V-Series Avalon-MM DMA Interface for PCIe Solutions User Guide

RSS

UG-01154 | 2018.07.31

Datasheet

V-Series Avalon-MM DMA Interface for PCIe Datasheet

Intel ^® V-Series FPGAs include a configurable, hardened protocol stack for PCI Express^® that is compliant with PCI Express Base Specification 2.1 or 3.0.

The V-Series Avalon^® Memory-Mapped (Avalon-MM) DMA for PCI Express removes some of the complexities associated with the PCIe^® protocol. For example, the IP core handles TLP encoding and decoding. In addition, it includes Read DMA and Write DMA engines. If you have already architected your own DMA system with the Avalon-MM interface, you may want to continue to use it. However, you may want to take advantage of the simplicity of having the DMA engines already implemented. Intel recommends this variant for new users. Depending of the device you select, this variant is available in Qsys for 128- and 256-bit interfaces to the Application Layer. The Avalon-MM interface and DMA engines are implemented in FPGA soft logic.

Figure 1. V-Series PCIe Variant with Avalon-MM DMA InterfaceThe following figure shows the high-level modules and connecting interfaces for this variant.

Note: This variant was renamed in the Quartus^® II 14.0 release. The name in the Quartus II 13.1 release was Avalon-MM 256-bit Hard IP for PCI Express IP Core.

Table 1. PCI Express Data Throughput.
The following table shows the aggregate bandwidth of a PCI Express link for Gen1, Gen2, and Gen3 for 2, 4, and 8 lanes. The protocol specifies 2.5 giga-transfers per second for Gen1, 5.0 giga-transfers per second for Gen2, and 8.0 giga‑transfers per second for Gen3. This table provides bandwidths for a single transmit (TX) or receive (RX) channel. The numbers double for duplex operation. Gen1 and Gen2 use 8B/10B encoding which introduces a 20% overhead. In contrast, Gen3 uses 128b/130b encoding which introduces only a 1.5% overhead.

Units are Gigabits per second (Gbps).
	Link Width
	×2	×4	×8
PCI Express Gen1 (2.5 Gbps)	N/A	N/A	16 Gbps
PCI Express Gen2 (5.0 Gbps)	8 Gbps	16 Gbps	32 Gbps
PCI Express Gen3 (8.0 Gbps)	15.75 Gbps	31.51 Gbps	63Gbps

Features

New features in the Quartus^® Prime 16.1 software release:

Increased maximum DMA transfer size for the 128- and 256-bit interfaces to 1 megabyte (MB).

The V-Series Avalon-MM DMA for PCI Express supports the following features:

Complete protocol stack including the Transaction, Data Link, and Physical Layers implemented as hard IP.
Native support for Gen1 x8, Gen2 x4, Gen2 x8, Gen3 x2, Gen3 x4, Gen3 x8 for Endpoints. The variant downtrains when plugged into a lesser link width or changes to a different maximum link rate.
Dedicated 16 kilobyte (KB) receive buffer.
Optional hard reset controller for Gen2.
Support for 128- or 256-bit Avalon-MM interface to Application Layer with embedded DMA up to Gen3 ×8 data rate.
Support for 32- or 64-bit addressing for the Avalon-MM interface to the Application Layer.
Qsys design example demonstrating parameterization, design modules, and connectivity.
Extended credit allocation settings to better optimize the RX buffer space based on application type.
Optional end-to-end cyclic redundancy code (ECRC) generation and checking and advanced error reporting (AER) for high reliability applications.
Support for Configuration Space Bypass Mode (in Arria^® V GZ and Stratix^® V), allowing you to design a custom Configuration Space and support multiple functions.
Support for Gen3 PIPE simulation.
Easy to use:
- Flexible configuration.
- No license requirement.
- Design examples to get started.

Table 2. Feature Comparison for 128- and 256-Bit Avalon-MM with DMA Interface to the Application Layer
Feature	128-Bit Interface	256-Bit Interface
Gen1	x8	Not supported
Gen2	x4, x8	x8
Gen3	x4	x4, x8
Root Port	Not supported	Supported
Tags supported	16	16 or 256
Maximum descriptor size	1 MB	1 MB
Maximum payload size	128 or 256	128 or 256
Immediate write¹	Not supported	Supported

¹ The Immediate Write provides a fast mechanism to send a Write TLP upstream. The descriptor stores the 32-bit payload, replacing the Source Low Address field of the descriptor.

Comparison of Avalon-ST, Avalon-MM and Avalon-MM with DMA Interfaces for V-Series Devices

Table 3. Feature Comparison for all Hard IP for PCI Express IP Cores. The table compares the features of the three mainstream Hard IP for PCI Express IP Cores. Refer to the Stratix V Avalon-ST Interface with SR-IOV PCIe Solutions User Guide for the features of that variant.
Feature	Avalon-ST Interface	Avalon-MM Interface	Avalon-MM DMA
IP Core License	Free	Free	Free
Native Endpoint	Supported	Supported	Supported
Legacy Endpoint ²	Supported	Not Supported	Not Supported
Root port	Supported	Supported	Not Supported
Gen1	×1, ×2, ×4, ×8	×1, ×2, ×4, ×8	x8
Gen2	×1, ×2, ×4, ×8	×1, ×2, ×4, ×8	×4, ×8
Gen3	×1, ×2, ×4, ×8	×1, ×2, ×4	×4, ×8
64-bit Application Layer interface	Supported	Supported	Not supported
128-bit Application Layer interface	Supported	Supported	Supported
256‑bit Application Layer interface	Supported	Not Supported	Supported
Maximum payload size	128, 256, 512, 1024, 2048 bytes	128, 256 bytes	128, 256 bytes
Number of tags supported for non-posted requests	256	8 for 64-bit interface 16 for 128-bit interface	16 or 256
62.5 MHz clock	Supported	Supported	Not Supported
Automatically handle out-of-order completions (transparent to the Application Layer)	Not supported	Supported	Not Supported
Automatically handle requests that cross 4 KB address boundary (transparent to the Application Layer)	Not supported	Supported	Supported
Polarity Inversion of PIPE interface signals	Supported	Supported	Supported
ECRC forwarding on RX and TX	Supported	Not supported	Not supported
Number of MSI requests	1, 2, 4, 8, 16, or 32	1, 2, 4, 8, 16, or 32	1, 2, 4, 8, 16, or 32
MSI-X	Supported	Supported	Supported
Legacy interrupts	Supported	Supported	Supported
Expansion ROM	Supported	Not supported	Not supported

Table 4. TLP Support Comparison for all Hard IP for PCI Express IP Cores. The table compares the TLP types that the Hard IP for PCI Express IP Cores variants can transmit. Each entry indicates whether this TLP type is supported (for transmit) by Endpoints (EP), Root Ports (RP), or both (EP/RP). For the Avalon-MM DMA interface, a software application programs a descriptor controller to specify DMA transfers between host and IP memory. The Read DMA Avalon-MM Master port and Write DMA Avalon-MM Master port send read and write TLPs, respectively. The optional TX Slave module supports single, non-bursting Memory Write TLPs to send status updates to the host.
TLP (Transmit Support)	Avalon-ST Interface	Avalon-MM Interface	Avalon-MM DMA
Memory Read Request (`Mrd`)	EP/RP	EP/RP	EP/RP (Read DMA Avalon-MM Master)
Memory Read Lock Request (`MRdLk`)	EP/RP		Not supported
Memory Write Request (`MWr`)	EP/RP	EP/RP	EP/RP (Write DMA Avalon-MM Master) (TX Slave - optional)
I/O Read Request (`IORd`)	EP/RP	EP/RP	Not supported
I/O Write Request (`IOWr`)	EP/RP	EP/RP	Not supported
Config Type 0 Read Request (`CfgRd0`)	RP	RP	Not supported
Config Type 0 Write Request (`CfgWr0`)	RP	RP	Not supported
Config Type 1 Read Request (`CfgRd1`)	RP	RP	Not supported
Config Type 1 Write Request (`CfgWr1`)	RP	RP	Not supported
Message Request (`Msg`)	EP/RP	Not supported	Not supported
Message Request with Data (`MsgD`)	EP/RP	Not supported	Not supported
Completion (`Cpl`)	EP/RP	EP/RP	EP/RP (Read & Write DMA Avalon-MM Masters)
Completion with Data (`CplD`)	EP/RP	Not supported	EP/RP (Read & Write DMA Avalon-MM Masters)
Completion-Locked (`CplLk`)	EP/RP	Not supported	Not supported
Completion Lock with Data (`CplDLk`)	EP/RP	Not supported	Not supported
Fetch and Add AtomicOp Request (`FetchAdd`)	EP	Not supported	Not supported

The V-Series Avalon-MM DMA Interface for PCIe Solutions User Guide explains how to use this IP core and not the PCI Express protocol. Although there is inevitable overlap between these two purposes, use this document only in conjunction with an understanding of the PCI Express Base Specification.

² Not recommended for new designs.

Release Information

Table 5. Hard IP for PCI Express Release Information
Item	Description
Version	15.1
Release Date	May 2018
Ordering Codes	No ordering code is required
Product IDs	The Product ID and Vendor ID are not required because this IP core does not require a license.
Vendor ID

Intel verifies that the current version of the Intel^® Quartus^® Prime software compiles the previous version of each IP core, if this IP core was included in the previous release. Intel reports any exceptions to this verification in the Intel IP Release Notes or clarifies them in the Intel^® Quartus^® Prime IP Update tool. Intel does not verify compilation with IP core versions older than the previous release.

V-Series Device Family Support

Table 6. Device Family Support
Device Family	Support
Arria V, Arria V GZ, Cyclone V, Stratix V	Final. The IP core is verified with final timing models. The IP core meets all functional and timing requirements for the device family and can be used in production designs.
Other device families	Refer to the Related Information below for other device families:

Design Examples

Platform Designer example designs are available for the V-Series Avalon-MM DMA for PCI Express IP Core. You can download them from the <install_dir>/ip/altera/altera_pcie/altera_pcie_hip_256_avmm/example_design/<dev> directory.

Starting from the 18.0 release of the Intel^® Quartus^® Prime software, you can generate both Endpoint example designs and Root Port example designs.

Debug Features

Debug features allow observation and control of the Hard IP for faster debugging of system-level problems.

IP Core Verification

To ensure compliance with the PCI Express specification, Intel performs extensive verification. The simulation environment uses multiple testbenches that consist of industry‑standard bus functional models (BFMs) driving the PCI Express link interface. Intel performs the following tests in the simulation environment:

Directed and pseudorandom stimuli test the Application Layer interface, Configuration Space, and all types and sizes of TLPs
Error injection tests inject errors in the link, TLPs, and Data Link Layer Packets (DLLPs), and check for the proper responses
PCI-SIG^® Compliance Checklist tests that specifically test the items in the checklist
Random tests that test a wide range of traffic patterns

Intel provides the following two example designs that you can leverage to test your PCBs and complete compliance base board testing (CBB testing) at PCI-SIG.

Compatibility Testing Environment

Intel has performed significant hardware testing to ensure a reliable solution. In addition, Intel internally tests every release with motherboards and PCI Express switches from a variety of manufacturers. All PCI-SIG compliance tests are run with each IP core release.

Resource Utilization

Because the PCIe protocol stack is implemented in hardened logic, it uses no core device resources (no ALMs and no embedded memory).

The Avalon-MM with DMA V-Series variants include an Avalon-MM DMA bridge implemented in soft logic that operates as a front end to the hardened protocol stack. The following table shows the typical expected device resource utilization for selected configurations using the current version of the Quartus^® Prime software targeting an V-Series device. With the exception of M20K memory blocks, the numbers of ALMs and logic registers are rounded up to the nearest 50.

Table 7. Resource Utilization V-Series Avalon-MM DMA for PCI Express
Data Rate, Number of Lanes, and Interface Width	ALMs	M20K Memory Blocks	Logic Registers
Gen2 x4 128	4300	29	5800
Gen2 x8 128	12700	19	22300
Gen3 x8 256	18000	47	31450

Note: Soft calibration of the transceiver module requires additional logic. The amount of logic required depends upon the configuration.

V-Series Recommended Speed Grades

Altera recommends setting the Quartus^® Prime Analysis & Synthesis Settings Optimization Technique to Speed when the Application Layer clock frequency is 250 MHz. For information about optimizing synthesis, refer to Setting Up and Running Analysis and Synthesis in Quartus^® Prime Help. For more information about how to effect the Optimization Technique settings, refer to Area and Timing Optimization in volume 2 of the Quartus^® Prime Handbook.

Table 8. Arria V Recommended Speed Grades for All Link Widths, Link Widths, and Application Layer Clock Frequencies
Link Rate	Link Width	Interface Width	Application Clock Frequency (MHz)	Recommended Speed Grades
Gen1	×1	64 bits	62.5 ³,125	–4,–5,–6
	×2	64 bits	125	–4,–5,–6
	×4	64 bits	125	–4,–5,–6
	×8	128 bits	125	–4,–5,–6
Gen2	×1	64 bits	125	–4,–5
	×2	64 bits	125	–4,–5
	×4	128 bits	125	–4,–5

Table 9. Arria V GZ Recommended Speed Grades for All Widths, Link Widths, and Application Layer Clock Frequencies
Link Rate	Link Width	Interface Width	Application Clock Frequency (MHz)	Recommended Speed Grades
Gen1	x1	64 bits	62.5 ⁴,125	–1, –2, –3, –4
	x2	64 bits	125	–1, –2, –3, –4
	x4	64 bits	125	–1, –2, –3, –4
	x8	64 bits	250	–1, –2, –3 ⁵
	x8	128 Bits	125	–1, –2, –3, –4
Gen2	x1	64 bits	125	–1, –2, –3, –4
	x2	64 bits	125	–1, –2, –3, –4
	x4	64 bits	250	–1, –2, –3 ⁽⁵⁾
	x4	128 bits	125	–1, –2, –3, –4
	x8	128 bits	250	–1, –2, –3 ⁽⁵⁾
	x8	256 bits	125	–1, –2, –3, –4
Gen3	x1	64 bits	125	–1, –2, –3, –4
	x2	64 bits	250	–1, –2, –3, –4
	x2	128 bits	125	–1, –2, –3, –4
	x4	128 bits	250	–1, –2, –3⁽⁵⁾
	x4	256 bits	125	–1, –2, –3,–4
	x8	256 bits	250	–1, –2, –3⁽⁵⁾

Table 10. Cyclone V Recommended Speed Grades for All Link Widths, Link Widths, and Application Layer Clock Frequencies . The Gen2 data rate requires Cyclone V GT devices.
Link Rate	Link Width	Interface Width	Application Clock Frequency (MHz)	Recommended Speed Grades
Gen1	×1	64 bits	62.5 ⁶,125	–6, –7,–8
	×2	64 bits	125	–6, –7,–8
	×4	64 bits	125	–6, –7,–8
Gen2	×1	64 bits	125	–7
	×2	64 bits	125	–7
	×4	128 bits	125	–7

Table 11. Stratix V Recommended Speed Grades for All Widths, Link Widths, and Application Layer Clock Frequencies
Link Rate	Link Width	Interface Width	Application Clock Frequency (MHz)	Recommended Speed Grades
Gen1	x1	64 bits	62.5 ⁷,125	–1, –2, –3, –4
	x2	64 bits	125	–1, –2, –3, –4
	x4	64 bits	125	–1, –2, –3, –4
	x8	64 bits	250	–1, –2, –3 ⁵ ⁸
	x8	128 Bits	125	–1, –2, –3, –4
Gen2	x1	64 bits	125	–1, –2, –3, –4
	x2	64 bits	125	–1, –2, –3, –4
	x4	64 bits	250	–1, –2, –3 ⁵
	x4	128 bits	125	–1, –2, –3, –4
	x8	128 bits	250	–1, –2, –3 ⁵
	x8	256 bits	125	–1, –2, –3, –4
Gen3	x1	64 bits	125	–1, –2, –3, –4
	x2	64 bits	250	–1, –2, –3, –4
	x2	128 bits	125	–1, –2, –3, –4
	x4	128 bits	250	–1, –2, –3 ⁵
	x4	256 bits	125	–1, –2, –3,–4
	x8	256 bits	250	–1, –2, –3 ⁵

³ This is a power-saving mode of operation

⁴ This is a power-saving mode of operation

⁵ The -4 speed grade is also possible for this configuration; however, it requires significant effort by the end user to close timing.

⁶ This is a power-saving mode of operation

⁷ This is a power-saving mode of operation

⁸ The -4 speed grade is also possible for this configuration; however, it requires significant effort by the end user to close timing.

Creating a Design for PCI Express

Select the PCIe variant that best meets your design requirements.

Is your design an Endpoint or Root Port?
What Generation do you intend to implement?
What link width do you intend to implement?
What bandwidth does your application require?
Does your design require Configuration via Protocol (CvP)?

Note: The following steps only provide a high-level overview of the design generation and simulation process. For more details, refer to the Quick Start Guide chapter.

Select parameters for that variant.
For Intel^® Arria^® 10 devices, you can use the new Example Design tab of the component GUI to generate a design that you specify. Then, you can simulate this example and also download it to an Intel^® Arria^® 10 FPGA Development Kit. Refer to the Intel^® Arria^® 10/ Intel^® Cyclone^® 10 GX PCI Express^® IP Core Quick Start Guide for details.
For all devices, you can simulate using an Intel-provided example design. All static PCI Express example designs are available under <install_dir>/ip/altera/altera_pcie/altera_pcie_<dev>_ed/example_design/<dev> . Alternatively, create a simulation model and use your own custom or third-party BFM. The Platform Designer Generate menu generates simulation models. Intel supports ModelSim^® - Intel FPGA Edition for all IP. The PCIe cores support the Aldec RivieraPro^®, Cadence NCSim^®, Mentor Graphics ModelSim^®, and Synopsys VCS^® and VCS-MX^® simulators.
The Intel testbench and Root Port or Endpoint BFM provide a simple method to do basic testing of the Application Layer logic that interfaces to the variation. However, the testbench and Root Port BFM are not intended to be a substitute for a full verification environment. To thoroughly test your application, Intel suggests that you obtain commercially available PCI Express verification IP and tools, or do your own extensive hardware testing, or both.
Compile your design using the Quartus^® Prime software. If the versions of your design and the Quartus^® Prime software you are running do not match, regenerate your PCIe design.
Download your design to an Intel development board or your own PCB. Click on the All Development Kits link below for a list of Intel's development boards.
Test the hardware. You can use Intel's Signal Tap Logic Analyzer or a third-party protocol analyzer to observe behavior.
Substitute your Application Layer logic for the Application Layer logic in Intel's testbench. Then repeat Steps 3–6. In Intel's testbenches, the PCIe core is typically called the DUT (device under test). The Application Layer logic is typically called APPS.

Getting Started with the Avalon-MM DMA

You can download the Platform Designer design example, pcie_de_ep_dma_g3x8_integrated.qsys, from the <install_dir>/ ip/altera/altera_pcie/altera_pcie_hip_256_avmm/example_design/<dev> directory.

Note: This design example provides instructions for generating simulation and synthesis files, but does not not generate all the files necessary to download the design to hardware. Refer to AN 690: PCI Express Avalon-MM DMA Reference Design and AN 708: PCI Express DMA Reference Design Using External DDR3 Memory for Stratix V and Arria V GZ Devices This reference design for reference designs that include all files necessary to download your design to an FPGA Development Kit.

The design example includes the following components:

Avalon-MM DMA for PCI Express

This IP core includes highly efficient DMA Read and DMA Write modules. The DMA Read and Write modules effectively move large blocks of data between the PCI Express address domain and the Avalon-MM address domain using burst data transfers. Depending on the configuration you select, the DMA Read and DMA Write modules use either a 128- or 256-bit Avalon-MM datapath.

In addition to high performance data transfer, the DMA Read and DMA Write modules ensure that the requests on the PCI link adhere to the PCI Express Base Specification, 3.0. The DMA Read and Write engines also perform the following functions:

Divide the original request into multiple requests to avoid crossing 4KByte boundaries.
Divide the original request into multiple requests to ensure that the maximum payload size is equal to or smaller than the maximum payload size for write requests and maximum read request size for read requests.
Supports out-of-order completions when the original request is divided into multiple requests to adhere to the read request size. The Read Completions can come back in any order. The Read DMA Avalon-MM master port supports out-of-order Completions by writing the Read Completions to the correct locations. The Read DMA Avalon^® -MM master port does not have an internal reordering buffer.

Using the DMA Read and DMA Write modules, you can specify descriptor entry table entries with large payloads.

On-Chip Memory IP core

This IP core stores the DMA data. This memory has a 256-bit data width.

Descriptor Controller

The Descriptor Controller manages the Read DMA and Write DMA modules. Host software programs the Descriptor Controller internal registers with the location of the descriptor table. The Descriptor Controller instructs the Read DMA module to copy the entire table to its internal FIFO. It then pushes the table entries to DMA Read or DMA Write modules to transfer data. The Descriptor Controller also sends DMA status upstream via an Avalon-MM TX slave port.

In this example design the Descriptor Controller parameter, Instantiate internal descriptor controller, is on. Consequently, the Descriptor Controller is integrated into the Avalon-MM DMA bridge as shown in the figure below. Embedding the Descriptor Controller in the Avalon-MM DMA bridge simplifies the design. If you plan to replace the Descriptor Controller IP core with your own implementation, do not turn on the Instantiate internal descriptor controller in the parameter editor when parameterizing the IP core.

The Descriptor Controller supports the following features:

A single duplex channel.
Minimum transfer size of one dword (4 bytes).
Maximum transfer size of 1 M (1024 * 1024) - 4 bytes.
Note: Although the Descriptor Controller supports a maximum transfer size of (1 M (1024 * 1024) - 4 bytes), the on-chip memory in this design example is smaller. Consequently, this design example cannot handle the maximum transfer size.
Endpoints, only.
Provides status to host software by generating an MSI interrupt when the DMA transfer completes.

Transceiver Reconfiguration Controller IP Core

The Transceiver Reconfiguration Controller performs offset cancellation to compensate for variations due to process, voltage, and temperature (PVT).

The following provides a high‑level block diagram of the V-Series Avalon-MM DMA for PCI Express Design Example.

Intel PCIe Reconfig Driver IP Core

The PCIe Reconfig Driver drives the Transceiver Reconfiguration Controller. This driver is a plain text Verilog HDL file that you can modify if necessary to meet your system requirements.

Block Diagram of the Avalon-MM DMA for PCI Express Example Design

Design Example Limitations

This design example is intended to show basic DMA functionality. It is not a substitute for a robust verification testbench. If you modify this testbench, be sure to verify that the modifications result in the correct behavior.

Understanding the Avalon-MM DMA Ports

The Avalon^® -MM DMA bridge includes ports to implement the DMA functionality. The following figure and table below illustrate and describe these ports.

