Xilinx SDSoC：构建嵌入式系统的工具套件

5星 · 超过95%的资源需积分: 9 184 浏览量更新于2024-07-19 收藏 2.92MB PDF 举报

身份认证购VIP最低享 7 折!

领优惠券(最高得80元）

"SDSoC开发环境用户指南" SDSoC（Software-Defined System-on-Chip）环境是一个工具套装，它包含了一个基于Eclipse的集成开发环境（IDE），用于实现异构嵌入式系统的设计和开发。这个环境主要针对的是使用ARM Cortex处理器的Zynq-7000全可编程SoC和Zynq UltraScale+ MPSoC，同时也支持在所有Xilinx SoC和FPGA上的MicroBlaze处理器应用程序。 SDSoC环境的核心特点是它的系统编译器，这些编译器能够将C/C++程序转化为完整的硬件/软件系统，其中部分功能被编译到可编程逻辑中。这使得开发者可以充分利用硬件和软件的并行处理能力，提升系统的性能和效率。系统编译器通过分析程序，识别软件和硬件之间的数据流，然后生成一个应用特定的系统级芯片来实现该程序。为了达到高性能，每个硬件函数作为独立的线程运行。编译器会生成硬件和软件组件，确保硬件和软件线程之间的同步，同时允许计算和通信的流水线化执行。这样，应用程序可以包含多个硬件函数实例，以实现更复杂的并行处理任务。在SDSoC环境下，开发者可以利用高级语言（如C/C++）进行设计，而无需深入底层硬件细节。这种抽象层使得软件工程师也能参与到硬件优化中，降低了系统级设计的复杂性。此外，SDSoC还提供了调试工具和性能分析工具，帮助开发者理解和优化他们的设计方案。在硬件加速方面，SDSoC环境使得开发者能够创建定制的硬件加速器，这些加速器可以直接在FPGA上实现，从而加速特定的计算密集型任务。通过这种方式，系统性能可以显著提高，尤其是在处理大数据、图像处理、机器学习等需要高效并行计算的应用中。 SDSoC环境提供了一个全面的解决方案，它融合了软件和硬件开发的优点，让开发者能够轻松地创建高效的异构系统，充分利用Xilinx SoC和FPGA的硬件资源。无论是对于嵌入式系统开发者还是对硬件加速感兴趣的软件工程师，SDSoC都是一个强大的工具，可以帮助他们快速实现高性能、低功耗的系统设计。

资源详情

资源推荐

o The __SDSCC__ macro is defined and passed as a -D option to sub-tools

whenever sdscc or sds++ is used to compile source files, and can be used to guard code dependent on

whether it is compiled by sdscc/sds++ or by another compiler, for example a GNU host compiler.

o When sdscc or sds++ compiles source files targeted for hardware acceleration using Vivado HLS,

the __SDSVHLS__ macro is defined and passed as a -D option, and can be used to guard code dependent

on whether high-level synthesis is run or not.

o In 2017.2 running on the Windows operating system, you will typically have to use these macros

to guard code that will be synthesized within SDx with type long long (which should be 64 bits).

o Vivado HLS employs some 32-bit libraries irrespective of the host machine. Furthermore, the tool

does not provide a true cross-compilation.

All object code for the ARM CPUs is generated with the GNU toolchains, but the sdscc (and sds++) compiler,

built upon Clang/LLVM frameworks, is generally less forgiving of C/C++ language violations than the GNU

compilers. As a result, you might find that some libraries needed for your application cause front-end compiler

errors when using sdscc. In such cases, compile the source files directly through the GNU toolchain rather than

through sdscc, either in your makefiles or by setting the compiler to arm-linux-gnueabihf-g++ by right-clicking on

the file (or folder) in the Project Explorer and selecting C/C++ Build > Settings > SDSCC/SDS++ Compiler.

