Re: [SSI-devel] Re: SSI-1.9 oops update_load_info()

[Roger T. <rog...@gm...>]

Same suspicion. I added some code elsewhere to initialized node value and
made the tests more robust in update_load_info(). So far no oops after
making these changes. cross-fingers.

Roger


On 9/24/05, John Byrne <joh...@hp...> wrote:
>
> Roger Tsang wrote:
> > Here's what was traversed so far...
>
> It died at 0x60 and your eax is really too large from your original
> oops. I suspect the larray[j].node is returning a totally bogus value.
> I'm not sure why, though. Could something be sending bogus messages to
> the UDP port. There is basically no security.
>
> I'll check back again late this evening.
>
> John
>
>
> >
> > 0xc025e090 update_load_info: push %ebp
> > 0xc025e091 update_load_info+0x1: mov %esp,%ebp
> > 0xc025e093 update_load_info+0x3: push %edi
> > 0xc025e094 update_load_info+0x4: push %esi
> > 0xc025e095 update_load_info+0x5: push %ebx
> > 0xc025e096 update_load_info+0x6: sub $0x10,%esp
> > 0xc025e099 update_load_info+0x9: mov 0xc0628044,%eax
> > 0xc025e09e update_load_info+0xe: test %eax,%eax
> > 0xc025e0a0 update_load_info+0x10: je 0xc025e167 update_load_info+0xd7
> > 0xc025e0a6 update_load_info+0x16: mov 0xc(%ebp),%eax
> > 0xc025e0a9 update_load_info+0x19: xor %esi,%esi
> > 0xc025e0ab update_load_info+0x1b: dec %eax
> > 0xc025e0ac update_load_info+0x1c: mov %eax,0xfffffff0(%ebp)
> > 0xc025e0af update_load_info+0x1f: je 0xc025e15f update_load_info+0xcf
> > 0xc025e0b5 update_load_info+0x25: mov 0xc04b64c8,%edi
> > 0xc025e0bb update_load_info+0x2b: mov 0xc04b5fe0,%eax
> > kdb>
> > 0xc025e0c0 update_load_info+0x30: mov $0x1,%ebx
> > 0xc025e0c5 update_load_info+0x35: mov $0xc,%ecx
> > 0xc025e0ca update_load_info+0x3a: mov %edi,0xffffffe8(%ebp)
> > 0xc025e0cd update_load_info+0x3d: mov %eax,0xffffffec(%ebp)
> > 0xc025e0d0 update_load_info+0x40: jmp 0xc025e0db update_load_info+0x4b
> > 0xc025e0d2 update_load_info+0x42: inc %ebx
> > 0xc025e0d3 update_load_info+0x43: add $0xc,%ecx
> > 0xc025e0d6 update_load_info+0x46: cmp $0x7d,%ebx
> > 0xc025e0d9 update_load_info+0x49: jg 0xc025e140 update_load_info+0xb0
> > 0xc025e0db update_load_info+0x4b: mov 0x8(%ebp),%edi
> > 0xc025e0de update_load_info+0x4e: mov (%ecx,%edi,1),%edx
> > 0xc025e0e1 update_load_info+0x51: test %edx,%edx
> > 0xc025e0e3 update_load_info+0x53: je 0xc025e0d2 update_load_info+0x42
> > 0xc025e0e5 update_load_info+0x55: cmp %edx,0xffffffe8(%ebp)
> > 0xc025e0e8 update_load_info+0x58: je 0xc025e0d2 update_load_info+0x42
> > 0xc025e0ea update_load_info+0x5a: lea (%edx,%edx,2),%eax
> > kdb>
> > 0xc025e0ed update_load_info+0x5d: shl $0x3,%eax
> > 0xc025e0f0 update_load_info+0x60: mov 0xc06289c0(%eax),%edi
> > 0xc025e0f6 update_load_info+0x66: test %edi,%edi
> > 0xc025e0f8 update_load_info+0x68: jne 0xc025e117 update_load_info+0x87
> > 0xc025e0fa update_load_info+0x6a: mov %edx,0xc06289c0(%eax)
> > 0xc025e100 update_load_info+0x70: mov 0xffffffec(%ebp),%eax
> > 0xc025e103 update_load_info+0x73: cmp %eax,0xffffffe8(%ebp)
> > 0xc025e106 update_load_info+0x76: setne %al
> > 0xc025e109 update_load_info+0x79: movzbl %al,%eax
> > 0xc025e10c update_load_info+0x7c: add 0xc0627e28,%eax
> > 0xc025e112 update_load_info+0x82: mov %eax,0xc0627e28
> > 0xc025e117 update_load_info+0x87: lea (%edx,%edx,2),%eax
> > 0xc025e11a update_load_info+0x8a: mov 0x8(%ebp),%edi
> > 0xc025e11d update_load_info+0x8d: inc %esi
> > 0xc025e11e update_load_info+0x8e: shl $0x3,%eax
> > 0xc025e121 update_load_info+0x91: mov 0xc06289d4(%eax),%edx
> >
> > On 9/24/05, *Roger Tsang* <rog...@gm...
