valid TX_EN6 (input) CRS_DV (output) Receive Error No connection RX_ER (output) Receive Data Bits [1:0] TXD6_[1:0] (input) RXD[1:0] (output) 4.13.3 REDUCED GIGABIT MEDIA INDEPENDENT INTERFACE (RGMII) (PORT 6) RGMII provides a common interface between RGMII PHYs and MACs, and has the following key characteristics: * * * * Pin count is reduced from 24 pins for GMII to 12 pins for RGMII. All speeds (10Mbps, 100Mbps and 1000Mbps) are supported at both half- and full-duplex. Data transmission and reception are independent and belong to separate signal groups. Transmit data and receive data are each four bits wide - a nibble. In RGMII operation, the RGMII pins function as follows: * The MAC sources the transmit reference clock, TX_CLK6, at 125MHz for 1000Mbps, 25MHz for 100Mbps, and 2.5MHz for 10Mbps. * The PHY recovers and sources the receive reference clock, RX_CLK6, at 125MHz for 1000Mbps, 25MHz for 100Mbps, and 2.5MHz for 10Mbps. * For 1000BASE-T, the transmit data, TXD6_[3:0], is presented on both edges of TX_CLK6, and the received data, RXD6_[3:0], is clocked out on both edges of the recovered 125MHz clock, RX_CLK6. * For 10BASE-T/100BASE-TX, the MAC holds TX_CTL6 low until both the PHY and MAC operate at the same speed. During the speed transition, the receive clock is stretched on either a positive of negative pulse to ensure that no clock glitch is presented to the MAC. * TX_ER6 and RX_ER6 are combined with TX_EN6 and RX_DV6, respectively, to form TX_CTL6 and RX_CTL6. These two RGMII control signals are valid at the falling clock edge. After power-up or reset, the device is configured to RGMII mode if the appropriate configuration strap pins are set to one of the RGMII mode capability options. Refer to Section 3.2.1, "Configuration Straps," on page 17 for available options. Note that there is no mechanism for the RGMII interface to adapt its speed automatically to the speed of the connected RGMII device. A configuration strap option sets the speed of the RGMII interface at power-up to either 1000Mbps or 100Mbps. A control register can override the configuration strap option and set the RGMII speed to either 1000, 100 or 10Mbps. If a PHY is connected to the RGMII port, it should be ensured that the PHY link speed is fixed in order to avoid a mismatch to the RGMII speed. The device provides the option to add a minimum of 1.5ns internal delay to either TX_CLK6 or RX_CLK6, via the RGMII Internal Delay control bits in the XMII Port Control 1 Register. This can reduce or eliminate the need to add trace delay to the clock signals on the printed circuit board. RGMII_ID_ig enables delay on TX_CLK6, and the default is off. RGMII_ID_eg enables delay on RX_CLK6, and the default is on. Users should also be aware of any internal clock delay that may be added by the connected RGMII device. 2017 Microchip Technology Inc. DS00002391A-page 63 KSZ8567S TABLE 4-31: RGMII SIGNAL DESCRIPTIONS RGMII Signal Name (per spec) RGMII Signal (per KSZ8567S) Pin Direction (with respect to PHY, KSZ8567S) Pin Direction (with respect to MAC) TXC TX_CLK6 Input Output Transmit Reference Clock (125MHz for 1000Mbps, 25MHz for 100Mbps, 2.5MHz for 10Mbps) TX_CTL TX_CTL6 Input Output Transmit Control TXD[3:0] TXD6_[3:0] Input Output RXC Transmit Data [3:0] Receive Reference Clock (125MHz for 1000Mbps, 25MHz for 100Mbps, 2.5MHz for 10Mbps) RX_CLK6 Output RX_CTL RX_CTL6 Output Input Receive Control RXD[3:0] RXD6_[3:0] Output Input Receive Data [3:0] 4.13.4 Input RGMII Signal Description SERIAL GIGABIT MEDIA INDEPENDENT INTERFACE (SGMII) (PORT 7) The port 7 MAC has a Serial Gigabit Media Independent Interface (SGMII) for interfacing to an external 10/100/1000 BASE-T PHY that supports the SGMII MAC/PHY interface, or to another device with an SGMII interface. The interface has one receive differential pair and one transmit differential pair for sending and receiving data and control at a serial bit rate of 1.25 Gbaud. The SGMII block recovers the clock from the incoming data. Therefore, a separate input SGMII clock is not needed. Likewise, no output SGMII is provided, with the expectation that the connected device will also recover the clock from the receive data. SGMII uses auto-negotiation to establish the MAC - PHY connection. The PHY side device is sometimes referred to as the master, and the MAC side device is referred to as the slave. The PHY side device relays the following information about the 10/100/1000 BASE-T link to the MAC side device: * Speed * Duplex * Link up/down status The SGMII interface defaults to MAC side mode, with auto-negotiation enabled. The mode can be changed to PHY side for connection to a MAC side device. There is also the option to disable auto-negotiation. If auto-negotiation is disabled, it must also be disabled in both devices. Do not confuse "auto-negotiation" across the SGMII interface with auto-negotiation performed between the PHY and its far-end link partner. When the SGMII speed and duplex can be fixed, as when connected to another switch or a processor, auto-negotiation is optional. 10/100/1000 BASE-T PHYs with SGMII interfaces may be discrete devices or may be housed in SFP modules. Note, however, that not all SFP modules are SGMII. Specifically, SFP fiber optic modules such as 1000BASE-SX and 1000BASE-LX are not SGMII devices even though their electrical interface is compatible. AC coupling should be used on the SGMII differential pairs. AC coupling capacitors are included in SFP modules. SGMII registers are accessed indirectly via the Port SGMII Address Register and Port SGMII Data Register. Additional information on SGMII registers is provided in Section 5.5, "SGMII Registers (Indirect)," on page 201. DS00002391A-page 64 2017 Microchip Technology Inc. KSZ8567S 5.0 DEVICE REGISTERS The KSZ8567S has a rich set of registers for device management. The registers are accessed by the SPI or I2C interfaces, or by in-band management. Alternatively, the MIIM interface can be used to access the PHY registers only. The MIIM interface cannot access the switch registers. A 16-bit address is used to access the device registers. This address is split into three hierarchical spaces, as shown in Figure 5-1. These three spaces are used to designate the port/channel (4-bits), function (page) of the port (4-bits), and register of function (8-bits). The individual ports are numbered 1 through 7. In the port space, a value of 0 is used for global registers. Address bit 15 is always 0. FIGURE 5-1: Examples: Register Address REGISTER ADDRESS MAPPING 0 N N N 0 1 0 0 0 0 0 0 1 0 0 0 = 0xN408 0 0 0 0 0 0 1 1 0 0 0 1 0 0 1 0 = 0x0312 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 Port Space Function Space (Page) Register Space 0 = Global 0 = Operation 1 = I/O Interface 2 = PHY 3 = General 4 = Look-up Tables 5 = IEEE 1588 PTP 6-F = Reserved N = Port # 0 = Operation 1 = PHY 2 = SGMII 3 = RGMII / MII / RMII 4 = MAC 5 = MIB Counters 6 = ACL 7 = Reserved 8 = Ingress - Classification, Policing 9 = Egress - Shaping A = Queue management B = Address lookup engine C = IEEE 1588 PTP D-F = Reserved Values of N are 1 - 7 2017 Microchip Technology Inc. Control, Status, Storage, Etc. DS00002391A-page 65 KSZ8567S Register addressing is by bytes, and the management interface (SPI, I2C or in-band) transfers data by bytes. Where registers are shown as 16-bits or 32-bits, this is for descriptive purposes only. Data can always be written and read as individual bytes and in any order. For multi-byte registers, the data is addressed in a big-endian format, with the most significant byte at the lowest address, and the least significant byte at the highest address, as shown in Figure 5-2. FIGURE 5-2: BYTE ORDERING 16-bit register 32-bit register 0A0B0 0A0B0C0D 0A address a+1 0B 0A address a+1 0B address a+2 0C address a address a address a+3 memory 0D memory The global and port register address maps are detailed in Table 5-1 and Table 5-2, respectively. Table 1-3, "Register Nomenclature," on page 7 provides a list of register bit type notations. The remainder of this chapter is organized as follows: * * * * * Global Registers Port Registers Tables and MIB Counters (Access) MDIO Manageable Device (MMD) Registers (Indirect) SGMII Registers (Indirect) TABLE 5-1: GLOBAL REGISTER ADDRESS MAP Address Functional Group 0x0000 - 0x00FF Global Operation Control Registers (0x0000 - 0x00FF) 0x0100 - 0x01FF Global I/O Control Registers (0x0100 - 0x01FF) 0x0200 - 0x02FF Global PHY Control and Status Registers (0x0200 - 0x02FF) 0x0300 - 0x03FF Global Switch Control Registers (0x0300 - 0x03FF) 0x0400 - 0x04FF Global Switch Look Up Engine (LUE) Control Registers (0x0400 - 0x04FF) 0x0500 - 0x05FF Global Switch PTP Control Registers (0x0500 - 0x05FF) 0x0600 - 0x0FFF RESERVED DS00002391A-page 66 2017 Microchip Technology Inc. KSZ8567S TABLE 5-2: PORT N (1-7) REGISTER ADDRESS MAP Address Functional Group 0xN000 - 0xN0FF Port N: Port Operation Control Registers (0xN000 - 0xN0FF) 0xN100 - 0xN1FF Port N: Port Ethernet PHY Registers (0xN100 - 0xN1FF) 0xN200 - 0xN2FF Port N: Port SGMII Control Registers (0xN200 - 0xN2FF) 0xN300 - 0xN3FF Port N: Port RGMII/MII/RMII Control Registers (0xN300 - 0xN3FF) 0xN400 - 0xN4FF Port N: Port Switch MAC Control Registers (0xN400 - 0xN4FF) 0xN500 - 0xN5FF Port N: Port Switch MIB Counters Registers (0xN500 - 0xN5FF) 0xN600 - 0xN6FF Port N: Port Switch ACL Control Registers (0xN600 - 0xN6FF) 0xN700 - 0xN7FF RESERVED 0xN800 - 0xN8FF Port N: Port Switch Ingress Control Registers (0xN800 - 0xN8FF) 0xN900 - 0xN9FF Port N: Port Switch Egress Control Registers (0xN900 - 0xN9FF) 0xNA00 - 0xNAFF Port N: Port Switch Queue Management Control Registers (0xNA00 - 0xNAFF) 0xNB00 - 0xNBFF Port N: Port Switch Address Lookup Control Registers (0xNB00 - 0xNBFF) 0xNC00 - 0xNCFF Port N: Port Switch PTP Control Registers (0xNC00 - 0xNCFF) 0xND00 - 0xNFFF RESERVED Note: RESERVED address space must not be written under any circumstances. Failure to heed this warning may result in untoward operation and unexpected results. If it is necessary to write to registers which contain both writable and reserved bits in the same register, the user should first read back the reserved bits (RO or R/W), "OR" the desired settable bits with the value read, and then write back the "ORed" value to the register. 2017 Microchip Technology Inc. DS00002391A-page 67 KSZ8567S 5.1 Global Registers This section details the device's global registers. For an overview of the device's entire register map, refer to Section 5.0, "Device Registers". For details on the device's port registers, refer to Section 5.2, "Port Registers". 5.1.1 5.1.1.1 GLOBAL OPERATION CONTROL REGISTERS (0x0000 - 0x00FF) Global Chip ID 0 Register Address: Bits 7:0 5.1.1.2 0x0001 Size: Default RO 0x00 Type Default RO 0x85 Type Default RO 0x67 Type Default 8 bits Description Chip ID (MSB) Global Chip ID 2 Register Bits 5.1.1.4 Type Global Chip ID 1 Register Address: 7:0 8 bits Fixed Value Bits 5.1.1.3 Size: Description Address: 7:0 0x0000 0x0002 Size: 8 bits Description Chip ID (LSB) Global Chip ID 3 Register Address: Bits 0x0003 Size: 8 bits Description 7:4 Revision ID RO - 3:1 RESERVED RO - Global Software Reset Refer to the Switch Operation Register for another reset control bit. 0 = Normal operation 1 = Resets the data path and state machines, but not register values. R/W SC 0b 0 DS00002391A-page 68 2017 Microchip Technology Inc. KSZ8567S 5.1.1.5 PME Pin Control Register Address: Bits 0x0006 Size: 8 bits Type Default RESERVED RO - 1 PME Pin Output Enable 0 = Disabled 1 = Enabled R/W 0b 0 PME Pin Output Polarity 0 = PME is active low 1 = PME is active high R/W 0b 7:2 5.1.1.6 Description Global Interrupt Status Register Address: 0x0010 - 0x0013 Size: 32 bits This register provides the top level interrupt status for the LUE and GPIO trigger and timestamp functions. These interrupts are enabled in the Global Interrupt Mask Register. For port specific interrupts, refer to the Port Interrupt Status Register. Bits 31 Description Lookup Engine (LUE) Interrupt Status Type Default RO 0b RO 0b RO - Refer to the Address Lookup Table Interrupt Register for detailed LUE interrupt status bits. 0 = No interrupt 1 = Interrupt request 30 GPIO Pin Output Trigger and Timestamp Unit Interrupt Status Refer to the GPIO Status Monitor 1 Register to determine whether it is a Trigger Output Unit interrupt or a Timestamp Unit interrupt. For the interrupt status for the port-based time stamping of PTP egress frames, refer to the Port Interrupt Status Register and Port PTP Timestamp Interrupt Status Register. 0 = No interrupt 1 = Interrupt request 29:0 RESERVED 2017 Microchip Technology Inc. DS00002391A-page 69 KSZ8567S 5.1.1.7 Global Interrupt Mask Register Address: 0x0014 - 0x0017 Size: 32 bits This register enables the interrupts in the Global Interrupt Status Register. Bits 31 Description Lookup Engine (LUE) Interrupt Mask Type Default R/W 0b R/W 0b RO - 0 = Interrupt enabled 1 = Interrupt disabled 30 GPIO Pin Output Trigger and Timestamp Unit Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 29:0 RESERVED DS00002391A-page 70 2017 Microchip Technology Inc. KSZ8567S 5.1.1.8 Global Port Interrupt Status Register Address: 0x0018 - 0x001B Size: 32 bits This register provides the top level interrupt status for the individual ports. These interrupts are enabled in the Global Port Interrupt Mask Register. Refer to the Port Interrupt Status Register for detailed port interrupt status. Bits 31:7 6 Description Type Default RESERVED RO - Port 7 Interrupt Status RO 0b RO 0b RO 0b RO 0b RO 0b RO 0b RO 0b 0 = No interrupt 1 = Interrupt request 5 Port 6 Interrupt Status 0 = No interrupt 1 = Interrupt request 4 Port 5 Interrupt Status 0 = No interrupt 1 = Interrupt request 3 Port 4 Interrupt Status 0 = No interrupt 1 = Interrupt request 2 Port 3 Interrupt Status 0 = No interrupt 1 = Interrupt request 1 Port 2 Interrupt Status 0 = No interrupt 1 = Interrupt request 0 Port 1 Interrupt Status 0 = No interrupt 1 = Interrupt request 2017 Microchip Technology Inc. DS00002391A-page 71 KSZ8567S 5.1.1.9 Global Port Interrupt Mask Register Address: 0x001C - 0x001F Size: 32 bits This register enables the interrupts in the Global Port Interrupt Status Register. Bits 31:7 6 Description Type Default RESERVED RO - Port 7 Interrupt Mask R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b 0 = Interrupt enabled 1 = Interrupt disabled 5 Port 6 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 4 Port 5 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 3 Port 4 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 2 Port 3 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 1 Port 2 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled 0 Port 1 Interrupt Mask 0 = Interrupt enabled 1 = Interrupt disabled DS00002391A-page 72 2017 Microchip Technology Inc. KSZ8567S 5.1.2 5.1.2.1 GLOBAL I/O CONTROL REGISTERS (0x0100 - 0x01FF) Serial I/O Control Register Address: Bits 7:3 2 0x0100 Size: 8 bits Description Type Default RESERVED R/W 0100_0b MIIM Preamble Suppression R/W 0b R/W 1b R/W 0b This feature affects only the MIIM (MDIO / MDC) interface. When using SPI or I2C, this bit has no effect. 0 = Normal operation. The switch always expects the MIIM preamble. 1 = The switch will respond to MIIM commands even in the absence of a preamble. 1 Automatic SPI Data Out Edge Select When enabled, this feature automatically determines the edge of SCL that is used to clock out the SPI data on SDO. If SCL ~25MHz, SDO data is clocked by the rising edge of SCL. If SCL < ~25 MHz, SDO data is clocked by the falling edge of SCL. 0 = The automatic feature is disabled, and bit 0 determines the SCL clock edge used for SDO. 1 = The automatic feature is enabled, and bit 0 is ignored. 0 SPI Data Out Edge Select When bit 1 is zero, then this bit determines the clock edge used for SPI data out. When bit 1 is set to 1, this bit is ignored. 0 = SDO data is clocked by the falling edge of SCL 1 = SDO data is clocked by the rising edge of SCL 2017 Microchip Technology Inc. DS00002391A-page 73 KSZ8567S 5.1.2.2 Output Clock Control Register Address: 0x0103 Bits 7 Description Recovered Clock Ready (REC_CLK_RDY) Size: 8 bits Type Default RO - 0 = The selected recovered clock is not ready 1 = The selected recovered clock is ready 6:5 RESERVED RO 00b 4:2 SYNCLKO Source R/W 000b R/W 1b R/W 0b 000 = From crystal / clock input at XI pin 001 = From port 1 recovered clock 010 = From port 2 recovered clock 011 = From port 3 recovered clock 100 = From port 4 recovered clock 101 = From port 5 recovered clock 110 - 111 = Reserved 1 SYNCLKO Output Pin Enable 0 = Disabled 1 = Enabled 0 SYNCLKO Frequency 0 = 25 MHz 1 = 125 MHz DS00002391A-page 74 2017 Microchip Technology Inc. KSZ8567S 5.1.2.3 In-Band Management (IBA) Control Register Address: 0x0104 - 0x0107 Size: 32 bits This register controls the In-Band Access (IBA) feature. Bits 31 Description IBA Enable Type Default R/W Note 5-1 R/W 0b 0b The initial value is strapped in from the IBA pin. 0 = Disabled 1 = Enabled 30 IBA Destination MAC Address Match Enable Set this bit to enable checking of the destination MAC address in received IBA frames against the switch MAC address in the Switch MAC Address 0 Register through Switch MAC Address 5 Register. Non-matching frames are discarded. When not enabled, the MAC address is not checked. 29 IBA Reset Set this bit to initialize the IBA state machine. This bit is self-clearing. R/W SC 28:24 RESERVED RO 0x00 23:22 Priority Queue for IBA response R/W 01b Specifies the transmit priority queue for the IBA response frame. Typically this value is not changed. 21:19 RESERVED RO 00_0b 18:16 Port used for IBA communication R/W 110b R/W 0x40FE 000 = Port 1 001 = Port 2 010 = Port 3 011 = Port 4 100 = Port 5 101 = Port 6 110 = Port 7 111 = Reserved 15:0 Note 5-1 TPID (EtherType) value for IBA frame header The default value of this field is determined by the associated configuration strap value. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. 2017 Microchip Technology Inc. DS00002391A-page 75 KSZ8567S 5.1.2.4 I/O Drive Strength Register Address: Bits 7 6:4 0x010D Size: 8 bits Description Type Default RESERVED R/W 0b High Speed Drive Strength (24mA) R/W 110b RESERVED R/W 0b Low Speed Drive Strength (8mA) R/W 10b Type Default RO 0b RO 0b RO 0b RESERVED RO 0x0000 IBA MAC Address Mismatch Error RO 0b RO 0b RO 0b Controls drive strength of RGMII / MII / RMII (except TX_CLK / REFCLKI, COL and CRS) and SYNCLKO. 3 2:0 Controls drive strength of TX_CLK / REFCLKI, COL, CRS, LEDs, PME_N, INTRP_N, SDO and SDI/SDA/MDIO. 5.1.2.5 In-Band Management (IBA) Operation Status 1 Register Address: Bits 31 0x0110 - 0x0113 Size: 32 bits Description Good IBA Packet Detect 1 = A good IBA packet is received. 30 IBA Response Packet Transmit Done 1 = An IBA response packet is sent out. This bit is cleared when a packet with a matching IBA tag field is received. 29 IBA Execution Done 1 = All the commands in one IBA packet are completely executed. This bit is cleared when a packet with a matching IBA tag field is received. 28:15 14 This bit is active only when IBA_ENABLE (In-Band Management (IBA) Control Register, bit 30) is set. 1 = An IBA packet is received with an unmatched MAC address, unequal to the switch's MAC address. This bit is cleared when a packet with a matching IBA tag field is received. 13 IBA Access Format Error 1 = An IBA packet with a wrong access format (not equal to 0x9800) is received. This bit is cleared when a packet with a matching IBA tag field is received. 12 IBA Access Code Error 1 = An IBA packet with an unrecognized access code is received. (Valid access codes are 0x0001 and 0x0002.) This bit is cleared when a packet with a matching IBA tag field is received. DS00002391A-page 76 2017 Microchip Technology Inc. KSZ8567S Bits 11 Description IBA Access Command Error Type Default RO 0b RO 0b 1 = An IBA packet with an unrecognized command code is received. This bit is cleared when a packet with a matching IBA tag field is received. 10 IBA Oversize Packet Error 1 = An oversized IBA packet is received. The maximum IBA packet size is 320 bytes, including 8-byte zeros before FCS and the 4-byte FCS. No response packet is sent. This bit is cleared when a packet with a matching IBA tag field is received. 9:7 RESERVED RO 000b 6:0 IBA Access Code Error Location RO 0x000 Type Default RO 0x00000 When IBA Access Command Error (bit 11) is set, these bits indicate the address location of the wrong command code within the IBA packet. 5.1.2.6 LED Override Register Address: Bits 31:10 9:0 5.1.2.7 32 bits RESERVED Override LED These bits select whether each LEDx_0 and LEDx_1 pin will function as an LED or General Purpose Output (GPO). The LSB bit of this field represents LED1_0, followed by LED1_1, LED2_0, etc. When configured as a GPO, the GPO output is controlled via the LED Output Register. 0 = LEDx_y pin functions as an LED 1 = LEDx_y pin functions as a GPO 0000000000b LED Output Register Bits 9:0 Size: Description Address: 31:10 0x0120 - 0x0123 0x0124 - 0x0127 Size: 32 bits Description Type Default RESERVED RO 0x00000 GPO Output Control When configured as a GPO via the LED Override Register, the GPO output is controlled via this field. The LSB bit of this field represents LED1_0, followed by LED1_1, LED2_0, etc. 0 = LEDx_y pin outputs low 1 = LEDx_y pin outputs high R/W 0000000000b 2017 Microchip Technology Inc. DS00002391A-page 77 KSZ8567S 5.1.2.8 LED2_0/LED2_1 Source Register Address: Bits 0x0128 - 0x012B Size: 32 bits Type Default RESERVED RO 0x0000000 3 LED2_1 Source 0 = LED2_1 outputs as LED/GPO (configured via the LED Override Register) 1 = LED2_1 outputs the PTP Trigger Output 1 R/W 0b 2 LED2_0 Source 0 = LED2_0 outputs as LED/GPO (configured via the LED Override Register) 1 = LED2_0 outputs the PTP Trigger Output 0 R/W 0b RESERVED RO 00b Type Default RESERVED RO 00b PLL Power Down R/W 0b R/W 00b RO 000b Type Default RESERVED RO 0x000000 Configuration strap values of LED pins RO Note 5-2 31:4 1:0 5.1.3 5.1.3.1 Description GLOBAL PHY CONTROL AND STATUS REGISTERS (0x0200 - 0x02FF) Power Down Control 0 Register Address: Bits 7:6 5 0x0201 Size: 8 bits Description 0 = Normal operation. 1 = Disable PLL. This may be used in combination with EDPD mode - see below. 4:3 Power Management Mode 00 = Normal operation 01 = Energy Detect Power Down (EDPD) Mode 10 = Soft Power Down Mode 11 = invalid 2:0 5.1.3.2 RESERVED LED Configuration Strap Register Address: Bits 31:10 9:0 0x0210 - 0x0213 Size: 32 bits Description [LED5_1, LED5_0, LED4_1, LED4_0, LED3_1, LED3_0, LED2_1, LED2_0, LED1_1, LED1_0] DS00002391A-page 78 2017 Microchip Technology Inc. KSZ8567S Note 5-2 5.1.4 5.1.4.1 The default value of this field is determined by the associated configuration strap values. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. GLOBAL SWITCH CONTROL REGISTERS (0x0300 - 0x03FF) Switch Operation Register Address: Bits 7 0x0300 Size: 8 bits Description Type Default R/W 0b RESERVED RO 0x00 Soft Hardware Reset R/W SC 0b R/W Note 5-3 Double Tag Enable 1 = Double tagging is enabled 0 = Double tagging is disabled 6:2 1 When set to 1, all register settings, except configuration strap options, are reset to default values. 0 Start Switch 1 = Switch function is enabled 0 = Switch function is disabled; no traffic will be passed until this bit is set Note 5-3 5.1.4.2 The default value of this field is determined by the LED5_1 configuration strap value. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. Switch MAC Address 0 Register Address: Bits 7:0 0x0302 Size: 8 bits Description MAC Address [47:40] Type Default R/W 0x00 Type Default R/W 0x10 This register, along with the Switch MAC Address 1-5 Registers, define the switch's MAC address to be used as the source address in MAC pause control frames, and for self-address filtering. 5.1.4.3 Switch MAC Address 1 Register Address: Bits 7:0 0x0303 Description MAC Address [39:32] 2017 Microchip Technology Inc. Size: 8 bits DS00002391A-page 79 KSZ8567S 5.1.4.4 Switch MAC Address 2 Register Address: Bits 7:0 5.1.4.5 Size: Default R/W 0xA1 Type Default R/W 0xFF Type Default R/W 0xFF Type Default R/W 0xFF 8 bits MAC Address [23:16] Switch MAC Address 4 Register 0x0306 Size: 8 bits Description MAC Address [15:8] Switch MAC Address 5 Register Address: Bits 7:0 0x0305 Description Bits 5.1.4.7 Type Switch MAC Address 3 Register Address: 7:0 8 bits MAC Address [31:24] Bits 5.1.4.6 Size: Description Address: 7:0 0x0304 0x0307 Description MAC Address [7:0] DS00002391A-page 80 Size: 8 bits 2017 Microchip Technology Inc. KSZ8567S 5.1.4.8 Switch Maximum Transmit Unit Register Address: Bits 0x0308 - 0x0309 Size: 16 bits Description Type Default 15:14 RESERVED R/W 00b 13:0 Maximum Frame Length (MTU) R/W 0x07D0 Type Default R/W 0x9100 Type Default RESERVED RO 0x0000 Shaping Credit Accounting R/W 1b R/W 1b Specifies the maximum transmission unit (MTU), which is the maximum frame payload size. Frames which exceed this maximum are truncated. This value can be set as high as 9000 (= 0x2328) if jumbo frame support is required. Also refer to the Switch MAC Control 1 Register and Port MAC Control 0 Register. 5.1.4.9 Switch ISP TPID Register Address: Bits 15:0 0x030A - 0x030B Size: 16 bits Description ISP Tag TPID Default tag TPID (EtherType) for untagged incoming frames or the ISP frame tag TPID for the double tagging function. 5.1.4.10 AVB Credit Based Shaper Strategy Register Address: Bits 15:2 1 0x030E - 0x030F Size: 16 bits Description 1 = Shaper credit deduction occurs on both data and IPG + preamble 0 = Shaper credit deduction occurs on data only 0 Policing Credit Accounting 1 = Policing credit deduction occurs on both data and IPG + preamble 0 = Policing credit deduction occurs on data only 2017 Microchip Technology Inc. DS00002391A-page 81 KSZ8567S 5.1.4.11 Switch Lookup Engine Control 0 Register Address: Bits 7 0x0310 Size: 8 bits Description 802.1Q VLAN Enable Type Default R/W 0b R/W 1b R/W 10_0b R/W 0b R/W 01b This is the master enable for VLAN forwarding and filtering. Note that the VLAN Table must be set up before VLAN mode is enabled. 1 = VLAN mode enabled 0 = VLAN mode disabled 6 Drop Invalid VID 1 = All received packets with invalid VLAN ID are dropped. 0 = Received packets with invalid VLAN ID are forwarded to the host port. Note that the Unknown VID Forwarding feature (Unknown VLAN ID Control Register), if enabled, takes precedence over this bit. 5:3 Age Count This bit, in combination with the Age Period value (Switch Lookup Engine Control 3 Register), determines the aging time of dynamic entries in the address lookup table. This value is used for the Age Count field whenever a dynamic table entry is updated. 2 Reserved Multicast Lookup Enable 1 = Enable Reserved Multicast Table 0 = Disable Reserved Multicast Table 1:0 HASH_OPTION Defines the hashing option for mapping entries to the dynamic lookup table. 00, 11 = Entry is mapped directly using the 10 least significant bits of the destination address. 01 = The CRC hashing function is used. 10 = The XOR hashing function is used. Refer to Section 4.4.2.1, "Address Lookup (ALU) Table," on page 27 for additional information. DS00002391A-page 82 2017 Microchip Technology Inc. KSZ8567S 5.1.4.12 Switch Lookup Engine Control 1 Register Address: 0x0311 Bits 7 Size: 8 bits Description Unicast Learning Disable Type Default R/W 0b R/W 0b R/W SC 0b R/W SC 0b R/W 1b R/W 1b R/W 0b R/W 0b 1 = Unicast address learning is disabled 0 = Unicast address learning is enabled 6 Self-Address Filtering - Global Enable The source address of received packets is compared to the MAC address in registers Switch MAC Address 0 Register through Switch MAC Address 5 Register, and the packet is dropped if there is a match. Self-address filtering can be enabled on a port-by-port basis by setting the port enable bit in the Port Control 2 Register in addition to setting this bit. 1 = Enable self-address filtering globally for those ports whose port enable bit (Port Control 2 Register) is set. 0 = Do not filter self-addressed packets on any port. 5 Flush Address Lookup Table The Flush Option bit in the Switch Lookup Engine Control 2 Register determines whether flushing is performed on dynamic entries, static entries, or both. 1 = Trigger a flush of the entire address lookup table. The static address table is not flushed. 0 = Normal operation 4 Flush MSTP Address Entries (Address Lookup Table) The Flush Option bit in the Switch Lookup Engine Control 2 Register determines whether flushing is performed on dynamic entries, static entries, or both. 1 = Trigger a flush of the matched MSTP entries 0 = Normal operation 3 Multicast Source Address Filtering 1 = Forward packets with a multicast source address 0 = Drop packets with a multicast source address 2 Aging Enable 1 = Enable address table aging 0 = Disable address table aging 1 Fast Aging 1 = Enable fast aging 0 = Disable fast aging 0 Link Down Flush 1 = Link down will cause the entries of any link down port to be flushed 0 = Link down flush is disabled 2017 Microchip Technology Inc. DS00002391A-page 83 KSZ8567S 5.1.4.13 Switch Lookup Engine Control 2 Register Address: Bits 0x0312 Size: 8 bits Description Type Default 7 RESERVED R/W 0b 6 Double Tag Multicast Trap R/W 0b R/W 0b R/W 0b R/W 00b R/W 00b 1 = When double tagging mode is enabled, forward all reserved multicast packets to the host port only. 0 = Normal forwarding 5 Dynamic Entry Egress VLAN Filtering Egress VLAN filtering uses the forwarding port map from the VLAN table to restrict the forwarding ports determined from the address lookup. This is the recommended mode of operation when VLAN is enabled. The default value is 0 only for backwards compatibility with previous switches. 