[STM32 NUCLEO] Init MAC address

[jeromecoutant]

Description

This patch set a MAC address depending on the target unique Id of each MCU.

Status

READY

Tests with STM32 NUCLEO boards are OK (ARM/IAR/GCC)

[sg-]

Is this globally assigned to ST?

[jeromecoutant]

yes, this is the official ST MAC address.
http://www.macmonster.co.uk/macoui=0080E1

[sg-]

how unique is the UUID and MSB, LSB? I tried this years ago and found that MSB, LSB didn't provide a great amount of randomness where in a room of 100 boards there would still be 2-5 conflicts. Ended up having to use CRC on the 128 to create 48 unique. Just some memories of bad times to avoid if not already accounted for.

[jeromecoutant]

Unique Id is unique for each target :-)
But it's true it is 96 bits...
As I set the ST Id in the 3 first MAC bytes, I checked with boards around me, and using MSB looks quite OK to get some random Id...

[sg-]

/morph test

[sg-]

@mbed-bot: TEST

HOST_OSES=ALL
BUILD_TOOLCHAINS=ALL
TARGETS=ALL

[mbed-bot]

Result: SUCCESS

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 919

All builds and test passed!

[mbed-bot]

[Build 965]
SUCCESS: Building succeeded and tests were run! Be sure to check the test results

[0xc0170]

@jeromecoutant Please see the comments above about the uniqueness of mac address

[sg-]

@jeromecoutant @andreaslarssonublox

There is a duplicate mbed_mac_address symbol now that #2743 and #2728 have been merged. I'll revert this for now #2808 since it seems that mbed_mac_address should be defined by a board before generically for all device given there is a WEAK version of the function.

Build failures:
18:10:59   * UBLOX_C029::ARM::TESTS-MBED_DRIVERS-STATS
18:10:59         Building project stats (UBLOX_C029, ARM)
18:10:59         Scan: ARM
18:10:59         Scan: FEATURE_IPV4
18:10:59         Scan: stats
18:10:59         Compile [100.0%]: main.cpp
18:10:59         [Warning] greentea_serial.h@13,0:  #1-D: last line of file ends without a newline
18:10:59         Link: stats
18:10:59         Error: L6200E: Symbol mbed_mac_address multiply defined (by stm32xx_emac.o and stm32f4_eth_init.o).
18:10:59         Not enough information to list the image map.
18:10:59         Finished: 1 information, 0 warning and 1 error messages.

[jeromecoutant]

@andreaslarssonublox

Do I remove mbed_mac_address function in TARGET_STM/TARGET_STM32F4/TARGET_UBLOX_C029/stm32f4_eth_init.c

or you prefer to keep your own function, and I can add
#if !defined (TARGET_UBLOX_C029)
in the generic STM32 function ?

[andreaslarssonublox]

@jeromecoutant
We would prefer to use our own MAC address.

[jeromecoutant]

Hi

I understand.
New Pull Request is on going, but I remaned your function from mbed_mac_address to mbed_otp_mac_address.

And add this in emac.c file:

void mbed_mac_address(char *mac) {
if (mbed_otp_mac_address(mac)) {
return;
} else {
mbed_default_mac_address(mac);
}
return;
}

I think it should be OK for you ?