glinet ar300m-lite
转自glinet官网https://docs.gl-inet.com/en/3/hardware/ar300m/
openwrt官网相关:https://openwrt.org/toh/hwdata/gl.inet/gl.inet_gl-ar300m
GL-AR300M Series
Best Mini Router Ever
AR300M is the our best mini router. It is small and powerful.
It has several sub-models:
Sub-model | Nor flash | Nand flash | Antenna | 5G wifi | USB power control |
---|---|---|---|---|---|
AR300M | 16MB | 128MB | Internal | No | Yes |
AR300M-Ext | 16MB | 128MB | External | No | Yes |
AR300M16 | 16MB | No | Internal | No | Yes |
AR300M16-Ext | 16MB | No | External | No | Yes |
AR300M-Lite | 16MB | No | Internal | No | No |
AR300MD* | 16MB | 128MB | Internal | Yes | Yes |
*AR300MD 5Gwifi is suggested only to work at monitoring mode because of power consumption.
Hardware Specification
Model | GL-AR300M Series |
---|---|
CPU | QCA9531 650MHz |
Memory | DDRII 128MB |
Storage | 16MB Nor, 128 MB Nand (Optional) |
Interfaces | 1 WAN, 1 LAN, 1 USB2.0, 1 Micro USB (power), 1 Reset Button, PCIe, UART |
Frequency | 2.4GHz |
Transmission Rate | 300Mbps |
Max. Tx Power | 20dBm |
Protocol | 802.11 b/g/n |
External Drive Format Support | FAT32/NTFS/EXT4/EXT3/EXT2 |
Webcam Support | MJPEG, YUV (not support from firmware v2.27) |
DIY Features | UART, GPIO, PCIe, 3.3V & 5V power port |
External Antenna Support | Yes (optional) |
PoE Module Support | No |
Power Input | 5V/1A |
Power Consumption | <2W |
Dimension, Weight | 58mmX58mmX25mm, 39g |
PCIe headers | Yes |
PCB Pinout
Note: I2C is not working in some early version of the router.
By using 128MB Nand flash, space is not a problem and the writing speed is improved greatly. You can compile a firmware with tons of features and writing data is almost instant.
By adding PCIe interface, a 5G 802.11a/ac addon module can be connected and you can enjoy the wifi speed on a mini router.
Performance
By adding a fast Nand flash, you can feel how the performance is improved compared to a Nor flash router.
The performance comparision between AR300M (nand) with AR150 is below:
It only takes 6 seconds to upload and process 250 ovpn files. When installing packages to the router, it takes no more than 20% for the time spent than using nor flash, including the time to download resource from the Internet. So everything just happends in seconds.
Building Firmware
The router has two flash so we have two firmwares. One firmware is installed on the Nor flash and one is installed on the nand flash. The two firmware is independent from each other. The router will always try to boot from the nand flash. If it fails for 3 times it will boot from the Nor flash.
By using two firmwares, businesses can have dual boot options to ensure the system can backup, upgrade and boot without problems.
Building the generic firmware for nor flash
The simple instructions of building the firmware can be found here: https://github.com/gl-inet/lede-17.01 You need to have a Linux machine. In this example we use Ubuntu.
$ sudo apt-get update
$ sudo apt-get install subversion build-essential git-core libncurses5-dev zlib1g-dev gawk flex quilt libssl-dev xsltproc libxml-parser-perl mercurial bzr ecj cvs unzip git wget
$ git clone https://github.com/domino-team/openwrt-cc.git
$ cd openwrt-cc
$ ./scripts/feeds update -a
$ ./scripts/feeds install -a
$ make menuconfig..choose your compile target and packages
$ make
If you want to build the firmware for nor flash, choose GL-AR300M
as the target.
Build the nand firmware for nand flash
If you want to build the firmware for nand flash, first choose subtarget Generic devices with NAND flash
then choose GL-AR300M Nand
as the target.
The make process will takes around 1 hour for the first time because it needs to download a lot of data from t he Internet. After compiling finished, you will find the following firmware files in bin/ar71xx/
- openwrt-ar71xx-generic-gl-ar300m-squashfs-sysupgrade.bin, for nor flash
- openwrt-ar71xx-nand-gl-ar300m-squashfs-sysupgrade.tar, for nand flash firmware upgrade inside of a working Openwrt firmware
- openwrt-ar71xx-nand-gl-ar300m-ubi.img, for initial nand flash firmware build, this is used when upgrading using uboot.
Note
There is two file for nand firmware. The xxx-ubi.img
is used for first time writing to the nand flash. If you are using uboot to upgrade the firmware, please use this one. The xxx-sysupgrade.tar
is used for upgrading from openwrt. So if you use our stock firmware or already using an openwrt firmware, choose this one.
Using Firmware
The firmware has exactly the same UI as our other mini routers. To determine you are using nor flash or nand flash, please check the available space from the UI.
Control which firmware you are booting into
There are two methods to control which firmware you can boot.
Method 1: set boot count
The router will always try to boot into the nand firmware. But if it fails for 3 times it will try to boot into the nor firmware. If you want to boot into the nor firmware, you can tell the router that it failed to boot for 3 times already by setting the uboot env. To do so, you can use the following command in a SSH terminal.