1 = Enable. For successful lookup of a dynamic entry in the address table, the forwarding ports are determined from the AND function of the address table port map and the VLAN table port map. 0 = Disable. For successful lookup of a dynamic entry in the address table, the forwarding ports are determined from the address table only. 4 Static Entry Egress VLAN Filtering Egress VLAN filtering uses the forwarding port map from the VLAN table to restrict the forwarding ports determined from the address lookup. This is the recommended mode of operation when VLAN is enabled. The default value is 0 only for backwards compatibility with previous switches. 1 = Enable. For successful lookup of a static entry in the address table, the forwarding ports are determined from the AND function of the address table port map and the VLAN table port map. 0 = Disable. For successful lookup of a static entry in the address table, the forwarding ports are determined from the address table only. 3:2 Flush Option Determines which address lookup table entries may be flushed by either of the flush operations in the Switch Lookup Engine Control 1 Register. 00 = No flush or flush is done 01 = Flush only dynamic table entries 10 = Flush only static table entries 11 = Flush both static and dynamic table entries 1:0 MAC Address Priority 00 = MAC Address (MACA) priority for a packet is determined from the destination address (DA) lookup 01 = MACA priority for a packet is determined from the source address (SA) lookup 10 = MACA priority for a packet is determined from the higher of the DA and SA lookups 11 = MACA priority for a packet is determined from the lower of the DA and SA lookups DS00002391A-page 84 2017 Microchip Technology Inc. KSZ8567S 5.1.4.14 Switch Lookup Engine Control 3 Register Address: Bits 7:0 0x0313 Size: 8 bits Description Age Period Type Default R/W 0x4B This value, multiplied by the Age Count value in the entries of the Address Lookup Table, determines the aging time of dynamic entries in that table. The unit is seconds. 5.1.4.15 Address Lookup Table Interrupt Register Address: 0x0314 Size: 8 bits This register provides the detailed interrupt status for the Address Lookup Table. These interrupts are enabled in the Address Lookup Table Mask Register. The LUE interrupt status bit in the Global Interrupt Status Register is the OR of the status bits in this register. Bits 7:3 2 Description RESERVED Learn Fail Interrupt Status Type Default RO 0x00 R/WC 0b R/WC 0b R/WC 0b An Address Lookup Table entry was not learned because all entries in the bucket are static 1 Almost Full Interrupt Status Interrupt indicates that the Address Lookup Table bucket was almost full (2 or 3 valid entries) when a new static entry was written. 0 Write Fail Interrupt Status Interrupt indicates that the Address Lookup Table bucket is full and a write failed 2017 Microchip Technology Inc. DS00002391A-page 85 KSZ8567S 5.1.4.16 Address Lookup Table Mask Register Address: 0x0315 Size: 8 bits This register masks the Address Lookup Table interrupts in the Address Lookup Table Interrupt Register. Bits 7:3 2 Description Type Default RESERVED RO 0x00 Learn Fail Interrupt Mask R/W 1b R/W 1b R/W 1b Type Default 1 = Interrupt is disabled 0 = Interrupt is enabled 1 Almost Full Interrupt Mask 1 = Interrupt is disabled 0 = Interrupt is enabled 0 Write Fail Interrupt Mask 1 = Interrupt is disabled 0 = Interrupt is enabled 5.1.4.17 Address Lookup Table Entry Index 0 Register Address: 0x0316 - 0x0317 Bits Size: 16 bits Description 15:12 RESERVED RO 0x0 11:0 / 9:0 Almost Full Entry Index [11:0] RO 0x000 Type Default RESERVED RO 0000_00 Fail Learn Index RO 0x000 When a static entry is successfully written into the Address Lookup Table, but the table bucket is almost full (contains 2 or 3 static entries prior to the write), the entry address is reported here. Fail Write Index [9:0] When a static entry write failure occurs in the Address Lookup Table, the bucket address is reported here. 5.1.4.18 Address Lookup Table Entry Index 1 Register Address: Bits 15:10 9:0 0x0318 - 0x0319 Size: 16 bits Description When a destination address fails to be learned in the Address Lookup Table because the bucket contains 4 static entries, the bucket address is reported here. DS00002391A-page 86 2017 Microchip Technology Inc. KSZ8567S 5.1.4.19 Address Lookup Table Entry Index 2 Register Address: 0x031A - 0x031B Bits 15:10 9:0 Size: 16 bits Description Type Default RESERVED RO 0000_00 CPU Access Index RO 0x000 Whenever there is an external read or write to the Address Lookup Table, the bucket address of the access is reported here. 5.1.4.20 Unknown Unicast Control Register Address: 0x0320 - 0x0323 Size: 32 bits The following three registers control forwarding of packets with 1) unknown unicast destination address, 2) unknown multicast destination address, and 3) unknown VLAN ID. If a received packet falls into more than one of these categories, the precedence is: 1. 2. 3. Unknown VID Unknown Unicast Unknown Multicast Bits 31 Description Unknown Unicast Packet Forward Type Default R/W 0b 1 = Enable forwarding of unknown unicast packets to the ports specified below 0 = Disable unknown unicast packet forwarding 30:7 RESERVED RO 0x000000 6:0 Unknown Unicast Forwarding Ports R/W 000_0000b Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward unknown unicast packets to that port 0 = Do not forward to that port All ones = Forwarded to all ports All zeros = Forwarded to no ports 2017 Microchip Technology Inc. DS00002391A-page 87 KSZ8567S 5.1.4.21 Unknown Multicast Control Register Address: 0x0324 - 0x0327 Bits 31 Size: 32 bits Description Unknown Multicast Packet Forward Type Default R/W 0b 1 = Enable forwarding of unknown multicast packets to the ports specified below 0 = Disable unknown multicast packet forwarding 30:7 RESERVED RO 0x000000 6:0 Unknown Multicast Forwarding Ports R/W 000_0000b Type Default R/W 0b Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward unknown multicast packets to that port 0 = Do not forward to that port All ones = Forwarded to all ports All zeros = Forwarded to no ports 5.1.4.22 Unknown VLAN ID Control Register Address: 0x0328 - 0x032B Bits 31 Size: 32 bits Description Unknown VID Packet Forward 1 = Enable forwarding of unknown VLAN ID (VID) packets to the ports specified below 0 = Disable unknown VID packet forwarding 30:7 RESERVED RO 0x000000 6:0 Unknown VID Forwarding Ports R/W 000_0000b Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward unknown VID packets to that port 0 = Do not forward to that port All ones = Forwarded to all ports All zeros = Forwarded to no ports DS00002391A-page 88 2017 Microchip Technology Inc. KSZ8567S 5.1.4.23 Switch MAC Control 0 Register Address: Bits 7 0x0330 Size: 8 bits Description Type Default R/W 0b RESERVED R/W 000b Frame Length Field Check R/W 0b Alternate Back-off Mode The back-off mode applies to half-duplex only. This bit should be set if the No Excessive Collision Drop bit in the Switch MAC Control 1 Register is enabled. 1 = Enable alternate back-off mode 0 = Disable 6:4 3 This applies only when the EtherType/Length field is <1500. 1 = Discard any packet if the actual packet length does not match the frame length field. 0 = Do not check the packet length. 2 RESERVED R/W 1b 1 Flow Control Packet Drop Mode R/W 0b R/W 0b This bit controls which flow control packets may be forwarded or dropped. To enable forwarding of all flow control packets, refer to bit 0 of the Switch MAC Control 4 Register. Enabling of flow control is managed in the PHY AutoNegotiation Advertisement Register (for PHY ports) and the XMII Port Control 0 Register (for MAC ports). 1 = The switch will drop received packets with either EtherType = 0x8808 or destination address (DA) = 01-80-C2-00-00-01. 0 = The switch will drop received packets with both EtherType = 0x8808 and DA = 01-80-C2-00-00-01. 0 Aggressive Back-off Enable For use with half-duplex back pressure. This is not an IEEE standard. 1 = Enable aggressive back-off algorithm in half-duplex mode to enhance performance. 0 = Disable 2017 Microchip Technology Inc. DS00002391A-page 89 KSZ8567S 5.1.4.24 Switch MAC Control 1 Register Address: Bits 0x0331 Size: 8 bits Description Type Default 7 RESERVED R/W 1b 6 Multicast Storm Protection Disable R/W 1b R/W 1b R/W 1b R/W 0b R/W 0b R/W 0b R/W 0b 1 = Multicast packets are not included in Broadcast Storm Protection 0 = Multicast packets are included in Broadcast Storm Protection 5 Back Pressure Mode 1 = Use carrier sense based back pressure mode 0 = Use collision based back pressure mode. (This is the recommended mode.) 4 Flow Control and Back Pressure Fair Mode 1 = Enable fair mode. If a flow controlled ingress port and a non-flow controlled ingress port forward traffic to the same egress port, packets from the non-flow controlled port may be dropped. 0 = Disable fair mode. In this scenario the flow controlled port will be flow controlled, and the non-flow controlled port will be neither flow controlled nor packets dropped. 3 No Excessive Collision Drop If this bit is set, Alternate Back-odd Mode (bit 7 in the Switch MAC Control 0 Register) should also be set. 1 = The switch will not drop packets when 16 or more collisions occur. 0 = The switch will drop packets when 16 or more collisions occur 2 Jumbo Packet Support The programmable packet payload size limit is specified in register 0x0308 - 0x0309, up to a maximum of 9000 bytes. This bit overrides bit 1 of this register. 1 = Enable support for jumbo packets 0 = Disable 1 Legal Maximum Packet Size Check Disable 1 = Accept packets up to 2000 bytes in size. 0 = Accept only standard size packets, up to 1522 bytes for tagged packets, or 1518 bytes for untagged packets. Larger packets will be dropped. 0 Pass Short Packet 1 = Accept packets between 32 and 64 bytes in size. 0 = Accept only standard size packets, at least 64 bytes. Smaller packets are dropped. DS00002391A-page 90 2017 Microchip Technology Inc. KSZ8567S 5.1.4.25 Switch MAC Control 2 Register Address: 0x0332 Bits 7:4 3 Size: 8 bits Description Type Default RESERVED R/W 0x0 Null VID Replacement R/W 0b R/W 000b Type Default R/W 0x4A Type Default RESERVED RO 0000_000b Pass Flow Control Packets R/W 0b Describes the behavior when a packet is received with a null (zero) VID. 1 = Replace a null VID with the Port Default VID as defined in the Port Default Tag 0 Register and Port Default Tag 1 Register 0 = No replacement of null VID 2:0 Broadcast Storm Protection Rate bits [10:8] The remainder of this fields bits are in the Switch MAC Control 3 Register. Multiply this value by 64 to determine how many bytes of packet data are allowed on an input port in a preset period. The period is 5ms for a 1000Mbps port, 50ms for a 100Mbps port, or 500ms for a 10Mbps. The default is 1%. 5.1.4.26 Switch MAC Control 3 Register Address: Bits 7:0 0x0333 Size: 8 bits Description Broadcast Storm Protection Rate bits [7:0] The remainder of this fields bits are in the Switch MAC Control 2 Register. Multiply this value by 64 to determine how many bytes of packet data are allowed on an input port in a preset period. The period is 5ms for a 1000Mbps port, 50ms for a 100Mbps port, or 500ms for a 10Mbps. The default is 1%. 5.1.4.27 Switch MAC Control 4 Register Address: Bits 7:1 0 0x0334 Size: Description 8 bits 1 = Switch will forward 802.3x PAUSE flow control frames. 0 = Switch will filter PAUSE frames. 2017 Microchip Technology Inc. DS00002391A-page 91 KSZ8567S 5.1.4.28 Switch MAC Control 5 Register Address: Bits 0x0335 Size: 8 bits Description Type Default 0b 7 RESERVED RO 6 RESERVED R/W 0b Ingress Rate Limit Period R/W 01b R/W 0b RO 000b 5:4 00 = 16ms 01 = 64ms 1x = 256ms 3 Queue Based Egress Rate Limit Enable 1 = Queue-based egress rate limiting 0 = Port-based egress rate limiting 2:0 5.1.4.29 RESERVED Switch MIB Control Register Address: 0x0336 Size: 8 bits MIB counters are provided on a per-port basis. They are read and controlled via the Port N: Port Switch MIB Counters Registers (0xN500 - 0xN5FF). Bits 7 Description Type Default R/W SC 0b R/W 0b RO 00_0000b Type Default RESERVED RO 0b When the tag PCP field is 0x1, use this value for priority R/W 001b RESERVED RO 0b When the tag PCP field is 0x0, use this value for priority R/W 000b Flush MIB Counters 1 = Flush all MIB counters of enabled ports. Refer to the Port MIB Control and Status Register. 0 = Normal counter operation 6 Freeze MIB Counters 1 = Freeze MIB counters of enabled ports. Refer to the Port MIB Control and Status Register. 0 = Normal counter operation 5:0 5.1.4.30 RESERVED 802.1p Priority Mapping 0 Register Address: Bits 7 6:4 3 2:0 0x0338 Size: Description DS00002391A-page 92 8 bits 2017 Microchip Technology Inc. KSZ8567S 5.1.4.31 802.1p Priority Mapping 1 Register Address: Bits 7 6:4 3 2:0 5.1.4.32 3 2:0 5.1.4.33 Default RESERVED RO 0b When the tag PCP field is 0x3, use this value for priority R/W 011b RESERVED RO 0b When the tag PCP field is 0x2, use this value for priority R/W 010b Type Default RESERVED RO 0b When the tag PCP field is 0x5, use this value for priority R/W 101b RESERVED RO 0b When the tag PCP field is 0x4, use this value for priority R/W 100b Type Default RESERVED RO 0b When the tag PCP field is 0x7, use this value for priority R/W 111b RESERVED RO 0b When the tag PCP field is 0x6, use this value for priority R/W 110b 802.1p Priority Mapping 2 Register 3 2:0 Size: 8 bits 802.1p Priority Mapping 3 Register Bits 6:4 0x033A Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x0339 0x033B Size: Description 2017 Microchip Technology Inc. 8 bits DS00002391A-page 93 KSZ8567S 5.1.4.34 IP DiffServ Priority Enable Register Address: Bits 0x033E Size: 8 bits Description Type Default 7:2 RESERVED RO 0000_00b 1 RESERVED R/W 0b 0 DiffServ Priority Remap Enable R/W 0b Type Default RESERVED RO 0b 1 = Use the following registers to remap the DSCP (DiffServ) priority to a 3bit priority value 0 = Use DSCP bits [5:3] for priority 5.1.4.35 IP DiffServ Priority Mapping 0 Register Address: Bits 7 6:4 3 2:0 5.1.4.36 3 2:0 8 bits When IPv4 / IPv6 DSCP field is 0x01, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x00, use this value for priority R/W 000b Type Default RESERVED RO 0b IP DiffServ Priority Mapping 1 Register Bits 6:4 Size: Description Address: 7 0x0340 0x0341 Size: Description 8 bits When IPv4 / IPv6 DSCP field is 0x03, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x02, use this value for priority R/W 000b DS00002391A-page 94 2017 Microchip Technology Inc. KSZ8567S 5.1.4.37 IP DiffServ Priority Mapping 2 Register Address: Bits 7 6:4 3 2:0 5.1.4.38 3 2:0 5.1.4.39 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x05, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x04, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x07, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x06, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x09, use this value for priority R/W 000b IP DiffServ Priority Mapping 3 Register 3 2:0 5.1.4.40 3 2:0 0x0344 Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x08, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0B, use this value for priority R/W 000b IP DiffServ Priority Mapping 5 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 4 Register Bits 6:4 0x0343 Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x0342 0x0345 Size: 8 bits Description RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0A, use this value for priority R/W 000b 2017 Microchip Technology Inc. DS00002391A-page 95 KSZ8567S 5.1.4.41 IP DiffServ Priority Mapping 6 Register Address: Bits 7 6:4 3 2:0 5.1.4.42 3 2:0 5.1.4.43 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0D, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0C, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0F, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x0E, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x11, use this value for priority R/W 000b IP DiffServ Priority Mapping 7 Register 3 2:0 5.1.4.44 3 2:0 0x0348 Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x10, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x13, use this value for priority R/W 000b IP DiffServ Priority Mapping 9 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 8 Register Bits 6:4 0x0347 Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x0346 0x0349 Size: Description 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x12, use this value for priority R/W 000b DS00002391A-page 96 2017 Microchip Technology Inc. KSZ8567S 5.1.4.45 IP DiffServ Priority Mapping 10 Register Address: Bits 7 6:4 3 2:0 5.1.4.46 3 2:0 5.1.4.47 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x15, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x14, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x17, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x16, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x19, use this value for priority R/W 000b IP DiffServ Priority Mapping 11 Register 3 2:0 5.1.4.48 3 2:0 0x034C Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x18, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1B, use this value for priority R/W 000b IP DiffServ Priority Mapping 13 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 12 Register Bits 6:4 0x034B Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x034A 0x034D Size: 8 bits Description RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1A, use this value for priority R/W 000b 2017 Microchip Technology Inc. DS00002391A-page 97 KSZ8567S 5.1.4.49 IP DiffServ Priority Mapping 14 Register Address: Bits 7 6:4 3 2:0 5.1.4.50 3 2:0 5.1.4.51 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1D, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1C, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1F, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x1E, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x21, use this value for priority R/W 000b IP DiffServ Priority Mapping 15 Register 3 2:0 5.1.4.52 3 2:0 0x0350 Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x20, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x23, use this value for priority R/W 000b IP DiffServ Priority Mapping 17 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 16 Register Bits 6:4 0x034F Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x034E 0x0351 Size: Description 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x22, use this value for priority R/W 000b DS00002391A-page 98 2017 Microchip Technology Inc. KSZ8567S 5.1.4.53 IP DiffServ Priority Mapping 18 Register Address: Bits 7 6:4 3 2:0 5.1.4.54 3 2:0 5.1.4.55 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x25, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x24, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x27, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x26, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x29, use this value for priority R/W 000b IP DiffServ Priority Mapping 19 Register 3 2:0 5.1.4.56 3 2:0 0x0354 Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x28, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2B, use this value for priority R/W 000b IP DiffServ Priority Mapping 21 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 20 Register Bits 6:4 0x0353 Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x0352 0x0355 Size: 8 bits Description RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2A, use this value for priority R/W 000b 2017 Microchip Technology Inc. DS00002391A-page 99 KSZ8567S 5.1.4.57 IP DiffServ Priority Mapping 22 Register Address: Bits 7 6:4 3 2:0 5.1.4.58 3 2:0 5.1.4.59 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2D, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2C, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2F, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x2E, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x31, use this value for priority R/W 000b IP DiffServ Priority Mapping 23 Register 3 2:0 5.1.4.60 3 2:0 0x0358 Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x30, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x33, use this value for priority R/W 000b IP DiffServ Priority Mapping 25 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 24 Register Bits 6:4 0x0357 Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x0350 0x0359 Size: Description 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x32, use this value for priority R/W 000b DS00002391A-page 100 2017 Microchip Technology Inc. KSZ8567S 5.1.4.61 IP DiffServ Priority Mapping 26 Register Address: Bits 7 6:4 3 2:0 5.1.4.62 3 2:0 5.1.4.63 Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x35, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x34, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x37, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x36, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x39, use this value for priority R/W 000b IP DiffServ Priority Mapping 27 Register 3 2:0 5.1.4.64 3 2:0 0x035C Size: 8 bits RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x38, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3B, use this value for priority R/W 000b IP DiffServ Priority Mapping 29 Register Bits 6:4 8 bits Description Address: 7 Size: IP DiffServ Priority Mapping 28 Register Bits 6:4 0x035B Description Address: 7 8 bits Type Bits 6:4 Size: Description Address: 7 0x035A 0x035D Size: 8 bits Description RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3A, use this value for priority R/W 000b 2017 Microchip Technology Inc. DS00002391A-page 101 KSZ8567S 5.1.4.65 IP DiffServ Priority Mapping 30 Register Address: 0x035E Bits 7 6:4 3 2:0 5.1.4.66 Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3D, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3C, use this value for priority R/W 000b Type Default RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3F, use this value for priority R/W 000b RESERVED RO 0b When IPv4 / IPv6 DSCP field is 0x3E, use this value for priority R/W 000b IP DiffServ Priority Mapping 31 Register 0x035F Bits 6:4 3 2:0 5.1.4.67 8 bits Description Address: 7 Size: Size: 8 bits Description Global Port Mirroring and Snooping Control Register Address: 0x0370 Size: 8 bits This register contains global controls for port mirroring and IGMP and MLD snooping. Mirroring also requires additional register settings for the individual ports. Refer to the Port N: Port Switch Ingress Control Registers (0xN800 - 0xN8FF). Bits Description Type Default 7 RESERVED RO 0b 6 IGMP Snooping Enable (IPv4) R/W 0b RESERVED RO 00b MLD Snooping Option R/W 0b R/W 0b RO 0b 1 = IGMP snooping is enabled. All IGMP packets will be forwarded to the host port. 0 = IGMP snooping is disabled. 5:4 3 1 = Snooping criteria: IPv6 next header = 43, 44, 50, 51 or 60; or next header = 0 and hop-by-hop next header = 43, 44, 50, 51 or 60. 0 = Snooping criteria: IPv6 next header = 1 or 58; or next header = 0 and hop-by-hop next header = 1 or 58. 2 MLD Snooping Enable (IPv6) 1 = MLD snooping is enabled. All MLD packets will be forwarded to the host port. 0 = MLD snooping is disabled. 1 RESERVED DS00002391A-page 102 2017 Microchip Technology Inc. KSZ8567S Bits 0 Description Sniff Mode Select Type Default R/W 0b Type Default 1 = Sniffing filter is "Rx AND Tx". Both the source port and destination ports need to match. 0 = Sniffing filter is "Rx OR Tx". Either the source port of the destination port needs to match. This mode is used to implement Rx-only sniffing. 5.1.4.68 WRED DiffServ Color Mapping Register Address: Bits 0x0378 Size: 8 bits Description 7:6 RESERVED RO 00b 5:4 Red This field specifies the red Differentiated Services Code Point (DSCP) value. R/W 11b 3:2 Yellow This field specifies the yellow DSCP value. R/W 10b 1:0 Green This field specifies the green DSCP value. R/W 01b 5.1.4.69 PTP Event Message Priority Register Address: 0x037C Size: 8 bits Bits Description Type Default 7 PTP Event Message Priority Override 0 = PTP event messages are assigned priority based on packet QoS 1 = PTP event messages are forced to the priority assigned to the PTP Event Message Priority field of this register. R/W 0b 6:4 RESERVED RO 000b 3:0 PTP Event Message Priority PTP non-event messages are assigned to this priority on queuing when bit 7 of this register is 1. R/W 1111b 2017 Microchip Technology Inc. DS00002391A-page 103 KSZ8567S 5.1.4.70 PTP Non-Event Message Priority Register Address: Bits 0x037D Size: 8 bits Type Default PTP Non-Event Message Priority Override 0 = PTP non-event messages are assigned priority based on packet QoS 1 = PTP non-event messages are forced to the priority assigned to the PTP Non-Event Message Priority field of this register. R/W 0b 6:4 RESERVED RO 000b 3:0 PTP Non-Event Message Priority PTP non-event messages are assigned to this priority on queuing when bit 7 of this register is 1. R/W 1111b Type Default 7 5.1.4.71 Description Queue Management Control 0 Register Address: Bits 0x0390 - 0x0393 Size: 32 bits Description 31:8 RESERVED RO 0x000000 7:6 Priority_2Q R/W 10b RESERVED R/W 00_00b Unicast Port VLAN Membership Discard R/W 1b R/W 0b When the 2 queue configuration is selected, this determines how to map the 2-bit priority regeneration result from the Port Priority to Queue Mapping Register into 2 queues. 00 = Priorities 0, 1, 2 map to the Low priority queue. Priority 3 maps to the High priority queue. 01 = Not used. 10 = Priorities 0, 1 map to the Low priority queue. Priorities 2, 3 map to the High priority queue. 11 = Priority 0 maps to the Low priority queue. Priorities 1, 2, 3 map to the High priority queue. 5:2 1 This bit applies to the Port VLAN Membership function in registers 0xNA0407. 1 = All frames are restricted to the forwarding ports as defined in the Port Control 1 Register. 0 = Frames forwarding to a single destination port are not limited to the forwarding ports defined in the Port Control 1 Register. Note that when mirroring is enabled, a single-destination frames will be dropped if it is mirrored to another port. 0 RESERVED DS00002391A-page 104 2017 Microchip Technology Inc. KSZ8567S 5.1.5 GLOBAL SWITCH LOOK UP ENGINE (LUE) CONTROL REGISTERS (0x0400 - 0x04FF) The following registers are used for accessing the VLAN Table, Address Lookup Table, Static Address Table and the Reserved Multicast Address Table. The organizations of these tables, and instructions for accessing them, are provided in Section 5.3, "Tables and MIB Counters (Access)," on page 187. 5.1.5.1 VLAN Table Entry 0 Register Address: Bits 31 0x0400 - 0x403 Size: 32 bits Description Type Default R/W 0b RESERVED RO 000b FORWARD OPTION R/W 0b R/W 000b VALID This field in the VLAN Table specifies if the table entry is valid. 1 = Table entry is valid 0 = Table entry is invalid 30:28 27 This field in the VLAN Table specifies how the forwarding ports are determined. 1 = Forward to the VLAN Table port map (PORT FORWARD field). 0 = Forwarding ports are determined by other variables. It may be the ALU port map, the VLAN port map, the combination of the two, or the Unknown Unicast/Multicast feature. 26:24 PRIORITY This field in the VLAN Table specifies the priority level. 23:15 RESERVED RO 0000_0000_0b 14:12 MSTP INDEX R/W 000b This field in the VLAN Table specifies the Multiple Spanning Tree Protocol index. 11:7 RESERVED RO 0000_0b 6:0 FID R/W 000_0000b This field in the VLAN Table specifies the Filter ID. The FID value is normally combined with the destination address and then hashed to index the address lookup table. Note: Refer to Section 5.3.4, "VLAN Table," on page 194 for additional information on VLAN tables. 2017 Microchip Technology Inc. DS00002391A-page 105 KSZ8567S 5.1.5.2 VLAN Table Entry 1 Register Address: 0x0404 - 0x407 Bits Size: 32 bits Description Type Default 31:7 RESERVED R/O 0x0000000 6:0 PORT UNTAG R/W 0x00 This field in the VLAN Table specifies the untagging policy for each egress port. Bits [6:0] correspond to ports [7:1]. 1 = Untag packets upon egress at this port 0 = Do not untag upon egress at this port Note: 5.1.5.3 Refer to Section 5.3.4, "VLAN Table," on page 194 for additional information on VLAN tables. VLAN Table Entry 2 Register Address: 0x0408 - 0x040B Bits Size: 32 bits Description Type Default 31:7 RESERVED R/O 0x0000000 6:0 PORT FORWARD R/W 0x00 This field specifies the forwarding policy to each port. The policy is applied if the FO bit is set. Bits [6:0] correspond to ports [7:1]. 