[U-Boot] Bogus "Warning: failed to set MAC address" on Gumstix Overo
When booting master on a Gumstix Overo based system, I get the following messages during ethernet initialization:
Net: smc911x-0 Warning: failed to set MAC address
(Note that patch [1] is needed to move the Warning to a new line, otherwise it is after the previous line with no space)
In this system, the ethernet chip is loading the MAC address from an attached eeprom, so the warning is inaccurate. I've looked at the code some, and it is not clear to me the best way to suppress this message for the case where the hardware configures its MAC address on reset.
Philip
[1] http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/107639
Dear Philip Balister,
In message 4E68C67F.2080002@opensdr.com you wrote:
When booting master on a Gumstix Overo based system, I get the following messages during ethernet initialization:
Net: smc911x-0 Warning: failed to set MAC address
(Note that patch [1] is needed to move the Warning to a new line, otherwise it is after the previous line with no space)
In this system, the ethernet chip is loading the MAC address from an attached eeprom, so the warning is inaccurate. I've looked at the code
The warning gets printed when eth_write_hwaddr() fails.
What makes you claim that this is not correct? There _is_ a problem when eth_write_hwaddr() returns an error code.
Best regards,
Wolfgang Denk
On 09/08/2011 10:40 AM, Wolfgang Denk wrote:
Dear Philip Balister,
In message4E68C67F.2080002@opensdr.com you wrote:
When booting master on a Gumstix Overo based system, I get the following messages during ethernet initialization:
Net: smc911x-0 Warning: failed to set MAC address
(Note that patch [1] is needed to move the Warning to a new line, otherwise it is after the previous line with no space)
In this system, the ethernet chip is loading the MAC address from an attached eeprom, so the warning is inaccurate. I've looked at the code
The warning gets printed when eth_write_hwaddr() fails.
What makes you claim that this is not correct? There _is_ a problem when eth_write_hwaddr() returns an error code.
My point is the code shouldn't be trying to write the MAC address, when the MAC address is set in hardware. Earlier, (say 2011.6 but I am not being scientific here) u-boots did the write thing (IE not print the warning). This behavior has changed in the past few months.
Philip
Dear Philip Balister,
In message 4E68DA54.4000005@balister.org you wrote:
My point is the code shouldn't be trying to write the MAC address, when the MAC address is set in hardware. Earlier, (say 2011.6 but I am not being scientific here) u-boots did the write thing (IE not print the warning). This behavior has changed in the past few months.
Yes, we started unifying the code. Please see doc/README.drivers.eth
What is the problem with running eth_write_hwaddr() on this system? _This_ should be fixed.
Best regards,
Wolfgang Denk
participants (3)
-
Philip Balister -
Philip Balister -
Wolfgang Denk