Stratix V Avalon-MM Interface for PCIe Solutions: User Guide

Download PDF
ID 683411
Date 5/21/2017
Public
Document Table of Contents
Document Table of Contents x
1. Datasheet 2. Getting Started with the Avalon-MM Design Example 3. Parameter Settings 4. 64- or 128-Bit Avalon-MM Interface to the Endpoint Application Layer 5. Registers 6. Interrupts for Endpoints 7. Error Handling A. PCI Express Protocol Stack 8. Transceiver PHY IP Reconfiguration 9. Throughput Optimization 10. Design Implementation 11. Additional Features 12. Debugging B. Frequently Asked Questions for PCI Express C. Lane Initialization and Reversal D. Document Revision History
1. Datasheet x
1.1. Stratix V Avalon-MM Interface for PCIe Datasheet 1.2. Features 1.3. Release Information 1.4. Device Family Support 1.5. Configurations 1.6. Design Examples 1.7. IP Core Verification 1.8. Resource Utilization 1.9. Recommended Speed Grades 1.10. Creating a Design for PCI Express
1.7. IP Core Verification x
1.7.1. Compatibility Testing Environment
2. Getting Started with the Avalon-MM Design Example x
2.1. Running Qsys 2.2. Generating the Example Design 2.3. Understanding Simulation Log File Generation 2.4. Running a Gate-Level Simulation 2.5. Simulating the Single DWord Design 2.6. Generating Synthesis Files 2.7. Creating a Quartus® Prime Project 2.8. Compiling the Design 2.9. Programming a Device 2.10. Understanding Channel Placement Guidelines
3. Parameter Settings x
3.1. System Settings 3.2. Base Address Register (BAR) Settings 3.3. Device Identification Registers 3.4. PCI Express and PCI Capabilities Parameters 3.5. Avalon Memory‑Mapped System Settings
3.4. PCI Express and PCI Capabilities Parameters x
3.4.1. Device Capabilities 3.4.2. Error Reporting 3.4.3. Link Capabilities 3.4.4. MSI and MSI-X Capabilities 3.4.5. Power Management
4. 64- or 128-Bit Avalon-MM Interface to the Endpoint Application Layer x
4.1. 32-Bit Non-Bursting Avalon-MM Control Register Access (CRA) Slave Signals 4.2. Bursting and Non-Bursting Avalon® -MM Module Signals 4.3. 64- or 128-Bit Bursting TX Avalon-MM Slave Signals 4.4. Clock Signals 4.5. Reset 4.6. Interrupts for Endpoints when Multiple MSI/MSI-X Support Is Enabled 4.7. Hard IP Status Signals 4.8. Physical Layer Interface Signals
4.7. Hard IP Status Signals x
4.7.1. Hard IP Reconfiguration Interface
4.8. Physical Layer Interface Signals x
4.8.1. Hard IP Status Extension 4.8.2. Serial Data Signals 4.8.3. PIPE Interface Signals 4.8.4. Test Signals
4.8.2. Serial Data Signals x
4.8.2.1. Physical Layout of Hard IP in Stratix V GX/GT/GS Devices 4.8.2.2. Channel Placement in Arria V GZ and Stratix V GX/GT/GS Devices
5. Registers x
5.1. Correspondence between Configuration Space Registers and the PCIe Specification 5.2. Type 0 Configuration Space Registers 5.3. Type 1 Configuration Space Registers 5.4. PCI Express Capability Structures 5.5. Intel-Defined VSEC Registers 5.6. CvP Registers 5.7. 64- or 128-Bit Avalon-MM Bridge Register Descriptions 5.8. Programming Model for Avalon-MM Root Port 5.9. Uncorrectable Internal Error Mask Register 5.10. Uncorrectable Internal Error Status Register 5.11. Correctable Internal Error Mask Register 5.12. Correctable Internal Error Status Register
5.7. 64- or 128-Bit Avalon-MM Bridge Register Descriptions x
5.7.1. Avalon-MM to PCI Express Interrupt Registers
5.7.1. Avalon-MM to PCI Express Interrupt Registers x
5.7.1.1. Avalon-MM to PCI Express Interrupt Status Registers 5.7.1.2. Avalon-MM to PCI Express Interrupt Enable Registers 5.7.1.3. PCI Express Mailbox Registers 5.7.1.4. Avalon-MM-to-PCI Express Address Translation Table 5.7.1.5. PCI Express to Avalon-MM Interrupt Status and Enable Registers for Endpoints 5.7.1.6. Avalon-MM Mailbox Registers 5.7.1.7. Control Register Access (CRA) Avalon-MM Slave Port