1 = Forward to this port 0 = Do not forward to this port Note: 5.1.5.4 Refer to Section 5.3.4, "VLAN Table," on page 194 for additional information on VLAN tables. VLAN Table Index Register Address: Bits 0x040C - 0x040D Size: 16 bits Description Type Default 15:12 RESERVED RO 0x0 11:0 VLAN Index R/W 0x000 This value addresses the VLAN table, in combination with the VLAN table action in the following register. Note: Refer to Section 5.3.4, "VLAN Table," on page 194 for additional information on VLAN tables. DS00002391A-page 106 2017 Microchip Technology Inc. KSZ8567S 5.1.5.5 VLAN Table Access Control Register Address: Bits 7 0x040E Size: 8 bits Description Start VLAN Table Action 1 = Start the action defined below Type Default R/W SC 0b 0 = Action finished 6:2 RESERVED RO 000_00b 1:0 Action R/W 00b Specify the action to be taken for the VLAN table entry addressed in the index register 00 = No operation 01 = Write 10 = Read 11 = Clear all entries to zero Note: 5.1.5.6 Refer to Section 5.3.4, "VLAN Table," on page 194 for additional information on VLAN tables. ALU Table Index 0 Register Address: 0x0410 - 0x0413 Size: 32 bits The ALU Table Index 0 Register and ALU Table Index 1 Register contain the index values for searching and reading/ writing the address lookup table as specified by the action in the ALU Table Access Control Register. For additional information on ALU tables, refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187. Bits Description Type Default 31 RESERVED R/W 0b 30:23 RESERVED RO 000_0000_0b 22:16 FID Index [6:0] R/W 000_0000b R/W 0x0000 This is the FID value used to hash index the table 15:0 MAC Index [47:32] These are the upper 16 bits of the MAC addressed used to hash index the table 2017 Microchip Technology Inc. DS00002391A-page 107 KSZ8567S 5.1.5.7 ALU Table Index 1 Register Address: 0x0414 - 0x0417 Size: 32 bits The ALU Table Index 0 Register and ALU Table Index 1 Register contain the index values for searching and reading/ writing the address lookup table as specified by the action in the ALU Table Access Control Register. For additional information on ALU tables, refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187. Bits 31:0 Description MAC Index [31:0] Type Default R/W 0x00000000 These are the lower 32 bits of the MAC addressed used to hash index the address lookup table. If direct addressing is enabled, then bits [11:0] are used to directly index the address lookup table. 5.1.5.8 ALU Table Access Control Register Address: 0x0418 - 0x041B Size: 32 bits This register provides control and status for searching and reading or writing the ALU Table. The ALU Table Index 0 Register and ALU Table Index 1 Register contain the index values, while the ALU / Static Address Table Entry 1 Register, ALU / Static Address / Reserved Multicast Table Entry 2 Register, ALU / Static Address Table Entry 3 Register, and ALU / Static Address Table Entry 4 Register are used for the entry values. Refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187 for additional information on ALU tables. Bits Description Type Default 31:30 RESERVED RO 00b 29:16 VALID_COUNT RO 0x0000 RO 0x00 R/W, SC 0b RO 0b RO 0b R/W 00b Indicates the total number of valid entries in the table after the search finishes 15:8 7 RESERVED START_FINISH 1 = Start the action defined below 0 = Action finished 6 VALID For search operation. It is cleared when register 0x042F (ALU / Static Address Table Entry 4 Register) is read. 1 = Indicates that the next valid entry is ready 0 = Next valid entry is not ready 5 VALID_ENTRY_OR_SEARCH_END For search operation. It is intended for added convenience when accessing registers by in-band management (IBA) rather than SPI or I2C. It combines bits 6 and 7. 1 = Indicates either that the next valid entry is ready, or that the search has ended. 0 = Neither next valid entry ready nor search ended. 4:3 RESERVED DS00002391A-page 108 2017 Microchip Technology Inc. KSZ8567S Bits 2 Description DIRECT Type Default R/W 0b R/W 00b 1 = Access by direct addressing the ALU Table. This method is used only for debugging, if at all. 0 = Hashing function is used to index the table. This is the normal method. Refer to the Switch Lookup Engine Control 0 Register. 1:0 ACTION Specifies the action to be taken for the ALU Table entry access 00 = No operation 01 = Write 10 = Read 11 = Search 5.1.5.9 Static Address and Reserved Multicast Table Control Register Address: 0x041C - 0x041F Size: 32 bits This register provides control and index for reading and writing the Static Address Table and the Reserved Multicast Table. The ALU / Static Address Table Entry 1 Register, ALU / Static Address / Reserved Multicast Table Entry 2 Register, ALU / Static Address Table Entry 3 Register, and ALU / Static Address Table Entry 4 Register are used for the Static Address Table entry values, while only the ALU / Static Address / Reserved Multicast Table Entry 2 Register is used for the Reserved Multicast Table entry value. Refer to Section 5.3.2, "Static Address Table," on page 191 and Section 5.3.3, "Reserved Multicast Address Table," on page 193 for additional information on these tables. Bits Description Type Default 31:22 RESERVED RO 0x000 21:16 TABLE_INDEX R/W 00_0000b RO 0x00 R/W, SC 0b RESERVED R/W 000_00b TABLE_SELECT R/W 0b R/W 0b Bits [21:16] used to index the Reserved Multicast Table Bits [19:16] used to index the Static Address Table 15:8 7 RESERVED START_FINISH 1 = Start access 0 = Access is finished 6:2 1 Specifies which table is being accessed 1 = Access Reserved Multicast Table 0 = Access Static Address Table 0 ACTION Specifies the action to be taken for the table 1 = Read 0 = Write 2017 Microchip Technology Inc. DS00002391A-page 109 KSZ8567S 5.1.5.10 ALU / Static Address Table Entry 1 Register Address: 0x0420 - 0x0423 Size: 32 bits This register contains the table entry values for read and write operations to the Address Lookup Table and Static Address Table. The field definitions of this register differ dependent on the table type used, as defined in the following sections: * ALU Table Entry 1 Register on page 190 * Static Address Table Entry 1 Register on page 192 Refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187 and Section 5.3.2, "Static Address Table," on page 191 for additional information. 5.1.5.11 ALU / Static Address / Reserved Multicast Table Entry 2 Register Address: 0x0424 - 0x0427 Size: 32 bits The field definitions of this register differ dependent on the table type used, as defined in the following sections: * ALU Table Entry 2 Register on page 190 * Static Address Table Entry 2 Register on page 192 * Reserved Multicast Address Table Entry 2 Register on page 194 Refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187, Section 5.3.2, "Static Address Table," on page 191, and Section 5.3.3, "Reserved Multicast Address Table," on page 193 for additional information. 5.1.5.12 ALU / Static Address Table Entry 3 Register Address: 0x0428 - 0x042B Size: 32 bits The field definitions of this register differ dependent on the table type used, as defined in the following sections: * ALU Table Entry 3 Register on page 191 * Static Address Table Entry 3 Register on page 193 Refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187 and Section 5.3.2, "Static Address Table," on page 191 for additional information. 5.1.5.13 ALU / Static Address Table Entry 4 Register Address: 0x042C - 0x042F Size: 32 bits This register contains the table entry values for read and write operations to the Address Lookup Table and Static Address Table. The field definitions of this register differ dependent on the table type used, as defined in the following sections: * ALU Table Entry 4 Register on page 191 * Static Address Table Entry 4 Register on page 193 Refer to Section 5.3.1, "Address Lookup (ALU) Table," on page 187 and Section 5.3.2, "Static Address Table," on page 191 for additional information. DS00002391A-page 110 2017 Microchip Technology Inc. KSZ8567S 5.1.6 5.1.6.1 GLOBAL SWITCH PTP CONTROL REGISTERS (0x0500 - 0x05FF) Global PTP Clock Control Register Address: Bits 15 0x0500 - 0x0501 Size: 16 bits Description Type Default R/W 0b RESERVED RO 000_0000_0b PTP Clock Step Adjustment R/W SC 0b R/W 0b R/W SC 0b R/W SC 0b R/W 0b R/W 0b R/W SC 0b Disable Switch Frequency Adjustment 1 = The switch timers, shapers and policing logic will NOT adjust the frequency based on PTP adjustments. 0 = The switch timers, shapers and policing logic will adjust the frequency based on PTP adjustments. 14:7 6 Setting this bit will cause the time value in the Global PTP RTC Clock Nanosecond High Word Register and Global PTP RTC Clock Nanosecond High Word Register to be added to or subtracted from (based on the value of bit 5) the PTP clock. 5 PTP Step Direction Direction control for PTP step adjustment mode. 1 = Add the time value 0 = Subtract the time value 4 PTP Clock Read Setting this bit will cause the current PTP clock value to be copied into registers 0x0502 to 0x050B. 3 PTP Clock Load Setting this bit will cause the PTP clock to be loaded with the time value in registers 0x0502 to 0x050B. 2 PTP Clock Continuous Adjustment When continuous adjustment is enabled, the SUB-NS_RATE value in the Global PTP Clock Sub-Nanosecond Rate High Word Register and Global PTP Clock Sub-Nanosecond Rate Low Word Register is added to or subtracted from (based on the PTP_RATE_DIR bit in the Global PTP Clock Sub-Nanosecond Rate High Word Register) the PTP clock on every 25MHz clock cycle. 1 = Enable continuous adjustment 0 = Disable continuous adjustment 1 Enable PTP Clock 1 = Enable PTP clocking 0 = Disable PTP clocking 0 Reset PTP Clock Setting this bit will reset the PTP clock. 2017 Microchip Technology Inc. DS00002391A-page 111 KSZ8567S 5.1.6.2 Global PTP RTC Clock Phase Register Address: Bits 0x0502 - 0x0503 Size: 16 bits Description Type Default 15:3 RESERVED RO 0x0000 2:0 PTP Real Time Clock 8ns Phase R/W 000b This register indicates one of the 8n sub-cycle phases of the 40ns period PTP real time clock. 000 = 0ns (real time clock at the first 8ns phase in the 40ns period) 001 = 8ns (real time clock at the second 8ns phase in the 40ns period) 010 = 16ns (real time clock at the third 8ns phase in the 40ns period) 011 = 24ns (real time clock at the fourth 8ns phase in the 40ns period) 100 = 32ns (real time clock at the fifth 8ns phase in the 40ns period) 101 - 111 = not valid 5.1.6.3 Global PTP RTC Clock Nanosecond High Word Register Address: 0x0504 - 0x0505 Size: 16 bits Bits Description Type Default 15:0 PTP Real Time Clock Nanosecond [31:16] R/W 0x0000 Type Default R/W 0x0000 Type Default R/W 0x0000 This is the high word of the nanosecond value of the PTP real time clock. 5.1.6.4 Global PTP RTC Clock Nanosecond Low Word Register Address: Bits 15:0 0x0506 - 0x0507 Size: 16 bits Description PTP Real Time Clock Nanosecond [15:0] This is the low word of the nanosecond value of the PTP real time clock. 5.1.6.5 Global PTP RTC Clock Second High Word Register Address: Bits 15:0 0x0508 - 0x0509 Size: 16 bits Description PTP Real Time Clock Second [31:16] This is the high word of the second value of the PTP real time clock. DS00002391A-page 112 2017 Microchip Technology Inc. KSZ8567S 5.1.6.6 Global PTP RTC Clock Second Low Word Register Address: Bits 15:0 0x050A - 0x050B Size: 16 bits Description PTP Real Time Clock Second [15:0] Type Default R/W 0x0000 Type Default R/W 0b R/W 0b R/W 0x0000 Type Default R/W 0x0000 This is the low word of the second value of the PTP real time clock. 5.1.6.7 Global PTP Clock Sub-Nanosecond Rate High Word Register Address: Bits 15 0x050C - 0x050D Size: 16 bits Description PTP Rate Direction Rate direction control for PTP clock Temporary Adjustment and Continuous Adjustment modes. 1 = The PTP Clock Sub-nanosecond value (this register and Global PTP Clock Sub-Nanosecond Rate Low Word Register) will be added to the PTP time every 25MHz clock cycle. 0 = The PTP Clock Sub-nanosecond value will be subtracted from the PTP time every 25MHz clock cycle. 14 PTP Temporary Adjustment Mode 1 = Enable temporary incrementing or decrementing of the PTP clock by the PTP Clock Sub-nanosecond value (this register and the Global PTP Clock Sub-Nanosecond Rate Low Word Register) every 25MHz clock cycle, for the duration set in the Global PTP Clock Temp Adjustment Duration High Word Register and Global PTP Clock Temp Adjustment Duration Low Word Register. 0 = Stop temporary adjustment of the PTP clock 13:0 PTP Real Time Clock Sub-Nanosecond [29:16] This is the high word of the sub-nanosecond value of the PTP real time clock. 5.1.6.8 Global PTP Clock Sub-Nanosecond Rate Low Word Register Address: Bits 15:0 0x050E - 0x050F Size: 16 bits Description PTP Real Time Clock Sub-Nanosecond [15:0] This is the low word of the sub-nanosecond value of the PTP real time clock. It is used for Continuous Adjustment and Temporary Adjustment modes. 2017 Microchip Technology Inc. DS00002391A-page 113 KSZ8567S 5.1.6.9 Global PTP Clock Temp Adjustment Duration High Word Register Address: Bits 15:0 0x0510 - 0x0511 Size: 16 bits Description Type Default R/W 0x0000 Type Default R/W 0x0000 Type Default RESERVED RO 0x00 Enable IEEE 802.3AS Mode R/W 0b R/W 0b R/W 1b R/W 1b R/W 1b R/W 0b PTP Temporary Adjustment Duration [31:16] This sets the duration for the PTP clock temporary rate adjustment, in number of 25MHz clock cycles. 5.1.6.10 Global PTP Clock Temp Adjustment Duration Low Word Register Address: Bits 15:0 0x0512 - 0x0513 Size: 16 bits Description PTP Temporary Adjustment Duration [15:0] This sets the duration for the PTP clock temporary rate adjustment, in number of 25MHz clock cycles. 5.1.6.11 Global PTP Message Config 1 Register Address: Bits 15:8 7 0x0514 - 0x0515 Size: Description 16 bits 1 = enable 0 = disable 6 Enable IEEE 1588 PTP Mode 1 = enable 0 = disable 5 Enable Detection of IEEE 802.3 Ethernet PTP Messages 1 = enable 0 = disable 4 Enable Detection of IPv4/UDP PTP Messages 1 = enable 0 = disable 3 Enable Detection of IPv6/UDP PTP Messages 1 = enable 0 = disable 2 Selection of P2P or E2E 1 = Peer-to-peer (P2P) transparent clock mode 0 = End-to-end (E2E) transparent clock mode DS00002391A-page 114 2017 Microchip Technology Inc. KSZ8567S Bits 1 Description Type Default R/W 0b R/W 1b Type Default RESERVED RO 000b Enable Unicast PTP R/W 1b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b Selection of Master or Slave 1 = Host port is PTP master ordinary clock 0 = Host port is PTP slave ordinary clock 0 Selection of One-step or Two-step Operation 1 = One-step clock mode 0 = Two-step clock mode 5.1.6.12 Global PTP Message Config 2 Register Address: Bits 15:13 12 0x0516 - 0x0517 Size: 16 bits Description 1 = The unicast PTP frames can be recognized. If the packet UDP destination port is either 319 or 320 and the frame MAC/IP address is not the PTP reserved address, then the frame will be considered as a unicast PTP frame and the frame forwarding will be decided by regular lookup table. 0 = Only multicast PTP frames will be recognized. 11 Enable Alternate Master 1 = Alternate master clock is supported. The Sync/Delay_Req frames of the same domain received at non-host ports by active master clock of the same domain will be forwarded to the non-host ports. 0 = Alternate master clock is not supported. The Sync/Delay_Req frames of the same domain received at non-host ports by active master clock of the same domain will be discarded on the host port and will be forwarded to the non-host ports if Sync/Delay_Req is for other domains. 10 PTP Messages Priority TX Queue 1 = All PTP messages are assigned to the highest priority TX queue. 0 = Only the PTP event messages are assigned to the highest priority TX queue. 9 Enable Checking of Associated Sync and Follow_up PTP messages Setting this bit will associate Follow_up message with Sync message when it has the same domain, sequenceID and sourcePortID. The PTP frame will be forwarded to the host port if the ID matches. 8 Enable Checking of Associated Delay_Req and Delay Resp PTP Messages Setting this bit will associate Delay_Resp message with Delay_Req message when it has the same domain, sequenceID and sourcePortID. The PTP frame will be forwarded to the host port if the ID matches. 7 Enable Checking of Associated Pdelay_Req and Pdelay_Resp PTP Messages Setting this bit will associate Pdelay_Resp/Pdelay_Resp_Follow_Up message with Pdelay_Req message when it is with the same domain, sequenceID and sourcePortID. The PTP frame will be forwarded to the host port if ID matches. 6 RESERVED 2017 Microchip Technology Inc. DS00002391A-page 115 KSZ8567S Bits 5 Description Enable Dropping of Sync/Follow_Up and Delay_Req PTP Messages Type Default R/W 0b R/W 0b Setting this bit will enable the device to automatically drop these Sync/ Follow_up and Delay_Req PTP messages if the Best Master Clock (BMC) is not determined. 4 Enable Checking of Domain Field Setting this bit will enable the device to automatically check the domain field of the PTP message with the PTP_DOMAIN in the Global PTP Domain and Version Register. The PTP message will be forwarded to the host port if the domain field matches. Otherwise it will be dropped. 3 RESERVED R/W 0b 2 Enable IPv4/UDP Checksum Calculation for Egress Packets R/W 1b 1 = The switch will re-calculate and generate a 2-byte checksum value when the frame content changes. 0 = The checksum field is set to zero. If the IPv4/UDP checksum is zero, the checksum will remain zero regardless of this bit setting. For IPv6/UDP, the checksum is always updated. 1 RESERVED R/W 0b 0 RESERVED R/W 0b Type Default 5.1.6.13 Global PTP Domain and Version Register Address: Bits 0x0518 - 0x0519 Size: 16 bits Description 15:12 RESERVED RO 0x0 11:8 PTP Version R/W 0x2 R/W 0x00 This is the value of the PTP message version number field. All PTP messages will be captured when the receive message version matches the value in this field. All PTP packets will be dropped if the receive PTP message version does not match the value in this field, except for 7:0 PTP Domain This is the value of the PTP message domain number field. If domain checking is enabled (Global PTP Message Config 2 Register, bit 4), the timestamp capture will be enabled when the receive PTP message domain number matches the value in this field. If domain checking is not enabled, the domain number field will be ignored. DS00002391A-page 116 2017 Microchip Technology Inc. KSZ8567S 5.1.6.14 Global PTP Unit Index Register Address: 0x0520 - 0x0523 Size: 32 bits This register is used to index the timestamp unit and trigger unit for accesses to PTP registers in the address range of 0x052C to 0x05B3. The timestamp and trigger units selected in this register will be the units that are read/written when accessing these registers. Bits Type Default RESERVED RO 0x00000 Timestamp Unit Index Pointer (TS_PTR_INDEX) This bit points to the unit/set registers of the timestamp. 1 = Timestamp Unit 1 0 = Timestamp Unit 0 R/W 0b 7:2 RESERVED RO 000000b 1:0 Trigger Unit Index Pointer (TRIGGER_PTR_INDEX) This bit points to the unit/set registers of the trigger. 11 = RESERVED 10 = Trigger Unit 2 01 = Trigger Unit 1 00 = Trigger Unit 0 R/W 00b Type Default RO 0x000 R/W1C 000b RO 0x000 R/W1C 000b 31:9 8 5.1.6.15 Description GPIO Status Monitor 0 Register Address: Bits 0x0524 - 0x0527 Size: 32 bits Description 31:19 RESERVED 18:16 Event Trigger Output Error (TRIGGER_ERROR) (Bit 18 = Trigger Unit 2, Bit 17 = Trigger Unit 1, Bit 16 = Trigger Unit 0) 1 = The event trigger time is set earlier than the system time clock when the TRIGGER_NOTIFY bit is set to "1" in the Trigger Output Unit Control 1 Register and will generate an interrupt to the host. This bit can be cleared by resetting the TRIGGER_EN bit to "0" in the Timestamp Control and Status Register. 0 = No event trigger error. 15:3 RESERVED 2:0 Event Trigger Output Unit Done (TRIGGER_DONE) (Bit 2 = Trigger Unit 2, Bit 1 = Trigger Unit 1, Bit 0 = Trigger Unit 0) 1 = The event trigger output unit has been generated when the TRIGGER_NOTIFY bit is set to "1" in the Trigger Output Unit Control 1 Register (write "1" to clear this bit) and will generate an interrupt to the host. 0 = Event trigger output unit not done. 2017 Microchip Technology Inc. DS00002391A-page 117 KSZ8567S 5.1.6.16 GPIO Status Monitor 1 Register Address: 0x0528 - 0x052B Size: 32 bits This register provides the interrupt status for the trigger output unit and timestamp units used with the GPIO. The GPIO pin Output Trigger and Timestamp Unit interrupt status bit in the Global Interrupt Status Register is the OR of the status bits in this register. Bits Description 31:19 RESERVED 18:16 Trigger Output Unit Interrupt Status Type Default RO 0x0000 R/WC 000b RO 0x0000 R/WC 00b (Bit 18 = Trigger Unit 2, Bit 17 = Trigger Unit 1, Bit 16 = Trigger Unit 0) These three bits provide the interrupt status for the three Trigger Output Units. These interrupts are enabled by setting the TRIGGER_NOTIFY bit in the Trigger Output Unit Control 1 Register. This is done separately for each TOU. Refer to the GPIO Status Monitor 0 Register for the Trigger Output Unit status details. 1 = Interrupt detected 0 = No interrupt 15:2 RESERVED 1:0 Timestamp Unit Interrupt Status (GPIO input timestamp only) (Bit 1 = Timestamp Unit 1, Bit 0 = Timestamp Unit 0) These two bits provide the interrupt status for the two Timestamp Output Units for the GPIO. These interrupts are enabled by the Timestamp Unit Interrupt Enable bit in the Timestamp Control and Status Register. This is done separately for each TSU. Refer to the registers beginning at the Timestamp Status and Control Register for Timestamp interrupt details. For the interrupt status for the port-based timestamping of PTP egress frames, refer to the Port Interrupt Status Register and Port PTP Timestamp Interrupt Status Register. 1 = Interrupt detected 0 = No interrupt DS00002391A-page 118 2017 Microchip Technology Inc. KSZ8567S 5.1.6.17 Timestamp Control and Status Register Address: 0x052C - 0x052F Size: 32 bits Portions of this register are indexed by the Trigger Unit Index Pointer (bits [1:0]) and Timestamp Unit Index Pointer (bit 8) values in the Global PTP Unit Index Register. Bits Type Default RESERVED RO 0x000000 8 GPIO Output Source Select (GPIO_OUT_SEL) 1 = The output is a combinatorial result 0 = The output is from flopped output R/W 0b 7 GPIO Inputs Monitor (GPIO_IN) This field reflects the current value seen on the GPIO input. R - 6 GPIO Output Enable (GPIO_OEN) 1 = Enables the GPIO pin as a trigger input 0 = Enables the GPIO pin as a timestamp input R/W 0b 5 Timestamp Unit Interrupt Enable (TS_INT_ENB) The timestamp unit interrupt status bit is found in the GPIO Status Monitor 1 Register. 0 = Interrupt disabled 1 = Interrupt enabled R/W 0b R 0b R/W 0b R/W 0b R/W 0b 31:9 Description Note: 4 Event Trigger Output Unit Active (TRIGGER_ACTIVE) 1 = The event trigger output unit is enabled and active without error 0 = The event trigger output unit is finished and inactive Note: 3 This field is indexed by the Trigger Unit Index Pointer (bits 1:0) value in the Global PTP Unit Index Register. Event Trigger Output Unit Software Reset (TRIGGER_SW_RESET) 1 = Resets the trigger output unit to the inactive state and default settings. This reset can be used to stop the cascade mode in continuous operation and prepare this trigger unit for the next operation. 0 = Trigger output unit under normal operation Note: 1 This field is indexed by the Trigger Unit Index Pointer (bits 1:0) value in the Global PTP Unit Index Register. Event Trigger Output Unit Enable (TRIGGER_EN) 1 = Enables the selected event trigger output unit. Self-clearing when trigger output is generated. In cascade mode, only enable the head of the trigger unit. 0 = The event trigger out unit is disabled Note: 2 This field is indexed by the Timestamp Unit Index Pointer (bit 8) value in the Global PTP Unit Index Register. This field is indexed by the Trigger Unit Index Pointer (bits 1:0) value in the Global PTP Unit Index Register. Event Timestamp Input Unit Enable (TS_ENB) 1 = Enables the selected event timestamp input unit. Writing "1" to this bit will clear the TS_EVENT_DET_CNT of the associated unit. 0 = Disables the selected event timestamp input unit. Writing "0" to this bit will clear the TS_RDY and TS_DET_CNT_OVFL. Note: This field is indexed by the Timestamp Unit Index Pointer (bit 8) value in the Global PTP Unit Index Register. 2017 Microchip Technology Inc. DS00002391A-page 119 KSZ8567S Bits 0 Description Event Timestamp Input Unit Software Reset (TS_SW_RESET) 1 = Resets the timestamp unit to the inactive state and default settings 0 = Timestamp input unit under normal operation Note: 5.1.6.18 Type Default R/W 0b This field is indexed by the Timestamp Unit Index Pointer (bit 8) value in the Global PTP Unit Index Register. Trigger Output Unit Target Time Nanosecond Register Address: 0x0530 - 0x0533 Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31:30 RESERVED RO 000b 29:0 Trigger Target Time (ns) (TRIGGER_TARGET_TIME_NS) This field contains the PTP event trigger output target time in nanoseconds for the trigger unit indexed in the Global PTP Unit Index Register. R/W 0x00000000 5.1.6.19 Trigger Output Unit Target Time Second Register Address: 0x0534 - 0x0537 Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31:0 Trigger Target Time (s) (TRIGGER_TARGET_TIME_S) This field contains the PTP event trigger output target time in seconds for the trigger unit indexed in the Global PTP Unit Index Register. R/W 0x00000000 DS00002391A-page 120 2017 Microchip Technology Inc. KSZ8567S 5.1.6.20 Trigger Output Unit Control 1 Register Address: 0x0538 - 0x053B Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31 Enable Event Trigger Output Unit in Cascade Mode (CASCADE_MODE_ENB) 1 = Enables the event trigger output unit in cascade mode 0 = Disables the event trigger output unit in cascade mode R/W 0b 30 Cascade Mode Event Trigger Output Unit Tail Unit Indicator (CASCADE_MODE_TAIL) 1 = This event trigger output unit is the last unit of the chain in cascade mode 0 = This event trigger output unit is not the last unit of the chain in cascade mode. Note: when this bit it cleared to "0" for all units in cascade mode, the iteration count is ignored and becomes an infinite number. To stop the infinite loop, set the respective TRIG_SW_RESET bit in the Timestamp Control and Status Register. R/W 0b 29:28 RESERVED RO 00b 27:26 Cascade Mode Upstream Trigger Done Unit Select This field selects one of the 3 upstream trigger done input units in cascade mode. For example, if units 1, 2, and 3 (tail unit) are set up in cascade mode, then Unit 1 is set to 0x3, Unit 2 is set to 0x1, and Unit 3 is set to 0x2. R/W 00b 25 Trigger Now (TRIGGER_NOW) 1 = Immediately trigger even output if trigger target time is less than the system clock time 0 = Wait for trigger target time to trigger event output R/W 0b 24 Trigger Notify (TRIGGER_NOTIFY) 1 = Enables reporting both TRIG_DONE and TRIG_ERR status as well as interrupt to host if interrupt enable bit is set. 0 = Disables reporting of both TRIG_DONE and TRIG_ERR status. R/W 0b 23 Trigger Edge (TRIGGER_EDGE) 1 = Trigger output on negative edge of clock 0 = Trigger output on positive edge of clock R/W 0b 2017 Microchip Technology Inc. DS00002391A-page 121 KSZ8567S Bits Description Type Default 22:20 Trigger Event Output Signal Pattern (TRIGGER_PATTERN) To select a trigger event output when TRIG_EN=1 and the trigger target time has reached the system time: 000 = Generates negative edge (from default "H" -> "L" and stays "L") 001 = Generates positive edge (from default "L" -> "H" and stays "H") 010 = Generates negative pulse (from default "H" -> "L" pulse -> "H" and stays "H"). The pulse width is defined in the Trigger Output Unit Control 2 Register. 011 = Generates positive pulse (from default "L" -> "H" pulse -> "L" and stays "L"). The pulse width is defined in the Trigger Output Unit Control 2 Register. 100 = Generates negative periodic signal. The "L" pulse width is defined in the Trigger Output Unit Control 2 Register, the cycle width is defined in the Trigger Output Unit Control 3 Register and Trigger Output Unit Control 4 Register, and the number of cycles is defined in the Trigger Output Unit Control 5 Register (it is an infinite number if the number of cycles is zero). 101 = Generates positive periodic signal. The "H" pulse width is defined in the Trigger Output Unit Control 2 Register, the cycle width is defined in the Trigger Output Unit Control 3 Register and Trigger Output Unit Control 4 Register, and the number of cycles is defined in the Trigger Output Unit Control 5 Register (it is an infinite number if the number of cycles is zero). 110 = Generates output signal from the 16-bit pattern in the Trigger Output Unit Control 3 Register, which is shifted MSB fist and looped. Each bit width is defined in the Trigger Output Unit Control 3 Register and Trigger Output Unit Control 4 Register and the total number of bits to shift out is defined in the Trigger Output Unit Control 5 Register (it is infinite if this register value is zero). 