Intel 82567LM e1000e DHCP problem/question

Previous thread: [PATCH 2/2] r8169: revert "r8169: read MAC address from EEPROM on init (2nd attempt)" by Francois Romieu on Sunday, March 15, 2009 - 4:10 am. (4 messages)

Next thread: [PATCH] bonding: select current active slave when enslaving device for mode tlb and alb by Jiri Pirko on Sunday, March 15, 2009 - 9:02 am. (1 message)
From: Pelle Svensson
Date: Sunday, March 15, 2009 - 7:04 am

Hi,

Quick question.

Is there any functionality inside the driver for DHCP request.
I see in the source that there is some kind of DHCP filter f the tx stream.
Can the drive issue DHCP request inside it self when the Cable is unplugged/plugged.

I'm tracing a DHCP fault. 
When the net is brought down, ifdown eth0, the dhclient is killed.
Then when I unplugged/plugged the cable I get a DHCP request even that 
dhclient is down.

Also the Wireshark format of the DHCP request are different.
Incorrect requests when dhclient is killed dose not have above.


/Thanks

_________________________________________________________________
Dela foton på ett smidigt sätt med Windows Live™ Photos.
http://www.microsoft.com/windows/windowslive/products/photos.aspx--

From: Brandeburg, Jesse
Date: Monday, March 16, 2009 - 11:38 am

good idea to include e1000-devel@lists.sourceforge.net on these kinds of 
requests.





you apparently have the Intel AMT firmware (ME aka Management Engine) 
enabled in your BIOS, you can try to configure it by hitting CTRL-P during 
post (pre-boot) or at the prompt where it says to hit CTRL-P to enter 
system management setup.

The ME (management engine) firmware is trying to acquire its own dhcp 
address and will do so if the driver is either not loaded or not up.

Once the driver is loaded the firmware may still try to acquire a second 
dhcp address for the ME, depending on your configuration.  This whole 
setup works a lot like IPMI, but generally the ME is not configurable via 
IPMI interfaces.

Hope this helps,
  Jesse
--

From: Pelle Svensson
Date: Monday, March 16, 2009 - 2:31 pm

Yes the DHCP server cause the fault in a short answer.
In the long answer it depends on several things. 
By the 2 kinds of request made by the client to the server,
it issued new IP adresses for each of the type of request rather frequently.
This ends up, after a while, with almost an idential faults as this bug.

Your assumption about the Intel AMT seems right!

Log Tracking DHCP Request when plugged net cable in:

- Power on and load Fedora -> No DHCP Request from BIOS until Desktop up
- ifdown eth0 -> DHCP Request
- Restart Fedora -> DHCP Request during the hole Shutdown process
- Enter BIOS menu -> DHCP Request
- Boot up Fedora -> DHCP Request from BIOS until Desktop up

ifdown eth0 trigged the plugged net cable DHCP Request

I tried to hit Ctrl-P frequently during the boot process but it newer display any


management setup. I'm running Fedora 10 on Lenovo T500 bios 1.20 (6FET50WW).





Only AMT stuff I can find for Lenovo is related to Vista or XP, But that seems to


for the OS not the BIOS. Can I load the Windows stuff and turn the AMT DHCP

stuff off? I do have dual boot for XP on this PC.



/Thanks
Pelle





_________________________________________________________________
Dela foton på ett smidigt sätt med Windows Live™ Photos.
http://www.microsoft.com/windows/windowslive/products/photos.aspx--

From: Brandeburg, Jesse
Date: Monday, March 16, 2009 - 10:23 pm

Okay, maybe there is a bios setting you can change.

I am not experienced in configuring this on a t500, so I can't give you 
direct instructions.

I did find the instructions on the Intel website that someone else on the 
web said they were able to use to configure iAMT.  I think you're only 
looking to disable it.


I don't think that is what you want, or at least it might be more work 
than you want.

let us know how it goes.
--

From: Bjørn Mork
Date: Tuesday, March 17, 2009 - 2:27 am

Assuming it's similar to a Thinkpad X301:

 ThinkVantage -> F1 -> Config -> Intel(R) AMT -> Disable

Totally offtopic, but IBM used to have a nice support tool called BIOS
simulators: http://www-307.ibm.com/pc/support/site.wss/TPAD-SIMS.html
But I haven't been able to find such simulators for the newer Lenovo
Thinkpads.




Bjørn
--

From: Pelle Svensson
Date: Tuesday, March 17, 2009 - 5:17 am

I don't have that entry and it seems to be a mystery one.
This guy has 2 identical T500 and one dose not have AMT

Nice tool but it is not up to date with an T500 entry as you say!


I'm more convinced now that it is AMT and that I need to get it turned off.

I did a session checking the DHCP Request when the cable is inserted.

- Power on 
- Boot XP - No Request Until XP ready.
- Restart - Request start when XP shut down graphics.
- Reboot - Requests
- Enter Grub - Requests
- Boot Fedora - Requests
- Fedora up - No Request.

In this case XP shut down trigged the plugg Requests.
So either XP of Linux shutdown stasrt the request.
One thing notice is that the DHCP Transaction number for
XP and Linux has format 0x12345678 and Host option.
The one comming from AMT has format 0x123456
and no Host option.

The bottom line is that DHCP Server gives different IP
according to the Host name (same MAC). 
This is right but there is a fault when the different hosts
are mixed frequently. Each MAC/Host dose not get the
same IP and thing start to get messy.

Have to look into that and leave AMT for the moment.

/Thanks

 







_________________________________________________________________
Dela foton på ett smidigt sätt med Windows Live™ Photos.
http://www.microsoft.com/windows/windowslive/products/photos.aspx--

Previous thread: [PATCH 2/2] r8169: revert "r8169: read MAC address from EEPROM on init (2nd attempt)" by Francois Romieu on Sunday, March 15, 2009 - 4:10 am. (4 messages)

Next thread: [PATCH] bonding: select current active slave when enslaving device for mode tlb and alb by Jiri Pirko on Sunday, March 15, 2009 - 9:02 am. (1 message)