Visible to Intel only — GUID: dne1481129295794
Ixiasoft
1. Intel® Agilex™ Hard Processor System Technical Reference Manual Revision History
2. Introduction to the Hard Processor System
3. Cortex-A53 MPCore Processor
4. Cache Coherency Unit
5. System Memory Management Unit
6. System Interconnect
7. Bridges
8. DMA Controller
9. On-Chip RAM
10. Error Checking and Correction Controller
11. Clock Manager
12. System Manager
13. Reset Manager
14. Hard Processor System I/O Pin Multiplexing
15. NAND Flash Controller
16. SD/MMC Controller
17. Ethernet Media Access Controller
18. USB 2.0 OTG Controller
19. SPI Controller
20. I2C Controller
21. UART Controller
22. General-Purpose I/O Interface
23. Timers
24. Watchdog Timers
25. CoreSight Debug and Trace
A. Booting and Configuration
B. Accessing the Secure Device Manager Quad SPI Flash Controller through HPS
2.2.1. HPS Block Diagram
2.2.2. Cortex-A53 MPCore Processor
2.2.3. Cache Coherency Unit
2.2.4. System Memory Management Unit
2.2.5. HPS Interfaces
2.2.6. System Interconnect
2.2.7. On-Chip RAM
2.2.8. Flash Memory Controllers
2.2.9. System Modules
2.2.10. Interface Peripherals
2.2.11. CoreSight* Debug and Trace
2.2.12. Hard Processor System I/O Pin Multiplexing
3.5.1. Exception Levels
3.5.2. Virtualization
3.5.3. Memory Management Unit
3.5.4. Level 1 Caches
3.5.5. Level 2 Memory System
3.5.6. Snoop Control Unit
3.5.7. Cryptographic Extensions
3.5.8. NEON Multimedia Processing Engine
3.5.9. Floating Point Unit
3.5.10. ACE Bus Interface
3.5.11. Abort Handling
3.5.12. Cache Protection
3.5.13. Generic Interrupt Controller
3.5.14. Generic Timers
3.5.15. Debug Modules
3.5.16. Cache Coherency Unit
3.5.17. Clock Sources
5.4.1. Translation Stages
5.4.2. Exception Levels
5.4.3. Translation Regimes
5.4.4. Translation Buffer Unit
5.4.5. Translation Control Unit
5.4.6. Security State Determination
5.4.7. Stream ID
5.4.8. Quality of Service Arbitration
5.4.9. System Memory Management Unit Interrupts
5.4.10. System Memory Management Unit Reset
5.4.11. System Memory Management Unit Clocks
15.1. NAND Flash Controller Features
15.2. NAND Flash Controller Block Diagram and System Integration
15.3. NAND Flash Controller Signal Descriptions
15.4. Functional Description of the NAND Flash Controller
15.5. NAND Flash Controller Programming Model
15.6. NAND Flash Controller Address Map and Register Definitions
15.5.1.1. NAND Flash Controller Optimization Sequence
15.5.1.2. Device Initialization Sequence
15.5.1.3. Device Operation Control
15.5.1.4. ECC Enabling
15.5.1.5. NAND Flash Controller Performance Registers
15.5.1.6. Interrupt and DMA Enabling
15.5.1.7. Timing Registers
15.5.1.8. Registers to Ignore
16.4.2.5.1. Internal DMA Controller Descriptors
16.4.2.5.2. Internal DMA Controller Descriptor Address
16.4.2.5.3. Internal DMA Controller Descriptor Fields
16.4.2.5.4. Host Bus Burst Access
16.4.2.5.5. Host Data Buffer Alignment
16.4.2.5.6. Buffer Size Calculations
16.4.2.5.7. Internal DMA Controller Interrupts
16.4.2.5.8. Internal DMA Controller Functional State Machine†
16.4.3.1.1. Load Command Parameters
16.4.3.1.2. Send Command and Receive Response
16.4.3.1.3. Send Response to BIU
16.4.3.1.4. Driving P-bit to the CMD Pin
16.4.3.1.5. Polling the CCS
16.4.3.1.6. CCS Detection and Interrupt to Host Processor
16.4.3.1.7. CCS Timeout
16.4.3.1.8. Send CCSD Command
16.4.3.1.9. I/O transmission delay (NACIO Timeout)
16.5.1. Software and Hardware Restrictions†
16.5.2. Initialization
16.5.3. Controller/DMA/FIFO Buffer Reset Usage
16.5.4. Non-Data Transfer Commands
16.5.5. Data Transfer Commands
16.5.6. Transfer Stop and Abort Commands
16.5.7. Internal DMA Controller Operations
16.5.8. Commands for SDIO Card Devices
16.5.9. CE-ATA Data Transfer Commands
16.5.10. Card Read Threshold
16.5.11. Interrupt and Error Handling