111 = RESERVED. R/W 000b 19:16 RESERVED RO 000b 15:0 Trigger Output Iteration Count (TRIGGER_PATTERN_ITERATION) Defines the iteration count number to output the 16-bit trigger pattern defined in the Trigger Output Unit Control 3 Register in cascade mode at tail unit. For example, 0x0000 is 1 count and 0x000F is 16 counts. It is infinite if there is no tail unit in cascade mode. R/W 0x0000 Note: 5.1.6.21 The maximum output clock frequency is up to 12.5MHz. Trigger Output Unit Control 2 Register Address: 0x053C - 0x053F Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31:0 Trigger Output Cycle Width or Bit Width (TRIGGER_CYCLE_WIDTH) Defines the cycle width for generating periodic signals. Each unit value equals 1ns. For example, the cycle width is 80ns if this register value is 80 (0x50). R/W 0x00000000 DS00002391A-page 122 2017 Microchip Technology Inc. KSZ8567S 5.1.6.22 Trigger Output Unit Control 3 Register Address: 0x0540 - 0x0543 Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31:16 Trigger Output Cycle or Bit Count (TRIGGER_CYCLE) Defines the number of output cycles for generating periodic signals. It is infinite if this field is zero. R/W 0x0000 15:0 Trigger Output Bit Pattern (TRIGGER_BIT_PATTERN) Defines the output bit pattern for generating output signals. R/W 0x0000 5.1.6.23 Trigger Output Unit Control 4 Register Address: 0x0544 - 0x0547 Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. Bits Description Type Default 31:0 Trigger Output Iteration Cycle Time in Cascade Mode (CASCADE_INTERATION_CYCLE_TIME) Defines the iteration cycle time to go through all the trigger output units in cascade mode. This time will be added to the current trigger target time for the next trigger time. A unit value equals 1ns. For example, the cycle is 800ns if this field value is 800 (0x320). R/W 0x00000000 5.1.6.24 Trigger Output Unit Control 5 Register Address: 0x0548 - 0x054B Size: 32 bits This register is indexed by the Trigger Unit Index Pointer value in bits [1:0] of the Global PTP Unit Index Register. This register contains the PTP event trigger output PPS signal pulse width for unit 2 and path delay compensation for unit 1. Bits Description Type Default 31:24 RESERVED RO - 23:16 PPS Pulse Width for Event Trigger (PPS_PULSE_WIDTH) This field is the upper third byte (23:16) in conjunction with the trigger output pulse width to make this to make this register value for PPS pulse width up to 124ms. R/W 0x00 15:0 Trigger Output Pulse Width (TRIGGER_PULSE_WIDTH) Defines the width for generating a pulse or periodic signal. Each unit value equals 8ns. For example, the pulse width is 80ns if this register value is 10 (0xA). R/W 0x0000 2017 Microchip Technology Inc. DS00002391A-page 123 KSZ8567S 5.1.6.25 Timestamp Status and Control Register Address: 0x0550 - 0x0553 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:21 RESERVED RO 0x00 20:17 Number of Detected Event Count for Timestamp Input Unit (TS_EVENT_DET_CNT) Reports the number of detected events (either rising or falling edge) count. In single mode, it can detect up to 15 events at any single unit. The pulse or edges can be detected up to 25MHz. The pulse width can be measured by the difference between consecutive timestamps in the same unit. RO 0000b 16 Number of Detected Event Count Overflow for Timestamp Input Unit (TS_DET_EVENT_CNT_OVERFLOW) 1 = The detected events (either rising or falling edge) count has overflowed. In cascade mode, only the tail unit will set this bit when overflow occurs. The TS_EVENT_DET_CNT field will remain at 15 when overflow has occurred. 0 = The number of events (either rising or falling edge) count has not overflowed. RO 0b RESERVED RO 00000000b 7 Enable Rising Edge Detection (TS_RISING_EDGE_ENB) 1 = Enable rising edge detection 0 = Disable rising edge detection R/W 0b 6 Enable Falling Edge Detection (TS_FALLING_EDGE_ENB) 1 = Enable falling edge detection 0 = Disable falling edge detection R/W 0b 5 Tail Unit Indicator for Timestamp Input Unit Event in Cascade Mode (TS_CASCADE_MODE_TAIL) 1 = This event timestamp input unit is the last unit of the chain in cascade mode. 0 = This event timestamp input unit is not the last unit of the chain in cascade mode. R/W 0b RESERVED RO 00b 1 Select Upstream Timestamp Done Unit in Cascade Mode (TS_CASCADE_MODE_ENB) This bit selects the timestamps done input unit in cascade mode. 1 = Timestamp Unit 1 0 = Timestamp Unit 0 In the head unit, this should be set to "0" so that no upstream timestamp unit will be input to the head unit. R/W 0b 0 Enable This Event Timestamp Input Unit in Cascade Mode (TS_CASCADE_MODE_ENB) 1 = Enables this event timestamp input unit in cascade mode. 0 = Disables this event timestamp input unit in cascade mode. R/W 0b 15:8 4:2 DS00002391A-page 124 2017 Microchip Technology Inc. KSZ8567S 5.1.6.26 Timestamp 1st Sample Time Nanoseconds Register Address: 0x0554 - 0x0557 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 1st Sample Edge Indication (TS_SAMPLE_EDGE_1ST) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 1st Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_1ST) This field is the low-word of the first sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.27 Timestamp 1st Sample Time Seconds Register Address: 0x0558 - 0x055B Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.28 Description Event Timestamp Input of the 1st Sample Time in Seconds (TS_SAMPLE_TIME_S_1ST) This field is the first sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 1st Sample Time Phase Register Address: 0x055C - 0x055F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 1st Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_1ST) This field indicates one of the 8ns cycles for the first sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b 2017 Microchip Technology Inc. DS00002391A-page 125 KSZ8567S 5.1.6.29 Timestamp 2nd Sample Time Nanoseconds Register Address: 0x0560 - 0x0563 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 2nd Sample Edge Indication (TS_SAMPLE_EDGE_2ND) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 2nd Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_2ND) This field is the low-word of the second sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.30 Timestamp 2nd Sample Time Seconds Register Address: 0x0564 - 0x0567 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.31 Description Event Timestamp Input of the 2nd Sample Time in Seconds (TS_SAMPLE_TIME_S_2ND) This field is the second sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 2nd Sample Time Phase Register Address: 0x0568 - 0x056F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 2nd Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_2ND) This field indicates one of the 8ns cycles for the second sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b DS00002391A-page 126 2017 Microchip Technology Inc. KSZ8567S 5.1.6.32 Timestamp 3rd Sample Time Nanoseconds Register Address: 0x056C - 0x056F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 3rd Sample Edge Indication (TS_SAMPLE_EDGE_3RD) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 3rd Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_3RD) This field is the low-word of the third sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.33 Timestamp 3rd Sample Time Seconds Register Address: 0x0570 - 0x0573 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.34 Description Event Timestamp Input of the 3rd Sample Time in Seconds (TS_SAMPLE_TIME_S_3RD) This field is the third sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 3rd Sample Time Phase Register Address: 0x0574 - 0x0577 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 3rd Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_3RD) This field indicates one of the 8ns cycles for the third sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b 2017 Microchip Technology Inc. DS00002391A-page 127 KSZ8567S 5.1.6.35 Timestamp 4th Sample Time Nanoseconds Register Address: 0x0578 - 0x057B Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 4th Sample Edge Indication (TS_SAMPLE_EDGE_4TH) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 4th Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_4TH) This field is the low-word of the fourth sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.36 Timestamp 4th Sample Time Seconds Register Address: 0x057C - 0x057F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.37 Description Event Timestamp Input of the 4th Sample Time in Seconds (TS_SAMPLE_TIME_S_4TH) This field is the fourth sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 4th Sample Time Phase Register Address: 0x0580 - 0x0583 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 4th Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_4TH) This field indicates one of the 8ns cycles for the fourth sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b DS00002391A-page 128 2017 Microchip Technology Inc. KSZ8567S 5.1.6.38 Timestamp 5th Sample Time Nanoseconds Register Address: 0x0584 - 0x0587 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 5th Sample Edge Indication (TS_SAMPLE_EDGE_5TH) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 5th Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_5TH) This field is the low-word of the fifth sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.39 Timestamp 5th Sample Time Seconds Register Address: 0x0588 - 0x058B Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.40 Description Event Timestamp Input of the 5th Sample Time in Seconds (TS_SAMPLE_TIME_S_5TH) This field is the fifth sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 5th Sample Time Phase Register Address: 0x058C - 0x058F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 5th Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_5TH) This field indicates one of the 8ns cycles for the fifth sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b 2017 Microchip Technology Inc. DS00002391A-page 129 KSZ8567S 5.1.6.41 Timestamp 6th Sample Time Nanoseconds Register Address: 0x0590 - 0x0593 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 6th Sample Edge Indication (TS_SAMPLE_EDGE_6TH) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 6th Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_6TH) This field is the low-word of the sixth sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.42 Timestamp 6th Sample Time Seconds Register Address: 0x0594 - 0x0597 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.43 Description Event Timestamp Input of the 6th Sample Time in Seconds (TS_SAMPLE_TIME_S_6TH) This field is the sixth sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 6th Sample Time Phase Register Address: 0x0598 - 0x059B Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 6th Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_6TH) This field indicates one of the 8ns cycles for the sixth sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b DS00002391A-page 130 2017 Microchip Technology Inc. KSZ8567S 5.1.6.44 Timestamp 7th Sample Time Nanoseconds Register Address: 0x059C - 0x059F Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 7th Sample Edge Indication (TS_SAMPLE_EDGE_7TH) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 7th Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_7TH) This field is the low-word of the seventh sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.45 Timestamp 7th Sample Time Seconds Register Address: 0x05A0 - 0x05A3 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.46 Description Event Timestamp Input of the 7th Sample Time in Seconds (TS_SAMPLE_TIME_S_7TH) This field is the seventh sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 7th Sample Time Phase Register Address: 0x05A4 - 0x05A7 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 7th Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_7TH) This field indicates one of the 8ns cycles for the seventh sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b 2017 Microchip Technology Inc. DS00002391A-page 131 KSZ8567S 5.1.6.47 Timestamp 8th Sample Time Nanoseconds Register Address: 0x05A8 - 0x05AB Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31 RESERVED RO 0b 30 Event Timestamp Input of the 8th Sample Edge Indication (TS_SAMPLE_EDGE_8TH) 1 = Indicates the event timestamp input is a rising edge signal 0 = Indicates the event timestamp input is a falling edge signal RO 0b Event Timestamp Input of the 8th Sample Time in Nanoseconds (TS_SAMPLE_TIME_NS_8TH) This field is the low-word of the eighth sample time for the timestamp in nanoseconds. RO 0x00000000 29:0 5.1.6.48 Timestamp 8th Sample Time Seconds Register Address: 0x05AC - 0x05AF Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits 31:0 5.1.6.49 Description Event Timestamp Input of the 8th Sample Time in Seconds (TS_SAMPLE_TIME_S_8TH) This field is the eighth sample time for the timestamp in seconds. Type Default RO 0x00000000 Timestamp 8th Sample Time Phase Register Address: 0x05B0 - 0x05B3 Size: 32 bits This register is indexed by the Timestamp Unit Index Pointer value in bit 8 of the Global PTP Unit Index Register. Bits Description Type Default 31:3 RESERVED RO 0x0000000 2:0 Event Timestamp Input of the 8th Sample Time in Sub 8ns (TS_SAMPLE_TIME_SUB_8NS_8TH) This field indicates one of the 8ns cycles for the eighth sample time of the timestamp. 000 = 0ns (sample time at the first 8ns cycle in 25MHz/40ns) 001 = 8ns (sample time at the second 8ns cycle in 25MHz/40ns) 101 = 16ns (sample time at the third 8ns cycle in 25MHz/40ns) 011 = 24ns (sample time at the fourth 8ns cycle in 25MHz/40ns) 100 = 32ns (sample time at the fifth 8ns cycle in 25MHz/40ns) 101-111 = RESERVED RO 000b DS00002391A-page 132 2017 Microchip Technology Inc. KSZ8567S 5.2 Port Registers This section details the device's port registers. All settings are on a per-port basis. Address field "N" specifies the port number. Valid values for "N" are 1 to 7 for some registers, 6 to 7 for MAC port specific registers, and 0 to 5 for PHY specific registers. For an overview of the device's entire register map, refer to Section 5.0, "Device Registers". For details on the device's global registers, refer to Section 5.1, "Global Registers". 5.2.1 5.2.1.1 PORT N: PORT OPERATION CONTROL REGISTERS (0xN000 - 0xN0FF) Port Default Tag 0 Register Address: 0xN000 Port N: 1-7 Bits 7:0 Size: 8 bits Description Port Default 802.1Q Tag [15:8] Type Default R/W 0x00 Type Default R/W 0x01 Type Default RO 0000_0b RO/WC 0b Bits [7:5]: Priority Code Point (PCP) Bit [4]: Drop Eligible Indicator (DEI) Bits [3:0]: VLAN Identifier (VID) [11:8] 5.2.1.2 Port Default Tag 1 Register Address: Bits 7:0 0xN001 Port N: 1-7 Size: 8 bits Description Port Default 802.1Q tag [7:0] VLAN Identifier (VID) [7:0] 5.2.1.3 Port PME_WoL Event Register Address: Bits 7:3 2 0xN013 Port N: 1-7 Size: 8 bits Description RESERVED Magic Packet Detect To be detected, the destination address of the Magic Packet must match the Global Switch MAC Address in the Switch MAC Address 0 Register through Switch MAC Address 5 Register. 1 = Magic Packet is detected at this port 0 = Not detected 2017 Microchip Technology Inc. DS00002391A-page 133 KSZ8567S Bits 1 Description Link Up Detect Type Default RO/WC 0b RO/WC 0b Type Default RO 0000_0b RO/WC 0b RO/WC 0b RO/WC 0b Applicable only to ports with integrated PHY. 1 = Link up is detected at this port 0 = Not detected 0 Energy Detect Applicable only to ports with integrated PHY. 1 = Cable energy is detected at this port 0 = Not detected 5.2.1.4 Port PME_WoL Enable Register Address: Bits 7:3 2 0xN017 Port N: 1-7 Size: 8 bits Description RESERVED Magic Packet Detect Enable To be detected, the destination address of the Magic Packet must match the Global Switch MAC Address in the Switch MAC Address 0 Register through Switch MAC Address 5 Register. 1 = The PME pin will be asserted when a Magic Packet is detected at this port 0 = The PME pin will not be asserted by Magic Packet detection at this port 1 Link Up Detect Enable Applicable only to ports with integrated PHY. 1 = The PME pin will be asserted when link up is detected at this port 0 = The PME pin will not be asserted by link up detection at this port 0 Energy Detect Enable Applicable only to ports with integrated PHY. 1 = The PME pin will be asserted when cable energy is detected at this port 0 = The PME pin will not be asserted by cable energy detection at this port DS00002391A-page 134 2017 Microchip Technology Inc. KSZ8567S 5.2.1.5 Port Interrupt Status Register Address: 0xN01B Port N: 1-7 Size: 8 bits These registers provide interrupt status for the individual ports. these interrupts are enabled in the Port Interrupt Mask Register. For non-port specific interrupt status, refer to the Global Interrupt Status Register. Bits 7:4 3 Description Type Default RESERVED RO 0000b SGMII Auto-Negotiation Done Interrupt Status RO 0b RO 0b RO 0b RO 0b 0 = No interrupt 2 1 = Interrupt detected PTP Interrupt Status Applies only to PHY ports, not MAC port(s). Refer to the Port PTP Timestamp Interrupt Status Register for individual PTP interrupts. 0 = No interrupt 1 = Interrupt detected 1 PHY Interrupt Status Applies only to PHY ports, not MAC port(s). Refer to the Port Interrupt Control / Status Register for individual PHY interrupts. 0 = No interrupt 1 = Interrupt detected 0 ACL Interrupt Status ACL interrupts can be generated by configuring an entry in the ACL Table for counter mode. All ports have an ACL Table. To clear this bit, toggle the ACL Interrupt Mask bit in the Port Interrupt Mask Register. 0 = No interrupt 1 = Interrupt detected 2017 Microchip Technology Inc. DS00002391A-page 135 KSZ8567S 5.2.1.6 Port Interrupt Mask Register Address: 0xN01F Port N: 1-7 Size: 8 bits This register enables the interrupts in the Port Interrupt Status Register. Bits 7:4 3 Description Type Default RESERVED RO 0000b SGMII Auto-Negotiation Done Interrupt Enable R/W 0b R/W 0b R/W 0b R/W 0b 0 = Interrupt enabled 2 1 = Interrupt disabled PTP Interrupt Enable Applies only to PHY ports. Refer to the Port PTP Timestamp Interrupt Enable Register for individual PTP interrupt enables. 0 = Interrupt enabled 1 = Interrupt disabled 1 PHY Interrupt Enable Applies only to PHY ports. 0 = Interrupt enabled 1 = Interrupt disabled 0 ACL Interrupt Enable Applies to all ports. 0 = Interrupt enabled 1 = Interrupt disabled DS00002391A-page 136 2017 Microchip Technology Inc. KSZ8567S 5.2.1.7 Port Operation Control 0 Register Address: Bits 7 0xN020 Port N: 1-7 Size: 8 bits Description Type Default R/W 0b R/W 0b RESERVED RO 000b Tail Tag Enable R/W 0b R/W 00b Local Loopback Data going from the internal switch fabric to an egress port is looped back at that egress port and returned to the internal switch fabric. 1 = Local loopback 0 = Normal operation 6 Remote Loopback Data received at the external pins of a port is looped back and sent out the same port, without passing through the internal switch fabric. For PHY ports, data is received on the RX+/- pins and transmitted out the TX+/- pins. For xMII ports, data is received on the TXD pins and transmitted on the RXD pins. 1 = Remote loopback 0 = Normal operation 5:3 2 When tail tagging is enabled for a port, it designates that port to be the "host" or "CPU" port. Do not enable tail tagging for more than one port. See the tail tagging description for details. 1 = Enable tail tagging on this port 0 = Disable tail tagging for this port 1:0 Egress Queue Split Enable 11 = Reserved 10 = Four egress queues. Packets are assigned to a queue based on priority. 01 = Two egress queues. Packets are assigned to a queue based on priority. 00 = Single egress queue. There is no priority differentiation. 2017 Microchip Technology Inc. DS00002391A-page 137 KSZ8567S 5.2.1.8 Port Status Register Address: Bits 0xN030 Port N: 1-7 Size: 8 bits Description Type Default 7:5 RESERVED RO 000 4:3 Port Speed Status RO - RO 1 or - RO - RO - For non-PHY ports, these bits duplicate the speed setting in the XMII Port Control 1 Register bit 6 and XMII Port Control 0 Register bit 4. For PHY ports, these bits indicate the actual link speed, which is also available in the PHY Control Register. 00 = 10 Mb/s 01 = 100 Mb/s 10 = 1000 Mb/s 2 Port Duplex Status For non-PHY ports, this bit duplicates the duplex setting in the XMII Port Control 0 Register bit 6. For PHY ports, this bit indicates the actual link duplex, which is also available in the PHY Control Register. 1 = Full duplex 0 = Half duplex 1 Transmit Flow Control Enabled Status For non-PHY ports, this bit duplicates the Tx FC enable bit 5 in the XMII Port Control 0 Register. For PHY ports, this bit is set only when FC is enabled (PHY Auto-Negotiation Advertisement Register, bits 11:10), link is up and FC is established via autonegotiation. 1 = TX flow control is enabled 0 = Disabled 0 Receive Flow Control Enabled Status For non-PHY ports, this bit duplicates the Rx FC enable bit 3 in the XMII Port Control 0 Register. For PHY ports, this bit is set only when FC is enabled (PHY Auto-Negotiation Advertisement Register, bits 11:10), link is up and FC is established via autonegotiation. 1 = RX flow control is enabled 0 = Disabled DS00002391A-page 138 2017 Microchip Technology Inc. KSZ8567S 5.2.2 PORT N: PORT ETHERNET PHY REGISTERS (0xN100 - 0xN1FF) The registers in this section are for PHY ports only. Refer to the IEEE802.3 clause 22.3.4 for additional details. 5.2.2.1 PHY Basic Control Register Address: 0xN100 - 0xN101 Port N: 1-5 Size: 16 bits PHY register 0x00. Bits 15 14 Description Type Default 0b Set this bit to reset this PHY. Registers are not reset. This bit is self-clearing. R/W SC Local Loopback Mode R/W 0b R/W Note 5-4 R/W Note 5-4 R/W 0b R/W 0b R/W, SC 0b R/W 1b R/W 0b PHY Software Reset Data going from the internal switch fabric to an egress port is looped back at that egress port and returned to the internal switch fabric. 1 = Local Loopback mode 0 = Normal operation 13 Speed Select (LSB) This bit is ignored if auto-negotiation is enabled (bit 12 in this register). 1 = 100 Mb/s 0 = 10 Mb/s 12 Auto-Negotiation Enable 1 = Auto-negotiation is enabled 0 = Auto-negotiation is disabled The initial value of this bit is determined by a strapping option, but it may be overwritten. 11 Power Down 1 = Power-down mode 0 = Normal operation When this bit is set to `1', the link-down status might not get updated in the PHY status register. Software should note link is down and should not rely on the PHY status register link status. After this bit is changed from `1' to `0', an internal reset is performed. Wait a minimum of 1ms before read/write access to this PHY's registers. 10 Isolate 1 = Logical isolation of the PHY from the switch core 0 = Normal operation 9 Re-start Auto-Negotiation Set this bit to re-start auto-negotiation. This bit is self-clearing. 8 Duplex Mode This bit is ignored if auto-negotiation is enabled (bit 12 in this register). 1 = Full duplex 0 = Half duplex 7 Collision Test 1 = Enable COL test 0 = Disable COL test 2017 Microchip Technology Inc. DS00002391A-page 139 KSZ8567S Bits Description Type Default 6:0 RESERVED RO 0x00 Note 5-4 The default value of this field is determined by the associated configuration strap value. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. 5.2.2.2 PHY Basic Status Register Address: 0xN102 - 0xN103 Port N: 1-5 Size: 16 bits PHY register 0x01. Bits 15 Description Type Default RO 0b RO 1b RO 1b RO 1b RO 1b RESERVED RO 00b Extended Status RO 0b 100BASE-T4 0 = PHY is not 100BASE-T4 capable 14 100BASE-TX Full Duplex 1 = PHY is 100BASE-TX full-duplex capable 13 100BASE-TX Half Duplex 1 = PHY is 100BASE-TX half-duplex capable 12 10 Mb/s Full Duplex 1 = PHY is 10 Mb/s full-duplex capable 11 10 Mb/s Half Duplex 1 = PHY is 10 Mb/s half-duplex capable 10:9 8 1 = Extended status information available. Note: Since the device PHYs only support 10/100Mbps operation, extended status information is not available. 7 RESERVED RO 0b 6 MF Preamble Suppression RO 1b RO - RO LH - RO 1b RO LL - RO LH - 1 = PHY will accept management frames with preamble suppressed 5 Auto-negotiation Complete 1 = Auto-negotiation process is complete 0 = Auto-negotiation process is not complete 4 Remote (Far End) Fault 1 = Remote fault condition detected 0 = No remote fault detected 3 Auto-Negotiation Ability 1 = PHY is able to perform auto-negotiation 2 Link Status 1 = Link is up 0 = Link is down 1 Jabber Detect 1 = Jabber condition detected 0 = No jabber condition detected DS00002391A-page 140 2017 Microchip Technology Inc. KSZ8567S Bits 0 Description Extended Capability Type Default RO 1b Type Default RO 0x0022 Type Default RO 0x1631 1 = Supports extended capabilities register 0 = Basic register set capabilities only 5.2.2.3 PHY ID High Register Address: 0xN104 - 0xN105 Port N: 1-5 Size: 16 bits PHY register 0x02. Bits 15:0 5.2.2.4 Description PHY Identifier High Word PHY ID Low Register Address: 0xN106 - 0xN107 Port N: 1-5 Size: 16 bits PHY register 0x03. Bits 15:0 Description PHY Identifier Low Word 2017 Microchip Technology Inc. DS00002391A-page 141 KSZ8567S 5.2.2.5 PHY Auto-Negotiation Advertisement Register Address: 0xN108 - 0xN109 Port N: 1-5 Size: 16 bits PHY register 0x04. Bits 15 Description Next Page Type Default R/W 0b 1 = Next page capable 0 = No next page capability 14 Reserved RO 0b 13 Remote Fault R/W 0b RESERVED RO 0b Pause (Flow Control) Capability R/W 01b RO 0b R/W 1b R/W 1b R/W 1b R/W 1b R/W 0x01 1 = Remote fault capable 0 = No remote fault capability 12 11:10 11 = Both Symmetric PAUSE and Asymmetric PAUSE toward local device 10 = Asymmetric PAUSE toward link partner 01 = Symmetric PAUSE 00 = No PAUSE 9 100BASE-T4 1 = 100BASE-T4 capable 0 = No 100BASE-T4 capability 8 100BASE-TX Full Duplex 1 = 100BASE-TX full duplex capable 0 = No 100BASE-TX full duplex capability 7 100BASE-TX Half Duplex 1 = 100BASE-TX half duplex capable 0 = No 100BASE-TX half duplex capability 6 10BASE-T Full Duplex 1 = 10BASE-T full duplex capable 0 = No 10BASE-T full duplex capability 5 10BASE-T Half Duplex 1 = 10BASE-T half duplex capable 0 = No 10BASE-T half duplex capability 4:0 Selector Field 00001 = IEEE 802.3 DS00002391A-page 142 2017 Microchip Technology Inc. KSZ8567S 5.2.2.6 PHY Auto-Negotiation Link Partner Ability Register Address: 0xN10A - 0xN10B Size: Port N: 1-5 16 bits PHY register 0x05. Bits 15 Description Type Default RO - RO - RO - RESERVED RO - Pause (Flow Control) Capability RO - RO - RO - RO - RO - RO - RO - Next Page 1 = Next page capable 0 = No next page capability 14 Acknowledge 1 = Link code word received from partner 0 = Link code word not yet received 13 Remote Fault 1 = Remote fault detected 0 = No remote fault 12 11:10 11 = Both Symmetric PAUSE and Asymmetric PAUSE toward local device 10 = Asymmetric PAUSE toward link partner 01 = Symmetric PAUSE 00 = No PAUSE 9 100BASE-T4 1 = 100BASE-T4 capable 0 = No 100BASE-T4 capability 8 100BASE-TX Full Duplex 1 = 100BASE-TX full duplex capable 0 = No 100BASE-TX full duplex capability 7 100BASE-TX Half Duplex 1 = 100BASE-TX half duplex capable 0 = No 100BASE-TX half duplex capability 6 10BASE-T Full Duplex 1 = 10BASE-T full duplex capable 0 = No 10BASE-T full duplex capability 5 10BASE-T Half Duplex 1 = 10BASE-T half duplex capable 0 = No 10BASE-T half duplex capability 4:0 Selector Field 00001 = IEEE 802.3 2017 Microchip Technology Inc. DS00002391A-page 143 KSZ8567S 5.2.2.7 PHY Auto-Negotiation Expansion Status Register Address: 0xN10C - 0xN10D Size: Port N: 1-5 16 bits PHY register 0x06. Bits 15:5 4 Description Type Default RESERVED RO 0x000 Parallel Detection Fault RO LH - RO - RO 1b RO LH - RO - Type Default R/W 0b 1 = Fault detected by parallel detection 0 = No fault detected by parallel detection 3 Link Partner Next Page Able 1 = Link partner has next page capability 0 = Link partner does not have next page capability 2 Next Page Able 1 = Local device has next page capability 0 = Local device does not have next page capability 1 Page Received 1 = New page received 0 = New page not received 0 Link Partner Auto-Negotiation Able 1 = Link partner has auto-negotiation capability 0 =Link partner does not have auto-negotiation capability 5.2.2.8 PHY Auto-Negotiation Next Page Register Address: 0xN10E - 0xN10F Size: Port N: 1-5 16 bits PHY register 0x07. Bits 15 Description Next Page 1 = Additional next page(s) will follow 0 = Last page 14 Reserved RO 0b 13 Message Page R/W 1b R/W 0b 1 = Message page 0 = Unformatted page 12 Acknowledge 2 1 = Will comply with message 0 = Cannot comply with message DS00002391A-page 144 2017 Microchip Technology Inc. KSZ8567S Bits 11 Description Toggle Type Default RO 0b R/W 0x001 Type Default R/W - RO - RO - RO - RO - RO - 1 = Previous value of transmitted link code word equaled logic one 0 = Previous value of transmitted link code word equaled logic zero 10:0 5.2.2.9 Message Field PHY Auto-Negotiation Link Partner Next Page Ability Register Address: 0xN110 - 0xN111 Port N: 1-5 Size: 16 bits PHY register 0x08. Bits 15 Description Next Page 1 = Additional next page(s) will follow 0 = Last page 14 Acknowledge 1 = Successful receipt of link word 0 = No successful receipt of link word 13 Message Page 1 = Message page 0 = Unformatted page 12 Acknowledge 2 1 = Able to act on the information 0 = Not able to act on the information 11 Toggle 1 = Previous value of transmitted link code word equaled logic zero 0 = Previous value of transmitted link code word equaled logic one 10:0 Message Field 2017 Microchip Technology Inc. DS00002391A-page 145 KSZ8567S 5.2.2.10 PHY MMD Setup Register Address: 0xN11A - 0xN11B Port N: 1-5 Size: 16 bits PHY register 0x0D. Bits 15:14 Description MMD Operation Mode Type Default R/W 00b For the selected MMD device address (bits [4:0] of this register), these two bits select one of the following register or data operations and the usage for the PHY MMD Data Register. 00 = Register 01 = Data, no post increment 10 = Data, post increment on reads and writes 11 = Data, post increment on writes only 13:5 RESERVED R/W 0x000 4:0 MMD Device Address R/W 0x00 Type Default R/W 0x0000 5.2.2.11 PHY MMD Data Register Address: 0xN11C - 0xN11D Size: Port N: 1-5 16 bits PHY register 0x0E. Bits 15:0 Description MMD R/W Index/Data For the selected MMD device address (in the PHY MMD Setup Register, bits [4:0]): When the PHY MMD Setup Register bits [15:14] = 00, this register contains the read/write register access for the MDD device address. Otherwise, this register contains the read/write data value for the MMD device address and its selected register address Refer to the PHY MMD Setup Register bits [15:14] for descriptions of post increment reads and writes of this register for data operation. DS00002391A-page 146 2017 Microchip Technology Inc. KSZ8567S 5.2.2.12 PHY Remote Loopback Register Address: 0xN122 - 0xN123 Port N: 1-5 Size: 16 bits PHY register 0x11. Bits 15:9 8 Description Type Default RESERVED R/W 0x00 Remote Loopback R/W 0b Data received at the external pins of a port is looped back and sent out the same port, without passing through the internal switch fabric. For PHY ports, data is received on the RX+/- pins and transmitted out the TX+/- pins. For xMII ports, data is received on the TXD pins and transmitted on the RXD pins. 1 = Remote loopback 0 = Normal operation 7:2 RESERVED R/W 1111_01b 1 RESERVED R/W RC 0b 0 RESERVED RO 0b Type Default R/W, SC 0b RESERVED R/W 00b Cable Diagnostic Test Pair R/W 0b R/W 00b 5.2.2.13 PHY LinkMD Register Address: 0xN124 - 0xN125 Port N: 1-5 Size: 16 bits PHY register 0x12. Bits 15 Description Cable Diagnostic Test Enable Write value: 1 = Enable cable diagnostic test. After test has completed, this bit is selfcleared. 