STM32F427VI - HardFault_Handler because of ethernet configuration

[As.51]

Hi all, I moved a code that was working on STM32F427VI and somehow I am facing a new issue that I have never see before.

My code works in debug and release when the unit is compiled from Visual Studio and the Ethernet cable is connected or disconnected. But, after I reset the unit (power off/power on) the unit is working when the ethernet cable is connected. If the cable is disconnected go over the while true inside HardFault_Handler. I think that the problem is on that function:

bool configureNetwork(network_config& netconfig)
{
	struct ip4_addr ipaddr;
	struct ip4_addr netmask;
	struct ip4_addr gw;
	struct ip4_addr dns;
	
	network_config::setIPAddr(&ipaddr, netconfig.ip);
	network_config::setIPAddr(&netmask, netconfig.mask);
	network_config::setIPAddr(&gw, netconfig.gateway);
	network_config::setIPAddr(&dns, netconfig.dns);
	
	if (gnetStarted) {
		handleDHCP(netconfig.isDynamic());
		if (curDynamic == false) {
			netif_set_addr(&gnetif, &ipaddr, &netmask, &gw);			
			dns_setserver(0, &dns);
		}
	}
	else {
		STARTUP_TRACE.Write("before: netif_add\r\n", 1000);
		netif_add(&gnetif, &ipaddr, &netmask, &gw, NULL, &ethernetif_init, &ethernet_input);
		STARTUP_TRACE.Write("after: netif_add\r\n", 1000);
		/*  Registers the default network interface */
		netif_set_default(&gnetif);
		
		netif_set_status_callback(&gnetif, netstatus_callback);
  
		if (netif_is_link_up(&gnetif))
		{
		  /* When the netif is fully configured this function must be called */
			netif_set_up(&gnetif);
		}
		else
		{
		  /* When the netif link is down this function must be called */
			netif_set_down(&gnetif);
		}
		handleDHCP(netconfig.isDynamic());
		if (curDynamic == false) {
			dns_setserver(0, &dns);
		}
		
		netif_set_link_callback(&gnetif, ethernetif_update_config);
		
		gnetStarted = true;
		sys_timeout(LINK_CHECK_INTERVAL, second_timeout, NULL);
	}	
		
	return true;
		
}

I have this conclusion becasue when I remark this function, the hardFaultHandler is called even when the unit is in debug on visual studio. So the result is the same.

I don't understand what is the difference when the mcu is conencted to the debugger and why the handler is called when the cable is disconnected. I have this issues on the Ethernet for the first time.

Any idea?

[TDK]

Attach a debugger and see where the code is getting stuck. Instrument the hard fault handler to give information about the error state.

[As.51]

The code reachs here

extern "C" void HardFault_Handler(void)
{
  /* USER CODE BEGIN HardFault_IRQn 0 */
 
  /* USER CODE END HardFault_IRQn 0 */
  while (1)
  {
    /* USER CODE BEGIN W1_HardFault_IRQn 0 */
	STARTUP_TRACE.Write("HardFault_Handler:stm32f4xx_it\r\n", 1000);  	  
    /* USER CODE END W1_HardFault_IRQn 0 */
  }
}

I think that I have a problem when I set the lwip interface as dynamic (DHCP_ and the IP = 0 (when the IP is loaded from flash from previous runniung).

In this case the &ethernet_input is not init in the correct mode and then I got an error here: err = netif->input(p, netif); becasue it is null.

 if (p == NULL) return;
    
  /* entry point to the LwIP stack */
  err = netif->input(p, netif);
  if (err != ERR_OK)
  {
    LWIP_DEBUGF(NETIF_DEBUG, ("ethernetif_input: IP input error\n"));
    pbuf_free(p);
    p = NULL;    
  }
}

[Piranha]

ST has never had a working link detection code. If you're using RTOS, then your whole code is broken as it doesn't follow lwIP API rules. The same is true for ST's non-working HAL/Cube bloatware:

https://community.st.com/s/question/0D50X0000BOtfhnSQB/how-to-make-ethernet-and-lwip-working-on-stm32

[As.51]

The same code works excellent on my previous ST MCU...

[As.51]

I got those messages when the cable is disconnected on DHCP

Assertion "netif is not up, old style port?" failed at line 727 in C:/Users/alex_s/STM32Cube/Repository/STM32Cube_FW_F4_V1.24.1/Middlewares/Third_Party/LwIP/src/core/ipv4/dhcp.c

Assertion "unknown dns_table entry state:" failed at line 1079 in C:/Users/alex_s/STM32Cube/Repository/STM32Cube_FW_F4_V1.24.1/Middlewares/Third_Party/LwIP/src/core/dns.c

[As.51]

I understand your point but I need to use DHCP and use the Ethernt on runtime, I can't reset the unti every time that it lose connection.

I changed this code:

 if (hal_eth_init_status == HAL_OK)
  {   
    netif->flags |= NETIF_FLAG_LINK_UP;
  }
  

I just removed the IF because when the cable is disconnected the status is eqaul to timeout. The questions is if it is ok to always configure the link_up flag?

The code works now. I start the MCU with DHCP and the cable disconnected, then plug and unplug multipletimes and I see that the communication start/stop accordingly.