3.4 Hardware Function Argument Types

The SDSoC™ environment sdscc/sds++ system compilers support hardware function arguments with types

that resolve to a single or array of C99 basic arithmetic type (scalar), a struct or classwhose members flatten to

a single or array of C99 basic arithmetic type (hierarchical structs are supported), an array of struct whose

members flatten to a single C99 basic arithmetic type. Scalar arguments must fit in a 1024-bit container. The

SDSoC™ environment automatically infers hardware interface types for each hardware function argument based

on the argument type and the following pragmas:

#pragma SDS data copy|zero_copy

#pragma SDS data access_pattern

To avoid interface incompatibilities, you should only incorporate Vivado® HLS interface type directives and

pragmas in your source code whensdscc/sds++ fails to generate a suitable hardware interface directive.

 Vivado® HLS provides arbitrary precision types ap_fixed<int>, ap_int<int>, and an hls::stream class.

In the SDSoC environment, ap_fixed<int> types must be specified as having widths greater than 7 but

less than 1025 (7 < width < 1025). The hls::stream data type is not supported as the function argument

to any hardware function.

 By default, an array argument to a hardware function is transferred by copying the data, that is, it is

equivalent to using #pragma SDS data copy. As a consequence, an array argument must be either used as

an input or produced as an output, but not both. For an array that is both read and written by the hardware

function, you must use #pragma SDS data zero_copy to tell the compiler that the array should be kept in the

shared memory and not copied.

 To ensure alignment across the hardware/software interface, do not use hardware function arguments

that are an array of bool.

Important: Pointer arguments for a hardware function require special consideration. Hardware functions

operate on physical addresses, which typically are not available to userspace programs, so pointers cannot be

embedded in data structures passed to hardware functions.

Important:

By default, in the absence of any pragmas, a pointer argument is taken to be a scalar parameter, even

though in C/C++ it might denote a one-dimensional array type. The following are the permitted pragmas.

 This pragma provides pointer semantics using shared memory.

#pragma SDS data zero_copy

 This pragma maps the argument onto a stream, and requires that array elements are accessed in index

order. The data copy pragma is only required when the sdscc system compiler is unable to determine the

data transfer size and issues an error.

#pragma SDS data copy(p[0:<p_size>])

#pragma SDS data access_pattern(p:SEQUENTIAL)

Important: When you require non-sequential access to the array in the hardware function, you should

change the pointer argument to an array with an explicit declaration of its dimensions, for example, A[1024].

3.5 Hardware Function Call Guidelines

 Stub functions generated in the SDSoC™ environment transfer the exact number of bytes according the

compile-time determinable array bound of the corresponding argument in the hardware function declaration.

If a hardware function admits a variable data size, you can use the following pragma to direct the SDSoC

environment to generate code to transfer data whose size is defined by an arithmetic expression:

#pragma SDS data copy|zero_copy(arg[0:<C_size_expr>])

#pragma SDS data zero_copy(arg[0:<C_size_expr>])

where the <C_size_expr> must compile in the scope of the function declaration.

The zero_copy pragma directs the SDSoC environment to map the argument into shared memory.

Important: Be aware that mismatches between intended and actual data transfer sizes can cause the

system to hang at runtime, requiring laborious hardware debugging.

 Align arrays transferred by DMAs on cache-line boundaries (for L1 and L2 caches). Use

the sds_alloc() API provided with the SDSoC environment instead of malloc() to allocate these arrays.

 Align arrays to page boundaries to minimize the number of pages transferred with the scatter-gather DMA,

for example, for arrays allocated with malloc.

 You must use sds_alloc to allocate an array for the following two cases:

1. You are using zero-copy pragma for the array.

2. You are using pragmas to explicitly direct the system compiler to use Simple-DMA.

Note that in order to use sds_alloc() from sds_lib.h, it is necessary to include stdlib.h before

including sds_lib.h. stdlib.h is included to provide the size_t type.