16.5.12. Booting Operation for eMMC and MMC
16.5.12.1. Boot Operation by Holding Down the CMD Line
16.5.12.2. Boot Operation for eMMC Card Device
16.5.12.3. Boot Operation for Removable MMC4.3, MMC4.4 and MMC4.41 Cards
16.5.12.4. Alternative Boot Operation
16.5.12.5. Alternative Boot Operation for eMMC Card Devices
16.5.12.6. Alternative Boot Operation for MMC4.3 Cards
17.1. Features of the Ethernet MAC
17.2. EMAC Block Diagram and System Integration
17.3. Distributed Virtual Memory Support
17.4. EMAC Controller Signal Description
17.5. EMAC Internal Interfaces
17.6. Functional Description of the EMAC
17.7. Ethernet MAC Programming Model
17.8. Ethernet MAC Address Map and Register Definitions
17.6.1. Transmit and Receive Data FIFO Buffers
17.6.2. DMA Controller
17.6.3. Descriptor Overview
17.6.4. IEEE 1588-2002 Timestamps
17.6.5. IEEE 1588-2008 Advanced Timestamps
17.6.6. IEEE 802.3az Energy Efficient Ethernet
17.6.7. Checksum Offload
17.6.8. Frame Filtering
17.6.9. Clocks and Resets
17.6.10. Interrupts
17.6.8.1.1. Unicast Destination Address Filter
17.6.8.1.2. Multicast Destination Address Filter
17.6.8.1.3. Hash or Perfect Address Filter
17.6.8.1.4. Broadcast Address Filter
17.6.8.1.5. Unicast Source Address Filter
17.6.8.1.6. Inverse Filtering Operation (Invert the Filter Match Result at Final Output)
17.6.8.1.7. Destination and Source Address Filtering Summary
17.7.1. System Level EMAC Configuration Registers
17.7.2. EMAC FPGA Interface Initialization
17.7.3. EMAC HPS Interface Initialization
17.7.4. DMA Initialization
17.7.5. EMAC Initialization and Configuration
17.7.6. Performing Normal Receive and Transmit Operation
17.7.7. Stopping and Starting Transmission
17.7.8. Programming Guidelines for Energy Efficient Ethernet
17.7.9. Programming Guidelines for Flexible Pulse-Per-Second (PPS) Output
18.1. Features of the USB OTG Controller
18.2. Block Diagram and System Integration
18.3. Distributed Virtual Memory Support
18.4. USB 2.0 ULPI PHY Signal Description
18.5. Functional Description of the USB OTG Controller
18.6. USB OTG Controller Programming Model
18.7. USB 2.0 OTG Controller Address Map and Register Definitions
24.4.1. Setting the Timeout Period Values
24.4.2. Selecting the Output Response Mode
24.4.3. Enabling and Initially Starting a Watchdog Timers
24.4.4. Reloading a Watchdog Counter
24.4.5. Pausing a Watchdog Timers
24.4.6. Disabling and Stopping a Watchdog Timers
24.4.7. Watchdog Timers State Machine
25.1. Features of CoreSight Debug and Trace
25.2. ARM® CoreSight Documentation
25.3. CoreSight Debug and Trace Block Diagram
25.4. Functional Description of CoreSight Debug and Trace
25.5. CoreSight Debug and Trace Programming Model
25.6. CoreSight Debug and Trace Address Map and Register Definitions
25.4.1. Debug Access Port
25.4.2. CoreSight SoC-400 Timestamp Generator
25.4.3. System Trace Macrocell
25.4.4. Trace Funnel
25.4.5. CoreSight Trace Memory Controller
25.4.6. AMBA Trace Bus Replicator
25.4.7. Trace Port Interface Unit
25.4.8. NoC Trace Ports
25.4.9. Embedded Cross Trigger System
25.4.10. Embedded Trace Macrocell
25.4.11. HPS Debug APB Interface
25.4.12. FPGA Interface
25.4.13. Debug Clocks
25.4.14. Debug Resets
B.1. Features of the Quad SPI Flash Controller
B.2. Taking Ownership of Quad SPI Controller
B.3. Quad SPI Flash Controller Block Diagram and System Integration
B.4. Quad SPI Flash Controller Signal Description
B.5. Functional Description of the Quad SPI Flash Controller
B.6. Quad SPI Flash Controller Programming Model
B.7. Accessing the SDM Quad SPI Flash Controller Through HPS Address Map and Register Definitions
B.5.1. Overview
B.5.2. Data Slave Interface
B.5.3. SPI Legacy Mode
B.5.4. Register Slave Interface
B.5.5. Local Memory Buffer
B.5.6. Arbitration between Direct/Indirect Access Controller and STIG