Table 12. Avalon^® -MM DMA Platform Designer System Port Descriptions
Function	Port	Description
TXS	`Txs`	This is an Avalon-MM slave port. In a typical application, an Avalon^® -MM master uses this port to send memory reads or writes to the PCIe domain. The Descriptor Controller uses it to write DMA status back to descriptor space in the PCIe domain when the DMA completes its operation. The Descriptor Controller also uses this port to send MSI interrupts upstream.
Read Data Mover	`dma_rd_master`	This is an Avalon^® -MM master port. The Read Data Mover moves data from the PCIe domain to the on-chip memory during normal read DMA operation. The Read Data Mover also fetches the descriptors from the PCIe domain and writes them to the FIFO in the Descriptor Controller. There are two separate descriptor tables for the read and write write DMA descriptors.The `dma_rd_master` connects to `wr_dts_slave` port to load the write DMA descriptor FIFO and `rd_dts_slave` port to load the read DMA descriptor FIFO.
Write Data Mover	`dma_wr_master`	This is an Avalon^® -MM master port. The Write Data Mover reads data from the on-chip memory and then writes data to the PCIe domain.
Descriptor Controller FIFOs	`wr_dts_slave` and `rd_dts_slave`	This is an Avalon^® -MM slave port for the Descriptor Controller FIFOs. When the Read Data Mover fetches the descriptors from system memory, it writes the descriptors to the FIFO using this port. Because there are separate descriptor tables for read and write, there are two ports. The address range for the write DMA FIFO is `0x100_0000`—`0x100_1FFF`. The address range for the read DMA FIFO is `0x100_2000`—`0x100_3FFF`.
Control in the Descriptor Controller	`wr_dcm_master` and `rd_dcm_master`	The control block in the Descriptor Controller has one transmit and one receive port, one for read DMA and another one for write DMA. The receive port connects to the `RXM_BAR0` and the transmit port connects to the `Txs`. The receive path from the `RXM_BAR0` connects internally. It is not shown in the connections panel. For the transmit path, both read and write DMA ports connect to the `Txs` externally as shown in the connections panel.
RXM_BAR0	not shown in connections panel	This is an Avalon^® -MM master port. It passes the memory access from PCIe host to PCIe BAR0. The host uses this port to program the Descriptor Controller. Because this Platform Designer system uses an internal descriptor controller, the port connection is not shown in Platform Designer. The connection is inside the `a10_pcie_hip_0` module.
RXM_BAR4	`Rxm_BAR4`	This is an Avalon^® -MM master port. It passes the memory access from PCIe host to PCIe BAR4. In the Platform Designer system, it connects to the on-chip memory. The PCIe host accesses the memory through PCIe BAR4 In a typical application, system software controls this port to initialize random data in the on-chip memory. Software also reads the data back to verify correct operation.

Generating the Testbench

Copy the example design, pcie_de_ep_dma_g3x8_integrated.qsys, from the installation directory: <install_dir>/ip/altera/altera_pcie/altera_pcie_hip_256_avmm/example_design/ to your working directory.
Start Platform Designer, by typing the following command:

qsys-edit
Open pcie_de_ep_dma_g3x8_integrated.qsys.

The following figure shows the Platform Designer system.

Figure 2. V-Series Avalon-MM DMA for PCI Express Platform Designer System Design
Click Generate > Generate Testbench System.

Specify the following parameters:

Table 13. Parameters to Specify in the Generation Dialog Box
Parameter	Value
Testbench System
Create testbench Platform Designer system	Standard, BFMs for standard Platform Designer interfaces
Create testbench simulation model	Verilog
Allow mixed-language simulation	You can leave this option off.
Output Directory
Path	`<working_dir>`//pcie_de_ep_dma_g3x8_integrated

Click Generate.

Understanding the Simulation Generated Files

Table 14. Platform Designer Generation Output Files
Directory	Description
`<testbench_dir>/<variant_name>/`testbench	Includes testbench subdirectories for the Aldec, Cadence, Mentor, and Synopsys simulation tools with the required libraries and simulation scripts.
`<testbench_dir>/<variant_name>/`testbench/`<cad_vendor>`	Includes the HDL source files and scripts for the simulation testbench.
`<testbench_dir>/<variant_name>/`testbench/`<variant_namer>`_tb	Includes HDL design files

Understanding Simulation Log File Generation

Starting with the Quartus II 14.0 software release, simulation automatically creates a log file, altpcie_monitor_<dev>_dlhip_tlp_file_log.log in your simulation directory.

Table 15. Sample Simulation Log File Entries
Time	TLP Type	Payload (Bytes)	TLP Header
17989 RX	CfgRd0	0004	04000001_0000000F_01080008
17989 RX	MRd	0000	00000000_00000000_01080000
18021 RX	CfgRd0	0004	04000001_0000010F_0108002C
18053 RX	CfgRd0	0004	04000001_0000030F_0108003C
18085 RX	MRd	0000	00000000_00000000_0108000C

Simulating the Example Design in ModelSim

In a terminal window, change directory to <workingdir>/pcie_de_ep_dma_g3x8_integrated/testbench/mentor/ .
Start the ModelSim^® simulator.
To run the simulation, type the following commands in a terminal window:
1. do msim_setup.tcl
2. ld_debug
  The ld_debug command compiles all design files and elaborates the top-level design without any optimization.
3. run -all

The simulation performs the following operations:

Various configuration accesses after the link is initialized
Setup of the DMA controller to read data from the BFM’s shared memory
Setup of the DMA controller to write the same data back to the BFM’s shared memory
Data comparison and report of any mismatch

Running a Gate-Level Simulation

The PCI Express testbenches run simulations at the register transfer level (RTL). However, it is possible to create your own gate-level simulations. Contact your Intel Sales Representative for instructions and an example that illustrates how to create a gate-level simulation from the RTL testbench.

Generating Synthesis Files

On the Generate menu, select Generate HDL.
For Create HDL design files for synthesis, select Verilog.
You can leave the default settings for all other items.
Click Generate to generate files for synthesis.
Click Finish when the generation completes.

Compiling the Design

On the Quartus^® Prime Processing menu, click Start Compilation.
After compilation, expand the TimeQuest Timing Analyzer folder in the Compilation Report. Note whether the timing constraints are achieved in the Compilation Report.

If your design does not initially meet the timing constraints, you can find the optimal Fitter settings for your design by using the Design Space Explorer. To use the Design Space Explorer, click Launch Design Space Explorer on the Tools menu.

Creating a Quartus Prime Project

You can create a new Quartus^® Prime project with the New Project Wizard, which helps you specify the working directory for the project, assign the project name, and designate the name of the top-level design entity.

On the Quartus^® Prime File menu, click then New Project Wizard, then Next.
Click Next in the New Project Wizard: Introduction (The introduction does not appear if you previously turned it off.)
On the Directory, Name, Top-Level Entity page, enter the following information:
1. For What is the working directory for this project, browse to <project_dir>/pcie_de_ep_dma_g3x8_integrated/.
2. For What is the name of this project? browse to the <project_dir>/pcie_de_ep_dma_g3x8_integrated/synthesis directory and select pcie_de_ep_dma_g3x8_integrated.v.
3. Click Next.
For Project Type select Empty project.
Click Next.
On the Add Files page, add <project_dir>/pcie_de_ep_dma_g3x8_integrated/synthesis/ep_g3x8_avmm256_integrated.qip to your Quartus^® Prime project.
Click Next to display the Family & Device Settings page.
On the Device page, choose the following target device family and options:
Note: Currently, you cannot target an Intel^® Cyclone^® 10 GX Development Kit when generating an example design for Intel^® Cyclone^® 10 GX.
1. In the Family list, select Stratix V (GS/GT/GX/E).
2. In the Devices list, select Stratix V GX PCIe.
3. In the Available devices list, select 5SGXEA7K2F40C2.
Click Next to close this page and display the EDA Tool Settings page.
From the Simulation list, select ModelSim. From the Format list, select the HDL language you intend to use for simulation.
Click Next to display the Summary page.
Check the Summary page to ensure that you have entered all the information correctly.
Click Finish.
Save your project.

Parameter Settings

System Settings

Table 16. System Settings for PCI Express

Parameter

Value

Description

Number of Lanes

x1, x2, ×4, ×8

Specifies the maximum number of lanes supported. Avalon-MM Interface with DMA does not support x1 configurations.

Lane Rate

Gen1 (2.5 Gbps)

Gen2 (2.5/5.0 Gbps)

Gen3 (2.5/5.0/8.0 Gbps)

Specifies the maximum data rate at which the link can operate.

Application interface width

128-bit

256-bit

Specifies the width of the interface to the Application Layer. The following table indicates the possible combinations.

Application Interface Width	Configuration	Application Layer Clock Frequency
128 bits	Gen1 x8	125 MHz
128 bits	Gen2 x4	125 MHz
128 bits	Gen2 x8	250 MHz
256 bits	Gen2 x8	125 MHz
128 bits	Gen3 x4	250 MHz
256 bits	Gen3 x4	125 MHz
256 bits	Gen3 x8	250 MHz

RX Buffer credit allocation -performance for received requests

Minimum

Low

Balanced

Determines the allocation of posted header credits, posted data credits, non-posted header credits, completion header credits, and completion data credits in the 16 KByte RX buffer. The settings allow you to adjust the credit allocation to optimize your system. The credit allocation for the selected setting displays in the message pane.

Refer to the Throughput Optimization chapter in the Stratix V Avalon-ST Interface for PCIe Solutions User Guide for more information about optimizing performance.

The Message window dynamically updates the number of credits for Posted, Non‑Posted Headers and Data, and Completion Headers and Data as you change this selection.

Minimum RX Buffer credit allocation -performance for received requests )—configures the minimum PCIe specification allowed for non-posted and posted request credits, leaving most of the RX Buffer space for received completion header and data. Select this option for variations where application logic generates many read requests and only infrequently receives single requests from the PCIe link.
Low—configures a slightly larger amount of RX Buffer space for non-posted and posted request credits, but still dedicates most of the space for received completion header and data. Select this option for variations for which application logic generates many read requests and infrequently receives small bursts of requests from the PCIe link. This option is recommended for typical endpoint applications in which most of the PCIe traffic is generated by a DMA engine that is located in the endpoint application layer logic.
Balanced—configures approximately half the RX Buffer space to received requests and the other half of the RX Buffer space to received completions. Select this option for applications in which the received requests and received completions are roughly equal.

Reference clock frequency

100 MHz

125 MHz

The PCI Express Base Specification 3.0 requires a 100 MHz ±300 ppm reference clock. The 125 MHz reference clock is provided as a convenience for systems that include a 125 MHz clock source. For more information about Gen3 operation, refer to 4.3.8 Refclk Specifications for 8.0 GT/s in the specification.

For Gen3 operation, Altera recommends using a common reference clock (0 ppm) because when using separate reference clocks (non 0 ppm), the PCS occasionally must insert SKP symbols, potentially causes the PCIe link to go to recovery. Gen1 or Gen2 modes are not affected by this issue. Systems using the common reference clock (0 ppm) are not affected by this issue. The primary repercussion of this is a slight decrease in bandwidth. On Gen3 x8 systems, this bandwidth impact is negligible. If non 0 ppm mode is required, so that separate reference clocks are being used, please contact Altera for further information and guidance.

Instantiate internal descriptor controller

On/Off

When you turn this option on, the descriptor controller is included in the Avalon-MM bridge. When you turn this option off, the descriptor controller should be included as a separate external component. Turn this option on, if you plan to use the Altera-provided descriptor controller in your design. Turn this option off if you plan to modify or replace the descriptor controller logic in your design.

Enable Avalon-MM CRA Slave hard IP status port

On/Off

Allows read and write access to bridge registers from the interconnect fabric using a specialized slave port. This option is required for Requester/Completer variants and optional for Completer Only variants. Enabling this option allows read and write access to bridge registers, except in the Completer-Only single dword variations.

Enable burst capabilities for RXM BAR2 port

On/Off

When you turn on this option, the BAR2 RX Avalon-MM masters is burst capable. If BAR2 is 32 bits and Burst capable, then BAR3 is not available for other use. If BAR2 is 64 bits, the BAR3 register holds the upper 32 bits of the address.

Enable configuration via the PCIe link

On/Off

On, the Quartus^® Prime software places the Endpoint in the location required for configuration via protocol (CvP). For more information about CvP, click the Configuration via Protocol (CvP) link below

Use ATX PLL

On/Off

When you turn on this option, the Hard IP for PCI Express uses the ATX PLL instead of the CMU PLL. For other configurations, using the ATX PLL instead of the CMU PLL reduces the number of transceiver channels that are necessary. This option requires the use of the soft reset controller and does not support the CvP flow.

Enable Hard IP reset pulse at power-up when using the soft reset controller

On/Off

When you turn on this option, the soft reset controller generates a pulse at power up to reset the Hard IP. This pulse ensures that the Hard IP is reset after programming the device, regardless of the behavior of the dedicated PCI Express reset pin, perstn. This option is available for Gen2 and Gen3 designs that use a soft reset controller.

Base Address Register (BAR) Settings

The type and size of BARs available depend on port type.

Table 17. BAR Registers
Parameter	Value	Description
Type	Disabled 64-bit prefetchable memory 32-bit non-prefetchable memory 32-bit prefetchable memory I/O address space	If you select 64-bit prefetchable memory, 2 contiguous BARs are combined to form a 64-bit prefetchable BAR; you must set the higher numbered BAR to Disabled. A non-prefetchable 64‑bit BAR is not supported because in a typical system, the Root Port Type 1 Configuration Space sets the maximum non‑prefetchable memory window to 32 bits. The BARs can also be configured as separate 32‑bit memories. Defining memory as prefetchable allows contiguous data to be fetched ahead. Prefetching memory is advantageous when the requestor may require more data from the same region than was originally requested. If you specify that a memory is prefetchable, it must have the following 2 attributes: Reads do not have side effects such as changing the value of the data read Write merging is allowed
Size	N/A	Platform Designer automatically calculates the required size after you connect your components.

Device Identification Registers

Table 18. Device ID Registers. The following table lists the default values of the read‑only Device ID registers. You can use the parameter editor to change the values of these registers. Refer to Type 0 Configuration Space Registers for the layout of the Device Identification registers.
Register Name	Range	Default Value	Description
Vendor ID	16 bits	0x00001172	Sets the read-only value of the `Vendor ID` register. This parameter cannot be set to 0xFFFF per the PCI Express Specification. Address offset: 0x000.
Device ID	16 bits	0x0000000	Sets the read-only value of the `Device ID` register. Address offset: 0x000.
Revision ID	8 bits	0x00000001	Sets the read-only value of the `Revision ID` register. Address offset: 0x008.
Class code	24 bits	0x00000000	Sets the read-only value of the `Class Code` register. Address offset: 0x008.
Subsystem Vendor ID	16 bits	0x00000000	Sets the read-only value of the `Subsystem Vendor ID` register in the PCI Type 0 Configuration Space. This parameter cannot be set to 0xFFFF per the PCI Express Base Specification. This value is assigned by PCI-SIG to the device manufacturer. Address offset: 0x02C.
Subsystem Device ID	16 bits	0x00000000	Sets the read-only value of the `Subsystem Device ID` register in the PCI Type 0 Configuration Space. Address offset: 0x02C

PCI Express and PCI Capabilities Parameters

This group of parameters defines various capability properties of the IP core. Some of these parameters are stored in the PCI Configuration Space - PCI Compatible Configuration Space. The byte offset indicates the parameter address.

Device Capabilities

Table 19. Capabilities Registers
Parameter	Possible Values	Default Value	Description
Maximum payload size	128 bytes 256 bytes	128 bytes	Specifies the maximum payload size supported. This parameter sets the read-only value of the max payload size supported field of the Device Capabilities register (0x084[2:0]). Address: 0x084.

Error Reporting

Table 20. Error Reporting
Parameter	Value	Default Value	Description
Advanced error reporting (AER)	On/Off	Off	When On, enables the Advanced Error Reporting (AER) capability.
ECRC checking	On/Off	Off	When On, enables ECRC checking. Sets the read-only value of the ECRC check capable bit in the `Advanced Error Capabilities and Control Register`. This parameter requires you to enable the AER capability.
ECRC generation	On/Off	Off	When On, enables ECRC generation capability. Sets the read-only value of the ECRC generation capable bit in the `Advanced Error Capabilities and Control Register`. This parameter requires you to enable the AER capability.
Enable ECRC forwarding on the Avalon-ST interface	On/Off	Off	When On, enables ECRC forwarding to the Application Layer. On the Avalon-ST RX path, the incoming TLP contains the ECRC dword⁽¹⁾and the TD bit is set if an ECRC exists. On the transmit the TLP from the Application Layer must contain the ECRC dword and have the TD bit set.
Track RX completion buffer overflow on the Avalon- ST interface	On/Off	Off	When On, the core includes the `rxfx_cplbuf_ovf` output status signal to track the RX posted completion buffer overflow status

Link Capabilities

Table 21. Link Capabilities
Parameter	Value	Description
Link port number	0x01	Sets the read-only value of the port number field in the `Link Capabilities` register.
Slot clock configuration	On/Off	When On, indicates that the Endpoint or Root Port uses the same physical reference clock that the system provides on the connector. When Off, the IP core uses an independent clock regardless of the presence of a reference clock on the connector.

MSI and MSI-X Capabilities

Table 22. MSI and MSI-X Capabilities
Parameter	Value	Description
MSI messages requested	1, 2, 4, 8, 16	Specifies the number of messages the Application Layer can request. Sets the value of the `Multiple Message Capable` field of the `Message Control` register. Address: 0x050[31:16].
MSI-X Capabilities
Implement MSI-X	On/Off	When On, adds the MSI-X functionality.
	Bit Range
Table size	[10:0]	System software reads this field to determine the MSI-X Table size <n>, which is encoded as <n–1>. For example, a returned value of 2047 indicates a table size of 2048. This field is read-only. Legal range is 0–2047 (2¹¹). Address offset: 0x068[26:16]
Table Offset	[31:0]	Points to the base of the MSI-X Table. The lower 3 bits of the table BAR indicator (BIR) are set to zero by software to form a 64-bit qword-aligned offset. This field is read-only.
Table BAR Indicator	[2:0]	Specifies which one of a function’s BARs, located beginning at 0x10 in Configuration Space is used to map the MSI-X table into memory space. This field is read-only. Legal range is 0–5.
Pending Bit Array (PBA) Offset	[31:0]	Used as an offset from the address contained in one of the function’s Base Address registers to point to the base of the MSI-X PBA. The lower 3 bits of the PBA BIR are set to zero by software to form a 32-bit qword-aligned offset. This field is read-only.
PBA BAR Indicator	[2:0]	Specifies the function Base Address registers, located beginning at 0x10 in Configuration Space, that maps the MSI-X PBA into memory space. This field is read-only. Legal range is 0–5.

Power Management

Table 23. Power Management Parameters
Parameter	Value	Description
Endpoint L0s acceptable latency	Maximum of 64 ns Maximum of 128 ns Maximum of 256 ns Maximum of 512 ns Maximum of 1 us Maximum of 2 us Maximum of 4 us No limit	This design parameter specifies the maximum acceptable latency that the device can tolerate to exit the L0s state for any links between the device and the root complex. It sets the read-only value of the Endpoint L0s acceptable latency field of the `Device Capabilities Register` (0x084). This Endpoint does not support the L0s or L1 states. However, in a switched system there may be links connected to switches that have L0s and L1 enabled. This parameter is set to allow system configuration software to read the acceptable latencies for all devices in the system and the exit latencies for each link to determine which links can enable Active State Power Management (ASPM). This setting is disabled for Root Ports. The default value of this parameter is 64 ns. This is a safe setting for most designs.
Endpoint L1 acceptable latency	Maximum of 1 us Maximum of 2 us Maximum of 4 us Maximum of 8 us Maximum of 16 us Maximum of 32 us Maximum of 64 ns No limit	This value indicates the acceptable latency that an Endpoint can withstand in the transition from the L1 to L0 state. It is an indirect measure of the Endpoint’s internal buffering. It sets the read-only value of the Endpoint L1 acceptable latency field of the `Device Capabilities Register`. This Endpoint does not support the L0s or L1 states. However, a switched system may include links connected to switches that have L0s and L1 enabled. This parameter is set to allow system configuration software to read the acceptable latencies for all devices in the system and the exit latencies for each link to determine which links can enable Active State Power Management (ASPM). This setting is disabled for Root Ports. The default value of this parameter is 1 µs. This is a safe setting for most designs.

PCIe Address Space Settings

Table 24. PCIe Address Space Settings
Parameter	Value	Default Value	Description
Address width of accessible PCIe Memory space	20–64	32	Specifies the width of the TX Slave Module Avalon-MM address. This address is used unchanged as the PCIe address.

IP Core Interfaces

This chapter describes the top-level signals of the V-Series Hard IP for PCI Express^® using the Avalon^® -MM interface with DMA. The Avalon^® -MM DMA bridge includes high-performance, burst-capable Read DMA and Write DMA modules. The DMA Descriptor Controller that controls the Read DMA and Write DMA modules can be included in the Avalon^® -MM DMA bridge or separately instantiated. It uses 64-bit addressing, making address translation unnecessary. A separately instantiated Descriptor Controller manages the Read DMA and Write DMA modules. This variant is available for the following configurations:

Gen1 x8
Gen2 x4
Gen2 x8
Gen3 x4
Gen3 x8

V-Series DMA Avalon-MM DMA Interface to the Application Layer

This section describes the top-level interfaces in the PCIe variant when it includes the high-performance, burst-capable read data mover and write data mover modules.

Depending on the device, the interface to the Application Layer can be 128 or 256 bits.

Figure 3. Avalon-MM DMA Bridge with Internal Descriptor Controller

Figure 4. Avalon-MM DMA Bridge with Internal Descriptor Controller

Figure 5. Avalon-MM DMA Bridge with External Descriptor Controller

This section describes the interfaces that are required to implement the DMA. All other interfaces are described in the next section, Avalon-MM Interface to the Application Layer.

Avalon-MM DMA Interfaces when Descriptor Controller Is Internally Instantiated

This configuration results from selecting both Enable Avalon-MM DMA and Instantiate internal descriptor controller in the component GUI.

The following figure shows the Avalon-MM DMA Bridge, implemented in soft logic. It interfaces to the Hard IP for PCIe through Avalon^® -ST interfaces.

In the following figure, Avalon^® -ST connections and the connection from the BAR0 non-bursting master to the Descriptor Controller slaves are internal. Dashed black lines show these connections. Connections between the Descriptor Controller Masters and the non-bursting slave and the connections between the Read DMA Data Master and the Descriptor Table Slaves are made in the Platform Designer. Blue lines show these connections.

Note: In the following diagrams and text descriptions, the terms Read and Write are from the system memory perspective. Thus, a Read transaction reads data from the system memory and writes it to the local memory in Avalon^® -MM address space. A Write transaction writes the data that was read from the local memory in Avalon^® -MM address space to the system memory.

