基本运算单元的高层次综合:C/C++ to RTL
本文以加法为例:[code]
//----------------------------------------------------
//adder.c
//---------------------------------------------------
void adder(int a, int b, int *sum)
{
*sum = a + b;
}
[/code][size=3]
[/size]
HLS工具(AutoPilot)综合之后的结果:[code]
//---------------------------------------------------
//adder.v
//--------------------------------------------------
`timescale 1 ns / 1 ps
module adder (
a,
b,
sum
);
input [31:0] a;
input [31:0] b;
output [31:0] sum;
assign sum = (b + a);
endmodule //adder
[/code][size=3]
[/size][code]
//---------------------------------------------------
//adder.vhd
//---------------------------------------------------
library IEEE;
use IEEE.std_logic_1164.all;
use IEEE.numeric_std.all;
library work;
use work.AESL_components.all;
entity adder is
port (
a : IN STD_LOGIC_VECTOR (31 downto 0);
b : IN STD_LOGIC_VECTOR (31 downto 0);
sum : OUT STD_LOGIC_VECTOR (31 downto 0) );
end;
architecture behav of adder is
begin
sum <= esl_add(b, a);
end behav;
[/code][size=3]
[/size]
备注:加减乘除、位运算、逻辑运算等等基本的C/C++运算都可以很方便的用AutoPilot综合成对应的RTL代码(verilog/vhdl)[/size]
相关文档:
1.C语言中,long被存储为四个字节的补码。写一个程序,分别将这四个字节的内容取出,以16进制的方式显示在屏幕上。程序所需的long由用户从键盘输入,0表示输入结束。
程序运行效果如下:
input n: 12345678<回车>
hex: 00 BC 61 4E
input
n: -12345678<回车>
hex: FF 43 9E B2
input n: 0<回车& ......
快速幂取模理论基础: 计算 a^b mod c ?
由(a x b) mod c=((a mod c) x b) mod c.
我们可以将 b先表示成就:
b=at2^t+at-1 2^t-1+……a02^0. (ai=[0,1]).
这样我们由 a^b mod c=(a^(at2^t+at-12^t-1+…a02^0)mod c.
然而我们求 a^(2^(i+1)) ......
pid_t pid=fork()
it has 3 situation for the return result pid
0 child
>0 parent process
<0 fork fail
fork create a new process and it parent live alse when the child process had been created ......
pid_t pid=fork()
it has 3 situation for the return result pid
0 child
>0 parent process
<0 fork fail
fork create a new process and it parent live alse when the child process had been created ......
http://www.edn.com/article/457428-Can_C_beat_RTL_.php
With the appearance of higher speeds and more DSP macrocells in low-cost FPGAs, more and more design teams are seeing the configurable chips not as glue but as a way to accelerate the inner loops of numerical algorithms, either in conjun ......