Re: [E1000-devel] e100e driver hangs when setting an IP with ipconfig

[Reiner P. <pr...@ps...>]

Hello,

I have tried the 7.6.15.5 driver which causes in segment fault in 
ifconfig during settin the IP address.

invalid operand: 0000
CPU:    0
EIP:    0010:[<f8cd1006>]    Not tainted
EFLAGS: 00010206
eax: f8cd1000   ebx: f5e1e000   ecx: 00000000   edx: 000049d4
esi: c02fcfd4   edi: f5be5f20   ebp: c02fcf60   esp: f59dde7c
ds: 0018   es: 0018   ss: 0018
Process ifconfig (pid: 1563, stackpage=f59dd000)
Stack: f8cd1000 f5be5f20 c02fcfd4 c02fcf60 f59ddea0 f5e1e000 000049d4 
00000000
       f8cd1000 c02165c4 f5e1e000 f5be5f20 f653fef4 c02fd39c f653ff30 
f5e1e000
       f5be5f20 c023fccf f5e1e000 c02fcf60 00000000 f7964a60 f5e1e000 
bffff4c0
Call Trace:    [<c02165c4>] [<c023fccf>] [<c0240190>] [<c0240a16>] 
[<c02428f7>]
  [<c020a5d5>] [<c0142d69>] [<c01089d7>] [<c01087ab>]

Code: c5 f8 d0 af cf f8 14 c6 02 00 00 00 00 00 01 00 00 00 dd 00

Reiner
 

Tantilov, Emil S schrieb:
> Hi Reiner,
>
> So it seems that the issue is somewhat related to the specific kernel you are trying to use. 
> I'm sure we can figure something out but we have to be able to reproduce it first before we can root-cause it.
>
> I think I have your kernel config and will try to look at it when I get a chance, but with issues like these sometimes it's important to have the exact same combination of HW and SW in order to reproduce it.
>
> Can you please try the old e1000 driver (before the e1000/e1000e split)? You can download it from e1000.sf.net - the driver is under the e1000 download section and it is version 7.6.15.5. This driver should support your device.
>
> Also could you please copy the e1000-devel mailing list in your replies? There are others who monitor the list that maybe able to help.
>
> Thanks,
> Emil 
>
> -----Original Message-----
> From: Reiner Pröls [mailto:pr...@ps...] 
> Sent: Friday, October 31, 2008 1:32 AM
> To: Tantilov, Emil S
> Subject: Re: [E1000-devel] e100e driver hangs when setting an IP with ipconfig
>
> Hello,
>
> I have compiled your e1000e driver using the 2.4.36.8 kernel. It 
> compiles modprobe works with the following messages
>
> e1000e: Intel(R) PRO/1000 Network Driver - 0.4.1.12-NAPI
> e1000e: Copyright (c) 1999-2008 Intel Corporation.
> PCI: Setting latency timer of device 03:00.0 to 64
> 03:00.0: : Failed to initialize MSI interrupts.  Falling back to legacy 
> interrupts.
> 03:00.0: eth0: (PCI Express:2.5GB/s:Width x1) 00:0e:8c:9e:a7:a8
> 03:00.0: eth0: Intel(R) PRO/1000 Network Connection
> 03:00.0: eth0: MAC: 3, PHY: 2, PBA No: ffffff-0ff
> PCI: Setting latency timer of device 04:00.0 to 64
> 04:00.0: : Failed to initialize MSI interrupts.  Falling back to legacy 
> interrupts.
> 04:00.0: eth1: (PCI Express:2.5GB/s:Width x1) 00:0e:8c:9e:a7:a7
> 04:00.0: eth1: Intel(R) PRO/1000 Network Connection
> 04:00.0: eth1: MAC: 3, PHY: 2, PBA No: ffffff-0ff
> 03:00.0: eth0: Link is Up 100 Mbps Full Duplex, Flow Control: RX/TX
>
> and I can bring up eth0 and assign an IP address. So it works with 
> 2.4.36.8. My problem is thst we need kernel 2.4.24 because we use a 
> realtime extension. The computer is an industrial PC from Siemens with 
> two ethernet connectors onboard.
> Do you see any possibility to get it running (a temporary solution would 
> be to plug in a very old Realtec ethernet card we used in the past, but 
> I don't like such solutions!)?
>
> Thanks
> Reiner
>
> Tantilov, Emil S schrieb:
>   
>> Reiner Pröls wrote:
>>   
>>     
>>> Hello,
>>>
>>> the System is based on Suse 9.0 but patched to use a realtime
>>> extension (but not loaded at the moment I have done my tests)
>>> Linux testwidth 2.4.24-rtl3.2-pre3 #5 Mon Jul 24 10:12:48 CEST 2006
>>> i686 i686 i386 GNU/Linux
>>>
>>> I have tried it with the vanilla kernel 2.4.36.8 and there the
>>> supplied driver with the kernel (e1000) works.
>>>     
>>>       
>> Does e1000e 0.4.1.12 fails with this kernel as well?
>>  
>>   
>>     
>>> I have also tried to compile the actual driver with the original Suse
>>> 9.0 kernel 2.4.21 but it does not work
>>>     
>>>       
>> This is a known problem. See the following link for fix:
>> http://sourceforge.net/tracker/index.php?func=detail&aid=1620185&group_id=42302&atid=447449
>>
>> Thanks,
>> Emil
>>   
>>