k64 datasheet学习笔记45---10/100-Mbps Ethernet MAC(ENET)之功能描述
1.前言
本文是对K64 datasheet 之ENET部分的功能描述,将对每个部分进行详细说明
2.Ethernet MAC frame formats
- MAC帧组成格式
(1)7字节前导码:如按最低位先传则为0x55 0x55 0x55 0x55 0x55 0x55;
(2)起始帧定界符(SFD):如按最低位先传则为0XD5;
(3)2个地址域:包括源MAC地址和目的MAC地址;
(4)VLAN-tagged :可选,占4字节
(5)长度或类型域:代表有效负荷的长度或者类型,以太网802.3组帧为例,如果小于0X05DC(1500字节)则为长度信息,如果大于0X0600(1536字节)则为类型信息
常见的类型为:IPV4,ARP等
(6)有效载荷:最小为42 octets(带有VALAN-tagged),46 octets(不带有VALAN-tagged),最大载荷为1500 octets
(7)填充域:有效载荷不足42 octets(带有VALAN-tagged),46 octets(不带有VALAN-tagged),则填充
(8)CRC值:4字节循环冗余校验
注:
[1]帧长度不包括Preamble和SFD,帧最小长度为64字节,最大长度为1518字节
[2]除去 6 octets的 Destination address+6 octsts Source address+2 octets Length/type+4octets FCS = 18 octets,因此有效载荷需要至少46 octets,需填充0~46octets
[3]可选的是VLAN-tagged 可以插入到MACSource address和Length/type之间4字节,构成VLAN-tagged MAC 帧
- MAC帧定义
- Pause Frames
接收端产生一个pause frame指示当前传输遇到拥堵,发送端需要停止发送数据
- Magic packets
用于实现远程唤醒,当侦测到magic packet,主机会被唤醒
3.IP and higher layers frame format
举例来说IP数据报指定了Ethernet帧的有效负荷,TCP数据报又在IP数据报中指定了有效负荷
- Ethernet types
- IPv4 datagram format
- IPv6 datagram format
- Internet Control Message Protocol (ICMP) datagram format
- User Datagram Protocol (UDP) datagram format
- TCP datagram format
4. IEEE 1588 message formats
- Adjustable timer module
ENET时间戳时钟源既可以来自外部ENET_1588_CLKIN PAD ,也可以来自内部ENET PLL 。最好选择整形值如3ns产生精准计时
通过专用校准逻辑定时器被校准来同步远端master时钟作为本地系统的参考时钟
Adjustable timer implementation
- Transmit timestamping
- Receive timestamping
- Time synchronization
5. MAC receive
- Collision detection in half-duplex mode
- Preamble processing
MAC core允许任意长度的preample,也允许是0长度,如果preample后SFD的首个字节不是0x55也不是0xd5则丢弃
- MAC address check
目的地址bit0区分了不同的地址:
(1)If bit 0 is 0, the MAC address is an individual (unicast) address.
(2)If bit 0 is 1, the MAC address defines a group (multicast) address.
(3)If all 48 bits of the MAC address are set, it indicates a broadcast address.
- Frame length/type verification: payload length check
If the length/type is less than 0x600 and NLC is set, the MAC checks the payload length and reports any error in the frame status word and interrupt bit PLR
If the length/type is greater than or equal to 0x600, the MAC interprets the field as a type and no payload length check is performed
The length check is performed on VLAN and stacked VLAN frames. If a padded frame is received, no length check can be performed
- Frame length/type verification: frame length check
When the receive frame length exceeds MAX_FL bytes, the BABR interrupt is generated and the RxBD[LG] bit is set. - VLAN frames processing
VLAN frames have a length/type field set to 0x8100 immediately followed by a 16-Bit VLAN control information field.
- Pause frame termination
The receive engine terminates pause frames and does not transfer them to the receive FIFO
- CRC check
The CRC-32 field is checked and forwarded to the core FIFO interface if ENETn_RCR[CRCFWD] is cleared and ENETn_RCR[PADEN] is set
- Frame padding removal
- Recieve flushing
RX flushing prevents frames in the RX FIFO from being blocked
6 MAC transmit
- Frame payload padding
The IEEE specification defines a minimum frame length of 64 bytes
- MAC address insertion
- CRC-32 generation
- Inter-packet gap (IPG)
- Collision detection and handling — half-duplex operation only
A collision occurs on a half-duplex network when concurrent transmissions from two or more nodes take place.
7. Full-duplex flow control operation
8 Magic packet detection
Magic packet detection wakes a node that is put in power-down mode by the node management agent
9. IP accelerator functions
- Checksum calculation
The IP and ICMP, TCP, UDP checksums are calculated with one's complement arithmetic summing up 16-bit values
- Additional padding processing
- 32-bit Ethernet payload alignment
The data FIFOs allow inserting two additional arbitrary bytes in front of a frame. This extends the 14-byte Ethernet header to a 16-byte header, which leads to alignment of the
Ethernet payload
- IPv4 fragments
- IPv6 support
10. Resets and stop controls
- Hardware reset
- Soft reset
- Hardware freeze
- Graceful stop
During a graceful stop, any currently ongoing transactions are completed normally and no further frames are accepted
11 IEEE 1588 functions
- Adjustable timer module
- Transmit timestamping
- Receive timestamping
- Time synchronization
12 FIFO thresholds
- Receive FIFO
- Transmit FIFO
13 Loopback options
14 Legacy buffer descriptors
To support the Ethernet controller on previous Freescale devices, legacy FEC buffer descriptors are available
15 Enhanced buffer descriptors
This section provides a description of the enhanced operation of the driver/DMA via the buffer descriptors
16 Client FIFO application interface
The FIFO interface is completely asynchronous from the Ethernet line, and the transmit and receive interface can operate at a different clock rate
17 FIFO protection
- Transmit FIFO underflow
- Transmit FIFO overflow
- Receive FIFO overflow
18 Reference clock
The input clocks to the ENET module must meet the specifications in the following table
19 PHY management interface
- MDIO frame format
- MDIO clock generation
The MDC clock is generated from the internal bus clock divided by the value programmed in ENETn_MSCR[MII_SPEED].