0 = Disable cable diagnostic test. Read value: 1 = Cable diagnostic test is in progress. 0 = Indicates cable diagnostic test (if enabled) has completed and the status information is valid for read. 14:13 12 Select the differential pair for testing: 0 = Differential pair TXP/TXM 1 = Differential pair RXP/RXM 11:10 RESERVED 2017 Microchip Technology Inc. DS00002391A-page 147 KSZ8567S Bits 9:8 Description Type Default RO 00b RO 0x00 Type Default RESERVED RO LH 0x0000 100BASE-TX Link Status RO - Type Default RO RC 0x0000 Cable Diagnostic Status 00 = Normal cable condition (no fault detected) 01 = Open cable fault detected 10 = Short cable fault detected 11 = Reserved 7:0 5.2.2.14 RESERVED PHY Digital PMA/PCS Status Register Address: 0xN126 - 0xN127 Port N: 1-5 Size: 16 bits PHY register 0x13. Bits 15:1 0 Description 1 = Link is up 0 = Link is down 5.2.2.15 Port RXER Count Register Address: 0xN12A - 0xN12B Size: Port N: 1-5 16 bits PHY register 0x15. Bits 15:0 Description RXER Counter Count of receive frames with one or more symbol errors. DS00002391A-page 148 2017 Microchip Technology Inc. KSZ8567S 5.2.2.16 Port Interrupt Control / Status Register Address: 0xN136 - 0xN137 Port N: 1-5 Size: 16 bits PHY register 0x1B. Bits 15 Description Jabber Interrupt Enable Type Default R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b RO RC 0b RO RC 0b RO RC 0b RO RC 0b 1 = Enable jabber interrupt 0 = Disable jabber interrupt 14 Receive Error Interrupt Enable 1 = Enable receive error interrupt 0 = Disable receive error interrupt 13 Page Received Interrupt Enable 1 = Enable page received interrupt 0 = Disable page received interrupt 12 Parallel Detect Fault Interrupt Enable 1 = Enable parallel detect fault interrupt 0 = Disable parallel detect fault interrupt 11 Link Partner Acknowledge Interrupt Enable 1 = Enable link partner acknowledge interrupt 0 = Disable link partner acknowledge interrupt 10 Link Down Interrupt Enable 1 = Enable link down interrupt 0 = Disable link down interrupt 9 Remote Fault Interrupt Enable 1 = Enable remote fault interrupt 0 = Disable remote fault interrupt 8 Link Up Interrupt Enable 1 = Enable link up interrupt 0 = Disable link up interrupt 7 Jabber Interrupt 1 = Jabber occurred 0 = Jabber did not occur 6 Receive Error Interrupt 1 = Receive error occurred 0 = Receive error did not occur 5 Page Received Interrupt 1 = Page receive occurred 0 = Page receive did not occur 4 Parallel Detect Fault Interrupt 1 = Parallel detect fault occurred 0 = Parallel detect fault did not occur 2017 Microchip Technology Inc. DS00002391A-page 149 KSZ8567S Bits 3 Description Type Default RO RC 0b RO RC 0b RO RC 0b RO RC 0b Type Default RESERVED R/W 0x00 MDI Set R/W 0b R/W 0b R/W 0x00 Link Partner Acknowledge Interrupt 1 = Link partner acknowledge occurred 0 = Link partner acknowledge did not occur 2 Link Down Interrupt 1 = Link down occurred 0 = Link down did not occur 1 Remote Fault Interrupt 1 = Remote fault occurred 0 = Remote fault did not occur 0 Link Up Interrupt 1 = Link up occurred 0 = Link up did not occur 5.2.2.17 PHY Auto MDI / MDI-X Register Address: 0xN138 - 0xN139 Port N: 1-5 Size: 16 bits PHY register 0x1C. Bits 15:8 7 Description When Swap-Off bit is one, 1 = PHY is set to operate in MDI mode 0 = PHY is set to operate in MDI-X mode This bit has no function when Swap-Off is zero. 6 Swap-Off 1 = Disable Auto MDI / MDI-X function 0 = Enable Auto MDI / MDI-X function (normal operation) 5:0 RESERVED DS00002391A-page 150 2017 Microchip Technology Inc. KSZ8567S 5.2.2.18 PHY Control Register Address: 0xN13E - 0xN13F Size: Port N: 1-5 16 bits PHY register 0x1F. Bits Description Type Default 15:12 RESERVED R/W 0x0 11:10 RESERVED RO LH RC 00b Enable Jabber R/W 1b 9 1 = Enable jabber counter 0 = Disable jabber counter 8:7 RESERVED R/W 10b 6 RESERVED RO - 5 Speed Status 100BASE-TX RO - RO - RO - 1 = PHY final speed status is 100BASE-TX 4 Speed Status 10BASE-T 1 = PHY final speed status is 10BASE-T 3 Duplex Status 1 = Full duplex 0 = Half duplex 2 RESERVED RO - 1 RESERVED R/W RC 0b 0 RESERVED RO RC - 2017 Microchip Technology Inc. DS00002391A-page 151 KSZ8567S 5.2.3 PORT N: PORT SGMII CONTROL REGISTERS (0xN200 - 0xN2FF) The Port 7 SGMII registers are accessed indirectly using the following two registers. The indirect SGMII registers, along with details on how to access them, are described in Section 5.5, "SGMII Registers (Indirect)," on page 201. Steps for SGMII register reads: 1. 2. Write the SGMII register address to the Port SGMII Address Register. Read the SGMII register data from the Port SGMII Data Register. Steps for SGMII register write: 1. 2. Write the SGMII register address to the Port SGMII Address Register. Write the SGMII register data to the Port SGMII Data Register. 5.2.3.1 Port SGMII Address Register Address: 0xN200 - 0xN203 Port N: 7 Bits 31:24 23 Size: 32 bits Description Type Default RESERVED RO 0x00 Auto-Increment Enable (AUTO_INC_ENB) R/W 0b Auto Increment can be enabled if the addresses are in sequence. 1 = Address will auto-increment 0 = Address will not auto-increment 22:21 RESERVED RO 00b 20:0 SGMII Address (SGMII_ADDR) R/W 0x1F0000 Type Default R/W 0x0000 This address will increment automatically after each red or write if AutoIncrement Enable is set to one. The upper 5 bits [20:16] are device address. Bit [15]: 0 = standard register 1 = vendor specific register The lowest 8 bits [7:0] are offset register address. 5.2.3.2 Port SGMII Data Register Address: Bits 15:0 0xN206 - 0xN207 Port N: 7 Description SGMII Data (SGMII_DATA) DS00002391A-page 152 Size: 16 bits 2017 Microchip Technology Inc. KSZ8567S 5.2.4 5.2.4.1 PORT N: PORT RGMII/MII/RMII CONTROL REGISTERS (0xN300 - 0xN3FF) XMII Port Control 0 Register Address: Bits 7 0xN300 Port N: 6-7 Size: 8 bits Description RESERVED Type RO Default 1b for port 7 0b for port 6 6 MAC Port Duplex R/W 1b R/W Note 5-5 R/W 1b R/W Note 5-5 R/W 000b 1 = Port MAC interface operates in full-duplex mode 0 = Port MAC interface operates in half-duplex mode 5 MAC Port Transmit Flow Control Enable 1 = Enable transmit flow control on this port 0 = Disable transmit flow control on this port 4 MAC Port Speed 10/100 This bit is ignored if bit 6 in the XMII Port Control 1 Register selects 1000 Mb/ s. If bit 6 in the XMII Port Control 1 Register is one, then this bit determines the port speed: 1 = Port operates at 100 Mb/s 0 = Port operates at 10 Mb/s 3 MAC Port Receive Flow Control Enable 1 = Enable receive flow control on this port 0 = Disable receive flow control on this port 2:0 Note 5-5 RESERVED The default value of this field is determined by the LED1_1 configuration strap value. Refer to Section 3.2.1, "Configuration Straps," on page 17 for additional information. 2017 Microchip Technology Inc. DS00002391A-page 153 KSZ8567S 5.2.4.2 XMII Port Control 1 Register Address: 0xN301 Port N: 6-7 Bits Size: 8 bits Description Type Default 7 RESERVED R/W 0b 6 Port Speed 1000 (Port 6 only) R/W Port 7: 0 Port 6: Config. strap value from RXD6_0 1 = RGMII operates at 10 or 100 Mb/s, as determined by bit 4 in the XMII Port Control 0 Register 0 = RGMII operates at 1000 Mb/s 5 RESERVED R/W 0b 4 RGMII Ingress Internal Delay (RGMII_ID_ig) R/W 0b R/W 1b R/W Port 7: 0 1 = Minimum 1.5 ns delay is added to ingress RGMII clock 0 = No delay is added 3 RGMII Egress Internal Delay (RGMII_ID_eg) 1 = Minimum 1.5 ns delay is added to egress RGMII clock 0 = No delay is added 2 MII / RMII Modes (Port 6 only) Port 6: Config. strap value from RXD6_1 For MII interface: 1 = The MII interface operates as a MAC device (receives clocks, etc.) 0 = The MII interface operates as a PHY device (drives clocks, etc.) For RMII interface: 1 = The 50 MHz RMII REFCLK is received at the RXC pin 0 = A 50 MHz RMII REFCLK is generated from the RXC pin 1:0 Port Interface Type Select (Port 6 only) 00 = Interface is RGMII 01 = Interface is RMII 10 = Interface is MII 11 = Interface is MII DS00002391A-page 154 R/W Port 7: 0 Port 6: Config. strap value from RXD6_[3:2] 2017 Microchip Technology Inc. KSZ8567S 5.2.5 5.2.5.1 PORT N: PORT SWITCH MAC CONTROL REGISTERS (0xN400 - 0xN4FF) Port MAC Control 0 Register Address: Bits 0xN400 Port N: 1-7 Size: 8 bits Description Type Default 0x0 7:4 RESERVED RO 3 RESERVED R/W 0b 2 RESERVED RO 0b 1 Broadcast Storm Protection Enable R/W 0b R/W 0b Type Default 1 = Enable broadcast storm protection for ingress traffic 0 = Disable broadcast storm protection for ingress traffic 0 Jumbo Packet Enable 1 = Accept packets sizes up to 9000 bytes payload (excludes header and CRC) 0 = Standard payload limit of 1500 bytes applies 5.2.5.2 Port MAC Control 1 Register Address: Bits 0xN401 Port N: 1-7 Size: 8 bits Description 7 RESERVED R/W 0b 6 RESERVED RO 0b 5:4 RESERVED R/W 00b Back pressure Enable R/W 0b Reserved R/W 00b Pass All Frames R/W 0b 3 Back pressure is used only for half duplex. Full duplex uses PAUSE frame flow control which is enabled separately. 1 = Enable port back pressure 0 = Disable port back pressure 2:1 0 Bad frames (CRC error, too large, too small) are normally dropped. This feature allows them to be forwarded for mirroring purposes only. This bit does not affect filtering of flow control frames. To disable filtering of flow control frames, refer to the Switch MAC Control 4 Register. 1 = Enable 0 = Disable 2017 Microchip Technology Inc. DS00002391A-page 155 KSZ8567S 5.2.5.3 Port Ingress Rate Limit Control Register Address: Bits 0xN403 Port N: 1-7 Size: 8 bits Description Type Default 7 RESERVED RO 0b 6 Port or Priority Based Ingress Rate Limiting R/W 0b R/W 0b R/W 0b R/W 00b R/W 0b R/W 0b 1 = Ingress rate limiting is port based 0 = Ingress rate limiting is priority based 5 Ingress PPS Mode 1 = Ingress rate limiting is based on number of packets (PPS mode enabled) 0 = Ingress rate limiting is based on number of bits (PPS mode disabled) 4 Ingress Rate Limit Flow Control Enable To use this feature, flow control must also be enabled for this port. Refer to the XMII Port Control 0 Register for non-PHY ports, and the PHY AutoNegotiation Advertisement Register for PHY ports. 1 = Flow control is asserted if the port's receive rate is exceeded 0 = Flow control is not asserted based on ingress rate limits 3:2 Ingress Limit Mode These bits determine what kind of frames are limited and counted against ingress rate limiting. 00 = Count and limit all frames 01 = Count and limit broadcast, multicast and flooded unicast frames only 10 = Count and limit broadcast and multicast frames only 11 = Count and limit broadcast frames only 1 Count IFG Bytes 1 = Each frame's minimum inter-frame gap (IFG) bytes (12 per frame) are included in ingress rate limiting calculations 0 = IFG byte count is not included 0 Count Preamble Bytes Not valid when PPS mode is enabled (bit 5) 1 = Each frame's preamble bytes (8 per frame) are included in ingress rate limiting calculations 0 = Preamble byte count is not included DS00002391A-page 156 2017 Microchip Technology Inc. KSZ8567S 5.2.5.4 Port Priority 0 Ingress Limit Control Register Address: 0xN410 Port N: 1-7 Bits 7 6:0 Size: 8 bits Description Type Default Reserved RO 0b Ingress Data Rate Limit for Priority 0 Frames R/W 0x00 Reference Table 5-3 for code values. Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. TABLE 5-3: DATA RATE SELECTION TABLE FOR INGRESS AND EGRESS RATE LIMITING Code 10Mbps 100Mbps 1000Mbps 7 bits, decimal PPS BPS PPS BPS PPS BPS 7d'0 19.2k 10Mbps 192k 100Mbps 1.92M 1000Mbps 7d'1 - 7d'10 1.92k * code 1Mbps * code 1.92k * code 1Mbps * code 19.2k * code 1Mbps * code 7d'11 - 7d'100 - 10Mbps 1.92k * code 1Mbps * code 19.2k * code 10Mbps * code 7d'101 64 64kbps 64 640kbps 640 640kbps 7d'102 128 128kbps 128 1280kbps 1280 1280kbps 7d'103 256 192kbps 256 1920kbps 2560 1920kbps 7d'104 384 256kbps 384 256kbps 3840 2560kbps 7d'105 512 320kbps 512 320kbps 5120 3200kbps 7d'106 640 384kbps 640 384kbps 6400 3840kbps 7d'107 768 448kbps 768 448kbps 7680 4480kbps 7d'108 896 512kbps 896 512kbps 8960 5120kbps 7d'109 1024 576kbps 1024 576kbps 10,240 5760kbps 7d'110 1152 640kbps 1152 640kbps 11,520 6400kbps 7d'111 1280 704kbps 1280 704kbps 12,800 7040kbps 7d'112 1408 768kbps 1408 768kbps 14,080 7680kbps 7d'113 1536 832kbps 1536 832kbps 15,360 8320kbps 7d'114 1664 896kbps 1664 896kbps 16,640 8960kbps 7d'115 1792 960kbps 1792 960kbps 17,920 9600kbps Note: PPS = Packets Per Second, BPS = Bits Per Second. 2017 Microchip Technology Inc. DS00002391A-page 157 KSZ8567S 5.2.5.5 Port Priority 1 Ingress Limit Control Register Address: Bits 7 6:0 0xN411 Port N: 1-7 Size: 8 bits Description Type Default RESERVED RO 0b Ingress Data Rate Limit for Priority 1 Frames R/W 0x00 Type Default RESERVED RO 0b Ingress Data Rate Limit for Priority 2 Frames R/W 0x00 Type Default Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. 5.2.5.6 Port Priority 2 Ingress Limit Control Register Address: Bits 7 6:0 0xN412 Port N: 1-7 Size: 8 bits Description Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. 5.2.5.7 Port Priority 3 Ingress Limit Control Register Address: Bits 7 6:0 0xN413 Port N: 1-7 Size: 8 bits Description RESERVED RO 0b Ingress Data Rate Limit for Priority 3 Frames R/W 0x00 Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. DS00002391A-page 158 2017 Microchip Technology Inc. KSZ8567S 5.2.5.8 Port Priority 4 Ingress Limit Control Register Address: Bits 7 6:0 0xN414 Port N: 1-7 Size: 8 bits Description Type Default RESERVED RO 0b Ingress Data Rate Limit for Priority 4 Frames R/W 0x00 Type Default RESERVED RO 0b Ingress Data Rate Limit for Priority 5 Frames R/W 0x00 Type Default Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. 5.2.5.9 Port Priority 5 Ingress Limit Control Register Address: Bits 7 6:0 0xN415 Port N: 1-7 Size: 8 bits Description Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. 5.2.5.10 Port Priority 6 Ingress Limit Control Register Address: Bits 7 6:0 0xN416 Port N: 1-7 Size: 8 bits Description RESERVED RO 0b Ingress Data Rate Limit for Priority 6 Frames R/W 0x00 Type Default RESERVED RO 0b Ingress Data Rate Limit for Priority 7 Frames R/W 0x00 Note that the update will not take effect until the Port Priority 7 Ingress Limit Control Register is written. 5.2.5.11 Port Priority 7 Ingress Limit Control Register Address: Bits 7 6:0 0xN417 Port N: 1-7 Description 2017 Microchip Technology Inc. Size: 8 bits DS00002391A-page 159 KSZ8567S 5.2.5.12 Port Queue 0 Egress Limit Control Register Address: Bits 7 6:0 0xN420 Port N: 1-7 Size: 8 bits Description Type Default RESERVED RO 0b Egress Data Rate Limit for Queue 0 Frames R/W 0x00 Type Default RESERVED RO 0b Egress Data Rate Limit for Queue 1 Frames R/W 0x00 Type Default RESERVED RO 0b Egress Data Rate Limit for Queue 2 Frames R/W 0x00 When egress rate limiting is configured to "port-based" via bit 3 of the Switch MAC Control 5 Register, only this register is used for setting. The Port Queue 1-3 Egress Limit Control Registers are used only for queue-based rate limiting. Note that the update will not take effect until the Port Queue 3 Egress Limit Control Register is written. 5.2.5.13 Port Queue 1 Egress Limit Control Register Address: Bits 7 6:0 0xN421 Port N: 1-7 Size: 8 bits Description Note that the update will not take effect until the Port Queue 3 Egress Limit Control Register is written. 5.2.5.14 Port Queue 2 Egress Limit Control Register Address: Bits 7 6:0 0xN422 Port N: 1-7 Size: 8 bits Description Note that the update will not take effect until the Port Queue 3 Egress Limit Control Register is written. DS00002391A-page 160 2017 Microchip Technology Inc. KSZ8567S 5.2.5.15 Port Queue 3 Egress Limit Control Register Address: 0xN423 Port N: 1-7 Bits 7 6:0 5.2.6 Size: 8 bits Description Type Default RESERVED RO 0b Egress Data Rate Limit for Queue 3 Frames R/W 0x00 PORT N: PORT SWITCH MIB COUNTERS REGISTERS (0xN500 - 0xN5FF) Thirty MIB counters are provided for each port. They are accessed via the following two indirect registers. MIB Counters are read-clear. Refer to Section 5.3.6, "Management Information Base (MIB) Counters" for additional details. Note: 5.2.6.1 The Switch MIB Control Register contains two additional bits for global control of MIB counters. Port MIB Control and Status Register Address: 0xN500 - 0xN503 Port N: 1-7 Bits 31 Size: 32 bits Description Type Default RO 0b RESERVED RO 0x00 MIB Read Enable / Count Valid R/W SC 0b R/W 0b MIB Counter Overflow Indication 1 = Counter has overflowed 0 = Counter has not overflowed 30:26 25 Set this bit to `1' to initiate a counter read. When the counter read is complete and the count value is available in the MIB Counter Value, this bit will automatically clear to `0'. 24 MIB Flush and Freeze Enable Refer to the Switch MIB Control Register for the associated global MIB control bits for flushing and freezing port counters. 1 = Enable MIB counter flush and freeze function for this port 0 = Disable MIB counter flush and freeze function for this port 23:16 MIB Index R/W 0x00 15:4 RESERVED RO 0x000 3:0 MIB Counter Value [35:32] RO 0x0 2017 Microchip Technology Inc. DS00002391A-page 161 KSZ8567S 5.2.6.2 Port MIB Data Register Address: Bits 31:0 5.2.7 0xN504 - 0xN507 Port N: 1-7 Size: 32 bits Description MIB Counter Value [31:0] Type Default RO 0x0000 PORT N: PORT SWITCH ACL CONTROL REGISTERS (0xN600 - 0xN6FF) An Access Control List (ACL) Table is provided for each port. It is accessed via the following two indirect registers. Refer to Section 5.3.5, "Access Control List (ACL) Table" for additional details. 5.2.7.1 Port ACL Access 0 Register Address: Bits 0xN600 Port N: 1-7 Size: 8 bits Description Type Default 7:4 RESERVED RO 0x0 3:0 First Rule Number (FRN) R/W 0x0 Type Default 5.2.7.2 Port ACL Access 1 Register Address: Bits 0xN601 Port N: 1-7 Size: Description 8 bits 7:6 RESERVED R/W 00b 5:4 Mode R/W 00b Enable R/W 00b Source / Destination R/W 0b R/W 0b 00 = No action taken 01 = Layer 2 MAC header filtering 10 = Layer 3 IP address filtering 11 = Layer 4 TCP port number / IP protocol filtering 3:2 1 1 = Source 0 = Destination 0 Compare Equal 1 = Match if the compared values are equal 0 = Match if the compared values are not equal DS00002391A-page 162 2017 Microchip Technology Inc. KSZ8567S 5.2.7.3 Port ACL Access 2 Register Address: Bits 7:0 5.2.7.4 R/W 0x00 Type Default R/W 0x00 Type Default R/W 0x00 Type Default R/W 0x00 Type Default R/W 0x00 8 bits 0xN604 Port N: 1-7 Size: 8 bits Description MAC Address [31:24] Port ACL Access 5 Register 0xN605 Port N: 1-7 Size: 8 bits Description MAC Address [23:16] Port ACL Access 6 Register Address: Bits 7:0 Default Port ACL Access 4 Register Bits 5.2.7.7 Size: MAC Address [39:32] Address: 7:0 0xN603 Port N: 1-7 Description Bits 5.2.7.6 Type Port ACL Access 3 Register Address: 7:0 8 bits MAC Address [47:40] Bits 5.2.7.5 Size: Description Address: 7:0 0xN602 Port N: 1-7 0xN606 Port N: 1-7 Description MAC Address [15:8] 2017 Microchip Technology Inc. Size: 8 bits DS00002391A-page 163 KSZ8567S 5.2.7.8 Port ACL Access 7 Register Address: 0xN607 Port N: 1-7 Bits 7:0 5.2.7.9 MAC Address [7:0] 0xN608 Port N: 1-7 Default R/W 0x00 Type Default R/W 0x00 Type Default R/W 0x00 Type Default 8 bits EtherType [15:8] Port ACL Access 9 Register 0xN609 Port N: 1-7 Bits 5.2.7.11 Size: Description Address: 7:0 Type Port ACL Access 8 Register Bits 5.2.7.10 8 bits Description Address: 7:0 Size: Size: 8 bits Description EtherType [7:0] Port ACL Access A Register Address: 0xN60A Port N: 1-7 Bits Size: 8 bits Description 7:6 Priority Mode (PM) R/W 00b 5:3 Priority R/W 00_0b Remark Priority Enable (RPE) R/W 0b Remark Priority [2:1] R/W 00b 2 1:0 The two MSB of the 3-bit Remark Priority field. Refer to the Port ACL Access B Register for the LSB. DS00002391A-page 164 2017 Microchip Technology Inc. KSZ8567S 5.2.7.12 Port ACL Access B Register Address: 0xN60B Port N: 1-7 Bits 7 Size: 8 bits Description Remark Priority [0] Type Default R/W 0b The LSB of the 3-bit Remark Priority field. Refer to the Port ACL Access B Register for the MSB. 6:5 Map Mode (MM) R/W 00b 4:0 RESERVED R/W 0b Type Default R/W 0x00 Type Default RESERVED R/W 0b Port Forward Map R/W 000_0000b 5.2.7.13 Port ACL Access C Register Address: 0xN60C Port N: 1-7 Bits 7:0 5.2.7.14 RESERVED Port ACL Access D Register 0xN60D Port N: 1-7 Bits 6:0 8 bits Description Address: 7 Size: Description Size: 8 bits Each bit corresponds to a forwarding port. Bit 0 corresponds to port 1, Bit 1 corresponds to port 2, etc. 1 = enable forwarding to that port 0 = do not forward to that port 2017 Microchip Technology Inc. DS00002391A-page 165 KSZ8567S 5.2.7.15 Port ACL Access E Register Address: Bits 7:0 5.2.7.16 Type Default R/W 0x00 Type Default R/W 0x00 Type Default R/W 0x00 Port ACL Access F Register 0xN60F Port N: 1-7 Size: 8 bits Description Ruleset [7:0] Port ACL Byte Enable MSB Register Address: Bits 7:0 8 bits Ruleset [15:8] Bits 5.2.7.17 Size: Description Address: 7:0 0xN60E Port N: 1-7 0xN610 Port N: 1-7 Size: 8 bits Description Byte Enable [15:8] Each bit enables accessing one of the ACL bytes when a read or write is initiated by writing to the Port ACL Byte Enable LSB Register. Bit 0 applies to the Port ACL Access 7 Register Bit 1 applies to the Port ACL Access 6 Register, etc. Bit 7 applies to the Port ACL Access 0 Register 1 = Byte is selected for read/write 0 = Byte is not selected DS00002391A-page 166 2017 Microchip Technology Inc. KSZ8567S 5.2.7.18 Port ACL Byte Enable LSB Register Address: 0xN611 Port N: 1-7 Bits 7:0 Size: 8 bits Description Byte Enable [7:0] Type Default R/W 0x00 Type Default Each bit enables accessing one of the ACL bytes when a read or write is initiated by writing to the Port ACL Byte Enable LSB Register. Bit 0 applies to the Port ACL Access F Register Bit 1 applies to the Port ACL Access E Register, etc. Bit 7 applies to the Port ACL Access 8 Register 1 = Byte is selected for read/write 0 = Byte is not selected 5.2.7.19 Port ACL Access Control 0 Register Address: 0xN612 Port N: 1-7 Bits Size: 8 bits Description 7 RESERVED R/O 0b 6 Write Status RO 1b RO 1b R/W 0b R/W 0x0 1 = Write operation is complete 0 = Write operation is not complete 5 Read Status 1 = Read operation is complete 0 = Read operation is not complete; continue polling until this bit is set before reading results registers 4 Write / Read 1 = Write 0 = Read 3:0 ACL Index 2017 Microchip Technology Inc. DS00002391A-page 167 KSZ8567S 5.2.8 5.2.8.1 PORT N: PORT SWITCH INGRESS CONTROL REGISTERS (0xN800 - 0xN8FF) Port Mirroring Control Register Address: 0xN800 Port N: 1-7 Size: 8 bits This register contains the port controls for port mirroring. The Global Port Mirroring and Snooping Control Register must also be properly configured. Bits Description Type Default 7 RESERVED RO 0b 6 Receive Sniff R/W 0b R/W 0b 1 = All packets received on this port are designated as "monitored packets" and will be forwarded to the designated "sniffer port". 0 = No receive monitoring. 5 Transmit Sniff 1 = All packets transmitted on this port are designated as "monitored packets" and will be forwarded to the designated "sniffer port". 0 = No transmit monitoring. 4:2 RESERVED RO 0_00b 1 Sniffer Port R/W 0b RO 0b Type Default R/W 0b R/W 0b 1 = This port is designated as the sniffer port and will transmit monitored packets. 0 = Not a sniffer port. Normal operation. 0 5.2.8.2 RESERVED Port Priority Control Register Address: 0xN801 Port N: 1-7 Size: 8 bits To achieve the desired functionality, do not set more than one bit at a time in this register. Bits 7 Description Highest Priority 1 = Highest priority selected 0 = Highest priority not selected 6 OR'ed Priority 1 = All available priority OR'ed 0 = All available priority not OR'ed 5 RESERVED RO 0b 4 MAC Address Priority Classification R/W 0b 1 = Enable MAC address priority classification for ingress packets on port 0 = Disable MAC address classification DS00002391A-page 168 2017 Microchip Technology Inc. KSZ8567S Bits 3 Description Type Default R/W 0b R/W 0b R/W 0b R/W 0b Type Default R/W 0b RESERVED RO 0b Discard Untagged Packets R/W 0b R/W 0b R/W 000b VLAN Priority Classification 1 = Enable VLAN priority classification for ingress packets on port 0 = Disabled VLAN classification 2 802.1p Priority Classification 1 = Enable 802.1p priority classification for ingress packets on port 0 = Disable 802.1p priority classification 1 Diffserv Priority Classification 1 = Enable Diffserv priority classification for ingress packets on port 0 = Disable Diffserv priority classification 0 ACL Priority Classification 1 = Enable ACL priority classification for ingress packets on port 0 = Disable ACL priority classification 5.2.8.3 Port Ingress MAC Control Register Address: Bits 7 0xN802 Port N: 1-7 Size: 8 bits Description User Priority Ceiling 1 = If a packet's "user priority field" is greater than the "user priority field" in the port default tag register, replace the packet's user priority field with the user priority field in the port default tag from the Port Default Tag 0 Register. 0 = Do not replace the user priority field at this port. 6:5 4 This bit should be left as 0 for the host port. 1 = Any ingress packet at this port without an IEEE 802.1Q tag is discarded. 0 = Do not discard. 3 Discard Tagged Packets 1 = Any ingress packet at this port with an IEEE 802.1Q tag is discarded. 0 = Do not discard. 2:0 Port Default Priority Classification If Diffserv classification, 802.1p classification and VLAN classification are not enabled or fail to classify, ingress packets at this port will be classified with the priority specified here. Possible values are 0 to 7. 2017 Microchip Technology Inc. DS00002391A-page 169 KSZ8567S 5.2.8.4 Port Authentication Control Register Address: Bits 7:3 2 0xN803 Port N: 1-7 Size: 8 bits Description Type Default RESERVED RO 0000_0b Access Control List (ACL) Enable R/O 0b R/W 00b Type Default 1 = enable 0 = enable 1:0 Authentication Mode 00 = Pass Mode. Authentication is disabled. When ACL is enabled, all traffic that misses the ACL rules is forwarded; otherwise ACL actions apply. 01 = Block Mode. Authentication is enabled. When ACL is enabled, all traffic that misses the ACL rules is blocked; otherwise ACL actions apply. 10 = Trap Mode. Authentication is enabled. All traffic is forwarded to the host port. When ACL is enabled, all traffic that misses the ACL rules is blocked; otherwise ACL actions apply. 11 = Reserved 5.2.8.5 Port Pointer Register Address: Bits 0xN804 - 0xN807 Port N: 1-7 Size: Description 32 bits 31:19 RESERVED RO 0x0000 18:16 Port Index Used to point to the indirect mapping locations for the 7 ports R/W 000b 15:2 RESERVED RO 0x0000 1:0 Queue Pointer Used to point to the indirect mapping locations for the 4 queues R/W 00b DS00002391A-page 170 2017 Microchip Technology Inc. KSZ8567S 5.2.8.6 Port Priority to Queue Mapping Register Address: Bits 31:28 0xN808 - 0xN80B Port N: 1-7 Description Regenerated priority (queue) value for priority 7. Size: 32 bits Type Default R/W xx11b R/W xx11b R/W xx10b R/W xx10b R/W xx01b R/W xx01b R/W xx00b R/W xx00b The 2 most significant bits are reserved. 27:24 Regenerated priority (queue) value for priority 6. The 2 most significant bits are reserved. 23:20 Regenerated priority (queue) value for priority 5. 19:16 Regenerated priority (queue) value for priority 4. The 2 most significant bits are reserved. The 2 most significant bits are reserved. 15:12 Regenerated priority (queue) value for priority 3. The 2 most significant bits are reserved. 11:8 Regenerated priority (queue) value for priority 2. 7:4 Regenerated priority (queue) value for priority 1. The 2 most significant bits are reserved. The 2 most significant bits are reserved. 3:0 Regenerated priority (queue) value for priority 0. The 2 most significant bits are reserved. 