Figure 6. Avalon-MM DMA Bridge Block Diagram with Optional Internal Descriptor Controller (Showing DMA Write Flow)

The numbers in this figure describe the following steps in the DMA write flow:

The CPU writes registers in the Descriptor Controller Slave to start the DMA.
The Descriptor Controller instructs the Read Data Mover to fetch the descriptor table.
The Read Data Mover forwards the descriptor table to the PCIe Write Descriptor Table Slave.
The Descriptor Controller instructs the Write Data Mover to transfer data.
The Write Data Mover transfers data from FPGA to system memory.
The Write Data Mover notifies the Descriptor Controller of the completion of the data transfer using the done bit.
The Descriptor Controller Master updates the status of the descriptor table in system memory.
The Descriptor Controller Master sends an MSI interrupt to the host.

Figure 7. Avalon-MM DMA Bridge Block Diagram with Optional Internal Descriptor Controller (Showing DMA Read Flow)

The numbers in this figure describe the following steps in the DMA read flow:

The CPU writes registers in the Descriptor Controller Slave to start the DMA.
The Descriptor Controller instructs the Read Data Mover to fetch the descriptor table.
The Read Data Mover forwards the descriptor table to the PCIe Read Descriptor Table Slave.
The Descriptor Controller instructs the Read Data Mover to transfer data.
The Read Data Mover transfers data from system memory to FPGA.
The Read Data Mover notifies the Descriptor Controller of the completion of the data transfer using the done bit.
The Descriptor Controller Master updates the status of the descriptor table in system memory.
The Descriptor Controller Master sends an MSI interrupt to the host.

When the optional Descriptor Controller is included in the bridge, the Avalon-MM bridge includes the following Avalon^® interfaces to implement the DMA functionality:

PCIe Read DMA Data Master (rd_dma): This is a 256-bit wide write only Avalon-MM master interface which supports bursts of up to 16 cycles with the rd_dma* prefix. The Read Data Mover uses this interface to write at high throughput the blocks of data that it has read from the PCIe^® system memory space. This interface writes descriptors to the Read and Write Descriptor table slaves and to any other Avalon^® -MM connected slaves interfaces.
PCIe Write DMA Data Master (wr_dma): This read-only interface transfers blocks of data from the Avalon-MM domain to the PCIe system memory space at high throughput. It drives read transactions on its bursting Avalon-MM master interface. It also creates PCIe Memory Write (MWr) TLPs with data payload from Avalon^® -MM reads. It forwards the MWr TLPs to the Hard IP for transmission on the link. The Write Data Mover module decomposes the transfers into the required number of Avalon-MM burst read transactions and PCIe MWr TLPs. This is a bursting, 256-bit Avalon-MM interface with the wr_dma prefix.
PCIe Read Descriptor Table Slave (rd_dts): This is a 256-bit Avalon-MM slave interface that supports write bursts of up to 16 cycles. The PCIe Read DMA Data Master writes descriptors to this table. This connection is made outside the DMA bridge because the Read Data Mover also typically connects to other Avalon-MM slaves. The prefix for this interface is rd_dts.
PCIe Write Descriptor Table Slave (wr_dts): This is a 256-bit Avalon-MM slave interface that supports write bursts of up to 16 cycles. The PCIe Read DMA Data Master writes descriptors to this table. The PCIe Read DMA Data Master must connect to this interface outside the DMA bridge because the bursting master interface may also need to be connected to the destination of the PCIe Read Data Mover. The prefix for this interface is wr_dts.
Descriptor Controller Master (DCM): This is a 32-bit, non-bursting Avalon-MM master interface with write-only capability. It controls the non-bursting Avalon-MM slave that transmits single DWORD DMA status information to the host. The prefixes for this interface are wr_dcm and rd_dcm.
Descriptor Controller Slave (DCS): This is a 32-bit, non-bursting Avalon-MM slave interface with read and write access. The host accesses this interface through the BAR0 Non-Bursting Avalon-MM Master, to program the Descriptor Controller.
Note: This is not a top-level interface of the Avalon-MM Bridge. Because it connects to BAR0, you cannot use BAR0 to access any other Avalon-MM slave interface.

Read Data Mover

The Read Data module sends memory read TLPs. It writes the completion data to an external Avalon-MM interface through the high throughput Read Master port. This data mover operates on descriptors the IP core receives from the DMA Descriptor Controller.

The Read DMA Avalon-MM Master interface performs the following functions:

1. Provides the Descriptor Table to the Descriptor Controller

The Read Data Mover sends PCIe^® system memory read requests to fetch the descriptor table from PCIe^® system memory. This module then writes the returned descriptor entries in to the Descriptor Controller FIFO using this Avalon-MM interface.

2. Writes Data to Memory Located in Avalon-MM Space

After a DMA Read finishes fetching data from the source address in PCIe^® system memory, the Read Data Mover module writes the data to the destination address in Avalon-MM address space via this interface.

Table 25. Read DMA 256-Bit Avalon-MM Master Interface
Signal Name	Direction	Description
`RdDmaWrite_o`	Output	When asserted, indicates that the Read DMA module is ready to write read completion data to a memory component in the Avalon-MM address space.
`RdDmaAddress_o[63:0]`	Output	Specifies the write address in the Avalon-MM address space for the read completion data.
`RdDmaWriteData_o[127 or 255:0]`	Output	The read completion data to be written to the Avalon-MM address space.
`RdDmaBurstCount_o[4:0] or [5:0]`	Output	Specifies the burst count in 128- or 256-bit words. This bus is 5 bits for the 256-bit interface. It is 6 bits for the 128-bit interface.
`RdDmaByteEnable_o[15 or 31:0]`	Output	Specifies which DWORDs are valid.
`RdDmaWaitRequest_i`	Input	When asserted, indicates that the memory is not ready to receive data.

Figure 8. Read DMA Avalon-MM Master Writes Data to FPGA Memory

Write DMA Avalon-MM Master Port

The Write Data Mover module fetches data from the Avalon-MM address space using this interface before issuing memory write requests to transfer data to PCIe^® system memory.

Table 26. DMA Write 256-Bit Avalon-MM Master Interface
Signal Name	Direction	Description
`WrDMARead_o`	Output	When asserted, indicates that the Write DMA module is reading data from a memory component in the Avalon-MM address space to write to the PCIe address space.
`WrDmaAddress_o[63:0]`	Output	Specifies the address for the data to be read from a memory component in the Avalon-MM address space .
`WrDmaReadData_i[127 or 255:0]`	Input	Specifies the completion data that the Write DMA module writes to the PCIe address space.
`WrDmaBurstCount_o[4:0]or[5:0]`	Output	Specifies the burst count in 128- or 256-bit words. This bus is 5 bits for the 256-bit interface. It is 6 bits for the 128-bit interface
`WrDmaWaitRequest_i`	Input	When asserted, indicates that the memory is not ready to be read.
`WrDmaReadDataValid_i`	Input	When asserted, indicates that `WrDmaReadData_i` is valid.

Figure 9. Write DMA Avalon-MM Master Reads Data from FPGA Memory

RX Master Module

The RX Master module translates read and write TLPs received from the PCIe link to Avalon-MM requests for Platform Designer components connected to the interconnect. This module allows other PCIe components, including host software, to access other Avalon-MM slaves connected in the Platform Designer system.

If burst mode is not enabled, the RX Master module only supports 32-bit read or write request. All other requests received from the PCIe link are considered a violation of this device’s programming model, and are therefore handled with the PCIe Completer Abort status. You can enable burst mode for BAR2 using 32-bit addressing or BAR2 and BAR3 using 64-bit addressing. When enabled, the module supports dword, burst read, or write requests. When the Descriptor Controller is internally instantiated, the RX Master for BAR0 is used internally and not available for other uses.

Note: When you enable burst mode for BAR2, Flush reads (i.e. reads with all byte enables set to 0) are not supported.

Table 27. RX Master Control Interface Ports for BAR Access. Each BAR has one corresponding RX Master Control interface. In this table, *<n>* is the BAR number.
Signal Name	Direction	Description
`RxmRead_<n>_o`	Output	When asserted, indicates an Avalon-MM read request.
`RxmWrite_<n>_o`	Output	When asserted, indicates an Avalon-MM write request.
`RxmAddress_<n>_o[<w>-1:0]`	Output	Specifies the Avalon-MM byte address. Because all addresses are byte addresses, the meaningful bits of this address are [<w>-1:2]. Bits 1 and 0 have a value of 0. <w> can be 32 or 64.
`RxmBurstCount_<n>_o[5:0]`	Output	Specifies the burst count in dwords (32 bits). This optional signal is available for BAR2 when you turn on Enable burst capabilities for RXM BAR2 ports.
`RxmByteEnable_<n>_o[<w>-1:0]`	Output	Specifies the valid bytes of data to be written. <w> has the following values: 4: for the non-bursting RX Master 32: for the bursting 128-bit Avalon-MM interface 64: for the bursting 256-bit Avalon-MM interface
`RxmDataWrite_<n>_o[<w>-1:0]`	Output	Specifies the Avalon-MM write data. <w> has the following values: 32: for the non-bursting RX Master 128: for the bursting 128-bit Avalon-MM interface 256: for the bursting 256-bit Avalon-MM interface
`RxmReadData_<n>_i[<w>-1:0]`	Input	Specifies the Avalon-MM read data. <w> has the following values: 32: for the non-bursting RX Master 128: for the bursting 128-bit Avalon-MM interface 256: for the bursting 256-bit Avalon-MM interface
`RxmReadDataValid_ <n> _i[31:0]`	Input	`When asserted, indicates that RxmReadData_i[31:0]is valid.`
`RxmWaitRequest_<n>_i`	Input	When asserted indicates that the control register access Avalon-MM slave port is not ready to respond.

Figure 10. RXM Master Writes to Memory in the Avalon-MM Address Space

Non-Bursing Slave Module

The TX Slave module translates Avalon-MM read and write requests to PCI Express TLPs.

The slave module supports a single outstanding non-bursting request. It typically sends status updates to the host. This is a 32-bit Avalon-MM slave interface.

Table 28. TX Slave Control
Signal Name	Direction	Description
`TxsChipSelect_i`	Input	When asserted, indicates that this slave interface is selected. When `txs_chipselect_i` is deasserted, `txs_read_i` and `txs_write_i` signals are ignored.
`TxsRead_i`	Input	When asserted, specifies a Avalon-MM Ignored when the chip select is deasserted.
`TxsWrite_i`	Input	When asserted, specifies a Avalon-MM Ignored when the chip select is deasserted.
`TxsWriteData_i[31:0]`	Input	Specifies the Avalon-MM data for a write command.
`TxsAddress_i[<w>-1:0]`	Input	Specifies the Avalon-MM byte address for the read or write command. The width of this address bus is specified by the parameter Address width of accessible PCIe memory space.
`TxsByteEnable_i[3:0]`	Input	Specifies the valid bytes for a write command.
`TxsReadData_o[31:0]`	Output	Drives the read completion data.
`TxsReadDataValid_o`	Output	When asserted, indicates that read data is valid.
`TxsWaitRequest_o`	Output	When asserted, indicates that the Avalon-MM slave port is not ready to respond to a read or write request. The non-bursting Avalon-MM slave may assert`TxsWaitRequest_o` during idle cycles. An Avalon-MM master may initiate a transaction when `TxsWaitRequest_o` is asserted and wait for that signal to be deasserted.

Figure 11. Non-Bursting Slave Interface Sends Status to Host

32-Bit Control Register Access (CRA) Slave Signals

The CRA interface provides access to the control and status registers of the Avalon-MM bridge. This interface has the following properties:

32-bit data bus
Supports a single transaction at a time
Supports single-cycle transactions (no bursting)

Note: When Intel^® Stratix^® 10 is in Root Port mode, and the application logic issues a CfgWr or CfgRd via the CRA interface, it needs to fill the Tag field in the TLP Header with the value 0x10 to ensure that the corresponding Completion gets routed to the CRA interface correctly. If the application logic sets the Tag field to some other value, the Intel^® Stratix^® 10 Avalon^® -MM Hard IP for PCIe IP Core does not overwrite that value with the correct value.

Table 29. Avalon-MM CRA Slave Interface Signals
Signal Name	Direction	Description
`CraRead_i`	Input	Read enable.
`CraWrite_i`	Input	Write request.
`CraAddress_i[13:0]`	Input	An address space of 16 KB is allocated for the control registers. Avalon^® -MM slave addresses provide address resolution down to the width of the slave data bus. Because all addresses are byte addresses, this address logically goes down to bit 2. Bits 1 and 0 are 0. To read or write individual bytes of a DWORD, use byte enables. For example, to write bytes 0 and 1, set this signal to 4'b0011. An address space of 32 KB is allocated for the control registers. Avalon^® -MM slave addresses provide address resolution down to the width of the slave data bus. Because all addresses are byte addresses, this address logically goes down to bit 2. Bits 1 and 0 are 0. To read or write individual bytes of a DWORD, use byte enables. For example, to write bytes 0 and 1, set this signal to 4'b0011.
`CraWriteData_i[31:0]`	Input	Write data. The current version of the CRA slave interface is read-only. Including this signal as part of the Avalon-MM interface, makes future enhancements possible.
`CraReadData_o[31:0]`	Output	Read data lines.
`CraByteEnable_i[3:0]`	Input	Byte enable.
`CraWaitRequest_o`	Output	Wait request to hold off additional requests.
`CraChipSelect_i`	Input	Chip select signal to this slave.
`CraIrq_o`	Output	Interrupt request. A port request for an Avalon-MM interrupt.

Avalon-ST Descriptor Control Interface when Instantiated Separately

After fetching multiple descriptor entries from the Descriptor Table in the PCIe^® system memory, the Descriptor Controller uses its Avalon^® -ST Descriptor source interface to transfer 160-bit Descriptors to the Read or Write DMA Data Movers.

Table 30. Avalon^® -ST Descriptor Sink Interface. This interface sends instructions from Descriptor Controller to Read DMA Engine.
Signal Name	Direction	Description
`RdAstRxData_i[159:0]`	Input	Specifies the descriptors for the Read DMA module. Refer to DMA Descriptor Format table below for bit definitions.
`RdAstRxValid_i`	Input	When asserted, indicates that the data is valid.
`RdAstRxReady_o`	Output	When asserted, indicates that the Read DMA read module is ready to receive a new descriptor. The ready latency is 3 cycles. Consequently, interface can accept data 3 cycles after ready is asserted.

Table 31. Avalon^® -ST Descriptor Sink Interface . This interface sends instructions from Descriptor Controller to Write DMA Engine.
Signal Name	Direction	Description
`WrAstRxData_i[159:0]`	Input	Specifies the descriptors for the Write DMA module. Refer to DMA Descriptor Format table below for bit definitions.
`WrAstRxValid_i`	Input	When asserted, indicates that the data is valid.
`WrAstRxReady_o`	Output	When asserted, indicates that the Write DMA module engine is ready to receive a new descriptor. The ready latency for this signal is 3 cycles. Consequently, interface can accept data 3 cycles after ready is asserted.

Avalon-ST Descriptor Status Interface

When DMA module completes the processing for one Descriptor Instruction, it returns DMA Status to the Descriptor Controller via the following interfaces.

Table 32. Read DMA Status Interface from Read DMA Engine to Descriptor Controller
Signal Name	Direction	Description
`RdAstTxData_o[31:0]`	Output	Drives status information to the Descriptor Controller component. Refer to DMA Status Bus table below for more information
`RdAstTxValid_o`	Output	When asserted, indicates that the data is valid.

Table 33. Write DMA Status Interface from Write DMA Engine to Descriptor Controller
Signal Name	Direction	Description
`WrAstTxData_o[31:0]`	Output	Drives status information to the Descriptor Controller component. Refer to DMA Status Bus table below for more information about this bus.
`WrAstTxValid_o`	Output	When asserted, indicates that the data is valid.

Table 34. DMA Descriptor Format
Bits	Name	Description
[31:0]	`Source Low Address`	Low-order 32 bits of the DMA source address. The address boundary must align to the 32 bits so that the 2 least significant bits are 2'b00. For the Read DMA module, the source address is the PCIe domain address. For the Write DMA module, the source address is the Avalon-MM domain address.
[63:32]	`Source High Address`	High-order 32 bits of the source address.
[95:64]	`Destination Low Address`	Low-order 32 bits of the DMA destination address. The address boundary must align to the 32 bits so that the 2 least significant bits have the value of 2'b00. For the Read DMA module, the destination address is the Avalon-MM domain address. For the Write DMA module the destination address is the PCIe domain address.
[127:96]	`Destination High Address`	High-order 32 bits of the destination address.
[145:128]	`DMA Length`	Specifies DMA length in dwords. The length must be greater than 0. The maximum length is 1 MB - 4 bytes.
[153:146]	`DMA Descriptor ID`	Specifies up to 128 descriptors.
[158:154]	`Reserved`	—
[159]	`Immediate Write`	When set to 1'b1, the Write DMA Engine performs an `Immediate Write`. The `Immediate Write` provides a fast mechanism to send a Write TLP upstream. The descriptor stores the 32-bit payload, replacing the `Source Low Address` field of the descriptor.

Avalon -ST Descriptor Status Sources

Read DMA and Write DMA modules report status to the Descriptor Controller on the RdDmaTxData_o[31:0] or WrDmaTxData_o[31:0] bus when a descriptor completes successfully.

The following table shows the mappings of the triggering events to the DMA descriptor status bus:

Table 35. DMA Status Bus
Bits	Name	Description
[31:9]	Reserved	—
[8]	`Done`	When asserted, a single DMA descriptor has completed successfully.
[7:0]	`Descriptor ID`	The ID of the descriptor whose status is being reported.

Descriptor Controller Interfaces when Instantiated Internally

The Descriptor Controller controls the Read DMA and Write DMA Data Movers. It provides a 32-bit Avalon-MM slave interface to control and manage data flow from PCIe^® system memory to Avalon^® -MM memory and in the reverse direction.

The Descriptor Controller includes two, 128-entry FIFOs to store the read and write descriptor tables. The Descriptor Controller forwards the descriptors to the Read DMA and Write DMA Data Movers.

The Data Movers send completion status to the Read Descriptor Controller and Write Descriptor Controller. The Descriptor Controller forwards status and MSI to the host using the TX slave port.

Read Descriptor Controller Avalon-MM Master interface

The Read Descriptor Controller Avalon-MM master interface drives the non-bursting Avalon-MM slave interface. The Read Descriptor Controller uses this interface to write descriptor status to the PCIe domain and possibly to MSI when MSI messages are enabled. This Avalon-MM master interface is only available for variants with the internally instantiated Descriptor Controller.

By default MSI interrupts are enabled. You specify the Number of MSI messages requested on the MSI tab of the parameter editor. The MSI Capability Structure is defined in Section 6.8.1 MSI Capability Structure of the PCI Local Bus Specification.

Table 36. Read Descriptor Controller Avalon-MM Master interface
Signal Name	Direction	Description
`RdDCMAddress_o[63:0]`	Output	Specifies the descriptor status table or MSI address.
`RdDCMByteEnable_o[3:0]`	Output	Specifies which data bytes are valid.
`RdDCMReadDataValid_i`	Input	When asserted, indicates that the read data is valid.
`RdDCMReadData_i[31:0]`	Input	Specifies the read data of the descriptor status table entry addressed.
`RdDCMRead_o`	Output	When asserted, indicates a read transaction. Currently, this is a write-only interface so that this signal never asserts.
`RdDCMWaitRequest_i`	Input	When asserted, indicates that the connected Avalon-MM slave interface is busy and cannot accept a transaction.
`RdDCMWriteData_o[31:0]`	Output	Specifies the descriptor status or MSI data..
`RdDCMWrite_o`	Output	When asserted, indicates a write transaction.

Write Descriptor Controller Avalon-MM Master Interface

The Avalon-MM Descriptor Controller Master interface is a 32-bit single-DWORD master with wait request support. The Write Descriptor Controller uses this interface to write status back to the PCI-Express domain and possibly MSI when MSI messages are enabled. This Avalon-MM master interface is only available for the internally instantiated Descriptor Controller.

Table 37. **Write Descriptor Controller Avalon-MM Master interface**
Signal Name	Direction	Description
`WrDCMAddress_o[63:0]`	Output	Specifies the descriptor status table or MSI address.
`WrDCMByteEnable_o[3:0]`	Output	Specifies which data bytes are valid.
`WrDCMReadDataValid_i`	Input	When asserted, indicates that the read data is valid.
`WrRdDCMReadData_i[31:0]`	Output	Specifies the read data for the descriptor status table entry addressed.
`WrDCMRead_o`	Output	When asserted, indicates a read transaction.
`WrDCMWaitRequest_i`	Input	When asserted, indicates that the Avalon-MM slave device is not ready to respond.
`WrDCMWriteData_o[31:0]`	Output	Specifies the descriptor status table or MSI address.
`WrDCMWrite_o`	Output	When asserted, indicates a write transaction.

Read Descriptor Table Avalon-MM Slave Interface

This interface is available when you select the internal Descriptor Controller. It receives the Read DMA descriptors which are fetched by the Read Data Mover. Connect the interface to the Read DMA Avalon-MM master interface.

Table 38. Read Descriptor Table Avalon-MM Slave Interface
Signal Name	Direction	Description
`RdDTSAddress_i[7:0]`	Input	Specifies the descriptor table address.
`RdDTSBurstCount_i[4:0] or [5:0]`	Input	Specifies the burst count of the transaction in words.
`RdDTSChipSelect_i`	Input	When asserted, indicates that the read targets this slave interface.
`RdDTSWriteData_i[255:0] or [127:0]`	Input	Specifies the descriptor.
`RdDTSWrite_i`	Input	When asserted, indicates a write transaction.
`RdDTSWaitRequest_o`	Output	When asserted, indicates that the Avalon-MM slave device is not ready to respond.

Write Descriptor Table Avalon-MM Slave Interface

This interface is available when you select the internal Descriptor Controller. This interface receives the Write DMA descriptors which are fetched by Read Data Mover. Connect the interface to the Read DMA Avalon-MM master interface.

Table 39. Write Descriptor Table Avalon-MM Slave Interface
Signal Name	Direction	Description
`WrDTSAddress_i[7:0]`	Input	Specifies the descriptor table address.
`WrDTSBurstCount_i[4:0]`	Input	Specifies the burst count of the transaction in words.
`WrDTSChipSelect_i`	Input	When asserted, indicates that the write is for this slave interface.
`WrDTSWaitRequest_o`	Output	When asserted, indicates that this interface is busy and is not ready to respond.
`WrDTSWriteData_i[255:0] or [127:0]`	Input	Drives the descriptor table entry data.
`WrDTSWrite_i`	Input	When asserted, indicates a write transaction.

