Visible to Intel only — GUID: nik1410564961504
Ixiasoft
1. Datasheet
2. Getting Started with the Arria V Hard IP for PCI Express
3. Parameter Settings
4. Interfaces and Signal Descriptions
5. Registers
6. Interrupts
7. Error Handling
8. IP Core Architecture
9. Transaction Layer Protocol (TLP) Details
10. Throughput Optimization
11. Design Implementation
12. Additional Features
13. Hard IP Reconfiguration
14. Transceiver PHY IP Reconfiguration
15. Testbench and Design Example
16. Debugging
A. Transaction Layer Packet (TLP) Header Formats
B. Lane Initialization and Reversal
C. Document Revision History
1.1. Arria V Avalon-ST Interface for PCIe Datasheet
1.2. Features
1.3. Release Information
1.4. Device Family Support
1.5. Configurations
1.6. Example Designs
1.7. Debug Features
1.8. IP Core Verification
1.9. Performance and Resource Utilization
1.10. Recommended Speed Grades
1.11. Creating a Design for PCI Express
4.1. Arria V Hard IP for PCI Express with Avalon-ST Interface to the Application Layer
4.2. Clock Signals
4.3. Reset Signals
4.4. Hard IP Status
4.5. Error Signals
4.6. ECRC Forwarding
4.7. Interrupts for Endpoints
4.8. Interrupts for Root Ports
4.9. Completion Side Band Signals
4.10. Transaction Layer Configuration Space Signals
4.11. LMI Signals
4.12. Power Management Signals
4.13. Physical Layer Interface Signals
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. Uncorrectable Internal Error Mask Register
5.8. Uncorrectable Internal Error Status Register
5.9. Correctable Internal Error Mask Register
5.10. Correctable Internal Error Status Register
15.6.1. ebfm_barwr Procedure
15.6.2. ebfm_barwr_imm Procedure
15.6.3. ebfm_barrd_wait Procedure
15.6.4. ebfm_barrd_nowt Procedure
15.6.5. ebfm_cfgwr_imm_wait Procedure
15.6.6. ebfm_cfgwr_imm_nowt Procedure
15.6.7. ebfm_cfgrd_wait Procedure
15.6.8. ebfm_cfgrd_nowt Procedure
15.6.9. BFM Configuration Procedures
15.6.10. BFM Shared Memory Access Procedures
15.6.11. BFM Log and Message Procedures
15.6.12. Verilog HDL Formatting Functions
15.7.1. Changing Between Serial and PIPE Simulation
15.7.2. Using the PIPE Interface for Gen1 and Gen2 Variants
15.7.3. Viewing the Important PIPE Interface Signals
15.7.4. Disabling the Scrambler for Gen1 and Gen2 Simulations
15.7.5. Disabling 8B/10B Encoding and Decoding for Gen1 and Gen2 Simulations
15.7.6. Changing between the Hard and Soft Reset Controller
Visible to Intel only — GUID: nik1410564961504
Ixiasoft
8.2. Transaction Layer
The Transaction Layer is located between the Application Layer and the Data Link Layer. It generates and receives Transaction Layer Packets. The following illustrates the Transaction Layer. The Transaction Layer includes three sub-blocks: the TX datapath, Configuration Space, and RX datapath.
Tracing a transaction through the RX datapath includes the following steps:
- The Transaction Layer receives a TLP from the Data Link Layer.
- The Configuration Space determines whether the TLP is well formed and directs the packet based on traffic class (TC).
- TLPs are stored in a specific part of the RX buffer depending on the type of transaction (posted, non-posted, and completion).
- The TLP FIFO block stores the address of the buffered TLP.
- The receive reordering block reorders the queue of TLPs as needed, fetches the address of the highest priority TLP from the TLP FIFO block, and initiates the transfer of the TLP to the Application Layer.
- When ECRC generation and forwarding are enabled, the Transaction Layer forwards the ECRC DWORD to the Application Layer.
Tracing a transaction through the TX datapath involves the following steps:
- The Transaction Layer informs the Application Layer that sufficient flow control credits exist for a particular type of transaction using the TX credit signals. The Application Layer may choose to ignore this information.
- The Application Layer requests permission to transmit a TLP. The Application Layer must provide the transaction and must be prepared to provide the entire data payload in consecutive cycles.
- The Transaction Layer verifies that sufficient flow control credits exist and acknowledges or postpones the request. If there is insufficient space in the retry buffer, the Transaction Layer does not accept the TLP.
- The Transaction Layer forwards the TLP to the Data Link Layer.
Figure 64. Architecture of the Transaction Layer: Dedicated Receive Buffer