Hello!
I have STM32H743 project running lwIP v.2.1.2 (release) supplied by STM with STM32H7 package v.1.9.1
There were some problems with H7 Ethernet HAL driver, but after some fixes, it looks reasonably stable.
The next problem I face is lwIP stops receiving data after some time. Depending on network activity it could be minutes or hours. While debugging my application I found that the tcpip_thread task hangs and does not fetch messages from the queue anymore.
Other tasks (Modbus, HTTPS client, UPnP, etc - about 10) are running. All user tasks have priority 24 (normal). tcpip_thread used to have priority 24 too. Increasing it up to 48 (real-time) did not help.
In order to force this situation I run:
- 3 x Modbus TCP get Holding Registers[0..13] @10Hz (30 TCP requests/s)
- 8 x ping -i 0.1 ip.of.my.mcu @10Hz (80 ICMP requests/s)
Plus there are Wireguard client and HTTPS client working in the background, but they are low on activity.
I agree that is quite large traffic, and I am ready to have some packets lost, but most critical for me is that after traffic goes down network doesn't restore its operation (hung until reboot).
Can somebody help me with this situation?
Thank you!
Kirill.