|
LX4180
Features
- 32-bit R3000-class embedded processor core
- Operates at 33 MHz in an EPF10K200E or larger device
- Supports MIPS1 instruction set architecture (see note 1 below)
- Supports MIPS16 code compression
- Includes an optional Multiply-Accumulate (MAC) engine
- Includes optional EJTAG on-chip software debug functionality
- Supported by industry standard development tools and real-time operating systems
- Evaluation system board available
Block Diagram
Figure 1. shows the block diagram for the LX4180 megafunction.
Figure 1. Block Diagram |
|
General Description
The LX4180 is a 32-bit reduced instruction set computer (RISC) processor which executes the MIPSI instruction set (see Note 1 below). It allows designers to take advantage of a wide array of software development tools available from commercial suppliers.
The LX4180 can be configured to execute MIPS16 compressed code. Multiplication and division operations can be performed in hardware by a configurable MAC engine. On-chip code debug can be performed on the LX4180 with the configurable EJTAG block.
Architecture Overview
The LX4180 megafunction is based on Lexra's
high-performance LX4180 embedded processor architecture. Following are
characteristics of the LX4180 CPU:
Table 1. Architectural Features of the LX4180 Function |
Feature |
Description |
MIPSI Instruction Set Architecture (1) |
The LX4180 supports the MIPSI programming model. The instruction set incorporates a three-port register file. Two source operands can be supplied and one destination updated per cycle. The second operand is either a register or 16-bit immediate. The instruction set includes a wide selection of arithmetic logic unit (ALU) operations executed by the register arithmetic logic unit (RALU), Lexra's proprietary register-based ALU. The RALU also generates memory addresses for 8-bit, 16-bit and 32-bit register loads from and stores to memory by adding a register base to an immediate offset. Branches are based on comparisons between registers, rather than flags, and are therefore easy to relocate. Optional links following jump or branch instructions assist with subroutine programming. |
Pipeline |
Instructions are executed by a five-stage pipeline. The pipeline is designed so that all internal transactions and interfaces occur on the positive edge of the processor clock. Two-phase clocks are not used. |
Exception Handling |
The LX4180 supports the MIPSI exception handling model. Exceptions include instruction-synchronous traps, hardware, and software interrupts. All exceptions are prioritized. When an exception is taken, control is transferred to a user program located at the exception vector. The use program identifies the cause of the exception and transfers control to the application-specific handler. |
Co-processor Operations |
The LX4180 supports all 32-bit co-processor operations. These include moves to and from the co-processor general registers and control registers (MTCz, MFCz, CTCz, CFCz), co-processor loads and stores (LWCz, SWCz) and branches based on co-processor condition flags (BCzT, BCzF). The LX4180 includes a co-processor interface which can support all co-processor operations in a single cycle, without pipeline stalls. |
In addition to the LX4180 CPU, the LX4180 megafunction includes the following standard blocks:
- The Simple Memory Management Unit (SMUU)
is designed for embedded applications that use a single address space. Its
primary function is to provide memory protection between user space and
kernel space.
- Two Local Memory Interfaces (LMI). One
local memory interfaces to 2 Kbytes of instruction memory and the other
interfaces to 2 Kbytes of data memory.
- The Coprocessor Interface (CI). The
co-processor interface "eavesdrops" on the instruction. If a
co-processor load (LWCz) or move to (MTCz, CTCz) is decoded, data will be
enabled from the data bus into a CI register, then supplied to the
designer-defined co-processor. Similarly, if a co-processor store (SWCz) or
move from (MFCz, CFCz) is decoded, data will be fetched from the
co-processor and loaded into a CI register, then transferred onto the data
bus in the following cycle. The design interface includes a variable-width
data bus, five-bit address and independent read and write selects for
co-processor register and registers. The LX4180 pipeline and Harvard
architecture permit single cycle co-processor access and transfer.
Designer-defined co--processor condition flags (CpCondz) are synchronized by
the CI then passed to the sequencer for testing in branch instructions.
- The Custom Engine Interface (CEI), which the designer can use to extend the MIPS ALU opcodes with application-specific or proprietary operations. Similar to the standard ALU, the CEI supplies the custom engine with two input 32-bit operands, SRC1 and SRC2. One operand is selected from the register file. Depending on the opcode, the second operand is either selected from the register file or is a 16-bit sign-extended immediate. The opcode is locally decoded by the custom engine and, following execution by the custom engine, the result is returned on the 32-bit RES bus to the LX4180.
The LX4180 can be configured with the following optional features:
- The Lexra Bus Controller (LBC) connects
peripheral functions and secondary memories to the processor's own local
buses. It is a multiplexed, no-pipelined, and non-parity-checked bus that
provides the easiest protocol for design integration. On the processor side,
the LBC provides a four word deep write buffer, and control for byte and
half-word transfers. On the peripheral side, the LBC is designed to
interface easily with industry standard bus protocols, such as peripheral
component interconnect (PCI), universal serial bus (USB) and IEE Std. 1394.
- MIPS16 instruction compression reduces
program size by up to 40%. On-chip programs require less memory, resulting
in cost saving for system-on-chip designs.
- The optional MAC
engine completes one multiple or divide instruction every cycle, for
performance in signal processing applications.
- The LX4180 implements the
industry standard EJTAG 2.0.0 specification for full-speed debug with real time
instruction trace.
- Software Tool Support: The LX4180 is supported by development tools from Green Hills Software Inc., Wind River Systems, and Embedded Performance Inc.
Device Utilization Example
Table 2 lists the typical device utilization results for the megafunction.
Table 2. Typical Device Utilization for the Megafunction |
Device |
Speed Grade |
Utilization |
Performance
(fMAX) |
Parameter Setting |
Logic Cells |
EABs/ESBs (1) |
EP20K400E EP20K600E EP20K1000E EPF10K200E |
- |
4,000 - 12,000 |
2 - 200 |
33 MHz |
Contact Lexra |
Note:
- EABs: Embedded array blocks, ESBs: Embedded system blocks
Additional Information
The following items are available with the LX4180 megafunction:
- Netlist file for use with the MAX+PLUS® II software
- Encrypted register transfer level (RTL)
model for simulation model
- Encrypted register transfer level (RTL)
model for simulation with industry standard simulators
- Lexra Regression Test Kit
- Optional: LX4x80 Evaluation System Board
Note:
- Unaligned
load and store instructions are not supported in hardware. MIPS, R3000, MIPS
common-law marks are trademarks or registered trademarks of MIPS Technologies,
Inc. Lexra, Inc. is not associated with MIPS Technologies, Inc. in any
way.
Contact Information
For additional information, contact Lexra at:
Lexra Incorporated
2055 Gateway Place, Suite 150
San Jose, CA 95110
Tel: (408) 573-1890
Fax: (408) 573-1898
E-mail: info@lexra.com
WWW: http://www.lexra.com
|