B.5.7. Configuring the Flash Device
B.5.8. XIP Mode
B.5.9. Write Protection
B.5.10. Data Slave Sequential Access Detection
B.5.11. Clocks
B.5.12. Resets
B.5.13. Interrupts
Visible to Intel only — GUID: dne1481129295794
Ixiasoft
4.3. Cache Coherency Unit Transactions
The coherency interconnect in the CCU accepts both coherent and non-coherent transactions. These transactions are routed to the CMI.
The CCU handles transactions from the FPGA-to-HPS interface, TCU, and peripheral masters in the L3 interconnect as follows:
- Coherent read: The CMI sends the read to the coherency directory in the CCC to perform a lookup and issue a snoop to the Cortex® -A53 MPCore™ processor if required.
- If the access is a cache hit, data is routed from the cache.
- If the access is a cache miss, data is routed from the appropriate slave agent after cache operations have completed.
- Coherent write: The CMI sends the write to the coherency directory in the CCC to perform a lookup and issue a snoop.
- If the access is a cache hit, the cache is updated with the new data and the coherency directory continues to track the cache line.
- If the access is a cache miss, then the new data is written to the appropriate slave agent.
Note: You must configure the FPGA and I/O master TBUs to prevent coherent master transactions from accessing the HPS-to-SDM mailbox address range. Please refer to the System Memory Management Unit chapter for more details.
- Non-coherent transactions are handled differently depending on the master agent issuing the transaction.
- If the FPGA or TCU send a non-coherent access to the CCU, the CMI routes the access directly to the slave agent.
- If an HPS peripheral master issues a non-cacheable memory access to on-chip RAM or SDRAM, then the L3 interconnect routes the access to the CMI of the CCU. In turn, the CCU routes the access directly to the corresponding memory.
- If an HPS peripheral master issues a non-cacheable memory access to a peripheral slave agent, then the L3 interconnect routes the access directly to the slave, bypassing the CCU.
Some key points to remember about CCU transactions:
- You can program address ranges to be disabled, read-only, or write-only. During address decode, the CCU compares the transaction ARPROT or AWPROT with the access privilege programmed for an address range. A failed access check results in a decode error response for the transaction.
- Each address range can also be associated with hash functions that are used in the route lookup process.
- Master agents have no predefined priority. A master's L3 interconnect QoS level determines the associated coherency interconnect QoS priority for the L3 masters and slaves, as well as the SDRAM memory interface. The Cortex-A53 MPCore™ and FPGA-to-HPS interface priorities are configured in the System Manager and FPGA, respectively. You can configure the coherency interconnect QoS weights through the QoS Profile Data Register (p_0) registers.
- Fixed transactions are split into multiple single beat increments (INCRs).
- The CCU only accepts 16-, 32- or 64-byte WRAP transactions. All other cache line sizes generate a fatal error interrupt.
- Master and slave ports queue outstanding requests. The table below shows the maximum number of outstanding requests each agent supports.
Table 42. Maximum Outstanding Request Support Agent Outstanding Reads Outstanding Writes Cortex® -A53 MPCore™ processor 33 21 FPGA-to-HPS Interface 8 8 TCU 16 1 Peripheral masters 32 32 External SDRAM Memory 8 8 On-chip RAM 16 1 GIC 1 1 Peripheral slaves 33 21 SDRAM register group 32 32
- An unknown address or access privilege violation on the AR or AW channels causes a decode error. This error stalls the command channels until the decode error (DECERR) response can be issued on the R or B channel, respectively.