[okl4-developer] oklinux on gta01

Geoffrey Lee glee at ok-labs.com
Fri May 23 02:11:36 EST 2008


On Fri, May 23, 2008 at 12:00:47AM +0800, Tsai Tung-Chieh wrote:

Hi Tung-Chieh

It seems that your stack pointer register is corrupted, this
should not happen.  Newer QEMU versions for the OpenMoko 
loads the binary at 0x30100000, while the program gets decoded 
starting at 0x30000000, which only leaves 1MB before the ELF decode 
will run over the ELF binary.

Can you try editing the QEMU source so that the binary
gets put at a higher address (say, 0x31000000), and see
if that fixes the problem for you?

	-gl


> Dear Geoffrey,
> 
> Thanks your advice, I put it below :
> 
> 00108978 <__sys_entry>:
>   108978:       e52de004        str     lr, [sp, #-4]!
>   10897c:       e59f2018        ldr     r2, [pc, #24]   ; 10899c <.text+0x899c>
>   108980:       e1a03000        mov     r3, r0
>   108984:       e590101c        ldr     r1, [r0, #28]
>   108988:       e5823000        str     r3, [r2]
>   10898c:       e5900018        ldr     r0, [r0, #24]
>   108990:       eb000003        bl      1089a4 <__malloc_init>
>   108994:       e49de004        ldr     lr, [sp], #4
>   108998:       eaffeb3f        b       10369c <main>
>   10899c:       0011458c        andeqs  r4, r1, ip, lsl #11
> 
> I also put the whole result on :
> http://tsaitungchieh.googlepages.com/iguana_server.objdump
> 
> 
> Best Regards,
> Tsai Tung-Chieh
> 
> On Thu, May 22, 2008 at 11:13 PM, Geoffrey Lee <glee at ok-labs.com> wrote:
> > On Thu, May 22, 2008 at 08:14:55PM +0800, Tsai Tung-Chieh wrote:
> >
> > Hi Tung-Chieh
> >
> >
> >> Dear Geoffrey,
> >>
> >> Sorry I'm not sure how to get the objdump and target function,
> >> would you please give me some advice about how to get these
> >> information. However, I put the memdump below, hope it's useful.
> >> Thanks.
> >
> > You can obtain a disassembly output with the following command:
> >
> > arm-linux-objdump -dp -S build_gta01/iguana_server/bin/ig_server
> >
> > This will give you assembly output instead of raw opcodes, as well
> > as which function that fault occurred in.
> >
> >        -gl
> >
> >
> >>
> >> Best Regards,
> >> Tsai Tung-Chieh
> >>
> >> > memdump
> >> Dump address [0x1089]: 00108800
> >> 00108800  22822008 e3510010 21a01221 22822004   . ."..Q� !..!. ."
> >> 00108810  e3510004 82822003 908220a1 e1a00230   ..Q�. .. � ..0..
> >> 00108820  e1a0f00e e52de004 eb000034 e3a00000   .�.�.�-� 4..�...
> >> 00108830  e49df004 e2512001 3a00002c 11500001   .�.�. Q� ,..:..P.
> >> 00108840  03a00000 81110002 00000002 91a0f00e   ........ .....
> >> 00108850  e3a02000 e3510201 31510000 31a01201   . .�..Q� ..Q1...1
> >> 00108860  32822004 3afffffa e3510102 31510000   . .2���: ..Q�..Q1
> >> 00108870  31a01081 32822001 3afffffa e2522003   ...1. .2 ���:. R
> >> 00108880  ba00000e e1500001 20400001 e15000a1   ...�..P� ..@ �.P
> >> 00108890  204000a1 e1500121 20400121 e15001a1   �.@ !.P� !.@ �.P
> >> 001088a0  204001a1 e3500001 e1a01221 a2522004   �.@ ..P� !..�. R�
> >> 001088b0  aafffff3 e3120003 13300000 0a00000a   ����...� ..0.....
> >> 001088c0  e3720002 ba000006 0a000002 e1500001   ..r�...� ......P