Clock Signals

Table 40. Clock Signals
Signal	Direction	Description
`refclk`	Input	Reference clock for the IP core. It must have the frequency specified under the System Settings heading in the parameter editor. This is a dedicated free running input clock to the dedicated `REFCLK` pin. If your design meets the following criteria: Enables CvP Includes an additional transceiver PHY connected to the same Transceiver Reconfiguration Controller then you must connect `refclk` to the `mgmt_clk_clk` signal of the Transceiver Reconfiguration Controller and the additional transceiver PHY. In addition, if your design includes more than one Transceiver Reconfiguration Controller on the same side of the FPGA, they all must share the `mgmt_clk_clk` signal.
`coreclkout`	Output	This is a fixed frequency clock used by the Data Link and Transaction Layers. To meet PCI Express link bandwidth constraints, this clock has minimum frequency requirements as listed in Application Layer Clock Frequency for All Combination of Link Width, Data Rate and Application Layer Interface Width in the Reset and Clocks chapter .

Reset, Status, and Link Training Signals

Refer to Reset and Clocks for more information about the reset sequence and a block diagram of the reset logic.

Table 41. Reset Signals
Signal	Direction	Description
`npor`	Input	Active low reset signal. In the Intel hardware example designs, `npor` is the `OR` of `pin_perst` and `local_rstn` coming from the software Application Layer. If you do not drive a soft reset signal from the Application Layer, this signal must be derived from `pin_perst`. You cannot disable this signal. Resets the entire IP Core and transceiver. Asynchronous. In systems that use the hard reset controller, this signal is edge, not level sensitive; consequently, you cannot use a low value on this signal to hold custom logic in reset. For more information about the hard and soft reset controllers, refer to Reset.
`nreset_status`	Output	Active low reset signal. It is derived from `npor` or `pin_perstn`. You can use this signal to reset the Application Layer.
`pin_perst`	Input	Active low reset from the PCIe reset pin of the device. `pin_perst` resets the datapath and control registers. This signal is required for Configuration via Protocol (CvP). For more information about CvP refer to Configuration via Protocol (CvP). V-Series have 1 or 2 instances of the Hard IP for PCI Express. Each instance has its own `pin_perst` signal. You must connect the `pin_perst` of each Hard IP instance to the corresponding `nPERST` pin of the device. These pins have the following locations: For Arria V devices: `nPERSTL0`: bottom left Hard IP block and CvP blocks `nPERSTL1`: top left Hard IP For Cyclone V devices: `nPERSTL0`: top left Hard IP `nPERSTL1`: bottom left Hard IP and CvP blocks For Arria V GZ and Stratix V devices: `NPERSTL0`: bottom left Hard IP and CvP blocks `NPERSTL1`: top left Hard IP block `NPERSTR0`: bottom right Hard IP block `NPERSTR1`: top right Hard IP block For maximum use of the V-Series device, Intel recommends that you use the bottom left Hard IP first. This is the only location that supports CvP over a PCIe link. Refer to the appropriate device pinout for correct pin assignment for more detailed information about these pins. The PCI Express Card Electromechanical Specification 2.0 specifies this pin requires 3.3 V. You can drive this 3.3V signal to the `nPERST*` even if the V_VCCPGM of the bank is not 3.3V if the following 2 conditions are met: The input signal meets the V_IH and V_IL specification for LVTTL. The input signal meets the overshoot specification for 100°C operation as specified by the Maximum Allowed Overshoot and Undershoot Voltage in the device datasheet.

Figure 12. Reset and Link Training Timing Relationships

The following figure illustrates the timing relationship between npor and the LTSSM L0 state.

Note: To meet the 100 ms system configuration time, you must use the fast passive parallel configuration scheme with CvP and a 32-bit data width (FPP x32) or use the V-Series Hard IP for PCI Express in autonomous mode.

Table 42. Status and Link Training Signals
Signal	Direction	Description
`cfg_par_err`	Output	Indicates that a parity error in a TLP routed to the internal Configuration Space. This error is also logged in the Vendor Specific Extended Capability internal error register. You must reset the Hard IP if this error occurs. The signal is not available for Arria V and Cyclone V devices.
`derr_cor_ext_rcv`	Output	Indicates a corrected error in the RX buffer. This signal is for debug only. It is not valid until the RX buffer is filled with data. This is a pulse, not a level, signal. Internally, the pulse is generated with the 500 MHz clock. A pulse extender extends the signal so that the FPGA fabric running at 250 MHz can capture it. Because the error was corrected by the IP core, no Application Layer intervention is required. ⁹
`derr_cor_ext_rpl`	Output	Indicates a corrected ECC error in the retry buffer. This signal is for debug only. Because the error was corrected by the IP core, no Application Layer intervention is required. ⁹
`derr_rpl`	Output	Indicates an uncorrectable error in the retry buffer. This signal is for debug only. ⁹ The signal is not available for Arria V and Cyclone V devices.
`dlup_exit`	Output	This signal is asserted low for one `pld_clk` cycle when the IP core exits the DLCMSM DL_Up state, indicating that the Data Link Layer has lost communication with the other end of the PCIe link and left the Up state. When this pulse is asserted, the Application Layer should generate an internal reset signal that is asserted for at least 32 cycles.
`ev128ns`	Output	Asserted every 128 ns to create a time base aligned activity.
`ev1us`	Output	Asserted every 1µs to create a time base aligned activity.
`hotrst_exit`	Output	Hot reset exit. This signal is asserted for 1 clock cycle when the LTSSM exits the hot reset state. This signal should cause the Application Layer to be reset. This signal is active low. When this pulse is asserted, the Application Layer should generate an internal reset signal that is asserted for at least 32 cycles.
`int_status[3:0]`	Output	These signals drive legacy interrupts to the Application Layer as follows: int_status[0]: interrupt signal A int_status[1]: interrupt signal B int_status[2]: interrupt signal C int_status[3]: interrupt signal D
`ko_cpl_spc_data[11:0]`	Output	The Application Layer can use this signal to build circuitry to prevent RX buffer overflow for completion data. Endpoints must advertise infinite space for completion data; however, RX buffer space is finite. `ko_cpl_spc_data` is a static signal that reflects the total number of 16 byte completion data units that can be stored in the completion RX buffer.
`ko_cpl_spc_header[7:0]`	Output	The Application Layer can use this signal to build circuitry to prevent RX buffer overflow for completion headers. Endpoints must advertise infinite space for completion headers; however, RX buffer space is finite. `ko_cpl_spc_header` is a static signal that indicates the total number of completion headers that can be stored in the RX buffer.
`l2_exit`	Output	L2 exit. This signal is active low and otherwise remains high. It is asserted for one cycle (changing value from 1 to 0 and back to 1) after the LTSSM transitions from l2.idle to detect. When this pulse is asserted, the Application Layer should generate an internal reset signal that is asserted for at least 32 cycles.
`lane_act[3:0]`	Output	Lane Active Mode: This signal indicates the number of lanes that configured during link training. The following encodings are defined: 4’b0001: 1 lane 4’b0010: 2 lanes 4’b0100: 4 lanes 4’b1000: 8 lanes
`ltssmstate[4:0]`	Output	LTSSM state: The LTSSM state machine encoding defines the following states: 00000: Detect.Quiet 00001: Detect.Active 00010: Polling.Active 00011: Polling.Compliance 00100: Polling.Configuration 00101: Polling.Speed 00110: Config.Linkwidthstart 00111: Config.Linkaccept 01000: Config.Lanenumaccept 01001: Config.Lanenumwait 01010: Config.Complete 01011: Config.Idle 01100: Recovery.Rcvlock 01101: Recovery.Rcvconfig 01110: Recovery.Idle 01111: L0 10000: Disable 10001: Loopback.Entry 10010: Loopback.Active 10011: Loopback.Exit 10100: Hot.Reset 10101: L0s 11001: L2.transmit.Wake 11010: Speed.Recovery 11011: Recovery.Equalization, Phase 0 11100: Recovery.Equalization, Phase 1 11101: Recovery.Equalization, Phase 2 11110: Recovery.Equalization, Phase 3 11111: Recovery.Equalization, Done
`rx_par_err`	Output	When asserted for a single cycle, indicates that a parity error was detected in a TLP at the input of the RX buffer. This error is logged as an uncorrectable internal error in the VSEC registers. For more information, refer to Uncorrectable Internal Error Status Register. You must reset the Hard IP if this error occurs because parity errors can leave the Hard IP in an unknown state. The signal is not available for Arria V and Cyclone V devices.
`tx_par_err[1:0]`	Output	When asserted for a single cycle, indicates a parity error during TX TLP transmission. These errors are logged in the VSEC register. The following encodings are defined: 2’b10: A parity error was detected by the TX Transaction Layer. The TLP is nullified and logged as an uncorrectable internal error in the VSEC registers. For more information, refer to Uncorrectable Internal Error Status Register. 2’b01: Some time later, the parity error is detected by the TX Data Link Layer which drives 2’b01 to indicate the error. Reset the IP core when this error is detected. Contact Intel technical support if resetting becomes unworkable. Note that not all simulation models assert the Transaction Layer error bit in conjunction with the Data Link Layer error bit. The signal is not available for Arria V and Cyclone V devices.

⁹ Intel does not rigorously test or verify debug signals. Only use debug signals to observe behavior. Do not use debug signals to drive custom logic.

MSI Interrupts for Endpoints

The MSI interrupt notifies the host when a DMA operation has completed. After the host receives this interrupt, it can poll the DMA read or write status table to determine which entry or entries have the done bit set. This mechanism allows host software to avoid continuous polling of the status table done bits. Use this interface to receive information required to generate MSI or MSI-X interrupts to the Root Port via the TX Slave interface.

Table 43. MSI Interrupt
Signal	Direction	Description
`MSIIntfc_o[81:0]`	Output	This bus provides the following MSI address, data, and enabled signals: `MSIIntfc_o[81]`: Master enable `MSIIntfc_o[80]`: MSI enable `MSIIntfc_o[79:64]`: MSI data `MSIIntfc_o[63:0]`: MSI address
`MSIXIntfc_o[15:0]`	Output	Provides for system software control of MSI-X as defined in Section 6.8.2.3 Message Control for MSI-X in the PCI Local Bus Specification, Rev. 3.0. The following fields are defined: `MSIXIntfc_o[15]`: Enable `MSIXIntfc_o[14]`: Mask `MSIXIntfc_o[13:11]`: Reserved `MSIXIntfc_o[10:0]`: Table size
`MSIControl_o[15:0]`	Output	Provides system software control of the MSI messages as defined in Section 6.8.1.3 Message Control for MSI in the PCI Local Bus Specification, Rev. 3.0. The following fields are defined: `MSIControl_o[15:9]`: Reserved `MSIControl_o[8]`: Per-Vector Masking Capable `MSIControl_o[7]`: 64-Bit Address Capable `MSIControl_o[6:4]`: Multiple Message Enable `MSIControl_o[3:1]`: MSI Message Capable `MSIControl_o[0]`: MSI Enable

Physical Layer Interface Signals

Intel provides an integrated solution with the Transaction, Data Link and Physical Layers. The IP Parameter Editor generates a SERDES variation file, <variation>_serdes.v or .vhd , in addition to the Hard IP variation file, <variation>.v or .vhd. The SERDES entity is included in the library files for PCI Express.

Transceiver Reconfiguration

Dynamic reconfiguration compensates for variations due to process, voltage and temperature (PVT). Among the analog settings that you can reconfigure are V _OD, pre-emphasis, and equalization.

You can use the Intel Transceiver Reconfiguration Controller to dynamically reconfigure analog settings. For more information about instantiating the Intel Transceiver Reconfiguration Controller IP core refer to Hard IP Reconfiguration .

Table 44. Transceiver Control Signals. In this table, `<n>` is the number of interfaces required.
Signal Name	Direction	Description
`reconfig_from_xcvr[(<n>46)-1:0]`	Output	Reconfiguration signals to the Transceiver Reconfiguration Controller.
`reconfig_to_xcvr[(<n>70)-1:0]`	Input	Reconfiguration signals from the Transceiver Reconfiguration Controller.
`reconfig_clk_locked`	Output	When asserted, indicates that the PLL that provides the fixed clock required for transceiver initialization is locked. The Application Layer should be held in reset until `reconfig_clk_locked` is asserted.

The following table shows the number of logical reconfiguration and physical interfaces required for various configurations. The Quartus^® Prime Fitter merges logical interfaces to minimize the number of physical interfaces configured in the hardware. Typically, one logical interface is required for each channel and one for each PLL. The ×8 variants require an extra channel for PCS clock routing and control. The ×8 variants use channel 4 for clocking.

Table 45. Number of Logical and Physical Reconfiguration Interfaces
Variant	Logical Interfaces
Gen2 ×4	5
Gen1 and Gen2 ×8	10
Gen3 ×4	6
Gen3 ×8	11

For more information about the Transceiver Reconfiguration Controller, refer to the Transceiver Reconfiguration Controller chapter in the Intel Transceiver PHY IP Core User Guide .

Serial Data Signals

This differential, serial interface is the physical link between a Root Port and an Endpoint.

The PCIe IP Core supports 1, 2, 4, or 8 lanes. Each lane includes a TX and RX differential pair. Data is striped across all available lanes.

The PCIe IP Core supports 1, 2, or 4 lanes. Each lane includes a TX and RX differential pair. Data is striped across all available lanes.

Table 46. 1-Bit Interface Signals . In the following table `<n>` is the number of lanes.
Signal	Direction	Description
`tx_out[<n>-1:0]`	Output	Transmit output. These signals are the serial outputs of lanes `<n>-1`–0.
`rx_in[<n>-1:0]`	Input	Receive input. These signals are the serial inputs of lanes `<n>-1`–0.

Refer to Pin-out Files for Intel Devices for pin-out tables for all Intel devices in .pdf, .txt, and .xls formats.

Transceiver channels are arranged in groups of six. For GX devices, the lowest six channels on the left side of the device are labeled GXB_L0, the next group is GXB_L1, and so on. Channels on the right side of the device are labeled GXB_R0, GXB_R1, and so on. Be sure to connect the Hard IP for PCI Express on the left side of the device to appropriate channels on the left side of the device, as specified in the Pin-out Files for Intel Devices.

Physical Layout of Hard IP In Arria V GX/GX/SX/ST Devices

Arria V devices include one or two Hard IP for PCI Express IP cores. The following figures illustrate the placement of the PCIe IP core, transceiver banks, and channels.

Arria V devices include one or two Hard IP for PCI Express IP Cores. The following figures illustrates the placement of the Hard IP for PCIe IP cores, transceiver banks and channels for the largest Arria V devices. Note that the bottom left IP core includes the CvP functionality. Devices with a single Hard IP for PCIe IP Core only include the bottom left core.

Figure 13. Transceiver Bank and Hard IP for PCI Express IP Core Locations in Arria GX and GT Devices

Figure 14. Transceiver Bank and Hard IP for PCI Express IP Core Locations in Arria SX Devices

Figure 15. Transceiver Bank and Hard IP for PCI Express IP Core Locations in Arria ST Devices

Refer to Transceiver Architecture in Arria V Devices for comprehensive information on the number of Hard IP for PCIe IP cores available in various Arria V packages.

Channel Placement in Arria V Devices

Figure 16. Arria V Gen1 and Gen2 Channel Placement Using the CMU PLLIn the following figures the channels shaded in blue provide the transmit CMU PLL generating the high-speed serial clock.

You can assign other protocols to unused channels the if data rate and clock specification exactly match the PCIe configuration.

Physical Layout of Hard IP in Cyclone V Devices

Cyclone V devices include one or two Hard IP for PCI Express IP cores. The following figures illustrate the placement of the PCIe IP cores, transceiver banks, and channels. Note that the bottom left IP core includes the CvP functionality. The other Hard IP blocks do not include the CvP functionality. Transceiver banks include six channels. Within a bank, channels are arranged in 3-packs. GXB_L0 contains channels 0–2, GXB_L1 includes channels 3–5, and so on.

Figure 17. Cyclone V GX/GT/ST/ST Devices with 9 or 12 Transceiver Channels and 2 PCIe CoresIn the following figure, the x1 Hard IP for PCI Express uses channel 0 and channel 1 of GXB_L0 and channel 0 and channel 1 of GXB_L2.

Figure 18. Cyclone V GX/GT/ST/ST Devices with 6 Transceiver Channels and 2 PCIe Cores

For more comprehensive information about Cyclone V transceivers, refer to the Transceiver Banks section in the Transceiver Architecture in Cyclone V Devices.

Channel Placement in Cyclone V Devices

Figure 19. Cyclone V Gen1 and Gen2 Channel Placement Using the CMU PLLIn the following figures the channels shaded in blue provide the transmit CMU PLL generating the high-speed serial clock.

You can assign other protocols to unused channels the if data rate and clock specification exactly match the PCIe configuration.

Physical Layout of Hard IP in Arria V GZ Devices

Arria V GZ devices include one Hard IP for PCI Express IP core. The following figures illustrate the placement of the PCIe IP core, transceiver banks, and channels.

Figure 20. Physical Layout of Hard IP in Arria V GZ Devices

Refer to Transceiver Architecture in Arria V Devices for comprehensive information on the number of Hard IP for PCIe IP cores available in various Arria V GZ packages.

Refer to Channel Utilization for Data and Clock Routing in Arria V GZ and Stratix V Devices for additional information about channel and PLL utilization.

Physical Layout of Hard IP in Stratix V GX/GT/GS Devices

Stratix V devices include one, two, or four Hard IP for PCI Express IP cores. The following figures illustrate the placement of the PCIe IP cores, transceiver banks, and channels for the largest Stratix V devices. Note that the bottom left hard IP block includes the CvP functionality for flip chip packages. For other package types, the CvP functionality is in the bottom right block. All other Hard IP blocks do not include the CvP functionality.

Figure 21. Stratix V GX/GT/GS Devices with Four PCIe Hard IP Blocks

Smaller devices include the following PCIe Hard IP Cores:

One Hard IP for PCIe IP core - bottom left IP core with CvP, located at GX banks L0 and L1
Two Hard IP for PCIe IP cores - bottom left IP core with CvP and bottom right IP Core, located at banks L0 and L1, and banks R0 and R1

Refer to Stratix V GX/GT Channel and PCIe Hard IP (HIP) Layout for comprehensive information on the number of Hard IP for PCIe IP cores available in various Stratix V packages.

Channel Placement in Arria V GZ and Stratix V GX/GT/GS Devices

Figure 22. Arria V GZ and Stratix V GX/GT/GS Gen1 and Gen2 Channel Placement Using the CMU PLLIn the following figures the channels shaded in blue provide the transmit CMU PLL generating the high-speed serial clock.

Figure 23. Arria V GZ and Stratix V GX/GT/GS Gen3 Channel Placement Using the CMU and ATX PLLsGen3 requires two PLLs to facilitate rate switching between the Gen1, Gen2, and Gen3 data rates. Channels shaded in blue provide the transmit CMU PLL generating the high-speed serial clock. The ATX PLL shaded in blue is the ATX PLL used in these configurations.

Figure 24. Arria V GZ and Stratix V GX/GT/GS Gen1 and Gen2 Channel Placement Using the ATX PLLSelecting the ATX PLL has the following advantages over selecting the CMU PLL:

The ATX PLL saves one channel in Gen1 and Gen2 ×1, ×2, and ×4 configurations.
The ATX PLL has better jitter performance than the CMU PLL.

Note: You must use the soft reset controller when you select the ATX PLL and you cannot use CvP.

PIPE Interface Signals

These PIPE signals are available for Gen1, Gen2, and Gen3 variants so that you can simulate using either the serial or the PIPE interface. Note that Intel^® Arria^® 10 and Intel^® Cyclone^® 10 GX devices do not support the Gen3 PIPE interface. Simulation is faster using the PIPE interface because the PIPE simulation bypasses the SERDES model. By default, the PIPE interface is 8 bits for Gen1 and Gen2 and 32 bits for Gen3. You can use the PIPE interface for simulation even though your actual design includes a serial interface to the internal transceivers. However, it is not possible to use the Hard IP PIPE interface in hardware, including probing these signals using Signal Tap Embedded Logic Analyzer. These signals are not top-level signals of the Hard IP. They are listed here to assist in debugging link training issues.

These PIPE signals are available for Gen1 and Gen2 variants so that you can simulate using either the serial or the PIPE interface. Simulation is faster using the PIPE interface because the PIPE simulation bypasses the SERDES model. The PIPE interface is 8 bits for Gen1 and Gen2. You can use the PIPE interface for simulation even though your actual design includes a serial interface to the internal transceivers. However, it is not possible to use the hard IP PIPE interface in hardware, including probing these signals using Signal Tap Embedded Logic Analyzer. These signals are not top-level signals of the Hard IP. They are listed here to assist in debugging link training issues.

Note: The Intel Root Port BFM bypasses Gen3 Phase 2 and Phase 3 Equalization. However, Gen3 variants can perform Phase 2 and Phase 3 equalization if instructed by a third-party BFM.

