I want to use the ESP-IDF framework to make a program that will recieve UDP packets via ethernet. I am not using an ESP32-Ethernet-Kit, but an ESP32 alongside a PHY (DP83848) in order to gain access to ethernet.
Pretty much every example I can find about UDP and ESP32 is using Wifi instead of ethernet.
What I want to do is use the basic ethernet example available in the espressif esp-idf git repository as a starting point and then add the functionality to recieve the datagrams.
If I'm not wrong, what I should do is create a new event handler and use the esp_eth_update_input_path API present in the esp_eth_driver.h, but I am not sure about how to use it. Can anyone help me with this?
esp_err_t esp_eth_update_input_path(
esp_eth_handle_t hdl,
esp_err_t (*stack_input)(esp_eth_handle_t hdl, uint8_t *buffer, uint32_t length, void *priv),
void *priv);
I posted almost the same question in the espressif ESP-IDF forum before posting this one here and got an following answer by user ESP_ondrej.
I'm posting it here in case this can help someone.
All protocols examples located in the ESP-IDF supports both Wifi and
Ethernet. You just need to configure it via idf.py menuconfig.
Please check the README at https://github.com/espressif/esp-idf/bl ...
/README.md
If you want to start from scratch with your Ethernet UDP application,
you can do so. You don't need to call esp_eth_update_input_path.
This function is used to direct L2 Ethernet frames to specific
callback function which you don't want since you want to communicate
via IP protocol. Therefore keep it as is, i.e. Ethernet frames are to
be processed by lwIP stack and just use socket API as you might be
used to from other platforms.
Related
I was looking all over the internet to find a way to capture incoming packets from a certain network interface, then I came across PCAP, TCPDUMP, I believe the most commonly used networking library out there is Boost/Asio, so I wanted to use this library in order to capture traffic, but apparently there is no example for using Raw sockets or other classes to listen for incoming packets to a certain NI, I would appreciate any help or examples on this.
We eventually found out the best option for sniffing incoming packets was Libtins.
libtins is a high-level, multiplatform C++ network packet sniffing and crafting library.
Its main purpose is to provide the C++ developer an easy, efficient,
platform and endianness-independent way to create tools which need to
send, receive and manipulate network packets.
It uses a BSD-2 license and it's hosted at github.
actually I am on the path of learning usb . so I want to make my own usb stack for stm32f4 mcu. But I got stuck at one point . actually the problem is my framework layer accept and recognise the control request and the request is get descriptor but the wlength is 64 bytes instead of 18(size of device descriptor). And also I transfer the data to in endpoint 0 but the usb host never send the in token after the setup stage completion instead the setup stage is called again and after windows show the prompt message not able to recognise the device
Please help me what is that I am missing in my framework
Are there currently solutions where your server can act as the peer of a WebRTC connection?
The reason I am interested in WebRTC is not the peer-to-peer part of it, but because it enables you to use UDP. You could let players participate in a fast-paced game like Quake without needing any plugins.
It seems that essentially this same question was asked before, but surely things must now be quite different as 2 years have passed.
Yes, it is possible to deploy your WebRTC peer code on server. But since you need to run it on server, it's essentially different from how you run the WebRTC code within the browser - i.e. through a Java Script.
For server based WebRTC peer, you would need to use the WebRTC native code available on platforms - Windows, Mac OS X, Linux, Android and iOS. You can get the WebRTC native code from - https://webrtc.org/native-code/development/
Follow the instructions here to download and build the environment. Sample applications are also present in the repository at the locations - src/webrtc/examples and src/talk/examples
In summary you have use the WebRTC source code that is embedded in the browser in your application code and call the relevant methods / API for the WebRTC functionality.
I have answered similar question at: WebRTC Data Channel server to clients UDP communication. Is it currently possible?
We have implemented the exact same thing: a server/client way of using WebRTC. Besides we also implemented data port multiplexing, so that server would only need to expose one data port for all rtcdata channels.
Here is 2018 update for you: your off-the-shelf solutions are:
Red 5 Pro, Wowza, Kurento, Unreal Media Server, Flashphoner
Also note that in modern public networks TCP is not much slower than UDP;
but UDP may have a considerable packet loss, so try WebRTC-TCP for your Quake idea.
I don't know anything about wireless protocols. With this quick setup i want to get inside. I just want to write a windows desktop application which receives only a couple of bytes of data from a mcu+wifi module "NodeMCU" which is connected to the same router. I can write a simple desktop app with visual studio. But sending data over ip has many methods as far as i read.
What is the simplest way to achieve? (I'm experienced in STM32 mcu's and electronics except internet protocols.)
You can send the data from the embedded board (with the mcu) to a TTL-To-WiFi module. Then from your PC application you should open a socket and connect to the module IP to listen in the correct port the data from your board.. There are a lot of module in the market used as "bridge" or converter: you have only to setup in it the local IP address, the destination IP (your PC) and the TCP port where you want to send data. From point of view of the board you use a standard USART and the message is sent to the destination IP. In your PC application you should be able to create a socket and connect it to the wi-fi module.
On my mac I have two kinds of networks available - Ethernet, WiFi.
While making a server call, can I somehow control which network channel to use for making the server call? So, before making server call, I want to specify the network channel to be used for that call - Ethernet or WiFi.
How can this be achieved using objective C. I am working on a cocoa application.
I assume both NIC's are connected to Internet (so both have a IP):
I don't think you can solve it within code (not 100% sure). But what you could do is setup some local routes, configuring which traffic goes over what NIC.
Look at the route command ('man route').
This might help you:
https://serverfault.com/questions/100613/public-traffic-to-go-over-1-nic-and-private-traffic-to-use-another-nic
You can modify the routes available with the System Configuration framework. In scutil(8) you can see the routes that are presently installed in the State:/Network/Service/* dictionaries, and in order to manipulate these programmatically you have to us the SCDynamicStore framework, which is C.
However, if you were trying to just do some ad-hoc service on WLAN only, you could use the CoreWLAN framework, which is in Objective-C.