2017 Microchip Technology Inc. DS00002391A-page 171 KSZ8567S 5.2.8.7 Port Police Control Register Address: 0xN80C - 0xN80F Size: Port N: 1-7 Bits 32 bits Type Default RESERVED RO 0x00000 11 Dropped Color 1 = Color packet PMON holds dropped packets of that color 0 = Color packet PMON holds all packets of that color R/W 0b 10 Drop All 1 = All packets are dropped while max threshold is exceeded in PM WRED 0 = Drop packet based on WRED_PM_PROB_MULTIPLIER R/W 0b 9:8 Packet Type PMON packet type to be read for the connection from Queue Pointer of the port pointed by Port Index 11 = WRED_PMON holds the number of RED packets while read 10 = WRED_PMON holds the number of YELLOW packets while read 01 = WRED_PMON holds the number of GREEN packets while read 00 = WRED_PMON holds the number of dropped packets while read R/W 00b Port Based Policing 1 = Policing is per-port per-queue based 0 = Policing is per-queue based only, Port Index = 0 is used to set aggregated CIR, PIR for each queue R/W 0b NONDSCP_COLOR Color of non-IP frame for color aware R/W 01b 4 Color Mark Enable 1 = DSCP color mark enable 0 = DSCP color mark disable R/W 0b 3 Color Remap Enable 1 = DSCP color remap enable for color aware 0 = DSCP color remap disable, use DSCP color R/W 0b 2 Drop SRP 1 = Allow drop SRP packets while WRED is enabled 0 = Don't allow drop SRP packet while WRED is enabled R/W 0b 1 Police Mode This bit sets the policing modes for the queue if policing is enabled 1 = Color blind 0 = Color aware mode R/W 0b 0 Police Enable 1 = Enable policing and WRED 0 = Disable policing and WRED R/W 0b 31:12 7 6:5 Description DS00002391A-page 172 2017 Microchip Technology Inc. KSZ8567S 5.2.8.8 Port Police Queue Rate Register Address: Bits 0xN820 - 0xN823 Port N: 1-7 Size: 32 bits Description Type Default 31:16 Committed Information Rate Specifies the committed information rate for the connection from the Queue Pointer of the port pointed by the Port index. R/W 0x1000 15:0 Peak Information Rate Specifies the queue's peak information rate for the connection from the Queue Pointer of the port pointed by the Port index. R/W 0x2000 Type Default 5.2.8.9 Port Police Queue Burst Size Register Address: Bits 0xN824 - 0xN827 Port N: 1-7 Size: 32 bits Description 31:16 Committed Burst Size Specifies the queue's committed burst size in bytes that is supported for the connection from the Queue Pointer of the port pointed by the Port index. R/W 0x1000 15:0 Peak Information Rate Specifies the queue's peak burst size in bytes that is supported for the connection from the Queue Pointer of the port pointed by the Port index. R/W 0x3000 Type Default 5.2.8.10 Port WRED Packet Memory Control Register 0 Address: Bits 0xN830 - 0xN833 Port N: 1-7 Description Size: 32 bits 31:27 RESERVED RO 00000b 26:16 WRED Packet Memory Maximum Threshold R/W 0x400 15:11 RESERVED RO 00000b 10:0 WRED Packet Memory Minimum Threshold R/W 0x080 2017 Microchip Technology Inc. DS00002391A-page 173 KSZ8567S 5.2.8.11 Port WRED Packet Memory Control Register 1 Address: Bits 0xN834 - 0xN837 Port N: 1-7 Size: 32 bits Description Type Default 31:27 RESERVED RO 00000b 26:16 WRED Packet Memory Probability Multiplier R/W 0x020 15:11 RESERVED RO 00000b 10:0 WRED Packet Memory Average Queue Size R - Type Default 5.2.8.12 Port WRED Queue Control Register 0 Address: Bits 0xN840 - 0xN843 Port N: 1-7 Size: 32 bits Description 31:27 RESERVED RO 00000b 26:16 WRED Maximum Queue Threshold WRED maximum threshold for the connection from the Queue Pointer of the port pointed to by the Port index. R/W 0x080 15:11 RESERVED RO 00000b 10:0 WRED Minimum Queue Threshold WRED minimum threshold for the connection from the Queue Pointer of the port pointed to by the Port index. R/W 0x009 Type Default 5.2.8.13 Port WRED Queue Control Register 1 Address: Bits 0xN844 - 0xN847 Port N: 1-7 Size: 32 bits Description 31:27 RESERVED RO 00000b 26:16 WRED Queue Probability Multiplier Probability multiplier for the connection from the Queue Pointer of the port pointed to by the Port index. R/W 0x010 15:11 RESERVED RO 00000b 10:0 WRED Packet Memory Average Queue Size Average queue size for the connection from the Queue Pointer of the port pointed to by the Port index. R - DS00002391A-page 174 2017 Microchip Technology Inc. KSZ8567S 5.2.8.14 Port WRED Queue Performance Monitor Control Register Address: Bits 0xN848 - 0xN84B Port N: 1-7 Size: 32 bits Description Type Default 31 Random Drop Enable 1 = Random Drop enabled 0 = Random Drop disabled R/W 0b 30 PMON Flush 1 = PMON counters are flushed. This is a low to high edge triggered clear. 0 = No activity R/W 0x010 29 GYR Drop Disable 1 = Drop GREEN/YELLOW/RED is disabled 0 = Drop GREEN/YELLOW/RED is enabled R/W 0b 28 YR Drop Disable 1 = Drop YELLOW/RED is disabled 0 = Drop YELLOW/RED is enabled R/W 0b 27 R Drop Disable 1 = Drop RED is disabled 0 = Drop RED is enabled R/W 0b 26 Drop All 1 = Drop all packets wile the max threshold is exceeded 0 = Drop based on WRED Queue Probability Multiplier R/W 0b 25:24 RESERVED RO 00b 23:0 Packet Event Counter Packer event count for the connection from the Queue Pointer of the port pointed to by the Port index. R 0x00000 2017 Microchip Technology Inc. DS00002391A-page 175 KSZ8567S 5.2.9 5.2.9.1 PORT N: PORT SWITCH EGRESS CONTROL REGISTERS (0xN900 - 0xN9FF) Port Transmit Queue Index Register Address: 0xN900 - 0xN903 Port N: 1-7 Size: 32 bits This register holds an index value that is used when accessing several registers that follow. This register may be accessed as an 8-bit register, with bits [7:0] located at address 0xN903. Bits Description Type Default 31:2 RESERVED RO 0x00000000 1:0 Queue Index Points to the queue number for subsequent queue configuration registers. R/W 00 Type Default RESERVED RO 0x00000000 Port VID Replacement R/W 0b 5.2.9.2 Port Transmit Queue PVID Register Address: 0xN904 - 0x907 Port N: 1-7 Size: 32 bits This register may be accessed as an 8-bit register, with bits [7:0] located at address 0xN903. Bits 31:1 0 Description 1 = For any egress packet with a non-zero VLAN ID, replace the VID with the port default VID from the Port Default Tag 0 Register and Port Default Tag 1 Register. If double tagging is enabled, it is the ISP tag ID that will be replaced. 0 = Do not replace the VID. DS00002391A-page 176 2017 Microchip Technology Inc. KSZ8567S 5.2.9.3 Port Transmit Queue Control 0 Register Address: 0xN914 Port N: 1-7 Size: 8 bits This register is indexed. Settings are applied on a per-queue and per-port basis. Specify the target queue number in the Port Transmit Queue Index Register before accessing this register. Bits Description Type Default 7:6 Scheduler Mode Determines the egress scheduling policy when 2 or 4 transmit queues are enabled. 00 = Strict Priority. All packets in queue 3 are transmitted before any packets from any lower number queue will be transmitted. All packets in queue 2 are transmitted before any packets from any lower number queue, etc. 10 = Weighted Round Robin (WRR). As determined by the weight values in the (queue indexed) Port Transmit Queue Control 1 Register, a limited number of packets are transmitted from a queue before the next queue is serviced. All queues are serviced in turn. 01 and 11 = RESERVED R/W 10b 5:4 Shaper Mode Determines the egress traffic shaper used. 00 = No shaping. 01 = Credit based shaper (CBS) as defined in IEEE 802.1Qav for AVB 10 = Time aware shaper (TAS) per IEEE 802.1Qbv for TSN 11 = RESERVED R/W 00b 3:0 RESERVED RO 0000b 5.2.9.4 Port Transmit Queue Control 1 Register Address: 0xN915 Port N: 1-7 Size: 8 bits This register is indexed. Settings are applied on a per-queue and per-port basis. Specify the target queue number in the Port Transmit Queue Index Register before accessing this register. Bits 7 6:0 Description Type Default RESERVED RO 0b Queue Weight for WRR Scheduling During weighted round robin (WRR) scheduling, this value specifies the number of packets that may be transmitted from the particular queue before the next queue is serviced. These values are not relevant when the port is configured for a single egress queue. Do not assign 0 to any queue. Example values for a 4-queue configuration are 8 (for queue 3), 4 (for queue 2), 2 (for queue1) and 1 (for queue 0). R/W 0000001 2017 Microchip Technology Inc. DS00002391A-page 177 KSZ8567S 5.2.9.5 Port Transmit Credit Shaper Control 0 Register Address: 0xN916 - 0xN917 Port N: 1-7 Size: 16 bits This register is indexed. Settings are applied on a per-queue and per-port basis. Specify the target queue number in the Port Transmit Queue Index Register before accessing this register. Bits 15:0 5.2.9.6 Description Port Queue Credit High Water Mark Shaper credit high water mark in bytes Type Default R/W 0x0534 Port Transmit Credit Shaper Control 1 Register Address: 0xN918 - 0xN919 Port N: 1-7 Size: 16 bits This register is indexed. Settings are applied on a per-queue and per-port basis. Specify the target queue number in the Port Transmit Queue Index Register before accessing this register. Bits 15:0 5.2.9.7 Description Port Queue Credit Low Water Mark Shaper credit low water mark in bytes Type Default R/W 0x05F2 Port Transmit Credit Shaper Control 2 Register Address: 0xN91A - 0xN91B Size: Port N: 1-7 16 bits This register is indexed. Settings are applied on a per-queue and per-port basis. Specify the target queue number in the Port Transmit Queue Index Register before accessing this register. Bits 15:0 Description Port Queue Credit Increment Shaper credit increment, 12.5% DS00002391A-page 178 Type Default R/W 0x2000 2017 Microchip Technology Inc. KSZ8567S 5.2.9.8 Port Time Aware Shaper Control Register Address: Bits 0xN920 Port N: 1-7 Size: 8 bits Description Type Default 7 Cut-Through Enable 1 = Enable TAS cut-through 0 = Disable TAS cut-through R/W 0b 6 Restricted TAS 1 = TAS packets will not be allowed to transmit out until the OPEN (scheduled) period when this bit is set to 1. They (either store & forward or cut-through) will be queued up. 0 = TAS packets will be allowed to transmit at any time R/W 0b 5:2 RESERVED RO 0000b 1:0 Reference Time Select 11 = Start t0 when reference time is crossed 10 = Repeat t0 on PTP pps (pulse per second) 01 = Free-running, repeats t0 based on internal 1 second pulse 00 = No reference R/W 00b Type Default 5.2.9.9 Port Time Aware Shaper Event Index Register Address: Bits 0xN923 Port N: 1-7 Size: 8 bits Description 7 RESERVED RO 0b 6:0 Event Index R/W 0000000b Type Default 5.2.9.10 Port Time Aware Shaper Event Register Address: Bits 0xN924 - 0xN927 Port N: 1-7 Description Size: 32 bits 31:29 Event Event code: 111 = Repeat event 011-110 = RESERVED 010 = Scheduled open even 001 = Guard band start event 000 = Scheduled closed event R/W 000b 28:0 Time Cycle count of system clock R/W 0x000 2017 Microchip Technology Inc. DS00002391A-page 179 KSZ8567S 5.2.10 5.2.10.1 PORT N: PORT SWITCH QUEUE MANAGEMENT CONTROL REGISTERS (0xNA00 0xNAFF) Port Control 0 Register Address: Bits 0xNA00 - 0xNA03 Size: Port N: 1-7 32 bits Description Type Default 32:2 RESERVED RO 0x0000000 1:0 Drop Mode R/W 00b Type Default 00 = No drop, issue flow control to this port when destination is in congestion state. 01 = Drop packets with priority 0 sourced from this port when destination is in congestion state and disable the flow control to this port. 10 = Drop packets with priority 0, 1 sourced from this port when destination is in congestion state and disable the flow control to this port. 11 = Drop packets with priority 0, 1, 2 sourced from this port when destination is in congestion state and disable the flow control to this port. 5.2.10.2 Port Control 1 Register Address: Bits 0xNA04 - 0xNA07 Size: Port N: 1-7 32 bits Description 31:7 RESERVED RO 0x0000000 6:0 Port VLAN Membership R/W 0x7F Each bit corresponds to a device port. This feature does not utilize VLAN tags or the VLAN Table, and is unrelated to tag-based VLAN functions. Also refer to bit 1 in the Queue Management Control 0 Register. Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Frames may be forwarded to the corresponding port 0 = Frames are blocked from being forwarded to corresponding port DS00002391A-page 180 2017 Microchip Technology Inc. KSZ8567S 5.2.11 5.2.11.1 PORT N: PORT SWITCH ADDRESS LOOKUP CONTROL REGISTERS (0xNB00 - 0xNBFF) Port Control 2 Register Address: Bits 7 0xNB00 Port N: 1-7 Size: 8 bits Description Null VID Lookup Enable Type Default R/W 0b R/W 0b R/W 0b R/W 0b R/W 0b 0b How to handle tagged packets with VID = 0. 1 = Non-standard operation. VLAN Table lookup using VID = 0. 0 = Standard operation. VLAN Table lookup using the port default VID (Port Default Tag 0 Register and Port Default Tag 1 Register) 6 Ingress VLAN Filtering 1 = Discard packets whose VID port membership in the VLAN Table does not include the ingress port 0 = No ingress filtering 5 Discard Non-PVID Packet 1 = Discard packets whose VID does not match the ingress port default VID 0 = Do not compare VID to the port default VID 4 MAC Based 802.1X Enable 1 = Enable MAC based 802.1X authentication in lookup engine 0 = Only use ACL (if enabled) to perform MAC based authentication 3 Self-Address Filtering - Port Enable The source address of received packets is compared to the MAC address in the Switch MAC Address 0 Register through Switch MAC Address 5 Register, and the packet is dropped if there is a match. Both this port enable bit and the global enable bit in the Switch Lookup Engine Control 1 Register must be set to enable self-address filtering. 1 = Enable self-address filtering for this port. 0 = Disable self-address filtering for this port. 2 RESERVED RO 1 RESERVED R/W 0b 0 RESERVED RO 0b 2017 Microchip Technology Inc. DS00002391A-page 181 KSZ8567S 5.2.11.2 Port MSTP Pointer Register Address: Bits 0xNB01 Port N: 1-7 Size: 8 bits Description Type Default 7:3 RESERVED R/W 0000_0b 2:0 MSTP Pointer R/W 000b Points to one of the 8 MSTPs. It is used as an index when reading or writing the Port MSTP State Register. 5.2.11.3 Port MSTP State Register Address: 0xNB04 Port N: 1-7 Size: 8 bits There are eight MSTPs, and the MSTP Pointer in the Port MSTP Pointer Register is used as an index to select a particular MSTP when reading or writing this register. Bits 7:3 2 Description Type Default RESERVED RO 0000_0b Port Transmit Enable R/W 1b R/W 1b R/W 0b 1 = Enable packet transmission on the port 0 = Disable packet transmission on the port 1 Port Receive Enable 1 = Enable packet reception on the port 0 = Disable packet reception on the port 0 Port Learning Disable 1 = Disable switch address learning on the port 0 = Enable switch address learning on the port 5.2.12 5.2.12.1 PORT N: PORT SWITCH PTP CONTROL REGISTERS (0xNC00 - 0xNCFF) Port PTP Receive Latency Register Address: Bits 15:0 0xNC00 - 0xNC01 Size: Port N: 1-7 16 bits Description PTP Port N RX Latency in Nanoseconds (PTP_RX_LATENCY[15:0]) This field is used to set the fixed receive delay value from the port N wire to the RX timestamp reference point. The default value is 415ns. DS00002391A-page 182 Type Default R/W 0x019F 2017 Microchip Technology Inc. KSZ8567S 5.2.12.2 Port PTP Transmit Latency Register Address: 0xNC02 - 0xNC03 Size: Port N: 1-7 16 bits Bits Description Type Default 15:0 PTP Port N TX Latency in Nanoseconds (PTP_TX_LATENCY[15:0]) This field is used to set the fixed transmit delay value from the port N TX timestamp to the wire. The default value is 45ns. R/W 0x002D 5.2.12.3 Port PTP Asymmetry Correction Register Address: 0xNC04 - 0xNC05 Size: Port N: 1-7 16 bits Bits Description Type Default 15 PTP Port N Asymmetry Correction Sign Bit (PTP_ASYM_COR_SIGN) 1 = The magnitude in the PTP_ASM_COR field of this register is negative. 0 = The magnitude in the PTP_ASM_COR field of this register is positive. R/W 0b 14:0 PTP Port N Asymmetry Correction in Nanoseconds (PTP_ASYM_COR) This field is used to set the fixed asymmetry value to add in the correction field for ingress Sync and Pdelay_Resp or to subtract from the correction field for egress Delay_Req and Pdelay_Req. R/W 0x0000 Type Default RO 0x0000 Type Default RO 0x0000 5.2.12.4 Port PTP Egress Timestamp for Request and Delay High Word Register Address: Bits 15:0 5.2.12.5 16 bits Description PTP Port N Egress Timestamp for Pdelay_Req and Delay_Req in Nanoseconds (XDLY_REQ_TS[31:16]) This field contains the Port N egress timestamp high-word value for Pdelay_Req and Delay_Req frames in nanoseconds. Port PTP Egress Timestamp for Request and Delay Low Word Register Address: Bits 15:0 0xNC08 - 0xNC09 Size: Port N: 1-7 0xNC0A - 0xNC0B Size: Port N: 1-7 16 bits Description PTP Port N Egress Timestamp for Pdelay_Req and Delay_Req in Nanoseconds (XDLY_REQ_TS[15:0]) This field contains the Port N egress timestamp low-word value for Pdelay_Req and Delay_Req frames in nanoseconds. 2017 Microchip Technology Inc. DS00002391A-page 183 KSZ8567S 5.2.12.6 Port PTP Egress Timestamp for Sync High Word Register Address: 0xNC0C - 0xNC0D Size: Port N: 1-7 16 bits Bits Description Type Default 15:0 PTP Port N Egress Timestamp for Sync in Nanoseconds (SYNC_TS[31:16]) This field contains the Port N egress timestamp high-word value for the Sync frame in nanoseconds. RO 0x0000 5.2.12.7 Port PTP Egress Timestamp for Sync Low Word Register Address: 0xNC0E - 0xNC0F Size: Port N: 1-7 16 bits Bits Description Type Default 15:0 PTP Port N Egress Timestamp for Sync in Nanoseconds (SYNC_TS[15:16]) This field contains the Port N egress timestamp low-word value for the Sync frame in nanoseconds. RO 0x0000 5.2.12.8 Port PTP Egress Timestamp for PDelay_Resp High Word Register Address: 0xNC10 - 0xNC11 Size: Port N: 1-7 16 bits Bits Description Type Default 15:0 PTP Port N Egress Timestamp for Pdelay_Resp in Nanoseconds (PDLY_TS[31:16]) This field contains the Port N egress timestamp high-word value for the Pdelay_Resp frame in nanoseconds. RO 0x0000 Type Default RO 0x0000 5.2.12.9 Port PTP Egress Timestamp for PDelay_Resp Low Word Register Address: Bits 15:0 0xNC12 - 0xNC13 Size: Port N: 1-7 16 bits Description PTP Port N Egress Timestamp for Pdelay_Resp in Nanoseconds (PDLY_TS[15:0]) This field contains the Port N egress timestamp low-word value for the Pdelay_Resp frame in nanoseconds. DS00002391A-page 184 2017 Microchip Technology Inc. KSZ8567S 5.2.12.10 Port PTP Timestamp Interrupt Status Register Address: 0xNC14 - 0xNC15 Size: Port N: 1-7 16 bits Bits Description Type Default 15 Port N Egress Timestamp for Sync Frame Interrupt Status (TS_SYNC_INT_STATUS) When this bit is set to 1, it indicates that the egress timestamp is available from Port N for the Sync frame. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Status bit in the Port Interrupt Status Register. R/W1C 0b 14 Port N Egress Timestamp for Pdelay_Req and Delay_Req Frames Interrupt Status (TS_PDLY_REQ_INT_STATUS) When this bit is set to 1, it indicates that the egress timestamp is available from Port N for the Pdelay_Req and Delay_Req frames. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Status bit in the Port Interrupt Status Register. R/W1C 0b 13 Port N Egress Timestamp for Pdelay_Resp Frame Interrupt Status (TS_PDLY_RESP_INT_STATUS) When this bit is set to 1, it indicates that the egress timestamp is available from Port N for the Pdelay_Resp frame. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Status bit in the Port Interrupt Status Register. R/W1C 0b RO 0x000 12:0 RESERVED 2017 Microchip Technology Inc. DS00002391A-page 185 KSZ8567S 5.2.12.11 Port PTP Timestamp Interrupt Enable Register Address: 0xNC16 - 0xNC17 Size: Port N: 1-7 16 bits Bits Description Type Default 15 Port N Egress Timestamp for Sync Frame Interrupt Enable (TS_SYNC_INT_ENB) When this bit is set to 1, the egress timestamp interrupt from Port N for the Sync frame is enabled. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Enable bit in the Port Interrupt Mask Register. R/W 0b 14 Port N Egress Timestamp for Pdelay_Req and Delay_Req Frames Interrupt Enable (TS_PDLY_REQ_INT_ENB) When this bit is set to 1, the egress timestamp interrupt from Port N for the Pdelay_Req and Delay_Req frames is enabled. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Enable bit in the Port Interrupt Mask Register. R/W 0b 13 Port N Egress Timestamp for Pdelay_Resp Frame Interrupt Enable (TS_PDLY_RESP_INT_ENB) When this bit is set to 1, the egress timestamp interrupt from Port N for the Pdelay_Resp frame is enabled. This bit is logical OR'ed together with the rest of the bits in this register and the output is shown as the PTP Interrupt Enable bit in the Port Interrupt Mask Register. R/W 0b RESERVED RO 0x000 Type Default R/W 0x0000_0000 12:0 5.2.12.12 Port PTP Link Delay Register Address: Bits 31:0 0xNC18 - 0xNC1B Size: Port N: 1-7 32 bits Description PTP Port N Link Delay in nanoseconds This register is used to set the link delay value between port N and the link partner port. DS00002391A-page 186 2017 Microchip Technology Inc. KSZ8567S 5.3 Tables and MIB Counters (Access) Indirect address and data registers are used to access the various tables and counters: * * * * * * Address Lookup (ALU) Table Static Address Table Reserved Multicast Address Table VLAN Table Access Control List (ACL) Table Management Information Base (MIB) Counters 5.3.1 ADDRESS LOOKUP (ALU) TABLE The Address Lookup Table is the largest of three tables used for MAC address lookup. It supports both dynamic and static MAC address entries. In response to a destination address (DA) lookup, all tables are searched to make a packet forwarding decision. In response to a source address (SA) lookup, only this table is searched for aging, migration and learning of the dynamic entries. It is suggested that static address entries be programmed to the Static Address Table. When that table is full, additional static address entries may be programmed into this table. Static entries will not be aged out. A static DA lookup result (in either this table or the Static Address Table) takes precedence over the dynamic DA lookup result. This table is a 4-way associative memory, with 1K buckets, for a total of 4K entries. In normal operation, the MAC address (and optionally the FID) are hashed to generate a 10-bit index. The 10-bit index specifies a bucket of up to 4 entries, but the entries within each bucket are not individually addressable. A new entry can be added to a bucket if the bucket contains 3 or fewer valid entries. Once a bucket is full with 4 valid entries, any new entry will overwrite the leastrecently-used dynamic entry. If a bucket contains 2 or 3 static entries, adding an additional static entry to that bucket will generate an Almost Full interrupt. (Refer to the Address Lookup Table Interrupt Register and Global Interrupt Status Register). Once the interrupt is generated, the 12-bit absolute address of the new static entry will be available for reading in the Address Lookup Table Entry Index 0 Register bits [11:0]. If a bucket is full with 4 static entries, attempting to write an additional static entry will fail and will result in a Write Fail Interrupt. Once the interrupt is generated, the 10-bit index of the full bucket will be available for reading in the Address Lookup Table Entry Index 0 Register bits [9:0]. A bucket with 4 static entries will also block any dynamic entries from being learned. Such a failure will generate a Learn Fail Interrupt. Once the interrupt is generated, the 10-bit index of the full bucket will be available for reading in the Address Lookup Table Entry Index 1 Register bits [9:0]. This table is accessed indirectly via the ALU Table registers: * * * * * * * ALU Table Index 0 Register ALU Table Index 1 Register ALU Table Access Control Register ALU Table Entry 1 Register ALU Table Entry 2 Register ALU Table Entry 3 Register ALU Table Entry 4 Register All static and dynamic entries may be read, but only static entries may be written. Because most of the table entries are created automatically by hardware address learning, and because the table is 4-way associative, three different methods are available for accessing the memory: 1. 2. 3. Lookup by MAC address. This utilizes the same hash tag lookup circuitry that is used for ingress packet forwarding lookup. The hash tag algorithm is specified by the Hash Option bits in the Switch Lookup Engine Control 0 Register. This is normally how static entries should be written (Read or write). Direct addressing (Read or write). Sequential search, returning all valid entries (Read only). 2017 Microchip Technology Inc. DS00002391A-page 187 KSZ8567S FIGURE 5-3: 5.3.1.1 1. 2. 3. 4. ADDRESS LOOKUP TABLE CONFIGURATION Address Lookup Table Read Operation Write the index to the ALU Table Index 0 Register and ALU Table Index 1 Register. a) A MAC address is used for the hashing indexing function, which is the normal access mechanism for this table. b) A FID is also required if VLAN is enabled. c) If directly addressing the table, a 12-bit index is written to bits [11:0] of the MAC address field. Write to the ALU Table Access Control Register. a) Set the ACTION field to 10 to indicate a read operation. b) Select the addressing method via the DIRECT bit. c) Set the START_FINISH bit to initiate the operation. Read / poll the ALU Table Access Control Register. a) The START_FINISH bit transitions to 0 to indicate that the operation is complete. b) Once START_FINISH is 0, the VALID bit indicates whether or not a valid entry was found. If VALID is false, there is no need to proceed to step 4. c) The VALID_ENTRY_OR_SEARCH_END bit is an aggregate of the START_FINISH bit and the VALID bit. It is intended for use when accessing registers by in-band management (IBA), where polling multiple bits is less practical. Read the contents of the returned table entry from the ALU Table Entry 1 Register, ALU Table Entry 2 Register, ALU Table Entry 3 Register, and ALU Table Entry 4 Register. If no VALID entry is returned, these registers will contain all zeros. DS00002391A-page 188 2017 Microchip Technology Inc. KSZ8567S 5.3.1.2 Address Lookup Table Search Operation The second method to access the Address Lookup Table is through the search operation. The entire table is searched sequentially, revealing each valid entry. Invalid address entries are skipped, providing an efficient way to search the entire table. Setting the START_FINISH bit in the ALU Table Access Control Register begins the search from the top of the table. This bit is cleared when the search is complete. During the table search, the VALID bit in the ALU Table Access Control Register is cleared when the search starts. The VALID bit indicates when a found valid entry is available in the ALU Table Entry registers (0x0420 - 0x042F). After reading the last ALU Table Entry register (ALU Table Entry 4 Register), the search process automatically continues to seek the next valid entry in the address table. The START_FINISH bit remains set until all the valid entries in the table have been returned. The search can be stopped any time by setting the ALU Table Access Control Register START_FINISH bit to 0. 1. 2. 3. 4. 5. Write to the ALU Table Access Control Register. a) Set the ACTION field to 11 to indicate a search operation. b) Set the START_FINISH bit to initiate the operation. Poll the VALID bit until it is set. a) If register access is by in-band management (IBA) rather than SPI or I2C, poll VALID_ENTRY_OR_SEARCH_END instead. This bit goes high to indicate either a new valid entry is returned or the search is complete. Read the entry from the ALU Table Entry 1 Register, ALU Table Entry 2 Register, ALU Table Entry 3 Register, and ALU Table Entry 4 Register, in that order. a) If the search has ended and there are no more valid entries to read, these registers will return all zeros. In this case, go to step 5. If START_FINISH = 0, go to step 5; else go to step 2. Read VALID_COUNT to verify the number of valid entries. 5.3.1.3 1. 2. 3. Address Lookup Table Write Operation Perform a read operation to get the contents of the current entry. The values are kept in the ALU Table Entry 1 Register, ALU Table Entry 2 Register, ALU Table Entry 3 Register, and ALU Table Entry 4 Register. Modify the correct entry as necessary. Set the STATIC bit so that the entry is not aged out. Write to the ALU Table Access Control Register. a) Set the ACTION field to 01 to indicate a write operation. b) Select the addressing method via the DIRECT bit. c) Set the START_FINISH bit to initiate the operation. 2017 Microchip Technology Inc. DS00002391A-page 189 KSZ8567S 5.3.1.4 ALU Table Entry 1 Register Address: 0x0420 - 0x0423 Bits 31 Size: 32 bits Description STATIC Type Default R/W 0b R/W 0b R/W 0b R/W 0_00b 1 = Entry is static and updated by a host processor; will not be aged out 0 = Entry is dynamically learned and aged 30 SRC FILTER 1 = Drop packet if source address match during source learning 0 = Don't drop if source address match 29 DES FILTER 1 = Drop packet if destination address match during lookup 0 = Don't drop if destination address match 28:26 PRIORITY (for static entries) AGE COUNT (for dynamic entries) >0 = Entry has been accessed or learned since last aging process. A default value is reloaded every time the entry is learned or accessed. It is decremented during aging process. 0 = Entry has not been accessed or learned since last aging process. Entry is not valid if it's not static. 25:3 RESERVED RO 0x000000 2:0 MSTP R/W 000b Type Default R/W 0b Multiple Spanning Tree Protocol group ID for matching 5.3.1.5 ALU Table Entry 2 Register Address: Bits 31 0x0424 - 0x0427 Description OVERRIDE Size: 32 bits 1 = Enable overriding of port state 0 = Do not enable 30:7 RESERVED RO 0x000000 6:0 PORT FORWARD R/W 0x00 Each bit corresponds to a device port. Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward to the corresponding port 0 = Do not forward to the corresponding port DS00002391A-page 190 2017 Microchip Technology Inc. KSZ8567S 5.3.1.6 ALU Table Entry 3 Register Address: 0x0428 - 0x042B Bits Size: 32 bits Description Type Default 31:23 RESERVED RO 0x000 22:16 FID R/W 000_0000 R/W 0x0000 Type Default R/W 0x00000000 VLAN group ID for matching 15:0 5.3.1.7 MAC Address [47:32] ALU Table Entry 4 Register Address: Bits 31:0 5.3.2 0x042C - 0x042F Size: 32 bits Description MAC Address [31:0] STATIC ADDRESS TABLE The Static Address Table is one of three tables used for MAC address lookup. It can hold up to 16 static address entries, thereby minimizing the number of static entries that may need to be programmed into the Address Lookup Table, which is used primarily for dynamic entries. In response to a destination address (DA) lookup, all tables are searched to make a packet forwarding decision. Entries in this table are programmed by the host processor, and are never aged. A static DA lookup result (in either this table or the Address Lookup Table) takes precedence over the dynamic DA lookup result. The Static Address Table has 16 entries and is accessed indirectly. The Static Address and Reserved Multicast Table Control Register is used for indexing and read/write control. The following registers are used for the data fields: * * * * Static Address Table Entry 1 Register Static Address Table Entry 2 Register Static Address Table Entry 3 Register Static Address Table Entry 4 Register 5.3.2.1 1. 