Table 47. PIPE Interface Signals . In the following table, signals that include lane number 0 also exist for lanes 1-7. These signals are for simulation only. For Quartus^® Prime software compilation, these pipe signals can be left floating. In Platform Designer, the signals that are part of the PIPE interface have the prefix, *hip_pipe*. The signals which are included to simulate the PIPE interface have the prefix, *hip_pipe_sim_pipe*
Signal	Direction	Description
`eidleinfersel0[2:0]`	Output	Electrical idle entry inference mechanism selection. The following encodings are defined: 3'b0xx: Electrical Idle Inference not required in current LTSSM state 3'b100: Absence of COM/SKP Ordered Set the in 128 us window for Gen1 or Gen2 3'b101: Absence of TS1/TS2 Ordered Set in a 1280 UI interval for Gen1 or Gen2 3'b110: Absence of Electrical Idle Exit in 2000 UI interval for Gen1 and 16000 UI interval for Gen2 3'b111: Absence of Electrical idle exit in 128 us window for Gen1
`phystatus0`	Input	PHY status `<n>`. This signal communicates completion of several PHY requests.
`powerdown0[1:0]`	Output	Power down `<n>`. This signal requests the PHY to change its power state to the specified state (P0, P0s, P1, or P2).
`rxdata0[31:0]`	Input	Receive data. This bus receives data on lane `<n>`.
`rxdatak0[3:0]`	Input	Data/Control bits for the symbols of receive data. Bit 0 corresponds to the lowest-order byte of `rxdata`, and so on. A value of 0 indicates a data byte. A value of 1 indicates a control byte. For Gen1 and Gen2 only.
`rxelecidle0`	Input	Receive electrical idle `<n>`. When asserted, indicates detection of an electrical idle.
`rxpolarity0`	Output	Receive polarity `<n>`. This signal instructs the PHY layer to invert the polarity of the 8B/10B receiver decoding block.
`rxstatus0[2:0]`	Input	Receive status `<n>`. This signal encodes receive status and error codes for the receive data stream and receiver detection.
`rxvalid0`	Input	Receive valid `<n>`. This symbol indicates symbol lock and valid data on `rxdata` `<n>` and `rxdatak` `<n>`.
`sim_pipe_ltssmstate0[4:0]`	Input and Output	LTSSM state: The LTSSM state machine encoding defines the following states: 5’b00000: Detect.Quiet 5’b 00001: Detect.Active 5’b00010: Polling.Active 5’b 00011: Polling.Compliance 5’b 00100: Polling.Configuration 5’b00101: Polling.Speed 5’b00110: Config.LinkwidthsStart 5’b 00111: Config.Linkaccept 5’b 01000: Config.Lanenumaccept 5’b01001: Config.Lanenumwait 5’b01010: Config.Complete 5’b 01011: Config.Idle 5’b01100: Recovery.Rcvlock 5’b01101: Recovery.Rcvconfig 5’b01110: Recovery.Idle 5’b 01111: L0 5’b10000: Disable 5’b10001: Loopback.Entry 5’b10010: Loopback.Active 5’b10011: Loopback.Exit 5’b10100: Hot.Reset 5’b10101: L0s 5’b11001: L2.transmit.Wake 5’b11010: Recovery.Speed 5’b11011: Recovery.Equalization, Phase 0 5’b11100: Recovery.Equalization, Phase 1 5’b11101: Recovery.Equalization, Phase 2 5’b11110: Recovery.Equalization, Phase 3 5’b11111: Recovery.Equalization, Done
`sim_pipe_pclk_in`	Input	This clock is used for PIPE simulation only, and is derived from the `refclk`. It is the PIPE interface clock used for PIPE mode simulation.
`sim_pipe_rate[1:0]`	Input	Specifies the data rate. The 2-bit encodings have the following meanings: 2’b00: Gen1 rate (2.5 Gbps) 2’b01: Gen2 rate (5.0 Gbps) 2’b1X: Gen3 rate (8.0 Gbps)
`txcompl0`	Output	Transmit compliance `<n>`. This signal forces the running disparity to negative in compliance mode (negative COM character).
`txdata0[31:0]`	Output	Transmit data. This bus transmits data on lane `<n>`.
`txdatak0[3:0]`	Output	Transmit data control `<n>`. This signal serves as the control bit for `txdata` `<n>`. Bit 0 corresponds to the lowest-order byte of `rxdata`, and so on. A value of 0 indicates a data byte. A value of 1 indicates a control byte. For Gen1 and Gen2 only.
`txdataskip0`	Output	For Gen3 operation. Allows the MAC to instruct the TX interface to ignore the TX data interface for one clock cycle. The following encodings are defined: 1’b0: TX data is invalid 1’b1: TX data is valid
`txdeemph0`	Output	Transmit de-emphasis selection. The value for this signal is set based on the indication received from the other end of the link during the Training Sequences (TS). You do not need to change this value.
`txdetectrx0`	Output	Transmit detect receive `<n>`. This signal tells the PHY layer to start a receive detection operation or to begin loopback.
`txelecidle0`	Output	Transmit electrical idle `<n>`. This signal forces the TX output to electrical idle.
`txmargin0[2:0]`	Output	Transmit V_OD margin selection. The value for this signal is based on the value from the `Link Control 2` `Register`. Available for simulation only.
`txswing0`	Output	When asserted, indicates full swing for the transmitter voltage. When deasserted indicates half swing.
`txsynchd0` `[1:0]`	Output	For Gen3 operation, specifies the block type. The following encodings are defined: 2'b01: Ordered Set Block 2'b10: Data Block

Test Signals

Table 48. Test Interface Signals. The `test_in` bus provides run-time control and monitoring of the internal state of the IP core.
Signal	Direction	Description
`test_in[31:0]`	Input	The bits of the `test_in` bus have the following definitions: [0]: Simulation mode. This signal can be set to 1 to accelerate initialization by reducing the value of many initialization counters. [1]: Reserved. Must be set to 1’b0 [2]: Descramble mode disable. This signal must be set to 1 during initialization in order to disable data scrambling. You can use this bit in simulation for Gen1 and Gen2 Endpoints and Root Ports to observe descrambled data on the link. Descrambled data cannot be used in open systems because the link partner typically scrambles the data. [4:3]: Reserved. Must be set to 2’b01. [5]: Compliance test mode. Disable/force compliance mode. When set, prevents the LTSSM from entering compliance mode. Toggling this bit controls the entry and exit from the compliance state, enabling the transmission of Gen1, Gen2 and Gen3 compliance patterns. [6]: Forces entry to compliance mode when a timeout is reached in the polling.active state and not all lanes have detected their exit condition. [7]: Disable low power state negotiation. Intel recommends setting this bit. [31:8]: Reserved. Set to all 0s. For more information about using the `test_in` to debug, refer to the Knowledge Base Solution How can I observe the Hard IP for PCI Express PIPE interface signals for Arria V GZ and Stratix V devices? in the Related Links below.
`simu_pipe_mode`	Input	When 1'b1, counter values are reduced to speed simulation.
`currentspeed[1:0]`	Output	Indicates the current speed of the PCIe link. The following encodings are defined: 2b’00: Undefined 2b’01: Gen1 2b’10: Gen2 2b’11: Gen3

Registers

Correspondence between Configuration Space Registers and the PCIe Specification

Table 49. Address Map of Hard IP Configuration Space Registers. For the Type 0 and Type 1 Configuration Space Headers, the first line of each entry lists Type 0 values and the second line lists Type 1 values when the values differ.
Byte Address	Hard IP Configuration Space Register	Corresponding Section in PCIe Specification
0x000:0x03C	PCI Header Type 0 Configuration Registers	Type 0 Configuration Space Header
0x000:0x03C	PCI Header Type 1 Configuration Registers	Type 1 Configuration Space Header The Type 1 Configuration Space is not available for the Avalon-MM with DMA interface
0x040:0x04C	Reserved	N/A
0x050:0x05C	MSI Capability Structure	MSI Capability Structure
0x068:0x070	MSI-X Capability Structure	MSI-X Capability Structure
0x070:0x074	Reserved	N/A
0x078:0x07C	Power Management Capability Structure	PCI Power Management Capability Structure
0x080:0x0BC	PCI Express Capability Structure	PCI Express Capability Structure
0x0C0:0x0FC	Reserved	N/A
0x100:0x16C	Virtual Channel Capability Structure	Virtual Channel Capability
0x170:0x17C	Reserved	N/A
0x180:0x1FC	Virtual channel arbitration table	VC Arbitration Table
0x200:0x23C	Port VC0 arbitration table	Port Arbitration Table
0x240:0x27C	Port VC1 arbitration table	Port Arbitration Table
0x280:0x2BC	Port VC2 arbitration table	Port Arbitration Table
0x2C0:0x2FC	Port VC3 arbitration table	Port Arbitration Table
0x300:0x33C	Port VC4 arbitration table	Port Arbitration Table
0x340:0x37C	Port VC5 arbitration table	Port Arbitration Table
0x380:0x3BC	Port VC6 arbitration table	Port Arbitration Table
0x3C0:0x3FC	Port VC7 arbitration table	Port Arbitration Table
0x400:0x7FC	Reserved	PCIe spec corresponding section name
0x800:0x834	Advanced Error Reporting AER (optional)	Advanced Error Reporting Capability
0x838:0xFFF	Reserved	N/A
Overview of Configuration Space Register Fields
0x000	Device ID, Vendor ID	Type 0 Configuration Space Header Type 1 Configuration Space Header The Type 1 Configuration Space is not available for the Avalon-MM with DMA interface
0x004	Status, Command	Type 0 Configuration Space Header Type 1 Configuration Space Header The Type 1 Configuration Space is not available for the Avalon-MM with DMA interface
0x008	Class Code, Revision ID	Type 0 Configuration Space Header Type 1 Configuration Space Header The Type 1 Configuration Space is not available for the Avalon-MM with DMA interface
0x00C	BIST, Header Type, Primary Latency Timer, Cache Line Size	Type 0 Configuration Space Header Type 1 Configuration Space Header The Type 1 Configuration Space is not available for the Avalon-MM with DMA interface
0x010	Base Address 0	Base Address Registers
0x014	Base Address 1	Base Address Registers
0x018	Base Address 2 Secondary Latency Timer, Subordinate Bus Number, Secondary Bus Number, Primary Bus Number	Base Address Registers Secondary Latency Timer, Type 1 Configuration Space Header, Primary Bus Number
0x01C	Base Address 3 Secondary Status, I/O Limit, I/O Base	Base Address Registers Secondary Status Register ,Type 1 Configuration Space Header
0x020	Base Address 4 Memory Limit, Memory Base	Base Address Registers Type 1 Configuration Space Header
0x024	Base Address 5 Prefetchable Memory Limit, Prefetchable Memory Base	Base Address Registers Prefetchable Memory Limit, Prefetchable Memory Base
0x028	Reserved Prefetchable Base Upper 32 Bits	N/A Type 1 Configuration Space Header
0x02C	Subsystem ID, Subsystem Vendor ID Prefetchable Limit Upper 32 Bits	Type 0 Configuration Space Header Type 1 Configuration Space Header
0x030	I/O Limit Upper 16 Bits, I/O Base Upper 16 Bits	Type 0 Configuration Space Header Type 1 Configuration Space Header
0x034	Reserved, Capabilities PTR	Type 0 Configuration Space Header Type 1 Configuration Space Header
0x038	Reserved	N/A
0x03C	Interrupt Pin, Interrupt Line Bridge Control, Interrupt Pin, Interrupt Line	Type 0 Configuration Space Header Type 1 Configuration Space Header
0x050	MSI-Message Control Next Cap Ptr Capability ID	MSI and MSI-X Capability Structures
0x054	Message Address	MSI and MSI-X Capability Structures
0x058	Message Upper Address	MSI and MSI-X Capability Structures
0x05C	Reserved Message Data	MSI and MSI-X Capability Structures
0x068	MSI-X Message Control Next Cap Ptr Capability ID	MSI and MSI-X Capability Structures
0x06C	MSI-X Table Offset BIR	MSI and MSI-X Capability Structures
0x070	Pending Bit Array (PBA) Offset BIR	MSI and MSI-X Capability Structures
0x078	Capabilities Register Next Cap PTR Cap ID	PCI Power Management Capability Structure
0x07C	Data PM Control/Status Bridge Extensions Power Management Status & Control	PCI Power Management Capability Structure
0x080	PCI Express Capabilities Register Next Cap Ptr PCI Express Cap ID	PCI Express Capability Structure
0x084	Device Capabilities Register	PCI Express Capability Structure
0x088	Device Status Register Device Control Register	PCI Express Capability Structure
0x08C	Link Capabilities Register	PCI Express Capability Structure
0x090	Link Status Register Link Control Register	PCI Express Capability Structure
0x094	Slot Capabilities Register	PCI Express Capability Structure
0x098	Slot Status Register Slot Control Register	PCI Express Capability Structure
0x09C	Root Capabilities Register Root Control Register	PCI Express Capability Structure
0x0A0	Root Status Register	PCI Express Capability Structure
0x0A4	Device Capabilities 2 Register	PCI Express Capability Structure
0x0A8	Device Status 2 Register Device Control 2 Register	PCI Express Capability Structure
0x0AC	Link Capabilities 2 Register	PCI Express Capability Structure
0x0B0	Link Status 2 Register Link Control 2 Register	PCI Express Capability Structure
0x0B4:0x0BC	Reserved	PCI Express Capability Structure
0x800	Advanced Error Reporting Enhanced Capability Header	Advanced Error Reporting Enhanced Capability Header
0x804	Uncorrectable Error Status Register	Uncorrectable Error Status Register
0x808	Uncorrectable Error Mask Register	Uncorrectable Error Mask Register
0x80C	Uncorrectable Error Severity Register	Uncorrectable Error Severity Register
0x810	Correctable Error Status Register	Correctable Error Status Register
0x814	Correctable Error Mask Register	Correctable Error Mask Register
0x818	Advanced Error Capabilities and Control Register	Advanced Error Capabilities and Control Register
0x81C	Header Log Register	Header Log Register
0x82C	Root Error Command	Root Error Command Register
0x830	Root Error Status	Root Error Status Register
0x834	Error Source Identification Register Correctable Error Source ID Register	Error Source Identification Register

Type 0 Configuration Space Registers

Figure 25. Type 0 Configuration Space Registers - Byte Address Offsets and LayoutEndpoints store configuration data in the Type 0 Configuration Space. The Correspondence between Configuration Space Registers and the PCIe Specification lists the appropriate section of the PCI Express Base Specification that describes these registers.

Type 1 Configuration Space Registers

Figure 26. Type 1 Configuration Space Registers (Root Ports)

Note: Avalon-MM DMA for PCIe does not support Type 1 configuration space registers.

PCI Express Capability Structures

The layout of the most basic Capability Structures are provided below. Refer to the PCI Express Base Specification for more information about these registers.

Figure 27. MSI Capability Structure

Note: Refer to the Advanced Error Reporting Capability section for more details about the PCI Express AER Extended Capability Structure.

Intel-Defined VSEC Registers

Figure 28. VSEC RegistersThis extended capability structure supports Configuration via Protocol (CvP) programming and detailed internal error reporting.

Table 50. Intel‑Defined VSEC Capability Register, 0x200. The Intel‑Defined Vendor Specific Extended Capability. This extended capability structure supports Configuration via Protocol (CvP) programming and detailed internal error reporting.
Bits	Register Description	Value	Access
[15:0]	PCI Express Extended Capability ID. Intel-defined value for VSEC Capability ID.	0x000B	RO
[19:16]	Version. Intel-defined value for VSEC version.	0x1	RO
[31:20]	Next Capability Offset. Starting address of the next Capability Structure implemented, if any.	Variable	RO

Table 51. Intel‑Defined Vendor Specific Header . You can specify these values when you instantiate the Hard IP. These registers are read‑only at run‑time.
Bits	Register Description	Value	Access
[15:0]	`VSEC ID`. A user configurable VSEC ID.	User entered	RO
[19:16]	`VSEC Revision`. A user configurable VSEC revision.	Variable	RO
[31:20]	`VSEC Length`. Total length of this structure in bytes.	0x044	RO

Table 52. Intel Marker Register
Bits	Register Description	Value	Access
[31:0]	`Intel Marker`. This read only register is an additional marker. If you use the standard Intel Programmer software to configure the device with CvP, this marker provides a value that the programming software reads to ensure that it is operating with the correct VSEC.	A Device Value	RO

Table 53. JTAG Silicon ID Register
Bits	Register Description	Value	Access
[127:96]	`JTAG Silicon ID DW3`	Application Specific	RO
[95:64]	`JTAG Silicon ID DW2`	Application Specific	RO
[63:32]	`JTAG Silicon ID DW1`	Application Specific	RO
[31:0]	`JTAG Silicon ID DW0`. This is the JTAG Silicon ID that CvP programming software reads to determine that the correct SRAM object file (.sof) is being used.	Application Specific	RO

Table 54. User Device or Board Type ID Register
Bits	Register Description	Value	Access
[15:0]	Configurable device or board type ID to specify to CvP the correct .sof.	Variable	RO

CvP Registers

Table 55. CvP Status . The `CvP Status` register allows software to monitor the CvP status signals.
Bits	Register Description	Reset Value	Access
[31:26]	Reserved	0x00	RO
[25]	`PLD_CORE_READY`. From FPGA fabric. This status bit is provided for debug.	Variable	RO
[24]	`PLD_CLK_IN_USE`. From clock switch module to fabric. This status bit is provided for debug.	Variable	RO
[23]	`CVP_CONFIG_DONE`. Indicates that the FPGA control block has completed the device configuration via CvP and there were no errors.	Variable	RO
[22]	Reserved	Variable	RO
[21]	`USERMODE`. Indicates if the configurable FPGA fabric is in user mode.	Variable	RO
[20]	`CVP_EN`. Indicates if the FPGA control block has enabled CvP mode.	Variable	RO
[19]	`CVP_CONFIG_ERROR`. Reflects the value of this signal from the FPGA control block, checked by software to determine if there was an error during configuration.	Variable	RO
[18]	`CVP_CONFIG_READY`. Reflects the value of this signal from the FPGA control block, checked by software during programming algorithm.	Variable	RO
[17:0]	Reserved	Variable	RO

Table 56. CvP Mode Control . The `CvP Mode Control` register provides global control of the CvP operation.
Bits	Register Description	Reset Value	Access
[31:16]	Reserved.	0x0000	RO
[15:8]	`CVP_NUMCLKS`. This is the number of clocks to send for every CvP data write. Set this field to one of the values below depending on your configuration image: 0x01 for uncompressed and unencrypted images 0x04 for uncompressed and encrypted images 0x08 for all compressed images	0x00	RW
[7:3]	Reserved.	0x0	RO
[2]	`CVP_FULLCONFIG`. Request that the FPGA control block reconfigure the entire FPGA including the V-Series Hard IP for PCI Express, bring the PCIe link down.	1’b0	RW
[1]	`HIP_CLK_SEL`. Selects between PMA and fabric clock when `USER_MODE` = 1 and `PLD_CORE_READY` = 1. The following encodings are defined: 1: Selects internal clock from PMA which is required for `CVP_MODE`. 0: Selects the clock from soft logic fabric. This setting should only be used when the fabric is configured in `USER_MODE` with a configuration file that connects the correct clock. To ensure that there is no clock switching during CvP, you should only change this value when the Hard IP for PCI Express has been idle for 10 µs and wait 10 µs after changing this value before resuming activity.	1’b0	RW
[0]	`CVP_MODE`. Controls whether the IP core is in `CVP_MODE` or normal mode. The following encodings are defined: 1:`CVP_MODE` is active. Signals to the FPGA control block active and all TLPs are routed to the Configuration Space. This `CVP_MODE` cannot be enabled if `CVP_EN` = 0. 0: The IP core is in normal mode and TLPs are routed to the FPGA fabric.	1’b0	RW

Table 57. CvP Data Registers .
The following table defines the `CvP Data` registers. For 64-bit data, the optional `CvP Data2` stores the upper 32 bits of data. Programming software should write the configuration data to these registers. If you Every write to these register sets the data output to the FPGA control block and generates `<n>` clock cycles to the FPGA control block as specified by the `CVP_NUM_CLKS` field in the `CvP Mode` `Control` register. Software must ensure that all bytes in the memory write dword are enabled. You can access this register using configuration writes, alternatively, when in CvP mode, these registers can also be written by a memory write to any address defined by a memory space BAR for this device. Using memory writes should allow for higher throughput than configuration writes.
Bits	Register Description	Reset Value	Access
[31:0]	Upper 32 bits of configuration data to be transferred to the FPGA control block to configure the device. You can choose 32- or 64-bit data.	0x00000000	RW
[31:0]	Lower 32 bits of configuration data to be transferred to the FPGA control block to configure the device.	0x00000000	RW

Table 58. CvP Programming Control Register . This register is written by the programming software to control CvP programming.
Bits	Register Description	Reset Value	Access
[31:2]	Reserved.	0x0000	RO
[1]	`START_XFER`. Sets the CvP output to the FPGA control block indicating the start of a transfer.	1’b0	RW
[0]	`CVP_CONFIG`. When asserted, instructs that the FPGA control block begin a transfer via CvP.	1’b0	RW

Advanced Error Reporting Capability

Uncorrectable Internal Error Mask Register

Table 59. Uncorrectable Internal Error Mask Register . The `Uncorrectable Internal Error Mask` register controls which errors are forwarded as internal uncorrectable errors. With the exception of the configuration error detected in CvP mode, all of the errors are severe and may place the device or PCIe link in an inconsistent state. The configuration error detected in CvP mode may be correctable depending on the design of the programming software. The access code RWS stands for Read Write Sticky meaning the value is retained after a soft reset of the IP core.
Bits	Register Description	Reset Value	Access
[31:12]	Reserved.	1b’0	RO
[11]	Mask for RX buffer posted and completion overflow error.	1b’0	RWS
[10]	Reserved	1b’1	RO
[9]	Mask for parity error detected on Configuration Space to TX bus interface.	1b’1	RWS
[8]	Mask for parity error detected on the TX to Configuration Space bus interface.	1b’1	RWS
[7]	Mask for parity error detected at TX Transaction Layer error.	1b’1	RWS
[6]	Reserved	1b’1	RO
[5]	Mask for configuration errors detected in CvP mode.	1b’0	RWS
[4]	Mask for data parity errors detected during TX Data Link LCRC generation.	1b’1	RWS
[3]	Mask for data parity errors detected on the RX to Configuration Space Bus interface.	1b’1	RWS
[2]	Mask for data parity error detected at the input to the RX Buffer.	1b’1	RWS
[1]	Mask for the retry buffer uncorrectable ECC error.	1b’1	RWS
[0]	Mask for the RX buffer uncorrectable ECC error.	1b’1	RWS

Uncorrectable Internal Error Status Register

Table 60. Uncorrectable Internal Error Status Register . This register reports the status of the internally checked errors that are uncorrectable. When specific errors are enabled by the `Uncorrectable Internal Error Mask` register, they are handled as Uncorrectable Internal Errors as defined in the *PCI Express Base Specification 3.0*. This register is for debug only. It should only be used to observe behavior, not to drive custom logic. The access code RW1CS represents Read Write 1 to Clear Sticky.
Bits	Register Description	Access
[31:12]	Reserved.	RO
[11]	When set, indicates an RX buffer overflow condition in a posted request or Completion	RW1CS
[10]	Reserved.	RO
[9]	When set, indicates a parity error was detected on the Configuration Space to TX bus interface	RW1CS
[8]	When set, indicates a parity error was detected on the TX to Configuration Space bus interface	RW1CS
[7]	When set, indicates a parity error was detected in a TX TLP and the TLP is not sent.	RW1CS
[6]	When set, indicates that the Application Layer has detected an uncorrectable internal error.	RW1CS
[5]	When set, indicates a configuration error has been detected in CvP mode which is reported as uncorrectable. This bit is set whenever a `CVP_CONFIG_ERROR` rises while in `CVP_MODE`.	RW1CS
[4]	When set, indicates a parity error was detected by the TX Data Link Layer.	RW1CS
[3]	When set, indicates a parity error has been detected on the RX to Configuration Space bus interface.	RW1CS
[2]	When set, indicates a parity error was detected at input to the RX Buffer.	RW1CS
[1]	When set, indicates a retry buffer uncorrectable ECC error.	RW1CS
[0]	When set, indicates a RX buffer uncorrectable ECC error.	RW1CS

Correctable Internal Error Mask Register

Table 61. Correctable Internal Error Mask Register. The `Correctable Internal Error Mask` register controls which errors are forwarded as Internal Correctable Errors. This register is for debug only.
Bits	Register Description	Reset Value	Access
[31:8]	Reserved.	0	RO
[7]	Reserved.	1	RO
[6]	Mask for Corrected Internal Error reported by the Application Layer.	1	RWS
[5]	Mask for configuration error detected in CvP mode.	1	RWS
[4:2]	Reserved.	0	RO
[1]	Mask for retry buffer correctable ECC error.	1	RWS
[0]	Mask for RX Buffer correctable ECC error.	1	RWS

Correctable Internal Error Status Register

Table 62. Correctable Internal Error Status Register . The `Correctable Internal Error Status` register reports the status of the internally checked errors that are correctable. When these specific errors are enabled by the `Correctable Internal Error Mask`register, they are forwarded as Correctable Internal Errors as defined in the *PCI Express Base Specification 3.0*. This register is for debug only. Only use this register to observe behavior, not to drive logic custom logic.
Bits	Register Description	Access
[31:7]	Reserved.	RO
[6]	Corrected Internal Error reported by the Application Layer.	RW1CS
[5]	When set, indicates a configuration error has been detected in CvP mode which is reported as correctable. This bit is set whenever a `CVP_CONFIG_ERROR` occurs while in `CVP_MODE`.	RW1CS
[4:2]	Reserved.	RO
[1]	When set, the retry buffer correctable ECC error status indicates an error.	RW1CS
[0]	When set, the RX buffer correctable ECC error status indicates an error.	RW1CS

DMA Descriptor Controller Registers

The DMA Descriptor Controller manages Read and Write DMA operations. The DMA Descriptor Controller is available for use with Endpoint variations. The Descriptor Controller supports up to 128 descriptors each for Read and Write Data Movers. Read and Write are from the perspective of the FPGA. A read is from PCIe^® address space to the FPGA Avalon^® -MM address space. A write is to PCIe^® address space from the FPGA Avalon^® -MM space.

