From: Chad N. Tindel <ctindel@fa...> - 2002-03-01 22:41:18
I thought this would be relevant for the rest of the list as well.
---------- Forwarded message ----------
Date: Fri, 01 Mar 2002 14:37:16 -0800
From: Adam D Scislowicz <adams@...>
To: Chad N. Tindel <ctindel@...>
Subject: Re: [Bonding-devel] 2.4.18 w/ or w/o the bonding-20020226 patch
Doht! I had an older version of ifenslave. Now that Im
using the newer ifenslave this problem does not occur.
So all is well for me with 2.4.18-rc4 w/ bonding-2.4.18-20020226 :)
Thanx for your time.
/)dam.. . .
Chad N. Tindel wrote:
>>Oh, after tracing the ifconfig call I found it
>>was segfaulting while attempting to display
>>the 'Media' type, this was in ifconfig's src:
>> net-tools-1.60/lib/interface.c (search for 'Media')
>>Only the eth0 interface caused the segfault, as
>>it had the configurable media type flag set after
>>being bound as the primary slave(eth0's MAC
>>was automatically used), yet the media type
>>index was 8, which is out of bounds(6 is the max).
>>For now I just hacked interface.c to do a bounds
>>check, but its interesting because it occurs
>>only to the primary slave device after ifenslave
>>When I get into work tommorow Ill check the
>I think this is the same problem that Peter was reporting, only he was
>seeing corrupted ifconfig output as opposed to a sigsegv. I haven't been
>able to reproduce this here. If someone out there is ambitious enough to
>fix the problem I'd definitely accept the patch. David Miller just accepted
>the patch I gave him, so if we can get this fixed before 2.4.19 that'd be