[ipxe-devel] DHCP problems on Intel 82567LM

Christian Hesse list at eworm.de
Mon May 5 10:39:01 BST 2014


Christian Hesse <list at eworm.de> on Mon, 2014/05/05 10:26:
> Michael Brown <mcb30 at ipxe.org> on Fri, 2014/05/02 17:40:
> > On 02/05/14 11:01, Christian Hesse wrote:
> > >> Is the datapath broken in both directions?
> > >
> > > No. The Dell notebook (the one acting as client and having the problem)
> > > receives packets. I can not see a single ethernet frame on the network
> > > sent from the notebook.
> > >
> > >> Are you able to transmit ARP
> > >> requests or broadcast pings, or observe received data via tcpdump?
> > >
> > > tcpdump receives all types of packets on the notebook.
> > 
> > If you try sending a large number of packets (e.g. using arping or a 
> > broadcast ping: anything that won't wait for a reply before sending out 
> > the next packet) then what happens to the TX packet counts and error 
> > counts reported by ifconfig?
> 
> server# arping -I device 172.16.1.5
> ARPING 172.16.1.5 from 172.16.1.1 device
> 
> notebook# arping -I enp0s25 172.16.1.1
> ARPING 172.16.1.1 from 172.16.1.5 enp0s25
> 
> notebook# ip -s link show dev enp0s25
> 2: enp0s25 [...]
>     link/ether 01:23:45:c0:ff:ee brd ff:ff:ff:ff:ff:ff
>     RX: bytes  packets  errors  dropped overrun mcast   
>     2214       346      0       0       0       0      
>     TX: bytes  packets  errors  dropped carrier collsns 
>     2892       4        0       0       0       0 
> 
> RX packet and byte count increases, TX count does not.
> 
> Actually the system has a TX packet count of four, no idea where this comes
> from. I can not increase it any further.

Next reboot I had three TX packets. I could dump them on the other side, dump
file is attached. No idea what this is, though...

> > Does tcpdump running on the notebook see its own transmitted packets?
> 
> Yes, it does.
> 
> > > Did a warm reboot and tried to boot via UNDI. BIOS reported:
> > >
> > >> PXE-E05: The LAN adapter's configuration is corupted or has not been
> > >> initialized. The Boot agent cannot continue.
> > >
> > > Though this is not reproducible. After a cold reboot everything ist just
> > > fine again.
> > 
> > Good to know, but it's probably going to be easier to try to diagnose 
> > the problem from within Linux first.
> 
> Ok. :D
> 
> Does it help if I provide ssh access to a device suffering the issue? There
> is no sensitive data on it.

Looks like I do have network connectivity when booting Windows. So the
Windows driver does some magic to work around the problem.
Rebooting to Linux (without powering off completely) the problem still
persists.
-- 
main(a){char*c=/*    Schoene Gruesse                         */"B?IJj;MEH"
"CX:;",b;for(a/*    Chris           get my mail address:    */=0;b=c[a++];)
putchar(b-1/(/*               gcc -o sig sig.c && ./sig    */b/42*2-3)*42);}
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dump2.pcap
Type: application/vnd.tcpdump.pcap
Size: 1131 bytes
Desc: not available
URL: <http://lists.ipxe.org/pipermail/ipxe-devel/attachments/20140505/7b2f4a1c/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: not available
URL: <http://lists.ipxe.org/pipermail/ipxe-devel/attachments/20140505/7b2f4a1c/attachment.asc>


More information about the ipxe-devel mailing list