Long story short: after a power outage D2500CC won't boot anymore - black screen, so I enabled BIOS in recovery mode (jumper) and it booted into BIOS. After many experiments I realized that when I turn on PCIe ASPM D2500CC won't boot, when I disable it - it boots fine. Primary NIC on the MB is dead. When it's booting Linux says - "Invalid MAC Address: ff:d3:51:03:43:b2", in fact it is invalid, it's nowhere near what it used to be. In BIOS I can't see MAC for primary port - it's empty, secondary port works perfectly and I can see MAC in BIOS. In OS I can see both ports on the PCI device list, so it's not "fried", in fact led on the right side of port works, left doesn't.
The best thing - two weeks ago the warranty period ended, I have this board exactly 2 years and 2 weeks
Back to issue.
So I took e1000e source code and disable MAC check, now it boots w/o error and I can see the port in OS, ethx device appears with that invalid MAC. I didn't use the port and I'm not until I solve this.
So next what I did was I took a copy of EEPROM contents of each of port using "ethtool -e eth2 > file.txt" and "ethtool -e eth2 raw on > file.bin".
After comparing them and searching on the internet, secondary module looks fine from what I can tell (it's the working one, no surprise here ), first bytes are MAC address and some of the following bytes seems to show the same values what I have found on the internet. For the primary port - it's pretty much awful, nothing seems to be correct from what I know. MAC address in message "Invalid MAC address" is not the same as first bytes of EEPROM, so I assume EEPROM is full of invalid data now. Maybe because of this board won't boot with PCIe ASPM on and if I fix this, ASPM will work properly.
So the question is how to fix this.
One of the solutions I have found is get the content of EEPROM from working port, change the MAC address to broken port real MAC address and write byte-by-byte to the broken port using ethtool, so it will start working with supposedly correct EEPROM. This seems to be fine and as I have nothing to loose, I can try that.
Then I realized that maybe those ports are different, like primary and secondary differs one from other not just by MAC address. Can someone please get me a copy of EEPROM (using ethtool) of primary ethernet port from working D2500CC (you can replace MAC with XX and the send it so me)? So I can compare and see whether those are similar.
|Start of working ports EEPROM (changed MAC to XX)||Start of BROKEN ports EEPROM (invalid MAC is still shown)|
I tried reflashing BIOS to latest version - doesn't help.
Intel, please help me fix his.
Thanks for Your time.