> >> 001088d0  20400001 e1a010a1 e1500001 20400001   ..@ �..� ..P�..@
> >> 001088e0  e1a010a1 e1500001 20400001 e1a0f00e   �..�..P� ..@ .
> >> 001088f0  e52de004 eb000001 e3a00000 e49df004   .�-�...� ...�.
> >> Continue? (Continue/Quit) [continue]: continue
> >> 00108900  e92d4002 ef900014 e3700ffa 28bd8002   . at -�...� �.p�..�(
> >> 00108910  e3a01008 ef900025 e8bd8002 e52da004   ...�%..� ..��..-
> >> 00108920  e59fa044 e3500801 e08fa00a 2a00000a   D..�..P� ...�...*
> >> 00108930  e35000ff 83a0c008 93a0c000 e59f302c   �.P�.... ....,0.
> >> 00108940  e79a2003 e1a01c30 e7d20001 e080000c   . .�0..� ...�...
> >> 00108950  e2600020 e8bd0400 e1a0f00e e3500401    .`�..�� .�.�..P
> >> 00108960  23a0c018 33a0c010 eafffff3 0000bb34   ...#...3 ����4�..
> >> 00108970  0000000c eafffffe e52de004 e59f2018   ....���� .�-�. .
> >> 00108980  e1a03000 e590101c e5823000 e5900018   .0.�...� .0.�...
> >> 00108990  eb000003 e49de004 eaffeb3f 0011458c   ...�.�.� ?���.E..
> >> 001089a0  eaffeb3d e0601001 e2811001 e1a011a1   =���..`� ...��..
> >> 001089b0  e5801004 e2800008 ea000058 e92d4070   ...�...� X..�p at -
> >> 001089c0  e24dd008 e1a06000 eb000792 e3a04001   ..M�.`.� ...�. at .
> >> 001089d0  e1a05014 eb00078f e1a04014 e08651a5   .P.�...� . at .��Q.
> >> 001089e0  e1a041a4 e2455001 e2644000 e0056004   �A.�.PE� . at d�.`.
> >> 001089f0  e1a00186 e28d1004 e1a0200d ebffed47   ...�...� . .�G��
> >> Continue? (Continue/Quit) [continue]: continue
> >> 00108a00  e3500000 0a000007 e59d3000 e59d0004   ..P�.... .0.�...
> >> 00108a10  e1a031a3 e5803004 e2800008 eb00003f   �1.�.0.� ...�?..
> >> 00108a20  e59f3008 e5930000 e28dd008 e8bd8070   .0.�...� ...�p.�
> >> 00108a30  001144d8 e52de004 e1a0e000 e59f0020   .D...�-� .�.� ..
> >> 00108a40  e24dd008 e1a0c002 e5900000 e1a0200e   ..M�...� ...�. .
> >> 00108a50  e88d000a e59f100c e1a0300c eb000027   ...�...� .0.�'..
> >> 00108a60  ebffffc3 0011438c 0010c14c e92d4ff0   .���.C.. L...�O-
> >> 00108a70  e59db024 e3a07000 e1a05002 e1570002   $�.�.p.� .P.�..W
> >> 00108a80  e1a09000 e1a0a001 e1a08003 2a00000f   ...�...� ...�...*
> >> 00108a90  e0873005 e1a040a3 e026a498 e1a00009   .0.��@.� .�&�...
> >> 00108aa0  e1a01006 e1a0e00f e1a0f00b e3500000   ...�.�.� .�.�..P
> >> 00108ab0  b1a05004 ba000003 1a000001 e1a00006   .P.�...� .......
> >> 00108ac0  e8bd8ff0 e2847001 e1570005 eaffffee   �.��.p.� ..W����
> >> 00108ad0  e3a00000 e8bd8ff0 e92d4010 e0040091   ...��.�� . at -�...
> >> 00108ae0  e1a00004 eb00003f e1a02004 e2504000   ...�?..� . .�. at P
> >> 00108af0  e3a01000 1b000074 e1a00004 e8bd8010   ...�t... ...�..�
> >> Continue? (Continue/Quit) [continue]:
> >>
> >> On Thu, May 22, 2008 at 7:02 PM, Geoffrey Lee <glee at ok-labs.com> wrote:
> >> > On Thu, May 22, 2008 at 04:48:40PM +0800, Tsai Tung-Chieh wrote:
> >> >
> >> > Hi Tung-Chieh
> >> >
> >> > Can you please give us a objdump disassembly output of the instructions
> >> > around the faulting instruction pointer at 0x108978, as well as
> >> > what function this fault is in?
> >> >
> >> >        -gl
> >> >
> >
> > --
> >
> >

-- 




More information about the Developer mailing list