> > <mailto:rog...@gm...>> wrote:
> >
> > Hi,
> >
> > I keep on running into this oops on node 1. Any ideas?
> > update_load_info() code looks okay I think, so this is odd. node 2
> > is init node.
> >
> > Roger
> >
> >
> > Unable to handle kernel paging request at virtual address 42e7ce28
> > printing eip:
> > c025e0f0
> > *pde =3D 00000000
> > Oops: 0000 [#1]
> > Modules linked in: ipt_REJECT ipt_state ipt_multiport iptable_filter
> > iptable_nat ip_conntrack ip_tables uhci_hcd ehci_hcd usbcore drbd
> > sata_via libata sk98lin r8169 dm_mod
> > CPU: 0
> > EIP: 0060:[<c025e0f0>] Not tainted VLI
> > EFLAGS: 00010a92 (2.6.10-bk7-ssi5)
> > EIP is at update_load_info+0x60/0x150
> > eax: 82854468 ebx: 00000002 ecx: 00000018 edx: c570382f
> > esi: 00000000 edi: f26dd800 ebp: f6e0df74 esp: f6e0df58
> > ds: 007b es: 007b ss: 0068
> > Process ics_llunack_dae (pid: 65936, threadinfo=3Df6e0c000 task=3Df6dd8=
550)
> > Stack: 00000286 00000001 00000002 00000001 000005e8 00000604
> > 00000000 f6e0df88
> > c025e275 f26dd800 00000002 00000000 f6e0dfec c01f0a44
> > f26dd800 f6df001c
> > 000005e8 00000100 f6df001c f6df001c f26dd800 f6df0000
> > f6def4a0 f7fe14c0
> > Call Trace:
> > [<c0104a9f>] show_stack+0x7f/0xa0
> > [<c0104c39>] show_registers+0x159/0x220
> > [<c0104fb1>] die+0xd1/0x190
> > [<c011676d>] do_page_fault+0x45d/0x659
> > [<c010471b>] error_code+0x2b/0x30
> > [<c025e275>] loadinfo_received+0x25/0x50
> > [<c01f0a44>] ics_llunack_daemon+0x204/0x390
> > [<c0100995>] kernel_thread_helper+0x5/0x10
> > Code: b9 0c 00 00 00 89 7d e8 89 45 ec eb 09 43 83 c1 0c 83 fb 7d 7f
> > 65 8b 7d 08 8b 14 39 85 d2 74 ed 39 55 e8 74 e8 8d 04 52 c1 e0 03
> > <8b> b8 c0 89 62 c0 85 ff 75 1d 89 90 c0 89 62 c0 8b 45 ec 39 45
> >
> > Entering kdb (current=3D0xf6dd8550, pid 65936) Oops: Oops
> > due to oops @ 0xc025e0f0
> > eax =3D 0x82854468 ebx =3D 0x00000002 ecx =3D 0x00000018 edx =3D 0xc570=
382f
> > esi =3D 0x00000000 edi =3D 0xf26dd800 esp =3D 0xf6e0df58 eip =3D 0xc025=
e0f0
> > ebp =3D 0xf6e0df74 xss =3D 0x00000068 xcs =3D 0x00000060 eflags =3D 0x0=
0010a92
> > xds =3D 0x0000007b xes =3D 0x0000007b origeax =3D 0xffffffff &regs =3D
> > 0xf6e0df24
> > kdb> bt
> > Stack traceback for pid 65936
> > 0xf6dd8550 65936 2 1 0 R 0xf6dd8710 *ics_llunack_dae
