[Bug] CubeMX generates wrong ETH_PHY_ADDRESS for the Nucleo-144 F7

[carmine]

Posted on January 21, 2016 at 18:45

Hi,

I want just to signal that CubeMX generates a wrong ETH_PHY_ADDRESS inside the stm32f7xx_hal_conf.h file when a project for the Nucleo-144-F7 is generated. The address is declared as equal to 1:

...
#define DP83848_PHY_ADDRESS 0x01
...

while it should be equal to 0x0 as clearly stated inside the Nucleo-144-F7 documentation. This causes that the code inside thelow_level_init() function is wrong. #lwip-stm32f7-nucleo-144

[Nesrine M_O]

Posted on January 22, 2016 at 10:45

Hi cnoviello,

Thank you for your feedback. The issue has been reported internally.

-Syrine-

[stm32cube-t]

Posted on February 03, 2016 at 15:31

Dear user,

Thank you for reporting this issue.

STM32CubeMX 4.13 release will fix the Ethernet Phy Address to be zero for NucleoF7 boards.

Note the following should be removed from hal_conf.h as it is board specific and actually not a component on Nucleo boards:

* DP83848 PHY Address*/ 

#define DP83848_PHY_ADDRESS 0x01

Best regards

[alain2]

Posted on February 26, 2016 at 21:01

Hi,

It seems that the same issue can still occur for the 32F746GDISCOVERY board in STM32CubeMX 4.13.0. If I select the board, do not select ''Initialize all IP with their default mode'' and then select ETH and LWIP, the generated ethernetif.c contains:

  heth.Init.PhyAddress = 1;

With that line, I cannot ping the board.

If I change to:

  heth.Init.PhyAddress = 0;

rebuild and run, I can ping the board.

In an earlier project where I had selected ''Initialize all IP with their default mode'', the problem did not occur.

Best regards,

Alain Mosnier