- MDIO operation
To perform an MDIO access, set the MDIO command register (ENETn_MMFR) according to the description provided in MII Management Frame Register (ENETn_MMFR).
To check when the programmed access completes, read the ENETn_EIR[MII] field.
20 Ethernet interfaces
The following Ethernet interfaces are implemented:
-Fast Ethernet MII (Media Independent Interface)
-RMII 10/100 using interface converters/gaskets
The following table shows how to configure ENET registers to select each interface
- RMII interface
- MII Interface — transmit
MII interface — receive
k64 datasheet学习笔记45---10/100-Mbps Ethernet MAC(ENET)之功能描述的更多相关文章
- k64 datasheet学习笔记1---概述
1.前言 k64 datasheet描述了Freescale MCU的特性.架构和编程模型,主要是面向使用MCU的系统架构和软件应用开发人员. 2.模块划分 datasheet主要按功能对模块进行划分 ...
- k64 datasheet学习笔记45---10/100-Mbps Ethernet MAC(ENET)之概述
1.前言 k64 ENET CORE 实现了10M/100Mbps的Ethernet MAC,与IEEE802.3-2002标准兼容. MAC层与全双工/半双工的10M/100Mbps以太网兼容: M ...
- k64 datasheet学习笔记35---Analog-to-Digital Converter (ADC)
0.前言 本文主要介绍K64的ADC部分的特性,内部架构,信号描述及操作接口等 1.简介 1.1.ADC模块特性 线性逐次逼近算法,达16bit分辨率 达到4对差分和24个单端模拟量输入 输出模式 差 ...
- k64 datasheet学习笔记3---Chip Configuration之Analog
1.前言 本文主要讲述K64芯片配置,关于模拟部分的内容,主要包括:ADC, CMP, DAC, VREF 2.16bit SAR ADC 从上图可以看出ADC主要挂在外设总线0上,由于ADC的输入引 ...
- k64 datasheet学习笔记25--Multipurpose Clock Generator (MCG)
0.前言 MCG模块为MCU提供了几种可选时钟源.模块包含一个FLL和一个PLL.FLL使用内部或外部参考时钟是可控的,PLL受外部参考时钟控制 模块可以选择FLL或PLL输出时钟,或内/外部参考时钟 ...
- k64 datasheet学习笔记21--Direct Memory Access Multiplexer (DMAMUX)
0.前言 本章主要介绍DMA MUX的相关内容 1.简介 1.1 概述 用来路由DMA源到16路DMA通道: 1.2 特性 1.52个外设slot和10个常开slot可以路由到16路DMA通道 2.1 ...
- k64 datasheet学习笔记52---Universal Asynchronous Receiver/Transmitter (UART)
1.前言 UART实现与外设或CPU的通信 2. UART概述 2.1基本特性 (1)Full-duplex operation (2)Standard mark/space non-return-t ...
- k64 datasheet学习笔记10---Signal Multiplexing and Signal Descriptions
1.前言 本文主要描述多功能引脚相关的内容,包括5个32位的引脚的端口,每个32pin端口被指定一个中断 2.Signal Multiplexing Integration 3. Port contr ...
- k64 datasheet学习笔记4---Clock distribution
1.前言 本文主要讲述K64时钟配置相关的内容. 2.clock overview 2.1 clock diagram 2.2 对clock diagram的信号说明 (1)MCG决定哪个clock ...
随机推荐
- python 函数基础及装饰器
没有参数的函数及return操作 def test1(): print ("welcome") def test2(): print ("welcomt test2&qu ...
- Web API中的路由(一)——约定路由
一.Web API中的路由概念 路由的作用用一句话说明:通过request的uri找到处理该请求的Controller,Action,以及给Action的参数赋值. 一些路由的基本概念: route: ...
- .NET获取当前程序所在电脑的CPU和内存使用率
using System; using System.Diagnostics; using System.Text; using System.Runtime.InteropServices; nam ...
- Linux命令(十一)gcc
1. gcc -E 预处理 头文件展开 宏替换 2. gcc -S: 生成汇编指令 3. gcc - c: 生成二进制文件 4. gcc -I: (包含头文件) 5. gcc -o: 指定输出 6. ...
- java基础之反射---重要
java反射: 反射是框架设计的灵魂 (使用的前提条件:必须先得到代表的字节码的Class,Class类用于表示.class文件(字节码)): 1:获取Class字节码文件对象的三种方式: /** ...
- fork()和僵尸进程
2018-01-03@望京 关于fork()函数,Unix/Linux提供的fork()系统调用,fork()一次返回两次, 操作系统自动把当前进程(称为父进程)复制了一份(称为子进程),然后,分别在 ...
- Django多表操作
多表创建 创建模型 下面通过一个简单的图书管理系统,来阐述多表的创建和查询操作 在视图函数里里定义如下代码 from django.db import models class Book(models ...
- java的几个奇怪语法
=============================Java 双冒号 :: 语法的含义=============================下面的代码中有双冒号,初看上去很奇怪, list. ...
- Spring使用事务增加的注解实现方
以下是我的文件结构 步骤1:配置数据源 <bean id="datasource" class="com.mchange.v2.c3p0.ComboPooledDa ...
- [笔记]New in Chrome 66
原文 CSS Typed Object Model 使用CSS object model,返回的一切都是字符串 el.style.opacity = 0.3; console.log(typeof e ...