5.8. Programming Model for Avalon-MM Root Port x
5.8.1. Sending a Write TLP 5.8.2. Sending a Read TLP or Receiving a Non-Posted Completion TLP 5.8.3. Examples of Reading and Writing BAR0 Using the CRA Interface 5.8.4. PCI Express to Avalon-MM Interrupt Status and Enable Registers for Root Ports 5.8.5. Root Port TLP Data Registers
6. Interrupts for Endpoints x
6.1. Enabling MSI or Legacy Interrupts 6.2. Generation of Avalon-MM Interrupts 6.3. Interrupts for Endpoints Using the Avalon-MM Interface with Multiple MSI/MSI-X Support
7. Error Handling x
7.1. Physical Layer Errors 7.2. Data Link Layer Errors 7.3. Transaction Layer Errors 7.4. Error Reporting and Data Poisoning 7.5. Uncorrectable and Correctable Error Status Bits
A. PCI Express Protocol Stack x
A.1. Top-Level Interfaces A.2. Data Link Layer A.3. Physical Layer A.4. 32-Bit PCI Express Avalon-MM Bridge A.5. Completer Only Single Dword Endpoint
A.1. Top-Level Interfaces x
A.1.1. Avalon-MM Interface A.1.2. Clocks and Reset A.1.3. Transceiver Reconfiguration A.1.4. Interrupts A.1.5. PIPE
A.4. 32-Bit PCI Express Avalon-MM Bridge x
A.4.1. Avalon‑MM Bridge TLPs A.4.2. Avalon-MM-to-PCI Express Write Requests A.4.3. Avalon-MM-to-PCI Express Upstream Read Requests A.4.4. PCI Express-to-Avalon-MM Read Completions A.4.5. PCI Express-to-Avalon-MM Downstream Write Requests A.4.6. PCI Express-to-Avalon-MM Downstream Read Requests A.4.7. Avalon-MM-to-PCI Express Read Completions A.4.8. PCI Express-to-Avalon-MM Address Translation for 32-Bit Bridge A.4.9. Minimizing BAR Sizes and the PCIe Address Space A.4.10. Avalon® -MM-to-PCI Express Address Translation Algorithm for 32-Bit Addressing
A.5. Completer Only Single Dword Endpoint x
A.5.1. RX Block A.5.2. Avalon-MM RX Master Block A.5.3. TX Block A.5.4. Interrupt Handler Block
8. Transceiver PHY IP Reconfiguration x
8.1. Connecting the Transceiver Reconfiguration Controller IP Core 8.2. Transceiver Reconfiguration Controller Connectivity for Designs Using CvP
9. Throughput Optimization x
9.1. Throughput of Posted Writes 9.2. Throughput of Non-Posted Reads
10. Design Implementation x
10.1. Making Analog QSF Assignments Using the Assignment Editor 10.2. Making Pin Assignments to Assign I/O Standard to Serial Data Pins 10.3. Recommended Reset Sequence to Avoid Link Training Issues 10.4. SDC Timing Constraints
11. Additional Features x
11.1. Configuration over Protocol (CvP) 11.2. Autonomous Mode 11.3. ECRC
11.2. Autonomous Mode x
11.2.1. Enabling Autonomous Mode 11.2.2. Enabling CvP Initialization
11.3. ECRC x
11.3.1. ECRC on the RX Path 11.3.2. ECRC on the TX Path
12. Debugging x
12.1. Hardware Bring-Up Issues 12.2. Link Training 12.3. Use Third-Party PCIe Analyzer 12.4. BIOS Enumeration Issues
12.2. Link Training x
12.2.1. Debugging Link that Fails To Reach L0
D. Document Revision History x
D.1. Cyclone® V Avalon® Memory Mapped (Avalon-MM) Interface for PCIe* Solutions User Guide Revision History
1. Datasheet
2. Getting Started with the Avalon-MM Design Example
3. Parameter Settings
4. 64- or 128-Bit Avalon-MM Interface to the Endpoint Application Layer
5. Registers
6. Interrupts for Endpoints
7. Error Handling
A. PCI Express Protocol Stack
8. Transceiver PHY IP Reconfiguration
9. Throughput Optimization
10. Design Implementation
11. Additional Features
12. Debugging
B. Frequently Asked Questions for PCI Express
C. Lane Initialization and Reversal
D. Document Revision History

9. Throughput Optimization

The PCI Express Base Specification defines a flow control mechanism to ensure efficient transfer of TLPs.