You program the Descriptor Controller internal registers with the location and size of the descriptor table residing in the PCIe^® address space. The DMA Descriptor Controller instructs the Read Data Mover to copy the table to its own internal FIFO. When the DMA Descriptor Controller is instantiated as a separate component, it drives table entries on the RdDmaRxData_i[159:0] and WrDmaRxData_i[159:0] buses. When the DMA Descriptor Controller is embedded inside the Avalon-MM DMA bridge, it drives this information on internal buses. .

The Read Data Mover transfers data from the PCIe address space to Avalon-MM address space. It issues memory read TLPs on the PCIe link. It writes the data returned to a location in the Avalon^® -MM address space. The source address is the address for the data in the PCIe address space. The destination address is in the Avalon-MM address space.

The Write Data Mover reads data from the Avalon-MM address space and writes to the PCIe address space. It issues memory write TLPs on the PCIe link. The source address is in the Avalon-MM address space. The destination address is in the PCIe address space.

The DMA Descriptor Controller records the completion status for read and write descriptors in separate status tables. Each table has 128 consecutive DWORD entries that correspond to the 128 descriptors. The actual descriptors are stored immediately after the status entries at offset 0x200 from the values programmed into the RC Read Descriptor Base and RC Write Descriptor Base registers. The status and descriptor table must be located on a 32-byte boundary in Root Complex memory.

The Descriptor Controller writes a 1 to the done bit of the status DWORD to indicate successful completion. The Descriptor Controller also sends an MSI interrupt for the final descriptor. After receiving this MSI, host software can poll the done bit to determine status. The status table precedes the descriptor table in memory. The Descriptor Controller does not write the done bit nor send an MSI as each descriptor completes. It only writes the done bit or sends an MSI for the descriptor whose ID is stored in the RD_DMA_LAST PTR or WR_DMA_LAST_PTR registers. The Descriptor Controller supports out-of-order completions. Consequently, it is possible for the done bit to be set before all descriptors have completed.

Note: The following example clarifies this programming model. If 128 descriptors are specified and all of them execute, then descriptor ID 127 is written to the RD_DMA_LAST_PTR or WR_DMA_LAST_PTR register to start the DMA. The DMA Descriptor Controller only writes the done bit when descriptor 127 completes. To get intermediate status updates, host software should write multiple IDs into the last pointer register. For example, to get an intermediate status update when half of the 128 read descriptors have completed, host software should complete the following sequence:

Program the RD_DMA_LAST_PTR = 63.
Program the RD_DMA_LAST_PTR = 127.
Poll the status DWORD for read descriptor 63.
Poll the status DWORD for read descriptor 127.

Many commercial system Root Ports return out-of-order Read Completions based on optimized accesses to host memory channels. Consequently, the done status stored for descriptor <n> does not necessarily mean that descriptors <n-1> and <n -2> have also completed. You must request the completion status for every descriptor by writing the descriptor ID for every descriptor to RD_DMA_LAST_PTR or WR_DMA_LAST_PTR.

Note: Because the DMA Descriptor Controller uses FIFOs to store descriptor table entries, you cannot reprogram the DMA Descriptor Controller once it begins the transfers specified in the descriptor table.

Read DMA Descriptor Controller Registers

The following table describes the registers in the internal DMA Descriptor Controller. When the DMA Descriptor Controller is externally instantiated, these registers are accessed through a BAR. The offsets must be added to the base address for the read controller. When the Descriptor Controller is internally instantiated these registers are accessed through BAR0. The read controller is at offset 0x0000.

Address Offset	Register	Access	Description
0x0000	`RC Read Status and Descriptor Base (Low)`	R/W	Specifies the lower 32-bits of the base address of the read status and descriptor table in the Root Complex memory. This address must be on a 32-byte boundary. Software must program this register after programming the upper 32 bits at offset 0x4. To change the `RC Read Status and Descriptor Base (Low)`base address, all descriptors specified by the `RD_TABLE_SIZE` must be exhausted.
0x0004	`RC Read Status and Descriptor Base (High)`	R/W	Specifies the upper 32-bits of the base address of the read status and descriptor table in the Root Complex memory. Software must program this register before programming the lower 32 bits of this register.
0x0008	`EP Read Descriptor FIFO Base (Low)`	RW	Specifies the lower 32 bits of the base address of the read descriptor FIFO in Endpoint memory. The Read DMA fetches the descriptors from Root Complex memory. The address must be the Avalon-MM address of the Descriptor Controller's Read Descriptor Table Avalon-MM Slave Port as seen by the Read DMA Avalon-MM Master Port. You must program this register after programming the upper 32 bits at offset 0xC.
0x000C	`EP Read Descriptor FIFO Base (High)`	RW	Specifies the upper 32 bits of the base address of the read descriptor table in Endpoint Avalon-MM memory. The Read DMA fetches the descriptors from Root Complex memory and writes the descriptors to the FIFO at this location. This must be the Avalon-MM address of the descriptor controller's Read Descriptor Table Avalon-MM Slave Port as seen by the Read DMA Avalon-MM Master Port. You must program this register before programming the lower 32 bits of this register.
0x0010	`RD_DMA_LAST_PTR`	RW	When read, returns the ID of the last descriptor requested. If no DMA request is outstanding or the DMA is in reset, returns a value 0xFF. When written, specifies the ID of the last descriptor requested. The difference between the value read and the value written is the number of descriptors to be processed. For example, if the value reads 4, the last descriptor requested is 4. To specify 5 more descriptors, software should write a 9 into the `RD_DMA_LAST_PTR` register. The DMA executes 5 more descriptors. To have the read DMA record the `done` bit of every descriptor, program this register to transfer one descriptor at a time. The descriptor ID loops back to 0 after reaching `RD_TABLE_SIZE`. For example, if the `RD_DMA_LAST_PTR` value read is 126 and you want to execute three more descriptors, software must write 127, and then 1 into the `RD_DMA_LAST_PTR`register.
0x0014	`RD_TABLE_SIZE`	RW	Specifies the size of the Read descriptor table. Set to the number of descriptors - 1. By default, `RD_TABLE_SIZE` is set to 127.This value specifies the last `Descriptor ID`. To change the `RC Read Status and Descriptor Base (Low)`base address, all descriptors specified by the `RD_TABLE_SIZE` must be exhausted.
0x0018	`RD_CONTROL`	RW	[31:1] Reserved. [0]`Done`. When set, the Descriptor Controller writes the `Done` bit for each descriptor in the status table. When not set the Descriptor Controller writes the `Done` for the final descriptor, as specified by `RD_DMA_LAST_PTR`. In both cases, the Descriptor Controller sends a MSI to the host after the completion of the last descriptor along with the status update for the last descriptor.

Write DMA Descriptor Controller Registers

The following table describes the registers in the internal DMA Descriptor Controller. When the DMA Descriptor Controller is externally instantiated, these registers are accessed through a BAR. The offsets must be added to the base address for the write controller. When the Descriptor Controller is internally instantiated these registers are accessed through BAR0. The write controller is at offset 0x0100.

Address Offset	Register	Access	Description
0x0100	`RC Write Status and Descriptor Base (Low)`	R/W	Specifies the lower 32-bits of the base address of the write status and descriptor table in the Root Complex memory. This address must be on a 32-byte boundary. Software must program this register after programming the upper 32-bit register at offset 0x104. To change the `RC Write Status and Descriptor Base (Low)` base address, all descriptors specified by the `WR_TABLE_SIZE` must be exhausted.
0x0104	`RC Write Status and Descriptor Base (High)`	R/W	Specifies the upper 32-bits of the base address of the write status and descriptor table in the Root Complex memory. Software must program this register before programming the lower 32-bit register at offset 0x100.
0x0108	`EP Write Status and Descriptor FIFO Base (Low)`	RW	Specifies the lower 32 bits of the base address of the write descriptor table in Endpoint memory. The Write Descriptor Controller requests descriptors from Root Complex memory and writes the descriptors to this location. The address is the Avalon-MM address of the Descriptor Controller's Write Descriptor Table Avalon-MM Slave Port as seen by the Read DMA Avalon-MM Master Port. Software must program this register after programming the upper 32-bit register at offset 0x10C.
0x010C	`EP Write Status and Descriptor FIFO Base (High)`	RW	Specifies the upper 32 bits of the base address of the write descriptor table in Endpoint memory. The read DMA fetches the table from Root Complex memory and writes the table to this location. Software must program this register before programming the lower 32-bit register at offset 0x108.
0x0110	`WR_DMA_LAST_PTR`	RW	When read, returns the ID of the last descriptor requested. If no DMA request is outstanding or the DMA is in reset, returns a value 0xFF. When written, specifies the ID of the last descriptor requested. The difference between the value read and the value written is the number of descriptors to be processed. For example, if the value reads 4, the last descriptor requested is 4. To specify 5 more descriptors, software should write a 9 into the `RD_DMA_LAST_PTR` register. The DMA executes 5 more descriptors. To have the write DMA record the `done` bit of every descriptor, program this register to transfer one descriptor at a time. The `Descriptor ID` loops back to 0 after reaching `WR_TABLE_SIZE`. For example, if the `WR_DMA_LAST_PTR` value read is 126 and you want to execute three more descriptors, software must write 127, and then 1 into the `WR_DMA_LAST_PTR` register.
0x0114	`WR_TABLE_SIZE`	RW	Specifies the size of the Read descriptor table. Set to the number of descriptors - 1. By default, `WR_TABLE_SIZE` is set to 127. This value specifies the last `Descriptor ID`. To change the `RC Write Status and Descriptor Base (Low)` base address, all descriptors specified by the `WR_TABLE_SIZE` must be exhausted.
0x0118	`WR_CONTROL`	RW	[31:1] Reserved. [0]`Done`. When set, the Descriptor Controller writes the `Done` bit for each descriptor in the status table. The Descriptor Controller sends a single MSI interrupt after the final descriptor completes. When not set the Descriptor Controller generates `Done` for the final descriptor, as specified by `WR_DMA_LAST_PTR`. In both cases, the Descriptor Controller sends an MSI to the host after the completion of the last descriptor along with the status update for the last descriptor.

Read DMA and Write DMA Descriptor Format

Read and write descriptors are stored in separate descriptor tables in PCIe^® system memory. Each table can store up to 128 descriptors. Each descriptor is 8 DWORDs, or 32 bytes. The Read DMA and Write DMA descriptor tables start at a 0x200 byte offset from the addresses programmed into the Read Status and Descriptor Base and Write Status and Descriptor Base address registers.

Programming RD_DMA_LAST_PTR or WR_DMA_LAST_PTR registers triggers the Read or Write Descriptor Controller descriptor table fetch process. Consequently, writing these registers must be the last step in setting up DMA transfers.

Table 63. Read Descriptor Format. You must also use this format for the Read and Write Data Movers on their Avalon^® -ST when you use your own DMA Controller.
Address Offset	Register Name	Description
0x00	`RD_LOW_SRC_ADDR`	Lower DWORD of the read DMA source address. Specifies the address in PCIe^® system memory from which the Read Data Mover fetches data.
0x04	`RD_HIGH_SRC_ADDR`	Upper DWORD of the read DMA source address. Specifies the address in PCIe^® system memory from which the Read Data Mover fetches data.
0x08	`RD_CTRL_LOW_DEST_ADDR`	Lower DWORD of the read DMA destination address. Specifies the address in the Avalon-MM domain to which the Read Data Mover writes data.
0x0C	`RD_CTRL_HIGH_DEST_ADDR`	Upper DWORD of the read DMA destination address. Specifies the address in the Avalon-MM domain to which the Read Data Mover writes data.
0x10	`CONTROL`	Specifies the following information: [31:25] Reserved must be 0. [24:18] `ID`. Specifies the `Descriptor ID`. `Descriptor ID` 0 is at the beginning of the table. For descriptor tables of the maximum size, `Descriptor ID` 127 is at the end of the table. [17:0] `SIZE`. The transfer size in DWORDs. Must be non-zero. The maximum transfer size is (1 MB - 4 bytes). If the specified transfer size is less than the maximum, the transfer size is the actual size entered.
0x14 - 0x1C	Reserved	N/A

Table 64. Write Descriptor Format
Address Offset	Register Name	Description
0x00	`WR_LOW_SRC_ADDR`	Lower DWORD of the write DMA source address. Specifies the address in the Avalon^® MM domain from which the Write Data Mover fetches data.
0x04	`WR_HIGH_SRC_ADDR`	Upper DWORD of the write DMA source address. Specifies the address in the Avalon^® MM domain from which the Write Data Mover fetches data.
0x08	`WR_CTRL_LOW_DEST_ADDR`	Lower DWORD of the Write Data Mover destination address. Specifies the address in PCIe^® system memory to which the Write DMA writes data.
0x0C	`WR_CTRL_HIGH_DEST_ADDR`	Upper DWORD of the write DMA destination address. Specifies the address in PCIe^® system memory to which the Write Data Mover writes data.
0x10	`CONTROL`	Specifies the following information: [31:25]: Reserved must be 0. [24:18]:`ID`: Specifies the `Descriptor ID`. `Descriptor ID` 0 is at the beginning of the table. `Descriptor ID` is at the end of the table. [17:0] :`SIZE`: The transfer size in DWORDs. Must be non-zero. The maximum transfer size is (1 MB - 4 bytes). If the specified transfer size is less than the maximum, the transfer size is the actual size entered.
0x14 - 0x1C	Reserved	N/A

Read DMA Example

This example moves three data blocks from the system memory to the Avalon-MM address space. Host software running on an embedded CPU allocates the memory and creates the descriptor table in system memory.

This example uses the addresses in the Platform Designer design example, ep_g3x8_avmm256_integrated.qsys, available in the <install_dir>/ ip/altera/altera_pcie/altera_pcie_<dev>_ed/example_design/<dev> directory.

The following figures illustrate the location and size of the data blocks in the PCIe and Avalon-MM address spaces and the descriptor table format. In this example, the value of RD_TABLE_SIZE is 127.

Figure 29. Data Blocks to Transfer from PCIe to Avalon-MM Address Space Using Read DMA

Assume the descriptor table includes 128 entries. The status table precedes a variable number of descriptors in memory. The Read and Write Status and Descriptor Tables are at the address specified in the RC Read Descriptor Base Register and RC Write Descriptor Base Register, respectively.

Figure 30. Descriptor Table Format

Calculate the memory allocation required:
1. Each entry in the status table is 4 bytes. The 128 entries require 512 bytes of memory.
2. Each descriptor is 32 bytes. The three descriptors require 96 bytes of memory.
The total memory allocation for the status and descriptor tables is 608 bytes.
Allocate 608 bytes of memory in the PCI Express address space.
The start address of the allocated memory in this example is 0xF000_0000. Program this address into the Root Complex Read Status and Descriptor Base Address Registers.
Create the descriptor table in the PCI Express address space. Because the status table is stored before the descriptors, the first descriptor is stored at 0xF000_0000 + 0x200 = 0xF000_0200.
1. Program 0x0000_0000 into the source address 0xF000_0204 in descriptor 0.
  This is the upper 32 bits of the source address.
2. Program 0x1000_0000 into the source address 0xF000_0200 in descriptor 0.
  This is the lower 32 bits of the source address.
3. Program 0x0000_0000 into the destination address 0xF000_020C in descriptor 0.
  This is the upper 32 bits of the destination address.
4. Program 0x5000_0000 into the destination address 0xF000_0208 in descriptor 0.
  This is the lower 32 bits of the destination address.
  
  These four steps program the Avalon-MM destination address for the 64 KB block of memory into the Descriptor Table.
5. Program 0x0000_4000 to 0xF000_0210 to transfer 16K dwords (64 KB), of data for descriptor ID 0.
Repeat this procedure for the second data block:
1. Program 0x0000_0000 to source address 0xF000_0224.
2. Program 0x2000_0000 to source address 0xF000_0220.
3. Program 0x0000_0000 to destination address 0xF000_022C.
4. Program 0x0001_0000 to destination address 0xF000_0228.
5. Program 0x0004_2000 to 0xF000_0230 to transfer 8K dwords (32 KB) of data for descriptor ID 1.
Repeat this procedure for the third data block:
1. Program 0x0000_0001 to source address 0xF000_0244.
2. Program 0x2000_0000 to source address 0xF000_0240.
3. Program 0x0000_0000 to destination address 0xF000_024C.
4. Program 0x1000_0000 to destination address 0xF000_0248.
5. Program 0x0008_1000 to 0xF000_0250 to transfer 4K dwords (16 KB) of data for descriptor ID 2.
The following figure shows the values in the Descriptor Table after programming completes.
Figure 31. Descriptor Table Format
Program the DMA Descriptor Controller with the address of the status and descriptor table in the PCI Express system memory address space. When the DMA Descriptor Controller is internal, these registers are accessed through combined BAR0 and BAR1 because this example uses 64-bit addresses. The DMA read control registers start at offset 0x0000. The Write DMA control registers start at offset 0x0100
1. Program 0x0000_0000 to offset 0x0000_0004.
  This is the upper 32 bits of the PCIe system memory where the status and descriptor table is stored.
2. Program 0xF000_0000 to offset 0x0000_0000.
  This is the lower 32 bits of the address in PCIe memory that stores the status and descriptor tables. The Read DMA automatically adds an offset of 0x200 to this value to start the copy of the descriptors which follow the status table in memory.
Program the DMA Descriptor Controller with the on-chip FIFO address. This is the address to which the Descriptor Controller copies the status and descriptor table.
1. Program 0x0000_0000 to offset 0x0000_000C
  This is the upper 32 bits of the on-chip FIFO address in the Avalon-MM address domain.
2. Program 0x0100_0000 to offset 0x0000_0008.
  This is the lower 32 bits of the on-chip FIFO address. This is address of the internal on-chip FIFO that is a part of the Descriptor Controller as seen by the RX Master.
  Figure 32. Address of the On-Chip FIFO
Program the Descriptor Controller RD_DMA_LAST_PTR register.
This step starts the DMA. It also specifies the status dword to be updated when the three descriptors complete.
- To update a single done bit for the final descriptor, program 0x2 to offset 0x0000_0010. The Descriptor Controller processes all three descriptors and writes the done bit to 0xF000_0008 of the status table.
- To update the done bits for all three descriptors, program address 0x0000_0010 RD_DMA_LAST_PTR three times with the values 0, 1, and 2. The Descriptor Controller sets the done bits for addresses 0xF000_0000, 0xF000_0004, and 0xF000_0008. If the system returns Read Completions out-of-order, the Descriptor Controller may complete descriptors out of order. In such systems, you must use this method of requesting done status for each descriptor. Software must check for done status for every descriptor.

Software Program for Simultaneous Read and Write DMA

Program the following steps to implement a simultaneous DMA transfer:

Allocate PCIe^® system memory for the Read and Write DMA descriptor tables. If, for example, each table supports up to 128, eight-DWORD descriptors and 128, one-DWORD status entries for a total of 1152 DWORDs. Total memory for the Read and Write DMA descriptor tables is 2304 DWORDs.
Allocate PCIe^® system memory and initialize it with data for the Read Data Mover to read.
Allocate PCIe^® system memory for the Write Data Mover to write.
Create all the descriptors for the read DMA descriptor table. Assign the DMA Descriptor IDs sequentially, starting with 0 to a maximum of 127. For the read DMA, the source address is the memory space allocated in Step 2. The destination address is the Avalon‑MM address that the Read Data Mover module writes. Specify the DMA length in DWORDs. Each descriptor transfers contiguous memory. Assuming a base address of 0, for the Read DMA, the following assignments illustrate construction of a read descriptor:
1. RD_LOW_SRC_ADDR = 0x0000(The base address for the read descriptor table in the PCIe system memory.)
2. RD_HIGH_SRC_ADDR = 0x0004
3. RD_CTRL_LOW_DEST_ADDR 0x0008
4. RD_CTRL_HIGH_DEST_ADDR = 0x000C
5. RD_DMA_LAST_PTR = 0x0010
Writing the RD_DMA_LAST_PTR register starts operation.
For the Write DMA, the source address is the Avalon‑MM address that the Write Data Mover module should read. The destination address is the PCIe^® system memory space allocated in Step 3. Specify the DMA size in DWORDs. Assuming a base address of 0x100, for the Write Data Mover, the following assignments illustrate construction of a write descriptor:
1. WD_LOW_SRC_ADDR = 0x0100(The base address for the write descriptor table in the PCIe^® system memory.)
2. WD_HIGH_SRC_ADDR = 0x0104
3. WD_CTRL_LOW_DEST_ADDR 0x0108
4. WD_CTRL_HIGH_DEST_ADDR = 0x010C
5. WD_DMA_LAST_PTR = 0x0110
Writing the WD_DMA_LAST_PTR register starts operation.
To improve throughput, the Read DMA module copies the descriptor table to the Avalon-MM memory before beginning operation. Specify the memory address by writing to the Descriptor Table Base (Low) and (High) registers.
An MSI interrupt is sent for each WD_DMA_LAST_PTR or RD_DMA_LAST_PTR that completes. These completions result in updates to the done status bits. Host software can then read status bits to determine which DMA operations are complete.