4.Getting Started with Examples

All Xilinx SDx™ Environments are provided with examples designs:

 To help you quickly get started.

 To demonstrate useful coding styles.

 To highlight important optimization techniques.

Example designs are provided with the tool installation and additional examples may be downloaded from

the Xilinx® GitHub repository.

4.1 Installed Examples

The installed examples are provided through the Create SDx™ Project wizard. Select Create SDx

Project from the SDx Development Environment Welcome page to open the new project wizard. After selecting

your hardware platform and software platform, the final page of the wizard lists the available templates.

Note: Not all available platforms have an installed example.

You may select examples from the Templates page, as shown below.

Figure: Templates Page

The SDx Example Store dialog also lists the GitHub exmaples provided for specific SDx environments. For

example, the SDSoC folder lists all GitHub examples for the SDSoC environment.

Figure: SDx Examples for SDSoC

The GitHub examples also indicate if they are installed or not. Use the Refresh button to ensure you have

the latest update from the repository. Click Install to download and install the example design.

Once the example design has been installed, it may be accessed during new project creation in the same

manner as the installed examples.

4.3 Synthesizeable FIR Filter

Many of the functions in the Vivado HLS source code libraries included in the SDSoC environment do not

comply with the SDSoC environment Coding Guidelines. To use these libraries in the SDSoC environment, you

typically have to wrap the functions to insulate the SDSoC system compilers from non-portable data types or

unsupported language constructs.

The Synthesizeable FIR Filter example demonstrates a standard idiom to use such a library function that in

this case, computes a finite-impulse response digital filter. This example uses a filter class constructor and

operator to create and perform sample-based filtering. To use this class within the SDSoC environment, the

example wraps within a function wrapper as follows.

void cpp_FIR(data_t x, data_t *ret)

{

static CF<coef_t, data_t, acc_t> fir1;

*ret = fir1(x);

}

This wrapper function becomes the top-level hardware function that can be invoked from application code.

4.4 Matrix Multiplication

Matrix multiplication is a common compute-intensive operation for many application domains. The SDSoC

IDE provides template examples for all base platforms, and the code for these provide instructive use of SDSoC

environment system optimizations for memory allocation and memory access described in Improving System

Performance, and Vivado HLS optimizations like function inlining, loop unrolling and pipelining, and array

partitioning, described in Optimization Guidelines.

剩余87页未读，继续阅读

ArenCheung

粉丝: 0
资源: 3

Xilinx SDSoC：构建嵌入式系统的工具套件

ug1027-sdsoc-user-guide.pdf

Sdsoc优化指南

"Cadence SimVision Analysis Environment User Guide" 文献 哪里？

The USER environment variable is not set解决方法

EnvironmentNotWritableError: The current user does not have write permissions to the target environment. environment location:

createuser postgres -P zsh: command not found: createuser

报错EnvironmentNotWritableError: The current user does not have write permissions to the target environment.

EnvironmentNotWritableError: The current user does not have write permissions to the target environment. environment location

The current user does not have write permissions to the target environment.

java environment使用_Spring Environment的加载

EnvironmentNotWritableError: The current user does not have write permissions to the target environment. environment location: D:\Software\anaconda\Anaconda3 怎么解决

EnvironmentNotWritableError: The current user does not have write permissions to the target environment. environment location: E:\Anaconda

anaconda中出现EnvironmentNotWritableError: The current user does not have write permissions to the target environment.报错什么意思

选择当中的正确答案Operational environment defined on which life-cycle stage? development stage production stage

EnvironmentNotWritableError: The current user does not have write permissions to the target environment

nginx-1.24.0.tar

智能化病虫害标注系统前端.zip

Python 小游戏 （贪吃蛇、五子棋、扫雷、俄罗斯方块）

最新资源

"Cadence SimVision Analysis Environment User Guide" 文献哪里？

Python 小游戏（贪吃蛇、五子棋、扫雷、俄罗斯方块）