Just an update on my findings with the MAC Address problem.  It seems that
on first boot, the mac address is set in stone in the u-boot environment
variable ethaddr.  It cannot be manually changed by me if I went into the
u-boot configuration menu:
https://www.denx.de/wiki/view/DULG/UBootEnvVariables

So if you then move that SD card that has been booted in one particular
e310 to another, the mac address will follow.  That is the problem I have,
and I wanted to see if that also happens with others' e310s?  Or maybe it
is the particular sdimg file I have downloaded (v3.15.0.0).

I guess to solve this issue myself without digging in too deeply to the
nuances of bitbaking (which I've already started and in a little bit over
my head), I could just copy a fresh uboot image to the 7 other SDRs I have,
and theoretically on bootup u-boot would detect the hw address and set
their unique MAC address, which is what should happen.  But running md5sum
on u-boot.img and boot.bin, to my dismay I have found out the checksum
doesn't change after boot from a fresh sdimg file's u-boot.img and
boot.bin.  So where could the u-boot environment variables actually be
stored on the sd card?

Thanks,
Andrew

On Wed, Nov 25, 2020 at 2:21 PM Andrew Payne <[email protected]> wrote:

> Of the 8 e310s that I have tested with the same image, only one of them
> has the actual MAC address of 00:80:2f:19:4c:37, which I was able to find
> out via:
> hexdump /sys/bus/nvmem/devices/0-00510/nvmem
>
> Weird.  Glad I found the one device that had that MAC though. So that must
> have been patient 0. Then I copied that image to the others. But the
> question is why did they adopt the same MAC address instead of reading from
> the PROM at every boot up?  I'll have to dig deeper into U-Boot to see.
>
> On Wed, Nov 25, 2020 at 2:09 PM Marcus D Leech <[email protected]>
> wrote:
>
>> Near as I can tell the e310 gets its MAC from the motherboard i2c eeprom.
>>
>> Sent from my iPhone
>>
>> On Nov 25, 2020, at 12:55 PM, Andrew Payne via USRP-users <
>> [email protected]> wrote:
>>
>> 
>> So does the U-Boot environment determine the MAC address of eth0 prior to
>> booting the kernel?  Or is it during the make of the boot image?  I just
>> find it a stretch to believe that I am running an image on a number of
>> SDRs, and they're all using the same MAC address, possibly from the e310
>> for which the boot image was generated?
>>
>>
>>
>> On Wed, Nov 25, 2020 at 8:59 AM Philip Balister <[email protected]>
>> wrote:
>>
>>> Mine is different, but both seem to be from the National Instruments
>>> block (which is good):
>>>
>>> root@ni-e31x-30D280A:~# ifconfig
>>> eth0      Link encap:Ethernet  HWaddr 00:80:2F:25:02:37
>>>           inet addr:192.168.11.152  Bcast:192.168.11.255
>>>
>>> I see a warning early in boot about using a MAC address from ROM. I'm
>>> not sure where that came from.
>>>
>>> Philip
>>>
>>> On 11/25/20 8:36 AM, Andrew Payne via USRP-users wrote:
>>> > ---------- Forwarded message ---------
>>> > From: Andrew Payne <[email protected]>
>>> > Date: Wed, Nov 25, 2020 at 8:35 AM
>>> > Subject: Re: [USRP-users] E310: Change MAC Address?
>>> > To: Philip Balister <[email protected]>
>>> >
>>> >
>>> > It is: 00:80:2f:19:4c:37
>>> >
>>> > When is that .patch file ran, is it during compilation of the BSP?
>>> When
>>> > would I see that error message printf("I2C EEPROM MAC address read
>>> > failed\n");?
>>> >
>>> > How could I experiment with compiling a new BSP with the UHD 3.15
>>> > environment loaded, so then I may insert my own MAC Address or get the
>>> i2c
>>> > EEPROM read to work? Do I need
>>> > https://github.com/EttusResearch/meta-ettus
>>> > and
>>> > https://github.com/Xilinx/u-boot-xlnx
>>> > and may be use this guide:
>>> >
>>> https://opencpi.github.io/bsp_e310/Ettus_E3xx_Getting_Started_Guide.pdf
>>> > ?
>>> >
>>> > Thanks,
>>> > Andrew
>>> >
>>> >
>>> > On Wed, Nov 25, 2020 at 8:26 AM Philip Balister <[email protected]>
>>> wrote:
>>> >
>>> >> What is that mac address? I'd like to see if it matches mine.
>>> >>
>>> >> If I recall correctly, the mac address is flashed into the i2c eeprom,
>>> >> u-boot is supposed to read that and use it. Factory test should set it
>>> >> up uniquely. But my memories fade. Inspecting the u-boot source might
>>> >> show us ....
>>> >>
>>> >>
>>> >>
>>> https://github.com/EttusResearch/meta-ettus/blob/zeus/meta-e31x/recipes-bsp/u-boot/files/0001-ni-zynq-Add-support-for-NI-E31x-SG1-SG3-boards.patch#L397
>>> >>
>>> >> I didn't sift through the ifdef's though.
>>> >>
>>> >> Philip
>>> >>
>>> >> On 11/24/20 7:28 PM, Andrew Payne via USRP-users wrote:
>>> >>> Is there a way to change the mac address of my e310s?  I have imaged
>>> >>> several e310s using the image file in
>>> >>> https://files.ettus.com/binaries/cache/e3xx/meta-ettus-v3.15.0.0/,
>>> and
>>> >> they
>>> >>> all have the same MAC address.
>>> >>>
>>> >>> On some searching around the web I can see that it is likely that
>>> the MAC
>>> >>> address parameter is specified as a U-boot parameter.  I'd like to
>>> know
>>> >>> what my options are for changing it.
>>> >>>
>>> >>> Thanks,
>>> >>> Andrew
>>> >>>
>>> >>>
>>> >>> _______________________________________________
>>> >>> USRP-users mailing list
>>> >>> [email protected]
>>> >>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>> >>>
>>> >>
>>> >
>>> >
>>> > _______________________________________________
>>> > USRP-users mailing list
>>> > [email protected]
>>> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>> >
>>>
>> _______________________________________________
>> USRP-users mailing list
>> [email protected]
>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>
>>
_______________________________________________
USRP-users mailing list
[email protected]
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Reply via email to