Note: If the transfer size of the read DMA is greater than the maximum read request size, the Read DMA creates multiple read requests. For example, if maximum read request size is 512 bytes, the Read Data Mover breaks a 4 KB read request into 8 requests with 8 different tags. The Read Completions can come back in any order. The Read Data Mover's Avalon-MM master port writes the data received in the Read Completions to the correct locations in Avalon^® -MM memory, based on the tags in the same order as it receives the Completions. This order is not necessarily in increasing address order;. The data mover does not include an internal reordering buffer. If system allows out of order read completions, then status for the latest entry is latest only in number, but potentially earlier than other completions chronologically

Control Register Access (CRA) Avalon-MM Slave Port

Table 65. Configuration Space Register Descriptions . The optional CRA Avalon-MM slave port provides host access to selected Configuration Space and status registers. These registers are read only. For registers that are less than 32 bits, the upper bits are unused.
Byte Offset	Register	Access	Description
14'h0000	`cfg_dev_ctrl[15:0]`	RO	`cfg_devctrl[15:0]` is device control for the PCI Express capability structure.
14'h0004	`cfg_dev_ctrl2[15:0]`	RO	`cfg_dev2ctrl[15:0]` is device control 2 for the PCI Express capability structure.
14'h0008	`cfg_link_ctrl[15:0]`	RO	`cfg_link_ctrl[15:0]`is the primary Link Control of the PCI Express capability structure. For Gen2 or Gen3 operation, you must write a 1’b1 to Retrain Link bit (Bit[5] of the `cfg_link_ctrl)` of the Root Port to initiate retraining to a higher data rate after the initial link training to Gen1 L0 state. Retraining directs the LTSSM to the Recovery state. Retraining to a higher data rate is not automatic even if both devices on the link are capable of a higher data rate.
14'h000C	`cfg_link_ctrl2[15:0]`	RO	`cfg_link_ctrl2[31:16]` is the secondary Link Control register of the PCI Express capability structure for Gen2 operation. For Gen1 variants, the link bandwidth notification bit is always set to 0.For Gen2 variants, this bit is set to 1.
14'h0010	`cfg_prm_cmd[15:0]`	RO	Base/Primary Command register for the PCI Configuration Space.
14'h0014	`cfg_root_ctrl[7:0]`	RO	Root control and status register of the PCI-Express capability. This register is only available in Root Port mode.
14'h0018	`cfg_sec_ctrl[15:0]`	RO	Secondary bus Control and Status register of the PCI-Express capability. This register is only available in Root Port mode.
14'h001C	`cfg_secbus[7:0]`	RO	Secondary bus number. Available in Root Port mode.
14'h0020	`cfg_subbus[7:0]`	RO	Subordinate bus number. Available in Root Port mode.
14'h0024	`cfg_msi_addr_low[31:0]`	RO	`cfg_msi_add[31:0]` is the MSI message address.
14'h0028	`cfg_msi_addr_hi[63:32]`	RO	`cfg_msi_add[63:32]` is the MSI upper message address.
14'h002C	`cfg_io_bas[19:0]`	RO	The IO base register of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0030	`cfg_io_lim[19:0]`	RO	The IO limit register of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0034	`cfg_np_bas[11:0]`	RO	The non-prefetchable memory base register of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0038	`cfg_np_lim[11:0]`	RO	The non-prefetchable memory limit register of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h003C	`cfg_pr_bas_low[31:0]`	RO	The lower 32 bits of the prefetchable base register of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0040	`cfg_pr_bas_hi[43:32]`	RO	The upper 12 bits of the prefetchable base registers of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0044	`cfg_pr_lim_low[31:0]`	RO	The lower 32 bits of the prefetchable limit registers of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h0048	`cfg_pr_lim_hi[43:32]`	RO	The upper 12 bits of the prefetchable limit registers of the Type1 Configuration Space. This register is only available in Root Port mode.
14'h004C	`cfg_pmcsr[31:0]`	RO	`cfg_pmcsr[31:16]` is Power Management Control and `cfg_pmcsr[15:0]`is the Power Management Status register.
14'h0050	`cfg_msixcsr[15:0]`	RO	MSI-X message control register.
14'h0054	`cfg_msicsr[15:0]`	RO	MSI message control.
14'h0058	`cfg_tcvcmap[23:0]`	RO	Configuration traffic class (TC)/virtual channel (VC) mapping. The Application Layer uses this signal to generate a TLP mapped to the appropriate channel based on the traffic class of the packet. The following encodings are defined: `cfg_tcvcmap[2:0]`: Mapping for TC0 (always 0). `cfg_tcvcmap[5:3]`: Mapping for TC1. `cfg_tcvcmap[8:6]`: Mapping for TC2. `cfg_tcvcmap[11:9]`: Mapping for TC3. `cfg_tcvcmap[14:12]`: Mapping for TC4. `cfg_tcvcmap[17:15]`: Mapping for TC5. `cfg_tcvcmap[20:18]`: Mapping for TC6. `cfg_tcvcmap[23:21]`: Mapping for TC7.
14'h005C	`cfg_msi_data[15:0]`	RO	`cfg_msi_data[15:0]` is message data for MSI.
14'h0060	`cfg_busdev[12:0]`	RO	Bus/Device Number captured by or programmed in the Hard IP. The following fields are defined: cfg_busdev[12:5]: bus number cfg_busdev[4:0]: device number
14'h0064	`ltssm_reg[4:0]`	RO	Specifies the current LTSSM state. The LTSSM state machine encoding defines the following states: : 5'b: 00000: Detect.Quiet 5'b: 00001: Detect.Active 5'b: 00010: Polling.Active 5'b: 00011: Polling.Compliance 5'b: 00100: Polling.Configuration 5'b: 00101: Polling.Speed 5'b: 00110: config.Linkwidthstart 5'b: 00111: Config.Linkaccept 5'b: 01000: Config.Lanenumaccept 5'b: 01001: Config.Lanenumwait 5'b: 01010: Config.Complete 5'b: 01011: Config.Idle 5'b: 01100: Recovery.Rcvlock 5'b: 01101: Recovery.Rcvconfig 5'b: 01110: Recovery.Idle 5'b: 01111: L0 5'b: 10000: Disable 5'b: 10001: Loopback.Entry 5'b: 10010: Loopback.Active 5'b: 10011: Loopback.Exit 5'b: 10100: Hot.Reset 5'b: 10101: LOs 5'b: 11001: L2.transmit.Wake 5'b: 11010: Speed.Recovery 5'b: 11011: Recovery.Equalization, Phase 0 5'b: 11100: Recovery.Equalization, Phase 1 5'b: 11101: Recovery.Equalization, Phase 2 5'b: 11110: recovery.Equalization, Phase 3
14'h0068	`current_speed_reg[1:0]`	RO	Indicates the current speed of the PCIe link. The following encodings are defined: 2b’00: Undefined 2b’01: Gen1 2b’10: Gen2 2b’11: Gen3
14'h006C	`lane_act_reg[3:0]`	RO	Lane Active Mode: This signal indicates the number of lanes that configured during link training. The following encodings are defined: 4’b0001: 1 lane 4’b0010: 2 lanes 4’b0100: 4 lanes 4’b1000: 8 lanes

Reset and Clocks

V-Series Hard IP for PCI Express IP Core includes both a soft reset controller and a hard reset controller. Software selects the appropriate reset controller depending on the configuration you specify. Both reset controllers reset the IP core and provide sample reset logic in the example design. The figure below provides a simplified view of the logic that implements both reset controllers.

The pin_perst signal from the input pin of the FPGA resets the Hard IP for PCI Express IP Core. app_rstn which resets the Application Layer logic is derived from reset_status and pld_clk_inuse, which are outputs of the core. This reset controller is implemented in hardened logic. The figure below provides a simplified view of the logic that implements the reset controller.

Table 66. Use of Hard and Soft Reset Controllers
Reset Controller Used	Description
Hard Reset Controller	`pin_perst` from the input pin of the FPGA resets the Hard IP for PCI Express IP Core. `app_rstn`which resets the Application Layer logic is derived from `reset_status` and `pld_clk_inuse`, which are outputs of the core. This reset controller is supported for Gen 1 production devices.
Soft Reset Controller	Either `pin_perst` from the input pin of the FPGA or `npor`which is derived from `pin_perst` or `local_rstn` can reset the Hard IP for PCI Express IP Core. Application Layer logic generates the optional `local_rstn` signal. `app_rstn`which resets the Application Layer logic is derived from `npor`. This reset controller is supported for Gen2 and Gen3 production devices.

Figure 33. Reset Controller Block Diagram

Reset Sequence for Hard IP for PCI Express IP Core and Application Layer

Figure 34. Hard IP for PCI Express and Application Logic Reset SequenceYour Application Layer can instantiate a module with logic that implements the timing diagram shown below to generate app_rstn, which resets the Application Layer logic.

This reset sequence includes the following steps:

After pin_perst or npor is released, the Hard IP reset controller waits for pld_clk_inuse to be asserted.
csrt and srst are released 32 cycles after pld_clk_inuse is asserted.
The Hard IP for PCI Express deasserts the reset_status output to the Application Layer.
The altpcied_<device>v_hwtcl.sv deasserts app_rstn 32 pld_clkcycles after reset_status is released.

Figure 35. RX Transceiver Reset Sequence

The RX transceiver reset sequence includes the following steps:

After rx_pll_locked is asserted, the LTSSM state machine transitions from the Detect.Quiet to the Detect.Active state.
When the pipe_phystatus pulse is asserted and pipe_rxstatus[2:0]= 3, the receiver detect operation has completed.
The LTSSM state machine transitions from the Detect.Active state to the Polling.Active state.
The Hard IP for PCI Express asserts rx_digitalreset. The rx_digitalreset signal is deasserted after rx_signaldetect is stable for a minimum of 3 ms.

Figure 36. TX Transceiver Reset Sequence

The TX transceiver reset sequence includes the following steps:

After npor is deasserted, the IP core deasserts the npor_serdes input to the TX transceiver.
The SERDES reset controller waits for pll_locked to be stable for a minimum of 127 pld_clk cycles before deasserting tx_digitalreset.

For descriptions of the available reset signals, refer to Reset Signals, Status, and Link Training Signals.

Clocks

The Hard IP contains a clock domain crossing (CDC) synchronizer at the interface between the PHY/MAC and the DLL layers. The synchronizer allows the Data Link and Transaction Layers to run at frequencies independent of the PHY/MAC. The CDC synchronizer provides more flexibility for the user clock interface. Depending on parameters you specify, the core selects the appropriate coreclkout_hip. You can use these parameters to enhance performance by running at a higher frequency for latency optimization or at a lower frequency to save power.

In accordance with the PCI Express Base Specification, you must provide a 100 MHz reference clock that is connected directly to the transceiver.

As a convenience, you may also use a 125 MHz input reference clock as input to the TX PLL.

Clock Domains

Figure 37. Clock Domains and Clock Generation for the Application LayerThe following illustrates the clock domains when using coreclkout_hip to drive the Application Layer and the pld_clk of the IP core. The Intel-provided example design connects coreclkout_hip to the pld_clk. However, this connection is not mandatory. Inside the Hard IP for PCI Express^® , the blocks shown in white are in the pclk domain, while the blocks shown in yellow are in the coreclkout_hip domain.

As this figure indicates, the IP core includes the following clock domains: pclk, coreclkout_hip and pld_clk.

pclk

The transceiver derives pclk from the 100 MHz refclk signal that you must provide to the device.

The PCI Express Base Specification requires that the refclk signal frequency be 100 MHz ±300 PPM.

The transitions between Gen1, Gen2, and Gen3 should be glitchless. pclk can be turned off for most of the 1 ms timeout assigned for the PHY to change the clock rate; however, pclk should be stable before the 1 ms timeout expires.

The transitions between Gen1 and Gen2 should be glitchless. pclk can be turned off for most of the 1 ms timeout assigned for the PHY to change the clock rate; however, pclk should be stable before the 1 ms timeout expires.

Table 67. pclk Clock Frequency
Data Rate	Frequency
Gen1	250 MHz
Gen2	500 MHz

The CDC module implements the asynchronous clock domain crossing between the PHY/MAC pclk domain and the Data Link Layer coreclk domain. The transceiver pclk clock is connected directly to the Hard IP for PCI Express and does not connect to the FPGA fabric.

coreclkout_hip

Table 68. Application Layer Clock Frequency for All Combinations of Link Width, Data Rate and Application Layer Interface Widths. The `coreclkout_hip` signal is derived from `pclk`. The following table lists frequencies for `coreclkout_hip`, which are a function of the link width, data rate, and the width of the Application Layer to Transaction Layer interface. The frequencies and widths specified in this table are maintained throughout operation. If the link downtrains to a lesser link width or changes to a different maximum link rate, it maintains the frequencies it was originally configured for as specified in this table. (The Hard IP throttles the interface to achieve a lower throughput.)
Link Width	Max Link Rate	Avalon Interface Width	coreclkout_hip
×8	Gen1	128	125 MHz
×4	Gen2	128	125 MHz
×8	Gen2	128	250 MHz
×8	Gen2	256	125 MHz
×2	Gen3	128	125 MHz
×4	Gen3	128	250 MHz
×4	Gen3	256	125 MHz
×8	Gen3	256	250 MHz

pld_clk

coreclkout_hip can drive the Application Layer clock along with the pld_clk input to the IP core. The pld_clk can optionally be sourced by a different clock than coreclkout_hip. The pld_clk minimum frequency cannot be lower than the coreclkout_hip frequency. Based on specific Application Layer constraints, a PLL can be used to derive the desired frequency.

Note: For Gen3, Intel recommends using a common reference clock (0 ppm) because when using separate reference clocks (non 0 ppm), the PCS occasionally must insert SKP symbols, potentially causing the PCIe link to go to recovery. Gen1 or Gen2 modes are not affected by this issue. Systems using the common reference clock (0 ppm) are not affected by this issue. The primary repercussion of this issue is a slight decrease in bandwidth. On Gen3 x8 systems, this bandwidth impact is negligible. If non 0 ppm mode is required, so that separate reference clocks are used, please contact Intel for further information and guidance.

Clock Summary

Table 69. Clock Summary
Name	Frequency	Clock Domain
`coreclkout_hip`	62.5, 125 or 250 MHz	Avalon‑ST interface between the Transaction and Application Layers.
`pld_clk`	125 or 250 MHz	Application and Transaction Layers.
`refclk`	100 or 125 MHz	SERDES (transceiver). Dedicated free running input clock to the SERDES block.
`reconfig_xcvr_clk`	100 –125 MHz	Transceiver Reconfiguration Controller.

Error Handling

Each PCI Express compliant device must implement a basic level of error management and can optionally implement advanced error management. The IP core implements both basic and advanced error reporting. Error handling for a Root Port is more complex than that of an Endpoint.

Table 70. Error Classification. The *PCI Express Base Specification* defines three types of errors, outlined in the following table.
Type	Responsible Agent	Description
Correctable	Hardware	While correctable errors may affect system performance, data integrity is maintained.
Uncorrectable, non-fatal	Device software	Uncorrectable, non-fatal errors are defined as errors in which data is lost, but system integrity is maintained. For example, the fabric may lose a particular TLP, but it still works without problems.
Uncorrectable, fatal	System software	Errors generated by a loss of data and system failure are considered uncorrectable and fatal. Software must determine how to handle such errors: whether to reset the link or implement other means to minimize the problem.

Physical Layer Errors

Table 71. Errors Detected by the Physical Layer . The following table describes errors detected by the Physical Layer. Physical Layer error reporting is optional in the *PCI Express Base Specification*.
Error	Type	Description
Receive port error	Correctable	This error has the following 3 potential causes: Physical coding sublayer error when a lane is in L0 state. These errors are reported to the Hard IP block via the per lane PIPE interface input receive status signals, `rxstatus<lane_number>[2:0]` using the following encodings: 3'b100: 8B/10B Decode Error 3'b101: Elastic Buffer Overflow 3'b110: Elastic Buffer Underflow 3'b111: Disparity Error Deskew error caused by overflow of the multilane deskew FIFO. Control symbol received in wrong lane.

Data Link Layer Errors

Table 72. Errors Detected by the Data Link Layer
Error	Type	Description
Bad TLP	Correctable	This error occurs when a LCRC verification fails or when a sequence number error occurs.
Bad DLLP	Correctable	This error occurs when a CRC verification fails.
Replay timer	Correctable	This error occurs when the replay timer times out.
Replay num rollover	Correctable	This error occurs when the replay number rolls over.
Data Link Layer protocol	Uncorrectable(fatal)	This error occurs when a sequence number specified by the Ack/Nak block in the Data Link Layer (`AckNak_Seq_Num)` does not correspond to an unacknowledged TLP.

Transaction Layer Errors

Table 73. Errors Detected by the Transaction Layer
Error	Type	Description
Poisoned TLP received	Uncorrectable (non-fatal)	This error occurs if a received Transaction Layer Packet has the EP poison bit set. The received TLP is passed to the Application Layer and the Application Layer logic must take appropriate action in response to the poisoned TLP. Refer to “2.7.2.2 Rules for Use of Data Poisoning” in the PCI Express Base Specification for more information about poisoned TLPs.
ECRC check failed ⁽¹⁾	Uncorrectable (non-fatal)	This error is caused by an ECRC check failing despite the fact that the TLP is not malformed and the LCRC check is valid. The Hard IP block handles this TLP automatically. If the TLP is a non‑posted request, the Hard IP block generates a completion with completer abort status. In all cases the TLP is deleted in the Hard IP block and not presented to the Application Layer.
Unsupported Request for Endpoints	Uncorrectable (non-fatal)	This error occurs whenever a component receives any of the following Unsupported Requests: Type 0 Configuration Requests for a non-existing function. Completion transaction for which the Requester ID does not match the bus, device and function number. Unsupported message. A Type 1 Configuration Request TLP for the TLP from the PCIe link. A locked memory read (MEMRDLK) on native Endpoint. A locked completion transaction. A 64-bit memory transaction in which the 32 MSBs of an address are set to 0. A memory or I/O transaction for which there is no BAR match. A memory transaction when the Memory Space Enable bit (bit [1] of the PCI Command register at Configuration Space offset 0x4) is set to 0. A poisoned configuration write request (`CfgWr0`) In all cases the TLP is deleted in the Hard IP block and not presented to the Application Layer. If the TLP is a non-posted request, the Hard IP block generates a completion with Unsupported Request status.
Completion timeout	Uncorrectable (non-fatal)	This error occurs when a request originating from the Application Layer does not generate a corresponding completion TLP within the established time. It is the responsibility of the Application Layer logic to provide the completion timeout mechanism. The completion timeout should be reported from the Transaction Layer using the `cpl_err[0]` signal.
Completer abort ⁽¹⁾	Uncorrectable (non-fatal)	The Application Layer reports this error using the `cpl_err[2]`signal when it aborts receipt of a TLP.
Unexpected completion	Uncorrectable (non-fatal)	This error is caused by an unexpected completion transaction. The Hard IP block handles the following conditions: The Requester ID in the completion packet does not match the Configured ID of the Endpoint. The completion packet has an invalid tag number. (Typically, the tag used in the completion packet exceeds the number of tags specified.) The completion packet has a tag that does not match an outstanding request. The completion packet for a request that was to I/O or Configuration Space has a length greater than 1 dword. The completion status is Configuration Retry Status (CRS) in response to a request that was not to Configuration Space. In all of the above cases, the TLP is not presented to the Application Layer; the Hard IP block deletes it. The Application Layer can detect and report other unexpected completion conditions using the `cpl_err[2]` signal. For example, the Application Layer can report cases where the total length of the received successful completions do not match the original read request length.
Receiver overflow ⁽¹⁾	Uncorrectable (fatal)	This error occurs when a component receives a TLP that violates the FC credits allocated for this type of TLP. In all cases the hard IP block deletes the TLP and it is not presented to the Application Layer.
Flow control protocol error (FCPE) ⁽¹⁾	Uncorrectable (fatal)	This error occurs when a component does not receive update flow control credits with the 200 µs limit.
Malformed TLP	Uncorrectable (fatal)	This error is caused by any of the following conditions: The data payload of a received TLP exceeds the maximum payload size. The `TD` field is asserted but no TLP digest exists, or a TLP digest exists but the `TD` bit of the PCI Express request header packet is not asserted. A TLP violates a byte enable rule. The Hard IP block checks for this violation, which is considered optional by the PCI Express specifications. A TLP in which the `type` and `length` fields do not correspond with the total length of the TLP. A TLP in which the combination of format and type is not specified by the PCI Express specification. A request specifies an address/length combination that causes a memory space access to exceed a 4 KB boundary. The Hard IP block checks for this violation, which is considered optional by the PCI Express specification. Messages, such as Assert_INTX, Power Management, Error Signaling, Unlock, and Set Power Slot Limit, must be transmitted across the default traffic class. The Hard IP block deletes the malformed TLP; it is not presented to the Application Layer.
Note: Considered optional by the PCI Express Base Specification Revision.

Error Reporting and Data Poisoning

How the Endpoint handles a particular error depends on the configuration registers of the device.

Refer to the PCI Express Base Specification 3.0 for a description of the device signaling and logging for an Endpoint.

The Hard IP block implements data poisoning, a mechanism for indicating that the data associated with a transaction is corrupted. Poisoned TLPs have the error/poisoned bit of the header set to 1 and observe the following rules:

Received poisoned TLPs are sent to the Application Layer and status bits are automatically updated in the Configuration Space.
Received poisoned Configuration Write TLPs are not written in the Configuration Space.
The Configuration Space never generates a poisoned TLP; the error/poisoned bit of the header is always set to 0.

Poisoned TLPs can also set the parity error bits in the PCI Configuration Space Status register.

Table 74. Parity Error Conditions
Status Bit	Conditions
Detected parity error (status register bit 15)	Set when any received TLP is poisoned.
Master data parity error (status register bit 8)	This bit is set when the command register parity enable bit is set and one of the following conditions is true: The poisoned bit is set during the transmission of a Write Request TLP. The poisoned bit is set on a received completion TLP.

Poisoned packets received by the Hard IP block are passed to the Application Layer. Poisoned transmit TLPs are similarly sent to the link.

Uncorrectable and Correctable Error Status Bits

Figure 38. Uncorrectable Error Status RegisterThe default value of all the bits of this register is 0. An error status bit that is set indicates that the error condition it represents has been detected. Software may clear the error status by writing a 1 to the appropriate bit.

Figure 39. Correctable Error Status RegisterThe default value of all the bits of this register is 0. An error status bit that is set indicates that the error condition it represents has been detected. Software may clear the error status by writing a 1 to the appropriate bit.

PCI Express Protocol Stack

The V-Series Avalon-MM Hard IP for PCI Express implements the complete PCI Express protocol stack as defined in the PCI Express Base Specification. The protocol stack includes the following layers:

Transaction Layer—The Transaction Layer contains the Configuration Space, which manages communication with the Application Layer, the RX and TX channels, the RX buffer, and flow control credits.
Data Link Layer—The Data Link Layer, located between the Physical Layer and the Transaction Layer, manages packet transmission and maintains data integrity at the link level. Specifically, the Data Link Layer performs the following tasks:
- Manages transmission and reception of Data Link Layer Packets (DLLPs)
- Generates all transmission cyclical redundancy code (CRC) values and checks all CRCs during reception
- Manages the retry buffer and retry mechanism according to received ACK/NAK Data Link Layer packets
- Initializes the flow control mechanism for DLLPs and routes flow control credits to and from the Transaction Layer
Physical Layer—The Physical Layer initializes the speed, lane numbering, and lane width of the PCI Express link according to packets received from the link and directives received from higher layers.

