Skip to contentSkip to conversation

Crash in ip_conntrack on shutdown

DISCUSSION Stephen Hemminger 3 participants 4 messages Nov 2, 2006

Subsystems (2)
lore
Share & tools
lore
maintainer
#1 lore raw

From: Stephen Hemminger <[email protected]>
Date: Wed, 01 Nov 2006 16:56:35 -0800

Quoted 7 lines

My desktop machine (Fedora Core 6) crashed on during shutdown when running latest
2.6.19-rc4 kernel. Machine is UP Amd64.

It looks like a null dereference in ip_conntrack. See screenshot.

http://picasaweb.google.com/stephen.hemminger/LinuxCrashes/photo#4992593950061821970

Can some x86_64 expert decode this? I'm trying to figure out what object is being NULL defer'd, and after taking a quick look at this dump I can't figure it out.

I suspect there may also be a OOPS right before this one, I also wonder if this was actually a BUG_ON() trigger.

#2 lore raw
Quoted 15 lines · On Wed, 01 Nov 2006 17:49:31 -0800 (PST) David Miller <[email protected]>…

From: Stephen Hemminger <[email protected]>
Date: Wed, 01 Nov 2006 16:56:35 -0800

Quoted 7 lines

My desktop machine (Fedora Core 6) crashed on during shutdown when running latest
2.6.19-rc4 kernel. Machine is UP Amd64.

It looks like a null dereference in ip_conntrack. See screenshot.

http://picasaweb.google.com/stephen.hemminger/LinuxCrashes/photo#4992593950061821970

Can some x86_64 expert decode this? I'm trying to figure out what object is being NULL defer'd, and after taking a quick look at this dump I can't figure it out.

Quoted 6 lines

I suspect there may also be a OOPS right before this one, I also wonder if this was actually a BUG_ON() trigger.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Instruction decode of ip_conntrack:destroy_conntrack+0xfc = 0x21ce

    2177:       75 0a                   jne    2183 <destroy_conntrack+0xb1>
    2179:       0f 0b                   ud2a
    217b:       68 00 00 00 00          pushq  $0x0
    2180:       c2 56 01                retq   $0x156
    2183:       48 8b 41 08             mov    0x8(%rcx),%rax
    2187:       48 89 42 08             mov    %rax,0x8(%rdx)
    218b:       48 89 10                mov    %rdx,(%rax)
    218e:       48 c7 41 08 00 02 20    movq   $0x200200,0x8(%rcx)
    2195:       00
    2196:       48 c7 83 f8 00 00 00    movq   $0x100100,0xf8(%rbx)
    219d:       00 01 10 00
    21a1:       ff 05 00 00 00 00       incl   0(%rip)        # 21a7 <destroy_conntrack+0xd5>
    21a7:       bf 01 00 00 00          mov    $0x1,%edi
    21ac:       e8 00 00 00 00          callq  21b1 <destroy_conntrack+0xdf>
    21b1:       e8 00 00 00 00          callq  21b6 <destroy_conntrack+0xe4>
    21b6:       48 8b 53 58             mov    0x58(%rbx),%rdx
    21ba:       48 85 d2                test   %rdx,%rdx
    21bd:       74 0f                   je     21ce <destroy_conntrack+0xfc>
    21bf:       ff 0a                   decl   (%rdx)
    21c1:       0f 94 c0                sete   %al
    21c4:       84 c0                   test   %al,%al
    21c6:       74 06                   je     21ce <destroy_conntrack+0xfc>
    21c8:       48 89 d7                mov    %rdx,%rdi
    21cb:       ff 52 08                callq  *0x8(%rdx)
    21ce:       48 89 df                mov    %rbx,%rdi  <--- last visible spot in back trace
    21d1:       5b                      pop    %rbx
    21d2:       e9 00 00 00 00          jmpq   21d7 <invert_tuplepr>
#3 lore raw
Quoted 46 lines · On Wed, 1 Nov 2006 18:48:56 -0800 Stephen Hemminger <[email protected]> wrote:
Quoted 15 lines · On Wed, 01 Nov 2006 17:49:31 -0800 (PST) David Miller <[email protected]>…

From: Stephen Hemminger <[email protected]>
Date: Wed, 01 Nov 2006 16:56:35 -0800

Quoted 7 lines

My desktop machine (Fedora Core 6) crashed on during shutdown when running latest
2.6.19-rc4 kernel. Machine is UP Amd64.

It looks like a null dereference in ip_conntrack. See screenshot.

http://picasaweb.google.com/stephen.hemminger/LinuxCrashes/photo#4992593950061821970

Can some x86_64 expert decode this? I'm trying to figure out what object is being NULL defer'd, and after taking a quick look at this dump I can't figure it out.

Quoted 6 lines

I suspect there may also be a OOPS right before this one, I also wonder if this was actually a BUG_ON() trigger.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Instruction decode of ip_conntrack:destroy_conntrack+0xfc = 0x21ce

    2177:       75 0a                   jne    2183 <destroy_conntrack+0xb1>
    2179:       0f 0b                   ud2a
    217b:       68 00 00 00 00          pushq  $0x0
    2180:       c2 56 01                retq   $0x156
    2183:       48 8b 41 08             mov    0x8(%rcx),%rax
    2187:       48 89 42 08             mov    %rax,0x8(%rdx)
    218b:       48 89 10                mov    %rdx,(%rax)
    218e:       48 c7 41 08 00 02 20    movq   $0x200200,0x8(%rcx)
    2195:       00
    2196:       48 c7 83 f8 00 00 00    movq   $0x100100,0xf8(%rbx)
    219d:       00 01 10 00
    21a1:       ff 05 00 00 00 00       incl   0(%rip)        # 21a7 <destroy_conntrack+0xd5>
    21a7:       bf 01 00 00 00          mov    $0x1,%edi
    21ac:       e8 00 00 00 00          callq  21b1 <destroy_conntrack+0xdf>
    21b1:       e8 00 00 00 00          callq  21b6 <destroy_conntrack+0xe4>
    21b6:       48 8b 53 58             mov    0x58(%rbx),%rdx
    21ba:       48 85 d2                test   %rdx,%rdx
    21bd:       74 0f                   je     21ce <destroy_conntrack+0xfc>
    21bf:       ff 0a                   decl   (%rdx)
Quoted 9 lines
    21c1:       0f 94 c0                sete   %al
    21c4:       84 c0                   test   %al,%al
    21c6:       74 06                   je     21ce <destroy_conntrack+0xfc>
    21c8:       48 89 d7                mov    %rdx,%rdi
    21cb:       ff 52 08                callq  *0x8(%rdx)
    21ce:       48 89 df                mov    %rbx,%rdi  <--- last visible spot in back trace
    21d1:       5b                      pop    %rbx
    21d2:       e9 00 00 00 00          jmpq   21d7 <invert_tuplepr>

It might be in ip_conntrack_netbios?