2. 3. Write the content of the table entry to the Static Address Table Entry 1 Register, Static Address Table Entry 2 Register, Static Address Table Entry 3 Register, and Static Address Table Entry 4 Register. Write to the Static Address and Reserved Multicast Table Control Register. a) Write the TABLE_INDEX field with the 4-bit index value. b) Set the TABLE_SELECT bit to 0 to select the Static Address Table. c) Set the ACTION bit to 0 to indicate a write operation. d) Set the START_FINISH bit to 1 to initiate the operation. When the operation is complete, the START_FINISH bit will be cleared automatically. 5.3.2.2 1. Static Address Table Write Operation Static Address Table Read Operation Write to the Static Address and Reserved Multicast Table Control Register. a) Write the TABLE_INDEX field with the 4-bit index value. b) Set the TABLE_SELECT bit to 0 to select the Static Address Table. 2017 Microchip Technology Inc. DS00002391A-page 191 KSZ8567S 2. c) Set the ACTION bit to 1 to indicate a read operation. d) Set the START_FINISH bit to 1 to initiate the operation. When the operation is complete, the START_FINISH bit will be cleared automatically. a) Read the contents of the indexed entry from the Static Address Table Entry 1 Register, Static Address Table Entry 2 Register, Static Address Table Entry 3 Register, and Static Address Table Entry 4 Register. 5.3.2.3 Static Address Table Entry 1 Register Address: Bits 31 0x0420 - 0x0423 Size: 32 bits Description VALID Type Default R/W 0b R/W 0b R/W 0b R/W 0_00b 1 = Entry is valid 0 = Entry is not valid 30 SRC FILTER 1 = Drop packet if source address match during source learning 0 = Don't drop if source address match 29 DES FILTER 1 = Drop packet if destination address match during lookup 0 = Don't drop if destination address match 28:26 PRIORITY 25:3 RESERVED RO 0x000000 2:0 MSTP R/W 000b Type Default R/W 0b R/W 0b Multiple Spanning Tree Protocol group ID for matching 5.3.2.4 Static Address Table Entry 2 Register Address: Bits 31 0x0424 - 0x0427 Description OVERRIDE Size: 32 bits 1 = Enable overriding of port state 0 = Do not enable 30 USE FID Use FID on multicast packets for matching 29:7 RESERVED RO 0x000000 6:0 PORT FORWARD R/W 0x00 Each bit corresponds to a device port. Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward to the corresponding port 0 = Do not forward to the corresponding port DS00002391A-page 192 2017 Microchip Technology Inc. KSZ8567S 5.3.2.5 Static Address Table Entry 3 Register Address: Bits 0x0428 - 0x042B Size: 32 bits Description Type Default 31:23 RESERVED RO 0x000 22:16 FID R/W 000_0000b R/W 0x0000 Type Default R/W 0x00000000 VLAN group ID for matching 15:0 5.3.2.6 MAC Address [47:32] Static Address Table Entry 4 Register Address: Bits 31:0 5.3.3 0x042C - 0x042F Description MAC Address [31:0] Size: 32 bits RESERVED MULTICAST ADDRESS TABLE The Reserved Multicast Address Table determines the forwarding ports for 48 specific multicast addresses. The table is addressed by the least significant 6 bits of the multicast address, and the table contents are the bits (the PORT_FORWARD field) that represent each possible forwarding port of the device. It is not addressed by the group number in the first column of Table 4-6. Note that the 48 addresses are organized into 8 fixed groups, and changing a forwarding port for one address also makes the same change for all other addresses in the same group. The Reserved Multicast Table is accessed in the same manner as the Static Address Table, using the same indirect access registers. The Static Address and Reserved Multicast Table Control Register is used for indexing and read/write control, while the Reserved Multicast Address Table Entry 2 Register is used for the data fields. 5.3.3.1 1. 2. 3. Write the PORT_FORWARD value to the Reserved Multicast Address Table Entry 2 Register. Write to the Static Address and Reserved Multicast Table Control Register. a) Write the TABLE_INDEX field with the 6-bit index value. b) Set the TABLE_SELECT bit to 1 to select the Reserved Multicast Table. c) Set the ACTION bit to 0 to indicate a write operation. d) Set the START_FINISH bit to 1 to initiate the operation. When the operation is complete, the START_FINISH bit will be cleared automatically. 5.3.3.2 1. 2. Reserved Multicast Table Write Operation Reserved Multicast Table Read Operation Write to the Static Address and Reserved Multicast Table Control Register. a) Write the TABLE_INDEX field with the 6-bit index value. b) Set the TABLE_SELECT bit to 1 to select the Reserved Multicast Table. c) Set the ACTION bit to 1 to indicate a read operation. d) Set the START_FINISH bit to 1 to initiate the operation. When the operation is complete, the START_FINISH bit will be cleared automatically. a) Read the contents of the indexed entry from the Reserved Multicast Address Table Entry 2 Register. 2017 Microchip Technology Inc. DS00002391A-page 193 KSZ8567S 5.3.3.3 Reserved Multicast Address Table Entry 2 Register Address: 0x0424 - 0x0427 Bits Size: 32 bits Description Type Default R/W 00b 31:30 RESERVED 29:7 RESERVED RO 0x000000 6:0 PORT FORWARD R/W 0x00 Each bit corresponds to a device port. Bit 0 is for port 1 Bit 1 is for port 2, etc. 1 = Forward to the corresponding port 0 = Do not forward to the corresponding port 5.3.4 VLAN TABLE An internal VLAN Table is used for VLAN lookup. If 802.1Q VLAN mode is enabled (Switch Lookup Engine Control 0 Register), this table will be used to retrieve the VLAN information that is associated with the ingress packet. The table holds 4096 entries - one for each possible VLAN. The table must be set up before 802.1Q VLAN is enabled. The VLAN table is accessed one entry at a time using the following indirect registers: * * * * * VLAN Table Entry 0 Register VLAN Table Entry 1 Register VLAN Table Entry 2 Register VLAN Table Index Register VLAN Table Access Control Register The table data fields are described in Figure 5-4 and Table 5-4. FIGURE 5-4: VLAN TABLE STRUCTURE Entry # 0 Entry # 1 Entry # 2 Entry # 3 Entry # 4 V FO PRIORITY MSTP FID UNTAG FORWARD PORT FORWARD PORT UNTAG FILTER ID MSTP INDEX Entry # 4094 Entry # 4095 DS00002391A-page 194 PRIORITY FORWARD OPTION VALID 2017 Microchip Technology Inc. KSZ8567S TABLE 5-4: VLAN TABLE DATA FIELDS Field Size (bits) Description VALID 1 When 1, indicates that the table entry is valid. FORWARD OPTION 1 When 1, forward to VLAN port table (PORT FORWARD field). When 0, see Table 4-8, "VLAN Forwarding" for details. PRIORITY 3 Priority value for this VID. MSTP INDEX 3 Multiple Spanning Tree Protocol index. FID 7 Filter ID value. It is combined with destination address and hashed to index the Address Lookup Table. PORT UNTAG PORT FORWARD 7 (1 per port) When 1, untag at that egress port. 7 (1 per port) VLAN port membership list. There is one bit per port, starting with the LSB which corresponds to port 1. When 0, don't untag. A bit value of 1 indicates the associated port is included in the port membership list for that VID. When 0, that port is excluded. 5.3.4.1 1. 2. 3. Write to the VLAN Table Entry 0 Register, VLAN Table Entry 1 Register, and VLAN Table Entry 2 Register to set up the data fields as described in Figure 5-4 and Table 5-4. Write the VLAN Index value in the VLAN Table Index Register. This is the 12-bit index (address) to select the table entry. It is equivalent to the VID which indexes the table during lookup. Write the VLAN Table Access Control Register to specify a write operation, and set START (bit 7). When the operation is complete, bit 7 will be cleared automatically. 5.3.4.2 1. 2. 3. VLAN Table Write Operation VLAN Table Read Operation Write the VLAN Index value in the VLAN Table Index Register to select one of the 4k table entries. Write the VLAN Table Access Control Register to specify a read operation and set START (bit 7). When the operation is complete, bit 7 will be cleared automatically. Read the VLAN Table Entry 0 Register, VLAN Table Entry 1 Register, and VLAN Table Entry 2 Register to retrieve the read results from the VLAN table. 5.3.5 ACCESS CONTROL LIST (ACL) TABLE ACL filtering is implemented individually per-port. The ACL tables are accessed using the Port N: Port Switch ACL Control Registers (0xN600 - 0xN6FF). The 16 entries in each ACL table are addressed indirectly by an index register. Table 5-5 shows how the various fields of the ACL Table entries are mapped to data registers. The Port ACL Byte Enable MSB Register and Port ACL Byte Enable LSB Register make it possible to write or read any combination of bytes. This is useful for writing the Matching rule, Action rule and Process field separately. There are 16 bits in these byte enable registers, corresponding to the 16 data registers Port ACL Access 0 Register through Port ACL Access F Register. Note that the enable bits are applied in reverse order: Bit 0 for the Port ACL Access F Register Bit 1 for the Port ACL Access E Register ... Bit 14 for the Port ACL Access 1 Register Bit 15 for the Port ACL Access 0 Register Also note that the Port ACL Access C Register is not used, so byte enable bit 3 is a don't care. 2017 Microchip Technology Inc. DS00002391A-page 195 KSZ8567S TABLE 5-5: ACL FIELD REGISTER MAPPING MD = 01 ENB = 00 Count Mode MD = 01 ENB 00 Register Bits 0xN600 7:4 RESERVED 3:0 Process Field: FRN [3:0] 7:6 RESERVED 5:4 MD [1:0] 3:2 ENB [1:0] 1 S/D 0 EQ 0xN601 0xN602 7:0 0xN603 7:0 0xN604 7:0 0xN605 7:0 0xN606 7:3 0xN607 MAC ADDRESS [47:0] MD = 10 MD = 11 IP Address [31:0] MIN PORT [15:0] IP MASK [31:0] PC [1:0] 0 PRO [7:0] 7:1 0xN608 7:0 0xN609 7:0 0xN60A 7:6 FME TYPE [15:0] FMSK [7:0] FLAG [7:0] COUNT [10:3] 5:3 Action Rule: PM [1:0] Action Rule: P [2:0] 2 Action Rule: RPE 1:0 Action Rule: RP [2:1] 7 COUNT [2:0] 6:5 Action Field: RP [0] Action Field: MM [1:0] 4:0 RESERVED 0xN60C 7:0 RESERVED 0xN60D 7 6 RESERVED TU 5 CA 4:0 RESERVED Action Field: FORWARD [6:0] 0xN60E 7:0 Process Field: RuleSet [15:8] 0xN60F 7:0 Process Field: RuleSet [7:0] 5.3.5.1 1. 2. 3. RESERVED 2:1 0 0xN60B MAX PORT [15:0] ACL Table Read Write to the Port ACL Access Control 0 Register with the table entry number (0 to 15) in the ACL Index field, and the Write/Read bit 4 cleared to zero. This one write to this register initiates the read operation. Poll the Read Status bit in the Port ACL Access Control 0 Register to determine when the read operation is complete. When the operation is complete, data may be retrieved from the Port ACL Access 0 Register through Port ACL Access F Register. DS00002391A-page 196 2017 Microchip Technology Inc. KSZ8567S 5.3.5.2 1. 2. 3. 4. ACL Table Write Write the ACL table entry values to the Port ACL Access 0 Register through Port ACL Access F Register. Write the Port ACL Byte Enable MSB Register and Port ACL Byte Enable LSB Register to select which registers (Port ACL Access 0 Register through Port ACL Access F Register) are to be written into the ACL table. Write to the Port ACL Access Control 0 Register with the table entry number in the ACL Index field, and the Write/ Read bit 4 set to one. This one write to this register initiates the write operation. The Write Status bit in the Port ACL Access Control 0 Register may be polled to determine when the operation is complete. 5.3.6 MANAGEMENT INFORMATION BASE (MIB) COUNTERS There are 36 MIB counters per port. These counters accumulate a variety of statistics on ingress and egress traffic and events for network management. They are accessed indirectly using the Port MIB Control and Status Register and Port MIB Data Register. The Switch MIB Control Register provides global flush and freeze control of the MIB counters. TABLE 5-6: MIB Index MIB COUNTERS MIB Counter Size (bits) Description RxHiPriorityByte 30 RX high priority octet count, including bad packets. RxUndersizePkt 30 RX undersize packets with good CRC. 0x02 RxFragments 30 RX fragment packets with bad CRC, symbol errors or alignment errors. 0x03 RxOversize 30 RX oversize packets w/ good CRC (max: 1536 or 1522 bytes). 0x04 RxJabbers 30 RX packets longer than 1522 bytes with either CRC errors, alignment errors or symbol errors (depends on max packet size setting); or RX packets longer than 1916 bytes only. 0x05 RxSymbolError 30 RX packets with invalid data symbol; and legal preamble and packet size. 0x06 RxCRCerror 30 RX packets between 64 and 1522 bytes in size, with an integral number of bytes and a bad CRC. (Upper limit depends on max packet size setting.) 0x07 RxAlighmentError 30 RX packets between 64 and 1522 bytes in size, with a non-integral number of bytes and a bad CRC. (Upper limit depends on max packet size setting.) 0x08 RxControl8808Pkts 30 MAC control frames received with 0x8808 in the EtherType field. 0x09 RxPausePkts 30 PAUSE frames received. PAUSE is defined as EtherType (0x8808), DA, control opcode (0x0001), minimum 64 byte data length, and a valid CRC. 0x0A RxBroadcast 30 RX good broadcast packets. Does not include erred broadcast packets or valid multicast packets. 0x0B RXMulticast 30 RX good multicast packets. Does not include MAC control frames, erred multicast packets, or valid broadcast packets. 0x0C RxUnicast 30 RX good unicast packets. 0x0D Rx64Octets 30 RX packets (bad packets included) that are 64 bytes in length. 0x0E Rx65to127Octets 30 RX packets (bad packets included) that are 65 to 127 bytes in length. Rx128to255Octets 30 RX packets (bad packets included) that are 128 to 255 bytes in length. Rx256to511Octets 30 RX packets (bad packets included) that are 256 to 511 bytes in length. 0x11 Rx512to2023Octets 30 RX packets (bad packets included) that are 512 to 1023 bytes in length. 0x12 Rx1024to1522Octets 30 RX packets (bad packets included) that are 1024 to 1522 bytes in length. 0x13 Rx1523to2000Octets 30 RX packets (bad packets included) that are 1523 t0 2000 bytes in length. Rx2001+Octets 30 RX packets (bad packets included) that are between 2001 bytes and the upper limit in length. 0x00 0x01 0x0F 0x10 0x14 2017 Microchip Technology Inc. DS00002391A-page 197 KSZ8567S TABLE 5-6: MIB Index MIB COUNTERS (CONTINUED) MIB Counter Size (bits) Description 0x15 TxHiPriorityByte 30 TX high priority good octet count, including PAUSE packets. 0x16 TxLateCollision 30 Collision is detected later than 512 bit times into the transmission of a packet. 0x17 TxPausePkts 30 PAUSE frames transmitted. PAUSE is EtherType (0x8808), DA, control opcode (0x0001), minimum 64 byte data length, and a valid CRC. 0x18 TxBroadcastPkts 30 TX good broadcast packets. Does not include erred broadcast packets or valid multicast packets. 0x19 TxMulticastPkts 30 TX good multicast packets. Does not include MAC control frames, erred multicast packets, or valid broadcast packets. 0x1A TxUnicastPkts 30 TX good unicast packets. 0x1B TxDeferred 30 TX packets where the first transmit attempt is delayed due to the busy medium. 0x1C TxTotalCollision 30 TX total collisions. Half duplex only. 0x1D TxExcessiveCollision 30 TX fails due to excessive collisions. 0x1E TxSingleCollision 30 Successfully transmitted frames where transmission is inhibited by exactly one collision. 0x1F TxMultipleCollision 30 Successfully transmitted frames where transmission is inhibited by more than one collision. 0x80 RxByteCnt 36 RX byte count. TxByteCnt 36 TX byte count. 0x82 RxDropPackets 30 RX packets dropped due to lack of resources. 0x83 TXDropPackets 30 TX packets dropped due to lack of resources. 0x81 5.3.6.1 MIB Counter Read Operation Indirect access registers are used to read the MIB counters. Separate access registers are provided for each port via the Port MIB Control and Status Register and Port MIB Data Register. All MIB Counters are read-clear. The steps for reading a counter are as follows: 1. 2. 3. 4. Write the MIB Index to bits [23:16] of the Port MIB Control and Status Register. Set the MIB Read Enable in bit 25 of the Port MIB Control and Status Register. This step and the previous step may be done together. Read the MIB Read Enable / Count Valid in bit 25 of the Port MIB Control and Status Register. A '0' value indicates that the read is complete and the count is valid. Read the count value from the Port MIB Data Register. For 36-bit counters, counter bits [35:32] are read from the Port MIB Control and Status Register. The Counter Overflow bit is also found in the Port MIB Control and Status Register. DS00002391A-page 198 2017 Microchip Technology Inc. KSZ8567S 5.3.6.2 MIB Counter Freeze and Flush Functions Counter freeze and flush functions are available on a port-by-port basis. Freezing or flushing counters is initiated by setting the appropriate bit in the Switch MIB Control Register. The freeze or flush function will be applied to all ports for which the flush and freeze functions have been enabled. To enable flush and freeze for a port, set bit 24 in the Port MIB Control and Status Register. The following steps show an example of how flush and freeze are used to collect MIB statistics for all ports for a period of 1 second: 1. 2. 3. Set the MIB Flush and Freeze Enable bit 24 in the Port MIB Control and Status Register for all ports N. Write 0x40 to the Switch MIB Control Register to freeze the MIB counters for all enabled ports. Write 0xC0 to the Switch MIB Control Register to clear the MIB counters for all enabled ports (while continuing to also freeze the counters). At the beginning of the 1 second period, write 0x00 to the Switch MIB Control Register to enable the counters. At the end of the 1 second period, write 0x40 to the Switch MIB Control Register to freeze the counters. Read each counter for each port. 4. 5. 6. 5.4 MDIO Manageable Device (MMD) Registers (Indirect) MMD registers provide indirect read/write access to up to 32 MMD device addresses with each device supporting up to 65,536 16-bit registers, as defined in Clause 22 of the IEEE 802.3 Specification. However, the KSZ8567S uses only a small fraction of the available registers. Refer to Table 5-7, "MMD Register Map" for a list of accessible MMD device addresses and their associated register addresses. Detailed descriptions of the supported MMD registers are provided in the following subsections. Additional information on the MIIM interface is provided in Section 4.11.3, "MII Management (MIIM) Interface," on page 58. The following two standard port registers serve as the portal registers to access the indirect MMD registers. * PHY MMD Setup Register * PHY MMD Data Register TABLE 5-7: MMD REGISTER MAP Device Address (hex) Register Address (hex) Description 2h 00h MMD LED Mode Register 7h 3Ch MMD EEE Advertisement Register Example: MMD Register Write Write MMD - Device Address 2h, Register 00h = 0010h to enable single-LED mode. 1. 2. 3. 4. Write the PHY MMD Setup Register with 0002h // Set up register address for MMD - Device Address 2h. Write the PHY MMD Data Register with 0000h // Select Register 00h of MMD - Device Address 2h. Write the PHY MMD Setup Register with 4002h // Select register data for MMD - Device Address 2h, Reg. 00h. Write the PHY MMD Data Register with 0010h // Write value 0010h to MMD - Device Address 2h, Reg. 00h. Example: MMD Register Read Read MMD - Device Address 2h, Register 11h - 13h for the LED mode status. 1. 2. 3. 4. Write the PHY MMD Setup Register with 0002h // Set up register address for MMD - Device Address 2h. Write the PHY MMD Data Register with 0000h // Select Register 00h of MMD - Device Address 2h. Write the PHY MMD Setup Register with 4002h // Select register data for MMD - Device Address 2h, Reg. 00h. Read the PHY MMD Data Register // Read data in MMD - Device Address 2h, Reg. 00h. 2017 Microchip Technology Inc. DS00002391A-page 199 KSZ8567S 5.4.1 MMD LED MODE REGISTER MMD Address: Register: Bits 15:5 4 0x02 0x00 Size: 16 bits Description Type Default RESERVED RO 0x000 LED Mode R/W 0b RO 0001b Type Default 1 = Single-LED Mode 0 = Tri-color Dual-LED Mode 3:0 5.4.2 RESERVED MMD EEE ADVERTISEMENT REGISTER MMD Address: Register: Bits 15:2 0x07 0x3C Description Size: 16 bits RESERVED RO 0x000 1 100BASE-T EEE Enable 1 = 100 Mbps EEE capable 0 = No 100 Mbps EEE capability R/W 1b 0 RESERVED RO 0b DS00002391A-page 200 2017 Microchip Technology Inc. KSZ8567S 5.5 SGMII Registers (Indirect) This section details the indirectly addressable SGMII registers, which are summarized in Table 5-8, "SGMII Register Map (Indirect)". The indirect SGMII register address space is 21 bits. Detailed descriptions of the supported SGMII registers are provided in the following subsections. Refer to Section 5.2.3, "Port N: Port SGMII Control Registers (0xN200 - 0xN2FF)," on page 152 for additional information on accessing the indirect SGMII registers. The following two standard port registers serve as the portal registers to access the indirect SGMII registers: * Port SGMII Address Register * Port SGMII Data Register TABLE 5-8: SGMII REGISTER MAP (INDIRECT) Address Indirect SGMII Register 0x1F0000 SGMII Control Register 0x1F0001 SGMII Status Register 0x1F0002 SGMII PHY ID 1 Register 0x1F0003 SGMII PHY ID 2 Register 0x1F0004 SGMII Auto-Negotiation Advertisement Register 0x1F8000 SGMII Digital Control Register 0x1F8001 SGMII Auto-Negotiation Control Register 0x1F8002 SGMII Auto-Negotiation Status Register 5.5.1 SGMII CONTROL REGISTER Address: Bits 0x1F0000 Description Size: 16 bits Type Default 15 PHY Reset R/W 0b 14 PHY Loopback R/W 0b 13 Force 100Mbit Speed R/W 0b 1b [bit 6, bit 13] 00 = 10 Mbps 01 = 100 Mbps 10 = 1000 Mbps 11 = RESERVED 12 Auto-Negotiation Enable R/W 11 PHY Power Down R/W 0b 10 RESERVED RO 0b 9 Auto-Negotiation Restart R/WC 0b 8 Force Full Duplex R/W 0b 7 RESERVED RO 0b 6 Force 1000Mbit Speed R/W 0b RO 000000b Refer to bit 13 for details. 5:0 RESERVED The correct link speed should be manually set after auto-negotiation. The detected speed is reflected in the SGMII AutoNegotiation Status Register. 2017 Microchip Technology Inc. DS00002391A-page 201 KSZ8567S 5.5.2 SGMII STATUS REGISTER Address: 0x1F0001 Size: 16 bits These bits only indicate that this device is communicating with a connected device. For auto-negotiation results, such as the link status of the connected PHY, refer to the SGMII Auto-Negotiation Status Register. Bits 15:6 5 Description Type Default RESERVED RO 0000000110b SGMII Auto-Negotiation Complete RO 0b RESERVED RO 00b Device Detect RO 0b RO 00b Type Default RO 0x7996 Type Default RO 0xCED0 Indicates that "acknowledge" auto-negotiation codewords have been exchanged with the attached device. 4:3 2 Indicates that an SGMII device with auto-negotiation is attached. However, auto-negotiation is not necessarily complete. 0 = Auto-negotiation codeword not received from the connected device. 1 = Auto-negotiation codeword received from the connected device. 1:0 5.5.3 RESERVED SGMII PHY ID 1 REGISTER Address: 0x1F0002 Bits 15:0 5.5.4 16 bits Description PHY ID 1 SGMII PHY ID 2 REGISTER Address: Bits 15:0 Size: 0x1F0003 Description PHY ID 2 DS00002391A-page 202 Size: 16 bits 2017 Microchip Technology Inc. KSZ8567S 5.5.5 SGMII AUTO-NEGOTIATION ADVERTISEMENT REGISTER Address: Bits 0x1F0004 Size: 16 bits Description Type Default 00b 15:14 RESERVED RO 13:12 RESERVED R/W 00b 11:9 RESERVED RO 000b 8:5 RESERVED R/W 0001b 4:0 RESERVED RO 00000b Type Default 5.5.6 SGMII DIGITAL CONTROL REGISTER Address: Bits 0x1F8000 Size: 16 bits Description 15 RESERVED R/W 0b 14 Ingress to Egress Loopback Enable R/W 0b RESERVED R/W 0x2400 Type Default RESERVED RO 0000000b Link Up R/W 0b R/W 0b R/W 10b R/W 0b 13:0 5.5.7 SGMII AUTO-NEGOTIATION CONTROL REGISTER Address: Bits 15:5 4 0x1F8001 Size: 16 bits Description This is not a status bit. Set this bit when auto-negotiating and in PHY side SGMII mode. 3 Transmit Configuration Master 0 = Configure port as MAC-side SGMII (a.k.a Master Mode) 1 = Configure port as PHY-side SGMII (a.k.a. Slave Mode) 2:1 PCS Mode 00 = RESERVED 01 = RESERVED 10 = SGMII 11 = RESERVED 0 Auto-Negotiation Complete Interrupt Enable The Link Up and Transmit Configuration Master fields should be set when running as a master. 2017 Microchip Technology Inc. DS00002391A-page 203 KSZ8567S 5.5.8 SGMII AUTO-NEGOTIATION STATUS REGISTER Address: 0x1F8002 Size: 16 bits This register provides the results of the SGMII auto-negotiation. The SGMII Auto-Negotiation Control Register needs to be updated after the correct speed and duplex are detected in this register. Bits 15:5 4 3:2 Description Type Default RESERVED RO 0x000 Link Up RO 0b Link Speed RO 00b 00 = 10 Mbps 01 = 100 Mbps 10 = 1000 Mbps 11 = RESERVED 1 Full Duplex RO 0b 0 RESERVED RO 0b DS00002391A-page 204 2017 Microchip Technology Inc. KSZ8567S 6.0 OPERATIONAL CHARACTERISTICS 6.1 Absolute Maximum Ratings* Supply Voltage (AVDDL, DVDDL, VDDLS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.5 V to +1.8 V Supply Voltage (AVDDH, VDDIO, VDDHS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.5 V to +5.0 V Input Voltage (all inputs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.5 V to +5.0 V Output Voltage (all outputs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.5 V to +5.0 V Lead Temperature (soldering, 20 sec.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +260oC Storage Temperature (TS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .-65oC to +150oC Maximum Junction Temperature (TJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +125oC HBM ESD Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +/-6 kV *Stresses exceeding those listed in this section could cause permanent damage to the device. This is a stress rating only. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. Functional operation of the device at any condition exceeding those indicated in Section 6.2, "Operating Conditions**", Section 6.3, "Electrical Characteristics", or any other applicable section of this specification is not implied. 6.2 Operating Conditions** Supply Voltage (AVDDL, DVDDL, VDDLS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +1.14V to +1.26 V Supply Voltage (AVDDH @ 3.3V) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +3.135 V to +3.465 V Supply Voltage (AVDDH, VDDHS@ 2.5V) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +2.375 V to +2.625 V Supply Voltage (VDDIO @ 3.3V) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +3.135 V to +3.465 V Supply Voltage (VDDIO @ 2.5V) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +2.375 V to +2.625 V Supply Voltage (VDDIO @ 1.8V) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +1.71 V to +1.89 V Ambient Operating Temperature in Still Air (TA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Note 6-1 Junction to Ambient Resistance (JA) (Note 6-2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Note 6-3 Junction to Case Characterization (JT) (Note 6-2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.04oC/W Junction to Case Resistance (JC) (Note 6-2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Note 6-4 Note 6-1 -40oC to +105oC for extended version, . Note 6-2 JT and JA are under a 0 m/s air velocity. Note 6-3 11.3oC/W on a 6-layer PCB per JESD51, 14.4oC/W on a 4-layer PCB per JESD51. Note 6-4 1.5oC/W on a 6-layer PCB per JESD51, 1.21oC/W on a 4-layer PCB per JESD51. **Proper operation of the device is ensured only within the ranges specified in this section. 