The following figure provides a high‑level block diagram.

Figure 40. V-Series Avalon-MM DMA for PCI Express

Table 75. Application Layer Clock Frequencies
Lanes	Gen1	Gen2	Gen3
×2	N/A	125 MHz @ 128 bits	125 MHz @ 128 bits
×4	N/A	125 MHz @ 128 bits	250 MHz @ 128 bits or 125 MHz @ 256 bits
×8	125 MHz @ 128 bits	250 MHz @ 128 bits or 125 MHz @ 256 bits	250 MHz @ 256 bits

Top-Level Interfaces

Avalon-MM DMA Interface

An Avalon‑MM interface with DMA connects the Application Layer and the Transaction Layer. This interface includes high-performance, burst capable Read DMA and Write DMA modules. This variant is available for the following Endpoint configurations:

Gen1 x8
Gen2 x4, x8
Gen3 x2, x4, x8

Clocks and Reset

The PCI Express Base Specification requires an input reference clock, which is called refclk in this design. The PCI Express Base Specification stipulates that the frequency of this clock be 100 MHz.

The PCI Express Base Specification also requires a system configuration time of 100 ms. To meet this specification, IP core includes an embedded hard reset controller. This reset controller exits the reset state after the periphery of the device is initialized.

Transceiver Reconfiguration

The transceiver reconfiguration interface allows you to dynamically reconfigure the values of analog settings in the PMA block of the transceiver. Dynamic reconfiguration is necessary to compensate for process variations.

Interrupts

The Hard IP for PCI Express offers the following interrupt mechanisms:

Message Signaled Interrupts (MSI)— MSI uses the TLP single dword memory writes to to implement interrupts. This interrupt mechanism conserves pins because it does not use separate wires for interrupts. In addition, the single dword provides flexibility in data presented in the interrupt message. The MSI Capability structure is stored in the Configuration Space and is programmed using Configuration Space accesses.
MSI-X—The Transaction Layer generates MSI-X messages which are single dword memory writes. The MSI-X Capability structure points to an MSI-X table structure and MSI-X PBA structure which are stored in memory. This scheme is in contrast to the MSI capability structure, which contains all of the control and status information for the interrupt vectors.

PIPE

The PIPE interface implements the Intel‑designed PIPE interface specification. You can use this parallel interface to speed simulation; however, you cannot use the PIPE interface in actual hardware.

The simulation models support PIPE and serial simulation.

Data Link Layer

The Data Link Layer is located between the Transaction Layer and the Physical Layer. It maintains packet integrity and communicates (by DLL packet transmission) at the PCI Express link level.

The DLL implements the following functions:

Link management through the reception and transmission of DLL Packets (DLLP), which are used for the following functions:
- Power management of DLLP reception and transmission
- To transmit and receive ACK/NAK packets
- Data integrity through generation and checking of CRCs for TLPs and DLLPs
- TLP retransmission in case of NAK DLLP reception or replay timeout, using the retry (replay) buffer
- Management of the retry buffer
- Link retraining requests in case of error through the Link Training and Status State Machine (LTSSM) of the Physical Layer

Figure 41. Data Link Layer

The DLL has the following sub-blocks:

Data Link Control and Management State Machine—This state machine connects to both the Physical Layer’s LTSSM state machine and the Transaction Layer. It initializes the link and flow control credits and reports status to the Transaction Layer.
Power Management—This function handles the handshake to enter low power mode. Such a transition is based on register values in the Configuration Space and received Power Management (PM) DLLPs. All of the V-Series Hard IP for PCIe IP core variants do not support low power modes.
Data Link Layer Packet Generator and Checker—This block is associated with the DLLP’s 16-bit CRC and maintains the integrity of transmitted packets.
Transaction Layer Packet Generator—This block generates transmit packets, including a sequence number and a 32-bit Link CRC (LCRC). The packets are also sent to the retry buffer for internal storage. In retry mode, the TLP generator receives the packets from the retry buffer and generates the CRC for the transmit packet.
Retry Buffer—The retry buffer stores TLPs and retransmits all unacknowledged packets in the case of NAK DLLP reception. In case of ACK DLLP reception, the retry buffer discards all acknowledged packets.
ACK/NAK Packets—The ACK/NAK block handles ACK/NAK DLLPs and generates the sequence number of transmitted packets.
Transaction Layer Packet Checker—This block checks the integrity of the received TLP and generates a request for transmission of an ACK/NAK DLLP.
TX Arbitration—This block arbitrates transactions, prioritizing in the following order:
- Initialize FC Data Link Layer packet
- ACK/NAK DLLP (high priority)
- Update FC DLLP (high priority)
- PM DLLP
- Retry buffer TLP
- TLP
- Update FC DLLP (low priority)
- ACK/NAK FC DLLP (low priority)

Physical Layer

The Physical Layer is the lowest level of the PCI Express protocol stack. It is the layer closest to the serial link. It encodes and transmits packets across a link and accepts and decodes received packets. The Physical Layer connects to the link through a high‑speed SERDES interface running at 2.5 Gbps for Gen1 implementations, at 2.5 or 5.0 Gbps for Gen2 implementations, and at 2.5, 5.0 or 8.0 Gbps for Gen3 implementations.

The Physical Layer is responsible for the following actions:

Training the link
Scrambling/descrambling and 8B/10B encoding/decoding for 2.5 Gbps (Gen1), 5.0 Gbps (Gen2), or 128b/130b encoding/decoding of 8.0 Gbps (Gen3) per lane
Scrambling/descrambling and 8B/10B encoding/decoding for 2.5 Gbps (Gen1) and 5.0 Gbps (Gen2) per lane
Serializing and deserializing data
Equalization (Gen3)
Operating the PIPE 3.0 Interface
Implementing auto speed negotiation (Gen2 and Gen3)
Implementing auto speed negotiation (Gen2)
Transmitting and decoding the training sequence
Providing hardware autonomous speed control
Implementing auto lane reversal

Figure 42. Physical Layer Architecture

PHY Layer—The PHY layer includes the 8B/10B encode and decode functions for Gen1 and Gen2. The PHY also includes elastic buffering and serialization/deserialization functions.

The Physical Layer is subdivided by the PIPE Interface Specification into two layers (bracketed horizontally in above figure):

PHYMAC—The MAC layer includes the LTSSM and the scrambling/descrambling. byte reordering, and multilane deskew functions.
Media Access Controller (MAC) Layer—The MAC layer includes the LTSSM and the scrambling and descrambling and multilane deskew functions.
PHY Layer—The PHY layer includes the 8B/10B encode and decode functions for Gen1 and Gen2. It includes 128b/130b encode and decode functions for Gen3. The PHY also includes elastic buffering and serialization/deserialization functions.
PHY Layer—The PHY layer includes the 8B/10B encode and decode functions for Gen1 and Gen2. The PHY also includes elastic buffering and serialization/deserialization functions.

The Physical Layer integrates both digital and analog elements. Intel designed the PIPE interface to separate the PHYMAC from the PHY. The V-Series Hard IP for PCI Express complies with the PIPE interface specification.

Note: The internal PIPE interface is visible for simulation. It is not available for debugging in hardware using a logic analyzer such as Signal Tap. If you try to connect Signal Tap to this interface the design fails compilation.

Figure 43. Physical Layer Architecture

The PHYMAC block comprises four main sub-blocks:

MAC Lane—Both the RX and the TX path use this block.
- On the RX side, the block decodes the Physical Layer packet and reports to the LTSSM the type and number of TS1/TS2 ordered sets received.
- On the TX side, the block multiplexes data from the DLL and the Ordered Set and SKP sub-block (LTSTX). It also adds lane specific information, including the lane number and the force PAD value when the LTSSM disables the lane during initialization.
LTSSM—This block implements the LTSSM and logic that tracks TX and RX training sequences on each lane.
For transmission, it interacts with each MAC lane sub-block and with the LTSTX sub-block by asserting both global and per-lane control bits to generate specific Physical Layer packets.
- On the receive path, it receives the Physical Layer packets reported by each MAC lane sub-block. It also enables the multilane deskew block. This block reports the Physical Layer status to higher layers.
- LTSTX (Ordered Set and SKP Generation)—This sub-block generates the Physical Layer packet. It receives control signals from the LTSSM block and generates Physical Layer packet for each lane. It generates the same Physical Layer Packet for all lanes and PAD symbols for the link or lane number in the corresponding TS1/TS2 fields. The block also handles the receiver detection operation to the PCS sub-layer by asserting predefined PIPE signals and waiting for the result. It also generates a SKP Ordered Set at every predefined timeslot and interacts with the TX alignment block to prevent the insertion of a SKP Ordered Set in the middle of packet.
- Deskew—This sub-block performs the multilane deskew function and the RX alignment between the initialized lanes and the datapath. The multilane deskew implements an eight-word FIFO buffer for each lane to store symbols. Each symbol includes eight data bits, one disparity bit, and one control bit. The FIFO discards the FTS, COM, and SKP symbols and replaces PAD and IDL with D0.0 data. When all eight FIFOs contain data, a read can occur. When the multilane lane deskew block is first enabled, each FIFO begins writing after the first COM is detected. If all lanes have not detected a COM symbol after seven clock cycles, they are reset and the resynchronization process restarts, or else the RX alignment function recreates a 64-bit data word which is sent to the DLL.

V-Series Avalon-MM DMA for PCI Express

The V-Series Avalon-MM DMA for PCI Express IP Core includes highly efficient Read DMA, Write DMA, and DMA Descriptor Controller modules. The typical throughput for hardware systems using a 256-bit interface to the Application Layer is 6 GB/sec or higher. For hardware systems using a 128-bit interface to the Application Layer the throughput scales proportionately to 3 GB/sec.

Using a 64-byte payload, the maximum theoretical throughput is far less due to the increased proportion of the bandwidth taken by the TLP headers. The throughput for back-to-back TX memory write completions, RX read completions, and simultaneous reads and writes is 2 GB/sec.

Note: A 64-byte packet is the minimum packet size for Ethernet.

Understanding the Internal DMA Descriptor Controller

When you select Instantiate internal descriptor controller in the parameter editor, the Avalon-MM with DMA includes an internal DMA Descriptor Controller to manage read and write DMA operations. The DMA Descriptor Controller includes read and write data movers to perform local memory reads and writes. It supports up to 128 descriptors for read and write DMAs. Host software programs the DMA Descriptor Controller internal registers with the location and size of the descriptor table residing in the PCI Express main memory. The descriptor control logic directs the DMA read logic to copy the entire table to its local FIFOs.

Figure 44. Platform Designer Design Example with the Internal DMA Descriptor ControllerThis Platform Designer design example, ep_g3x8_avmm256_integrated.qsys, is available in the <install_dir>/ ip/altera/altera_pcie/altera_pcie_a10_ed/example_design/a10 directory. Refer to Getting Started with the V-Series Avalon-MM DMA for instructions on simulating and compiling this example design. This screen capture filters out some interface types for clarity.

Figure 45. Avalon-MM DMA Block Diagram with the Internal DMA Descriptor ControllerThis block diagram corresponds to the Platform Designer system shown in the previous figure.

This design uses BAR0 and BAR1 to create a 64-bit address to access the DMA Descriptor Controller. These BARs cannot connect to any other interface. If BAR0 must access a different interface, you must use an external DMA descriptor controller. Intel recommends that you select the internal DMA Descriptor Controller if you do not plan to modify this component.

The high-performance BAR2 or BAR2 and BAR3 for 64-bit addresses is available to receive data for other high performance functions.

Understanding the External DMA Descriptor Controller

Using the External DMA Descriptor Controller provides more flexibility. You can either modify or replace it to meet your system requirements.You may need to modify the DMA Descriptor Controller for the following reasons:

To implement multi-channel operation
To implement the descriptors as a linked list or to implement a custom DMA programming model
To store descriptors in a local memory, instead of system (host-side) memory

To interface to the DMA logic included in this variant, the custom DMA descriptor controller must implement the following functions:

It must communicate with the Write Mover and Read Mover to copy the descriptor table to local memory.
The Write Mover and Read Mover must execute the descriptors stored in local memory.
The DMA Avalon-MM write (WrDCM_Master) and read (RdDCM_Master) masters must be able to update status to the TX slave (TXS).

Figure 46. Avalon-MM DMA Block Diagram with External DMA Descriptor ControllerThis Platform Designer design example, ep_g3x8_avmm256.qsys, is available in the <install_dir>/ ip/altera/altera_pcie/altera_pcie_a10_ed/example_design/a10 directory. This screen capture filters out some interface types for clarity.

Figure 47. Avalon-MM DMA Block Diagram with External DMA Descriptor ControllerThis block diagram corresponds to the Platform Designer system shown in the previous figure. When the DMA Descriptor Controller is instantiated as a external component, it drives table entries on the RdDmaRxData_i[159:0] and WrDmaRxData_i[159:0] buses.

The DMA modules shown in the block diagram implements the following functionality:

Read DMA (Read Mover and dma_rd_master) –Transfers data from the host domain to the local domain. It sends Memory Read TLPs upstream and writes the completion data to external Avalon-MM components using its own high performance master port. It follows the PCI Express Base Specification rules concerning tags, flow control credits, read completion boundary, maximum read size, and 4 KB boundaries.
Write DMA (Write Mover and dma_wr_master) – Transfers data from the local domain to the host domain. It reads data from an Avalon-MM slave component using its own high performance master port. It sends data upstream using Memory Write TLPs. It follows the PCI Express Base Specification rules concerning tags, flow control credits, RX buffer completion rules, maximum payload size, and 4 KB boundaries.
DMA Descriptor Controller–Manages read and write DMA operations. Host software programs its internal registers with the location of the descriptor table residing in the PCI Express system memory. The descriptor control logic directs the DMA read logic to copy the entire table to the local FIFO. It then fetches the table entries from the FIFO one at a time. It directs the appropriate DMA to transfer the data between the local and host domains. It also sends DMA status upstream via the TX slave single dword port (TXS). For more information about the DMA Descriptor Controller registers, refer to DMA Descriptor Controller Registers.
RX Master (PCIe BAR0-1) –Allows the host to program internal registers of the DMA Descriptor Controller.
TX Slave (TXS) – The DMA Descriptor Controller reports status on each read and write descriptor to this Avalon-MM slave. It also uses this port to send MSI requests.

Transceiver PHY IP Reconfiguration

As silicon progresses towards smaller process nodes, circuit performance is affected by variations due to process, voltage, and temperature (PVT). Consequently, Gen3 designs require offset cancellation and adaptive equalization (AEQ) to ensure correct operation. Altera’s Qsys example designs all include Transceiver Reconfiguration Controller and Altera PCIe Reconfig Driver IP cores that automatically perform these functions during the LTSSM equalization states.

As silicon progresses towards smaller process nodes, circuit performance is affected by variations due to process, voltage, and temperature (PVT). Designs typically require offset cancellation to ensure correct operation. At Gen2 data rates, designs also require DCD calibration. Altera’s Qsys example designs all include Transceiver Reconfiguration Controller and Altera PCIe Reconfig Driver IP cores to perform these functions.

Connecting the Transceiver Reconfiguration Controller IP Core

The Transceiver Reconfiguration Controller IP Core is available for V-series devices and can be found in the Interface Protocols/Transceiver PHY category in the IP Catalog. When you instantiate the Transceiver Reconfiguration Controller the Enable offset cancellation block and Enable PLL calibration options are enabled by default. For Gen3 variants, you should also turn on Enable adaptive equalization (AEQ) block.

A software driver for the Transceiver Reconfiguration Controller IP Core, Altera PCIe Reconfig Driver IP core, is also available in the IP Catalog under Interface Protocols/PCIe. The PCIe Reconfig Driver is implemented in clear text that you can modify if your design requires different reconfiguration functions.

Note: You must include a software driver in your design to program the Transceiver Reconfiguration Controller IP Core.

Figure 48. Altera Transceiver Reconfiguration Controller Connectivity The following figure shows the connections between the Transceiver Reconfiguration Controller instance and the PHY IP Core for PCI Express instance for a ×4 variant.

As this figure illustrates, the reconfig_to_xcvr[ <n> 70-1:0] and reconfig_from_xcvr[ <n> 46-1:0] buses connect the two components. You must provide a 100–125 MHz free‑running clock to the mgmt_clk_clk clock input of the Transceiver Reconfiguration Controller IP Core.

Initially, each lane and TX PLL require a separate reconfiguration interface. The parameter editor reports this number in the message pane. You must take note of this number so that you can enter it as a parameter value in the Transceiver Reconfiguration Controller parameter editor. The following figure illustrates the messages reported for a Gen2 ×4 variant. The variant requires five interfaces: one for each lane and one for the TX PLL.

Figure 49. Number of External Reconfiguration Controller Interfaces

When you instantiate the Transceiver Reconfiguration Controller, you must specify the required Number of reconfiguration interfaces as the following figure illustrates.

Figure 50. Specifying the Number of Transceiver Interfaces for Arria V GZ and Stratix V Devices

Figure 51. Specifying the Number of Transceiver Interfaces for Arria V and Cyclone V Devices

The Transceiver Reconfiguration Controller includes an Optional interface grouping parameter. Transceiver banks include six channels. For a ×4 variant, no special interface grouping is required because all 4 lanes and the TX PLL fit in one bank.

Note: Although you must initially create a separate logical reconfiguration interface for each lane and TX PLL in your design, when the Quartus^® Prime software compiles your design, it reduces the original number of logical interfaces by merging them. Allowing the Quartus^® Prime software to merge reconfiguration interfaces gives the Fitter more flexibility in placing transceiver channels.

Note: You cannot use SignalTap to observe the reconfiguration interfaces.

Transceiver Reconfiguration Controller Connectivity for Designs Using CvP

If your design meets the following criteria:

It enables CvP
It includes an additional transceiver PHY that connect to the same Transceiver Reconfiguration Controller

then you must connect the PCIe refclk signal to the mgmt_clk_clk signal of the Transceiver Reconfiguration Controller and the additional transceiver PHY. In addition, if your design includes more than one Transceiver Reconfiguration Controller on the same side of the FPGA, they all must share the mgmt_clk_clk signal.

For more information about using the Transceiver Reconfiguration Controller, refer to the Transceiver Reconfiguration Controller chapter in the Altera Transceiver PHY IP Core User Guide.

Frequently Asked Questions for V-Series Avalon-MM DMA Interface for PCIe

The following miscellaneous facts might be of assistance in troubleshooting:

Only the Root Ports can be loopback masters.
Refer to Altera Solutions by searching in the Knowledge Base under Support on the Altera website.

V-Series Interface for PCIe Solutions User Guide Archive

If an IP core version is not listed, the user guide for the previous IP core version applies.

IP Core Version	User Guide
15.1	V-Series Avalon-MM DMA Interface for PCIe SolutionsUser Guide
14.1	V-Series Avalon-MM DMA Interface for PCIe SolutionsUser Guide

Document Revision History

Document Revision History for the V-Series Avalon -MM DMA Interface for PCIe Solutions User Guide

Date	Version	Changes Made
2018.06.15	18.0.1	Added note that Flush reads are not supported when burst mode for BAR2 is enabled. Updated the list of configurations supported by the Avalon-MM and Avalon-MM with DMA variants.
2018.05.07	18.0	Changed all references to Intel^® Cyclone^® 10 to Intel^® Cyclone^® 10 GX.
2017.10.06	17.1	Made the following change to the user guide: Added support for Intel^® Cyclone^® 10 GX devices. Added optional parameter to invert the RX polarity. Corrected Feature Comparison for all Hard IP for PCI Express IP Core table: The Avalon-MM DMA interface does not automatically handle out-of-order completions. Added missing sequence of programming steps in DMA Descriptor Controller Registers. Rebranded as Intel. Corrected minor errors and typos.
2017.05.26	17.0	Made the following changes to the user guide: Added note that starting with the Intel^® Quartus^® Prime Pro Edition Software, version 17.0, the QSF assignments in the following answer What assignments do I need for a PCIe Gen1, Gen2 or Gen3 design that targets an Intel^® Arria^® 10 ES2, ES3 or production device? are already included in the design.
2017.05.08	17.0	Made the following changes to the user guide: Corrected Comparison of Avalon-ST, Avalon-MM and Avalon-MM with DMA Interfaces table. Out-of-order Completions are not supported transparently for the Avalon-MM with DMA interface.
2017.03.15	16.1.1	Made the following changes: Rebranded as Intel.
2016.10.28	16.1	Made the following to the IP core changes: Increased the max DMA transfer to 1 MB for both the 128- and 256-bit interfaces. Made the following changes to the user guide: Corrected the number of tags supported in the Feature Comparison for all Hard IP for PCI Express IP Cores table. Added PCIe bifurcation to the Feature Comparison for all Hard IP for PCI Express IP Cores table. PCI bifurcation is not supported. Removed reference to a Linux software driver for the DMA modules which is not available. Added section covering design example limitations.
2016.05.02	16.0	Made the following changes: Redesigned the 128-bit interface to the Application Layer resulting in consistently high throughput for both on-chip and external memory. Added simulation support for Gen3 PIPE mode using the ModelSim, VCS, and NCSim simulators. Revised discussion of the DMA Descriptor Controller in the Avalon-MM with DMA IP Core Architecture. Revised Read DMA Example to reflect current maximum transfer size of 64 KB for 256-bit interface. The example now corresponds to an example design provided in the `<install_dir>`. Corrected description of Write Descriptor Table Avalon-MM Slave Port.
2015.11.30	15.1	Added description of the Altera PCIe Reconfig Driver in the Connecting the Transceiver Reconfiguration Controller IP Core topic. Added note explaining that the Getting Started design examples do not generate all the files necessary to download to an Altera FPGA Development Kit. Provided link to AN 690: PCI Express DMA Reference Design for Stratix V Devices and AN 708: PCI Express DMA Reference Design Using External DDR3 Memory for Stratix V and Arria V GZ Devices that includes all necessary files. Added a FAQ chapter. Removed `dlup` signal. This signal is no longer part of the `Hard IP Status` interface. Fixed minor errors and typos.
2014.12.18	14.1	Made the following changes: Updated the instructions to create a Quartus II project and compile the design in the Getting Started with the Avalon-MM DMA chapter. Added links to the PCI Express Multi-Channel DMA Interface Example Design User Guide and the Avalon-MM DMA FIFO Example Design User Guide.
2014.12.15	14.1	Made the following changes to the Intel^® Arria^® 10 user guide: In the Getting Started chapter, corrected directory path for the simulation. Added the fact that the RX Burst Master only support dword granularity. Added definitions for `test_in[2]`, `test_in[6]` and `test_in[7]`. Added instructions for Quartus II compilation.