Each transmitter, the write requester in this case, maintains a credit limit register and a credits consumed register. The credit limit register is the sum of all credits received by the receiver, the write completer in this case. The credit limit register is initialized during the flow control initialization phase of link initialization and then updated during operation by Flow Control (FC) Update DLLPs. The credits consumed register is the sum of all credits consumed by packets transmitted. Separate credit limit and credits consumed registers exist for each of the six types of Flow Control:

  • Posted Headers
  • Posted Data
  • Non-Posted Headers
  • Non-Posted Data
  • Completion Headers
  • Completion Data

Each receiver also maintains a credit allocated counter which is initialized to the total available space in the RX buffer (for the specific Flow Control class) and then incremented as packets are pulled out of the RX buffer by the Application Layer. The value of this register is sent as the FC Update DLLP value.

Figure 44. Flow Control Update Loop

The PCIe Hard IP maintains its own flow control logic, including a credit consumed register, and ensures that no TLP is sent that would use more credits than are available for that type of TLP. If you want optimum performance and granularity, you can maintain your own credit consumed register and flow control gating logic for each credit category (Header/Data, Posted/Non-posted/Completion). This allows you to halt the transmission of TLPs for a category that is out of credits, while still allowing TLP transmission for categories that have sufficient credits.

The following steps describe the Flow Control Update loop. The corresponding numbers in the figure show the general area to which they correspond.

  1. When the Application Layer has a packet to transmit, the number of credits required is calculated. If the required credits are less than or equal to the current value of available credits (credit limit - credits consumed so far), then the packet can be transmitted immediately. However, if the credit limit minus credits consumed is less than the required credits, then the packet must be held until the credit limit is increased to a sufficient value by an FC Update DLLP. This check is performed separately for the header and data credits; a single packet consumes only a single header credit.
  2. After the packet is selected for transmission the credits consumed register is incremented by the number of credits consumed by this packet. This increment happens for both the header and data credit consumed registers.
  3. The packet is received at the other end of the link and placed in the RX buffer.
  4. At some point the packet is read out of the RX buffer by the Application Layer. After the entire packet is read out of the RX buffer, the credit allocated register can be incremented by the number of credits the packet has used. There are separate credit allocated registers for the header and data credits.
  5. The value in the credit allocated register is used to create an FC Update DLLP.
  6. After an FC Update DLLP is created, it arbitrates for access to the PCI Express link. The FC Update DLLPs are typically scheduled with a low priority; consequently, a continuous stream of Application Layer TLPs or other DLLPs (such as ACKs) can delay the FC Update DLLP for a long time. To prevent starving the attached transmitter, FC Update DLLPs are raised to a high priority under the following three circumstances:
    1. When the last sent credit allocated counter minus the amount of received data is less than MAX_PAYLOAD and the current credit allocated counter is greater than the last sent credit counter. Essentially, this means the data sink knows the data source has less than a full MAX_PAYLOAD worth of credits, and therefore is starving.
    2. When an internal timer expires from the time the last FC Update DLLP was sent, which is configured to 30 µs to meet the PCI Express Base Specification for resending FC Update DLLPs.
    3. When the credit allocated counter minus the last sent credit allocated counter is greater than or equal to 25% of the total credits available in the RX buffer, then the FC Update DLLP request is raised to high priority.

      After arbitrating, the FC Update DLLP that won the arbitration to be the next item is transmitted. In the worst case, the FC Update DLLP may need to wait for a maximum sized TLP that is currently being transmitted to complete before it can be sent.

  7. The original write requester receives the FC Update DLLP. The credit limit value is updated. If packets are stalled waiting for credits, they can now be transmitted.
Note: You must keep track of the credits consumed by the Application Layer.

Section Content
Throughput of Posted Writes
Throughput of Non-Posted Reads

Level Two Title