2017 Microchip Technology Inc. DS00002391A-page 205 KSZ8567S 6.3 Electrical Characteristics TA = 25oC. TABLE 6-1: ELECTRICAL CHARACTERISTICS Symbol Parameter Conditions Min Typ Max Units Supply Current - Mixed 1000/100 Mbps Operation IDD_AH AVDDH supply current IDD_IO VDDIO supply current IDD_CA AVDDL supply current IDD_CD DVDDL supply current IDD_HS VDDHS supply current IDD_LS VDDLS supply current AVDDH @ 2.5V; VDDIO @ 3.3V Ports 1-5 in 100BASE-TX Ports 6 & 7 @ 1000 Mbps All ports 100% utilization 140 mA 40 mA 140 mA 350 mA 20 mA 15 mA 140 mA 25 mA 140 mA 350 mA 20 mA 15 mA Supply Current - Full 100 Mbps Operation IDD_AH AVDDH supply current IDD_IO VDDIO supply current IDD_CA AVDDL supply current IDD_CD DVDDL supply current IDD_HS VDDHS supply current IDD_LS VDDLS supply current AVDDH @ 2.5V; VDDIO @ 3.3V Ports 1-5 in 100BASE-TX Ports 6 & 7 @ 100 Mbps All ports 100% utilization Supply Current - Full 10 Mbps Operation IDD_AH AVDDH supply current 100 mA IDD_IO 20 mA 30 mA 150 mA IDD_HS VDDIO supply current (3.3V) AVDDH @ 2.5V; VDDIO @ 3.3V AVDDL supply current Ports 1-5 in 10BASE-T DVDDL supply current Ports 6 & 7 @ 10Mbps All ports 100% utilization VDDHS supply current 20 mA IDD_LS VDDLS supply current 15 mA IDD_CA IDD_CD Supply Current - Power Management - Energy Detect Mode IDD_AH AVDDH supply current 20 mA IDD_IO VDDIO supply current (3.3V) 30 mA IDD_CA AVDDL supply current 30 mA IDD_CD DVDDL supply current 150 mA IDD_HS VDDHS supply current 20 mA IDD_LS VDDLS supply current 15 mA DS00002391A-page 206 2017 Microchip Technology Inc. KSZ8567S TABLE 6-1: Symbol ELECTRICAL CHARACTERISTICS (CONTINUED) Parameter Conditions Min Typ Max Units Supply Current - Power Management - Global Soft Power Down Mode IDD_AH AVDDH supply current 2 mA IDD_IO VDDIO supply current (3.3V) 6 mA IDD_CA AVDDL supply current 0.01 mA IDD_CD DVDDL supply current 5 mA IDD_HS VDDHS supply current 20 mA IDD_LS VDDLS supply current 15 mA I Type CMOS Input Buffers (VDDIO = 3.3/2.5/1.8V) VIH Input High Voltage VIL Input Low Voltage IIN Input Current 2.1/1.7/1.3 VIN = GND ~ VDDIO -10 V 0.9/0.9/0.6 V 10 A O8 Type CMOS Output Buffers (VDDIO = 3.3/2.5/1.8V) VOH Output High Voltage IOH = 8/8/6 mA VOL Output Low Voltage IOL = 8/8/6 mA IOZ Output Tri-State Leakage 2.4/1.9/1.5 VIN = GND ~ VDDIO V 0.4/0.4/0.2 V 10 A O24 Type CMOS Output Buffers (VDDIO = 3.3/2.5/1.8V) VOH Output High Voltage IOH = 24/24/20 mA VOL Output Low Voltage IOL = 24/24/20 mA IOZ Output Tri-State Leakage 2.4/1.9/1.5 VIN = GND ~ VDDIO V 0.4/0.4/0.2 V 10 A SGMII Inputs VRXEYE Minimum RX Eye Height (differential peak-to-peak) 175 ZIN Differential Input Impedance 85 115 PPM Tolerance -350 350 ppm mV SGMII Outputs VDIFF Driver Output Voltage (peak-to-peak, differential amplitude) 0.8 1.2 V VCTM Transmit Common Mode Voltage 0.4 0.6 V ZD Driver Differential Output Impedance 85 115 2017 Microchip Technology Inc. DS00002391A-page 207 KSZ8567S TABLE 6-1: ELECTRICAL CHARACTERISTICS (CONTINUED) Symbol Parameter Conditions Min Typ Max Units I/O Pin Internal Pull-Up and Pull-Down Effective Resistance R1.8PU I/O Pin Effective Pull-Up Resistance R1.8PD I/O Pin Effective Pull-Down Resistance R2.5PU I/O Pin Effective Pull-Up Resistance R2.5PD I/O Pin Effective Pull-Down Resistance R3.3PU I/O Pin Effective Pull-Up Resistance R3.3PD I/O Pin Effective Pull-Down Resistance VDDIO = 1.8V VDDIO = 2.5V VDDIO = 3.3V 125 k 97 k 58 k 51 k 38 k 39 k 100BASE-TX Transmit (Measured Differentially After 1:1 Transformer) VO Peak Differential Output 100 termination on the differential output Vimb Output Voltage Imbalance 100 termination on the differential output tr , tf Rise/Fall Time Rise/Fall Time Imbalance 0.95 1.05 V 2 % 3 5 ns 0 0.5 ns 0.25 ns 5 % Duty cycle Distortion Overshoot VSET Reference Voltage of ISET (using 6.04k - 1% resistor) Output Jitter 1.21 Peak-to-Peak 0.7 5MHz Square Wave 400 V 1.4 ns 10BASE-T/Te Receive Vsq Squelch Threshold mV 10BASE-Te Transmit (Measured Differentially After 1:1 Transformer) Vp tr , tf Peak Differential Output Voltage 100 termination on the differential output Jitter Added 100 termination on the differential output (peak-to-peak) 1.54 Rise/Fall Time 1.75 1.96 V 3.5 ns 25 ns 10BASE-T Transmit (Measured Differentially After 1:1 Transformer) Vp tr , tf Peak Differential Output Voltage 100 termination on the differential output Jitter Added 100 termination on the differential output (peak-to-peak) Rise/Fall Time DS00002391A-page 208 2.2 2.5 25 2.8 V 3.5 ns ns 2017 Microchip Technology Inc. KSZ8567S 6.4 Timing Specifications This section details the various timing specifications of the device. Note: 6.4.1 The I2C interface timing adheres to the NXP I2C-Bus Specification (UM10204, Rev. 6) (high-speed mode and slower). Refer to the I2C-Bus Specification for additional information. RGMII TIMING Figure 6-1 illustrates the RGMII timing requirements. FIGURE 6-1: RGMII TIMING TCYC TX_CLK TSETUP THOLD TSETUP THOLD TX_CTL TXD[3:0] TCYC RX_CLK TSKEW RX_CTL RXD[3:0] TABLE 6-2: RGMII TIMING VALUES Symbol Description Min Typ Max Units TSETUP Data to clock input setup (Note 6-5) 2.2 ns THOLD Data to clock input hold (Note 6-5) 0 ns TSKEW Data to clock output skew (Note 6-6) 1.2 2.0 Clock cycle duration (Note 6-7) 7.2 8 8.8 ns Duty_G 1000Mbps duty cycle 45 50 55 % Duty_T 10/100Mbps duty cycle 40 50 60 % Note 6-8 ns TCYC Tr / Tf Rise / Fall time (20-80%) ns Note 6-5 For cases where there is no (or insufficient) skew between the input data and input clock, it is possible to add internal delay to the TX_CLK pinout by setting the RGMII Ingress Internal Delay bit in the XMII Port Control 1 Register register. This feature reduces the setup time requirement and increases the hold time requirement nominally by 1.3ns. Note 6-6 The RGMII interface adheres to the RGMII Specification Version 2.0, which specified that the driving device delay the output clock relative to the output data. This is the TSKEW parameter. This skew can be disabled by clearing the RGMII Egress Internal Delay bit in the XMII Port Control 1 Register register. Generally this is not recommended. Note 6-7 For 10Mbps and 100Mbps, TCYC will scale to 400ns +/- 40ns and 40ns +/- 4 ns, respectively. Note 6-8 0.75ns for VDDIO = 3.3V/2.5V, 1.0ns for VDDIO = 1.8V 2017 Microchip Technology Inc. DS00002391A-page 209 KSZ8567S 6.4.2 MII TIMING 6.4.2.1 MII Transmit Timing in MAC Mode Figure 6-2 illustrates a write operation from the KSZ8567S to a PHY or other device while operating the KSZ8567S in MAC Mode. FIGURE 6-2: MII TRANSMIT TIMING IN MAC MODE TABLE 6-3: MII TRANSMIT TIMING IN MAC MODE VALUES Symbol Description tP (100BASE-TX / Min Typ Max Units RX_CLK period 40/400 ns RX_CLK pulse width low 20/200 ns RX_CLK pulse width high 20/200 ns 16 ns 10BASE-T/Te) tWL (100BASE-TX / 10BASE-T/Te) tWH (100BASE-TX / 10BASE-T/Te) tOD RX_DV, RXD_[3:0] output delay from rising edge of RX_CLK DS00002391A-page 210 2017 Microchip Technology Inc. KSZ8567S 6.4.2.2 MII Receive Timing in MAC Mode Figure 6-3 illustrates a read operation by the KSZ8567S from a PHY or other device while operating the KSZ8567S in MAC Mode. FIGURE 6-3: TABLE 6-4: MII RECEIVE TIMING IN MAC MODE MII RECEIVE TIMING IN MAC MODE VALUES Symbol tP (100BASE-TX / Description Min Typ Max Units TX_CLK period 40/400 ns TX_CLK pulse width low 20/200 ns TX_CLK pulse width high 20/200 ns 10BASE-T/Te) tWL (100BASE-TX / 10BASE-T/Te) tWH (100BASE-TX / 10BASE-T/Te) tSU1 TXD_[3:0] setup time to rising edge of TX_CLK 10 ns tSU2 TX_EN, TX_ER setup time to rising edge of TX_CLK 10 ns tHD1 TXD_[3:0] hold time from rising edge of TX_CLK 10 ns tHD2 TX_EN, TX_ER hold time from rising edge of TX_CLK 10 ns 2017 Microchip Technology Inc. DS00002391A-page 211 KSZ8567S 6.4.2.3 MII Receive Timing in PHY Mode FIGURE 6-4: TABLE 6-5: MII RECEIVE TIMING IN PHY MODE MII RECEIVE TIMING IN PHY MODE VALUES Symbol tP (100BASE-TX / Description Min Typ Max Units RX_CLK period 40/400 ns RX_CLK pulse width low 20/200 ns RX_CLK pulse width high 20/200 ns 20 ns 10BASE-T/Te) tWL (100BASE-TX / 10BASE-T/Te) tWH (100BASE-TX / 10BASE-T/Te) tOD RX_DV, RXD_[3:0] output delay from rising edge of RX_CLK DS00002391A-page 212 2017 Microchip Technology Inc. KSZ8567S 6.4.2.4 MII Transmit Timing in PHY Mode FIGURE 6-5: MII TRANSMIT TIMING IN PHY MODE TABLE 6-6: MII TRANSMIT TIMING IN PHY MODE VALUES Symbol Description tP (100BASE-TX / Min Typ Max Units TX_CLK period 40/400 ns TX_CLK pulse width low 20/200 ns TX_CLK pulse width high 20/200 ns 10BASE-T/Te) tWL (100BASE-TX / 10BASE-T/Te) tWH (100BASE-TX / 10BASE-T/Te) tSU1 TXD_[3:0] setup time to rising edge of TX_CLK 10 ns tSU2 TX_EN, TX_ER setup time to rising edge of TX_CLK 10 ns tHD1 TXD_[3:0] hold time from rising edge of TX_CLK 0 ns tHD2 TX_EN, TX_ER hold time from rising edge of TX_CLK 0 ns 2017 Microchip Technology Inc. DS00002391A-page 213 KSZ8567S 6.4.3 RMII TIMING Figure 6-6 and Figure 6-7 illustrate the RMII timing requirements. FIGURE 6-6: RMII TRANSMIT TIMING FIGURE 6-7: RMII RECEIVE TIMING TABLE 6-7: RMII TIMING VALUES Symbol Description Min tcyc Clock cycle t1 Setup time t2 Hold time 2 tod Output delay 7 DS00002391A-page 214 Typ Max 20 Units ns 4 ns ns 9 13 ns 2017 Microchip Technology Inc. KSZ8567S 6.4.4 MIIM TIMING Figure 6-8 illustrates the MIIM timing requirements. FIGURE 6-8: TABLE 6-8: MIIM TIMING MIIM TIMING VALUES Symbol Description Min Typ Max Units tP MDC period tOD Output delay tSU MDIO setup time to rising edge of MDC 10 ns tHD MDIO hold time from rising edge of MDC 5 ns 2017 Microchip Technology Inc. 400 ns 200 ns DS00002391A-page 215 KSZ8567S 6.4.5 SPI TIMING Figure 6-9 and Figure 6-10 illustrate the SPI timing requirements. FIGURE 6-9: SPI DATA INPUT TIMING SCS_N SCL SDI SDO FIGURE 6-10: SPI DATA OUTPUT TIMING SCS_N SCL SDO SDO SDI TABLE 6-9: SPI TIMING VALUES Symbol fSCLK Description Min SCL clock frequency Typ Max Units 50 MHz t1 SCS_N active setup time 8 ns t2 SDI data input setup time 3 ns t3 SDI data input hold time 3 ns t4 SCS_N active hold time 8 ns t5 SCS_N disable high time 8 t6 SCL falling edge to SDO data output valid 2 t7 SCS_N inactive to SDO data input invalid 1 DS00002391A-page 216 ns 9 ns ns 2017 Microchip Technology Inc. KSZ8567S 6.4.6 AUTO-NEGOTIATION TIMING Figure 6-11 illustrates the Auto-Negotiation timing requirements. FIGURE 6-11: TABLE 6-10: AUTO-NEGOTIATION TIMING AUTO-NEGOTIATION TIMING VALUES Symbol tBTB tFLPW Description FLP burst to FLP burst Min Typ Max Units 8 16 24 ms FLP burst width 2 ms 100 ns tPW Clock/Data pulse width tCTD Clock pulse to data pulse 55.5 64 69.5 s tCTC Clock pulse to clock pulse 111 128 139 s Number of clock/data pulses per burst 17 2017 Microchip Technology Inc. 33 DS00002391A-page 217 KSZ8567S 6.4.7 TRIGGER OUTPUT UNIT AND TIMESTAMP INPUT UNIT TIMING Figure 6-12 provides details and constraints on various timing relationships within the twelve trigger output units and the timestamp input units. FIGURE 6-12: DS00002391A-page 218 TRIGGER OUTPUT UNIT AND TIMESTAMP INPUT UNIT TIMING 2017 Microchip Technology Inc. KSZ8567S TABLE 6-11: AUTO-NEGOTIATION TIMING VALUES Symbol Description Min Typ Max Units Trigger Output Unit Timing (Cascade Mode) tCASP1 In cascade mode for TRIGX_CFG_1[6:4] = 100, or 101, or 110 (Neg. Edge, Pos. Edge, and Shift Reg. Output signals). Minimum time between start of one TOU and the start of another TOU cascaded on the same GPIO pin. 80 ns tCASP2 In cascade mode for TRIGX_CFG_1[6:4] = 010, 011, 100, or 101 (Neg. Pulse, Pos. Pulse, Neg. Periodic, and Pos. Periodic Output signals). Minimum time between start of one TOU and the start of another TOU cascaded on the same GPIO pin. 120 ns tCYCCASP In cascade mode for TRIGX_CFG_1[6:4] = 010, and 011 (Neg. Pulse, Pos. Pulse Output signals). In cascade mode, the cycle time of the trigger output unit operating in the indicated modes. 80 32 + PWIDTH2 ns tCYCNC1 In cascade mode for TRIGX_CFG_1[6:4] = 100 or 101 (Neg. Periodic, Pos. periodic Output signals). Minimum cycle time for any trigger output unit operating in the indicated modes. 80 32 + PWIDTH2 ns tGAP23 In cascade mode for TRIGX_CFG_1[6:4] = 010, and 011 (Neg. Pulse, Pos. Pulse Output signals): Minimum gap time required between end of period of first trigger output unit to beginning of output of 2nd trigger output unit. 80 ns In cascade mode, the minimum low or high pulse width of the trigger output unit. 8 ns PWIDTH2 Trigger Output Unit Timing (Non-Cascade Mode) 32 + PWIDTH2 tCYCNC2 In non-cascade mode, the minimum cycle time for any trigger output unit. 80 ns tPOGAP In non-cascade mode, the minimum time between the end of the generated pulse to the start of the next pulse. 32 ns PWIDTH1 In non-cascade mode, the minimum low or high pulse width of the trigger output unit. 8 ns Timestamp Input Unit Timing IPHIGH Allowable high time of an incoming digital waveform on any GPIO pin. 24 ns IPLOW In non-cascade mode, the minimum time between the end of the generated pulse to the start of the next pulse. 24 ns IPCYC In non-cascade mode, the minimum time between the end of the generated pulse to the start of the next pulse. 48 ns 2017 Microchip Technology Inc. DS00002391A-page 219 KSZ8567S 6.4.8 POWER-UP AND RESET TIMING Figure 6-13 illustrates the power-up and reset timing requirements. FIGURE 6-13: POWER-UP AND RESET TIMING NOTE 1 TRANSCEIVER (AVDDH), DIGITAL I/Os (VDDIO) NOTE 3 CORE (AVDDL, DVDDL) NOTE 2 SUPPLY VOLTAGES tvr tpc tsr RESET_N tcs tch CONFIGURATION STRAP INPUT trc CONFIGURATION STRAP OUTPUT TABLE 6-12: POWER-UP AND RESET TIMING VALUES Symbol tvr Description Min Typ Max Units Supply voltage rise time (must be monotonic) 200 s tsr Stable supply voltages to de-assertion of reset 10 ms tcs Configuration strap input setup time 5 ns tch Configuration strap input hold time 5 ns trc De-assertion of reset to configuration strap pin output 6 ns tpc Supply voltages cycle off-to-on time 150 ms Note 1: The recommended powering sequence is to bring up all voltages at the same time. However, if that cannot be attained, then the recommended power-up sequence is to power-up the transceiver (AVDDH) and digital I/Os (VDDIO) voltages before the low voltage core (AVDDL and DVDDL). There is no power sequence requirement between transceiver (AVDDH) and digital I/Os (VDDIO) power rails. The power-up waveforms should be monotonic for all supply voltages. Note 2: After the de-assertion of reset, it is recommended to wait a minimum of 100s before starting to program the device through any interface. Note 3: The recommended power down sequence is to power down the low voltage core before powering down the transceiver and digital I/O voltages, or to have all supplies power down in unison. Before the next power-up cycle, all supply voltages to the device should reach less than 0.4V and there should be a minimum wait time of 150ms from power-off to power-on. DS00002391A-page 220 2017 Microchip Technology Inc. KSZ8567S 6.5 Clock Specifications A crystal or external clock source, such as an oscillator, is used to provide a 25MHz reference clock for the KSZ8567S. If an external clock source is used, the XO pin must be left floating. Since the XI/XO circuit is powered from AVDDH, the external clock source should also be powered from the same power rail. Figure 6-14 details the available connection methods. Table 6-13 details the recommended crystal specifications. FIGURE 6-14: INPUT REFERENCE CLOCK CONNECTION OPTIONS Ethernet Switch Ethernet Switch XI XO XI 25 MHz OSC +/-50ppm No Connect XO 25 MHz XTAL +/-50ppm TABLE 6-13: REFERENCE CRYSTAL CHARACTERISTICS Characteristic Min Typ Max Oscillation Mode Fundamental Frequency 25 Units MHz 50 ppm Effective Series Resistance (ESR) 50 Total period jitter (peak-to-peak) 100 ps Drive level 100 uW Frequency tolerance 2017 Microchip Technology Inc. DS00002391A-page 221 KSZ8567S 7.0 DESIGN GUIDELINES This section provides general design guidelines for the following: * Reset Circuit Guidelines * Magnetics Connection and Selection Guidelines 7.1 Reset Circuit Guidelines Figure 7-1 illustrates the recommended reset circuit for powering up the KSZ8567S if reset is triggered by the power supply. FIGURE 7-1: SIMPLE RESET CIRCUIT VDDIO D1: 1N4148 D1 Ethernet Switch R 10K RESET_N C 10uF Figure 7-2 illustrates a reset circuit recommended for applications where reset is driven by another device, such as a CPU. At power-on reset, R, C and D1 provide the necessary ramp rise time to reset the KSZ8567S. The RST_OUT_N from the CPU provides a warm reset after power-up. FIGURE 7-2: RESET CIRCUIT FOR CPU RESET INTERFACE VDDIO Ethernet Switch R 10K D1 CPU/FPGA RESET_N RST_OUT_N C 10uF D2 D1, D2: 1N4148 DS00002391A-page 222 2017 Microchip Technology Inc. KSZ8567S 7.2 Magnetics Connection and Selection Guidelines A 1:1 isolation transformer is required at the line interface. For designs exceeding FCC requirements, utilize one with integrated common-mode chokes. An optional auto-transformer stage following the chokes provides additional common-mode noise and signal attenuation. The KSZ8567S PHY port design incorporates voltage-mode transmit drivers and on-chip terminations. With the voltagemode implementation, the transmit drivers supply the common-mode voltages to the two differential pairs. Therefore, the two transformer center tap pins on the KSZ8567S chip side should not be connected to any power supply source on the board; rather, the center tap pins should be separated from one another and connected through separate 0.1F common-mode capacitors to ground. Separation is required because the common-mode voltage could be different between the differential pairs, depending on the connected speed mode. Figure 7-3 details a typical magnetic interface circuit for the KSZ8567S PHY port. FIGURE 7-3: TYPICAL MAGNETIC INTERFACE CIRCUIT Table 7-1 provides a list of recommended magnetic characteristics. TABLE 7-1: MAGNETICS SELECTION CRITERIA Parameter Turns ratio Open-circuit inductance (min.) Insertion loss (typ.) HIPOT (min.) 2017 Microchip Technology Inc. Value Test Condition 1 CT : 1 CT 350H 100mV, 100KHz, 8mA 1.0dB 100KHz to 100MHz 1500vrms DS00002391A-page 223 KSZ8567S 8.0 PACKAGE INFORMATION 8.1 Package Marking Information 128-TQFP-EP MICROCHIP KSZ8567STXt e3 Rnnn e3 VVVCO YYWWNNN Legend: Note: t R nnn e3 VVV CO YY WW NNN Temperature range designator (V = Extended) Product revision Internal code Pb-free JEDEC(R) designator for Matte Tin (Sn) Plant assembly Country of origin Year code (last two digits of calendar year) Week code (week of January 1 is week `01') Alphanumeric traceability code In the event the full Microchip part number cannot be marked on one line, it will be carried over to the next line, thus limiting the number of available characters for customer-specific information. * Standard device marking consists of Microchip part number, year code, week code and traceability code. For device marking beyond this, certain price adders apply. Please check with your Microchip Sales Office. For QTP devices, any special marking adders are included in QTP price. DS00002391A-page 224 2017 Microchip Technology Inc. KSZ8567S 8.2 Note: Package Drawings For the most current package drawings, please see the Microchip Packaging Specification located at http://www.microchip.com/packaging FIGURE 8-1: PACKAGE (DRAWING & DIMENSIONS) 2017 Microchip Technology Inc. DS00002391A-page 225 KSZ8567S APPENDIX A: TABLE A-1: DATA SHEET REVISION HISTORY REVISION HISTORY Revision DS00002391A (03-06-17) DS00002391A-page 226 Section/Figure/Entry Correction Initial Document Release 2017 Microchip Technology Inc. KSZ8567S THE MICROCHIP WEB SITE Microchip provides online support via our WWW site at www.microchip.com. This web site is used as a means to make files and information easily available to customers. Accessible by using your favorite Internet browser, the web site contains the following information: * Product Support - Data sheets and errata, application notes and sample programs, design resources, user's guides and hardware support documents, latest software releases and archived software * General Technical Support - Frequently Asked Questions (FAQ), technical support requests, online discussion groups, Microchip consultant program member listing * Business of Microchip - Product selector and ordering guides, latest Microchip press releases, listing of seminars and events, listings of Microchip sales offices, distributors and factory representatives CUSTOMER CHANGE NOTIFICATION SERVICE Microchip's customer notification service helps keep customers current on Microchip products. Subscribers will receive e-mail notification whenever there are changes, updates, revisions or errata related to a specified product family or development tool of interest. To register, access the Microchip web site at www.microchip.com. Under "Support", click on "Customer Change Notification" and follow the registration instructions. CUSTOMER SUPPORT Users of Microchip products can receive assistance through several channels: * * * * Distributor or Representative Local Sales Office Field Application Engineer (FAE) Technical Support Customers should contact their distributor, representative or field application engineer (FAE) for support. Local sales offices are also available to help customers. A listing of sales offices and locations is included in the back of this document. Technical support is available through the web site at: http://www.microchip.com/support 2017 Microchip Technology Inc. DS00002391A-page 227 KSZ8567S PRODUCT IDENTIFICATION SYSTEM To order or obtain information, e.g., on pricing or delivery, refer to the factory or the listed sales office. PART NO. XX Device Package X [XX]( 1) - Temp. Tape & Reel Range Option [XXX] Automotive Code Examples: a) b) Device: KSZ8567S = 6-Port Switch with 1 SGMII, 1 RGMII/MII/RMII Package: TX = 128-pin TQFP-EP Temperature Range: V = -40C to +105C Tape and Reel Option: Blank -TR = Standard packaging (tray) = Tape and Reel (Note 1) Automotive Code: Vxx = 3 character code with "V" prefix, specifying automotive product DS00002391A-page 228 KSZ8567STXV-VA0 128-pin TQFP-EP package, Extended automotive temperature, Standard packaging KSZ8567STXV-TR-VA0 128-pin TQFP-EP package, Extended automotive temperature, Tape and reel (Extended) Note 1: Tape and Reel identifier only appears in the catalog part number description. This identifier is used for ordering purposes and is not printed on the device package. Check with your Microchip Sales Office for package availability with the Tape and Reel option. 2017 Microchip Technology Inc. KSZ8567S Note the following details of the code protection feature on Microchip devices: * Microchip products meet the specification contained in their particular Microchip Data Sheet. * Microchip believes that its family of products is one of the most secure families of its kind on the market today, when used in the intended manner and under normal conditions. * There are dishonest and possibly illegal methods used to breach the code protection feature. All of these methods, to our knowledge, require using the Microchip products in a manner outside the operating specifications contained in Microchip's Data Sheets. Most likely, the person doing so is engaged in theft of intellectual property. * Microchip is willing to work with the customer who is concerned about the integrity of their code. * Neither Microchip nor any other semiconductor manufacturer can guarantee the security of their code. Code protection does not mean that we are guaranteeing the product as "unbreakable." Code protection is constantly evolving. We at Microchip are committed to continuously improving the code protection features of our products. Attempts to break Microchip's code protection feature may be a violation of the Digital Millennium Copyright Act. If such acts allow unauthorized access to your software or other copyrighted work, you may have a right to sue for relief under that Act. Information contained in this publication regarding device applications and the like is provided only for your convenience and may be superseded by updates. It is your responsibility to ensure that your application meets with your specifications. MICROCHIP MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WHETHER EXPRESS OR IMPLIED, WRITTEN OR ORAL, STATUTORY OR OTHERWISE, RELATED TO THE INFORMATION, INCLUDING BUT NOT LIMITED TO ITS CONDITION, QUALITY, PERFORMANCE, MERCHANTABILITY OR FITNESS FOR PURPOSE. Microchip disclaims all liability arising from this information and its use. Use of Microchip devices in life support and/or safety applications is entirely at the buyer's risk, and the buyer agrees to defend, indemnify and hold harmless Microchip from any and all damages, claims, suits, or expenses resulting from such use. No licenses are conveyed, implicitly or otherwise, under any Microchip intellectual property rights unless otherwise stated. Trademarks The Microchip name and logo, the Microchip logo, AnyRate, AVR, AVR logo, AVR Freaks, BeaconThings, BitCloud, CryptoMemory, CryptoRF, dsPIC, FlashFlex, flexPWR, Heldo, JukeBlox, KEELOQ, KEELOQ logo, Kleer, LANCheck, LINK MD, maXStylus, maXTouch, MediaLB, megaAVR, MOST, MOST logo, MPLAB, OptoLyzer, PIC, picoPower, PICSTART, PIC32 logo, Prochip Designer, QTouch, RightTouch, SAM-BA, SpyNIC, SST, SST Logo, SuperFlash, tinyAVR, UNI/O, and XMEGA are registered trademarks of Microchip Technology Incorporated in the U.S.A. and other countries. ClockWorks, The Embedded Control Solutions Company, EtherSynch, Hyper Speed Control, HyperLight Load, IntelliMOS, mTouch, Precision Edge, and Quiet-Wire are registered trademarks of Microchip Technology Incorporated in the U.S.A. Adjacent Key Suppression, AKS, Analog-for-the-Digital Age, Any Capacitor, AnyIn, AnyOut, BodyCom, chipKIT, chipKIT logo, CodeGuard, CryptoAuthentication, CryptoCompanion, CryptoController, dsPICDEM, dsPICDEM.net, Dynamic Average Matching, DAM, ECAN, EtherGREEN, In-Circuit Serial Programming, ICSP, Inter-Chip Connectivity, JitterBlocker, KleerNet, KleerNet logo, Mindi, MiWi, motorBench, MPASM, MPF, MPLAB Certified logo, MPLIB, MPLINK, MultiTRAK, NetDetach, Omniscient Code Generation, PICDEM, PICDEM.net, PICkit, PICtail, PureSilicon, QMatrix, RightTouch logo, REAL ICE, Ripple Blocker, SAM-ICE, Serial Quad I/O, SMART-I.S., SQI, SuperSwitcher, SuperSwitcher II, Total Endurance, TSHARC, USBCheck, VariSense, ViewSpan, WiperLock, Wireless DNA, and ZENA are trademarks of Microchip Technology Incorporated in the U.S.A. and other countries. SQTP is a service mark of Microchip Technology Incorporated in the U.S.A. Silicon Storage Technology is a registered trademark of Microchip Technology Inc. in other countries. GestIC is a registered trademark of Microchip Technology Germany II GmbH & Co. KG, a subsidiary of Microchip Technology Inc., in other countries. All other trademarks mentioned herein are property of their respective companies. (c) 2017, Microchip Technology Incorporated, All Rights Reserved. ISBN: 9781522414513 QUALITY MANAGEMENT SYSTEM CERTIFIED BY DNV == ISO/TS 16949 == 2017 Microchip Technology Inc. Microchip received ISO/TS-16949:2009 certification for its worldwide headquarters, design and wafer fabrication facilities in Chandler and Tempe, Arizona; Gresham, Oregon and design centers in California and India. The Company's quality system processes and procedures are for its PIC(R) MCUs and dsPIC(R) DSCs, KEELOQ(R) code hopping devices, Serial EEPROMs, microperipherals, nonvolatile memory and analog products. In addition, Microchip's quality system for the design and manufacture of development systems is ISO 9001:2000 certified. DS00002391A-page 229 Worldwide Sales and Service AMERICAS ASIA/PACIFIC ASIA/PACIFIC EUROPE Corporate Office 2355 West Chandler Blvd. Chandler, AZ 85224-6199 Tel: 480-792-7200 Fax: 480-792-7277 Technical Support: http://www.microchip.com/ support Web Address: www.microchip.com Asia Pacific Office Suites 3707-14, 37th Floor Tower 6, The Gateway Harbour City, Kowloon China - Xiamen Tel: 86-592-2388138 Fax: 86-592-2388130 Austria - Wels Tel: 43-7242-2244-39 Fax: 43-7242-2244-393 China - Zhuhai Tel: 86-756-3210040 Fax: 86-756-3210049 Denmark - Copenhagen Tel: 45-4450-2828 Fax: 45-4485-2829 India - Bangalore Tel: 91-80-3090-4444 Fax: 91-80-3090-4123 Finland - Espoo Tel: 358-9-4520-820 Atlanta Duluth, GA Tel: 678-957-9614 Fax: 678-957-1455 Hong Kong Tel: 852-2943-5100 Fax: 852-2401-3431 Australia - Sydney Tel: 61-2-9868-6733 Fax: 61-2-9868-6755 China - Beijing Tel: 86-10-8569-7000 Fax: 86-10-8528-2104 Austin, TX Tel: 512-257-3370 China - Chengdu Tel: 86-28-8665-5511 Fax: 86-28-8665-7889 Boston Westborough, MA Tel: 774-760-0087 Fax: 774-760-0088 China - Chongqing Tel: 86-23-8980-9588 Fax: 86-23-8980-9500 Chicago Itasca, IL Tel: 630-285-0071 Fax: 630-285-0075 Dallas Addison, TX Tel: 972-818-7423 Fax: 972-818-2924 Detroit Novi, MI Tel: 248-848-4000 Houston, TX Tel: 281-894-5983 Indianapolis Noblesville, IN Tel: 317-773-8323 Fax: 317-773-5453 Tel: 317-536-2380 Los Angeles Mission Viejo, CA Tel: 949-462-9523 Fax: 949-462-9608 Tel: 951-273-7800 Raleigh, NC Tel: 919-844-7510 New York, NY Tel: 631-435-6000 San Jose, CA Tel: 408-735-9110 Tel: 408-436-4270 Canada - Toronto Tel: 905-695-1980 Fax: 905-695-2078 DS00002391A-page 230 China - Dongguan Tel: 86-769-8702-9880 China - Guangzhou Tel: 86-20-8755-8029 China - Hangzhou Tel: 86-571-8792-8115 Fax: 86-571-8792-8116 China - Hong Kong SAR Tel: 852-2943-5100 Fax: 852-2401-3431 China - Nanjing Tel: 86-25-8473-2460 Fax: 86-25-8473-2470 China - Qingdao Tel: 86-532-8502-7355 Fax: 86-532-8502-7205 China - Shanghai Tel: 86-21-3326-8000 Fax: 86-21-3326-8021 China - Shenyang Tel: 86-24-2334-2829 Fax: 86-24-2334-2393 China - Shenzhen Tel: 86-755-8864-2200 Fax: 86-755-8203-1760 India - New Delhi Tel: 91-11-4160-8631 Fax: 91-11-4160-8632 India - Pune Tel: 91-20-3019-1500 Japan - Osaka Tel: 81-6-6152-7160 Fax: 81-6-6152-9310 Japan - Tokyo Tel: 81-3-6880- 3770 Fax: 81-3-6880-3771 Korea - Daegu Tel: 82-53-744-4301 Fax: 82-53-744-4302 Korea - Seoul Tel: 82-2-554-7200 Fax: 82-2-558-5932 or 82-2-558-5934 Malaysia - Kuala Lumpur Tel: 60-3-6201-9857 Fax: 60-3-6201-9859 Malaysia - Penang Tel: 60-4-227-8870 Fax: 60-4-227-4068 France - Paris Tel: 33-1-69-53-63-20 Fax: 33-1-69-30-90-79 France - Saint Cloud Tel: 33-1-30-60-70-00 Germany - Garching Tel: 49-8931-9700 Germany - Haan Tel: 49-2129-3766400 Germany - Heilbronn Tel: 49-7131-67-3636 Germany - Karlsruhe Tel: 49-721-625370 Germany - Munich Tel: 49-89-627-144-0 Fax: 49-89-627-144-44 Germany - Rosenheim Tel: 49-8031-354-560 Israel - Ra'anana Tel: 972-9-744-7705 Italy - Milan Tel: 39-0331-742611 Fax: 39-0331-466781 Italy - Padova Tel: 39-049-7625286 Philippines - Manila Tel: 63-2-634-9065 Fax: 63-2-634-9069 Netherlands - Drunen Tel: 31-416-690399 Fax: 31-416-690340 Singapore Tel: 65-6334-8870 Fax: 65-6334-8850 Norway - Trondheim Tel: 47-7289-7561 Taiwan - Hsin Chu Tel: 886-3-5778-366 Fax: 886-3-5770-955 Taiwan - Kaohsiung Tel: 886-7-213-7830 China - Wuhan Tel: 86-27-5980-5300 Fax: 86-27-5980-5118 Taiwan - Taipei Tel: 886-2-2508-8600 Fax: 886-2-2508-0102 China - Xian Tel: 86-29-8833-7252 Fax: 86-29-8833-7256 Thailand - Bangkok Tel: 66-2-694-1351 Fax: 66-2-694-1350 Poland - Warsaw Tel: 48-22-3325737 Romania - Bucharest Tel: 40-21-407-87-50 Spain - Madrid Tel: 34-91-708-08-90 Fax: 34-91-708-08-91 Sweden - Gothenberg Tel: 46-31-704-60-40 Sweden - Stockholm Tel: 46-8-5090-4654 UK - Wokingham Tel: 44-118-921-5800 Fax: 44-118-921-5820 2017 Microchip Technology Inc. 11/07/16