$ fw_printenv
$ fw_setenv bootcount 3
$ reboot
You need to do this each time if you want to boot into nor firmware, no matter you which firmware you are using. Otherwise after the firmware boot, it will reset bootcount
variable and next boot will always try to boot the nand firmware.
Method 2: using the switch
This is only valid in the new uboot (from Mar 2017).
First ssh to the router and set uboot env boot_dev
$ fw_printenv
$ fw_setenv boot_dev on
$ reboot
Then when the router boot, the router will choose which firmware to boot using the hardware switch. 1. Left side (near the reset button) --> Nand flash 2. Right side --> Nor flash
To disable using the switch to choose firmware function, clear boot_dev
variable.
$ fw_printenv
$ fw_setenv boot_dev
$ reboot
Debrick
If you brick your router and it cannot boot, you can try to save it using uboot.
Using uboot web UI
Follow this guide for a general instruction to get access of uboot web UI.
Here is a video guide for entering AR300M uboot failsafe.
In some early version of AR300M, the uboot boot very slowly. So be patient when you press and hold the reset button. The LEDs will start to flash around 5 seconds.
There is a bug in AR300M uboot and in some computers the connection will always reset when you upload firmware. Check this post to find the discussion. You can either try chaning the MTU or just try another computer.
Note: If you have an older version of uboot, the webUI will only be able to flash NAND flash. This is fixed in newer versions.
Note: When flash NAND firmware, only use the firmware with .img
in the name, not the one with .rar
in the name.
The following figure is the web ui of the new
uboot.
Using uboot console
If you want to use uboot console, you need to open the case and solder the serial connectors.
Then you need to set up a tftp server.
in uboot console, the command to flash nor flash is run lf
to flash nand firmware is run nlf
glinet ar300m-lite相关推荐
- Paddle Lite端侧部署
Paddle Lite端侧部署 端侧推理引擎的由来 随着深度学习的快速发展.特别是小型网络模型的不断成熟,原本应用到云端的深度学习推理,就可以放到终端上来做,比如手机.手表.摄像头.传感器.音响,也就 ...
- 超轻量AI引擎MindSpore Lite
超轻量AI引擎MindSpore Lite 揭秘一下端上的AI引擎:MindSpore Lite. MindSpore Lite是MindSpore全场景AI框架的端侧引擎,目前MindSpore L ...
- MindSpore图像分类模型支持(Lite)
MindSpore图像分类模型支持(Lite) 图像分类介绍 图像分类模型可以预测图片中出现哪些物体,识别出图片中出现物体列表及其概率. 比如下图经过模型推理的分类结果为下表: 使用MindSpore ...
- MindSpore Lite整体架构介绍
MindSpore Lite整体架构介绍 MindSpore Lite框架的总体架构如下所示: • 前端(Frontend): 负责模型生成,用户可以通过模型构建接口构建模型,将第三方模型和MindS ...
- 模型压缩95%:Lite Transformer,MIT韩松等人
模型压缩95%:Lite Transformer,MIT韩松等人 Lite Transformer with Long-Short Range Attention Zhanghao Wu, Zhiji ...
- GitHub上用于微控制器的TensorFlow Lite
这是TensorFlow Lite的实验端口,针对微控制器和其他只有千字节内存的设备.它不需要任何操作系统支持,任何标准的C或C ++库或动态内存分配,因此它的设计甚至可以移植到"裸机&qu ...
- TensorFlow Lite:TensorFlow在移动设备与嵌入式设备上的轻量级跨平台解决方案 | Google 开发者大会 2018...
Google 开发者大会 (Google Developer Days,简称 GDD) 是展示 Google 最新开发者产品和平台的全球盛会,旨在帮助你快速开发优质应用,发展和留住活跃用户群,充分利用 ...
- 基于Vue, Vuex 和 ElementUI 构建轻量单页Hexo主题Lite
Hexo Theme Lite Keep Calm, Lite and Writing. light single page blog application theme, using Vue, Vu ...
- 对标Mobileye!百度Apollo公布L4级自动驾驶纯视觉解决方案Apollo Lite
美国当地时间6月16日-20日,全球计算机视觉及模式识别领域顶级学术会议CVPR 2019(Conference on Computer Vision and Pattern Recognition) ...
最新文章
- python解一元二次方程步骤-使用Python解一元二次方程!
- 干货 | 架构师入门实战视频课程(一)
- 10种排序算法基础总结
- c++将int转换成string_Integer与int的区别 (== 与 equal)
- CPU缓存一致性协议MESI - 笔记
- c++中的类型转换--reinterpret_cast
- mysql003三表查询.三表查询.自链接查询
- android程序的入口点,常见android面试基础题
- python实现etl_为什么选择R而不是Python做ETL
- 菜鸟ING的博客终于开园了。
- python 读写文件 另存为_python读写文件(五)
- mysql 追加水印_php 使用GD库为页面增加水印示例代码
- Innodb 与自增长auto_increment
- SpringMVC 参数传递
- 一步完成 MySQL 向 Redis 迁移
- 风云激荡,成功者自强
- 2021-06-25
- 群晖 Drive 的团队文件夹显示 “无法取得“
- 用OBS直播时麦克风有杂音,求怎么取消。
- 读书笔记--《围城》