> > EBP EIP Function (args)
> > 0xf6e0df74 0xc025e0f0 update_load_info+0x60 (0xf26dd800, 0x2, 0x0)
> > 0xf6e0df88 0xc025e275 loadinfo_received+0x25 (0xf26dd800,
> > 0xf6df001c, 0x5e8, 0x100, 0xf6df001c)
> > 0xf6e0dfec 0xc01f0a44 ics_llunack_daemon+0x204
> > 0xc0100995 kernel_thread_helper+0x5
> > kdb> id 0xc025e0f0
> > 0xc025e0f0 update_load_info+0x60: mov 0xc06289c0(%eax),%edi
> > 0xc025e0f6 update_load_info+0x66: test %edi,%edi
> > 0xc025e0f8 update_load_info+0x68: jne 0xc025e117
> > update_load_info+0x87
> > 0xc025e0fa update_load_info+0x6a: mov %edx,0xc06289c0(%eax)
> > 0xc025e100 update_load_info+0x70: mov 0xffffffec(%ebp),%eax
> > 0xc025e103 update_load_info+0x73: cmp %eax,0xffffffe8(%ebp)
> > 0xc025e106 update_load_info+0x76: setne %al
> > 0xc025e109 update_load_info+0x79: movzbl %al,%eax
> > 0xc025e10c update_load_info+0x7c: add 0xc0627e28,%eax
> > 0xc025e112 update_load_info+0x82: mov %eax,0xc0627e28
> > 0xc025e117 update_load_info+0x87: lea (%edx,%edx,2),%eax
> > 0xc025e11a update_load_info+0x8a: mov 0x8(%ebp),%edi
> > 0xc025e11d update_load_info+0x8d: inc %esi
> > 0xc025e11e update_load_info+0x8e: shl $0x3,%eax
> > 0xc025e121 update_load_info+0x91: mov 0xc06289d4(%eax),%edx
> > 0xc025e127 update_load_info+0x97: add 0x4(%ecx,%edi,1),%edx
> > kdb>
> > 0xc025e12b update_load_info+0x9b: cmp 0xfffffff0(%ebp),%esi
> > 0xc025e12e update_load_info+0x9e: mov %edx,0xc06289c4(%eax)
> > 0xc025e134 update_load_info+0xa4: mov 0x8(%ecx,%edi,1),%edx
> > 0xc025e138 update_load_info+0xa8: mov %edx,0xc06289d0(%eax)
> > 0xc025e13e update_load_info+0xae: jl 0xc025e0d2
> > update_load_info+0x42
> > 0xc025e140 update_load_info+0xb0: movl $0x0,(%esp,1)
> > 0xc025e147 update_load_info+0xb7: call 0xc02888f0 mosix_calc_load
> > 0xc025e14c update_load_info+0xbc: cmpl $0x1,0xc0627e28
> > 0xc025e153 update_load_info+0xc3: jle 0xc025e15f
> > update_load_info+0xcf
> > 0xc025e155 update_load_info+0xc5: mov 0xc0627e2c,%ebx
> > 0xc025e15b update_load_info+0xcb: test %ebx,%ebx
> > 0xc025e15d update_load_info+0xcd: jne 0xc025e1ce
> > update_load_info+0x13e
> > 0xc025e15f update_load_info+0xcf: add $0x10,%esp
> > 0xc025e162 update_load_info+0xd2: pop %ebx
> > 0xc025e163 update_load_info+0xd3: pop %esi
> > 0xc025e164 update_load_info+0xd4: pop %edi
> > kdb>
> >
> >
>
>