I have the following entry in a Dr Watson log. What is the significance of the "ds:0023:003a3000=??" part of the entry to the right of the FAULT line?
*----> State Dump for Thread Id 0xdfc <----*
eax=00000000 ebx=00390320 ecx=0854ff48 edx=09e44bfc esi=00012ce1 edi=0854ff61
eip=00465c51 esp=0854ff30 ebp=00000000 iopl=0 nv up ei pl zr na po nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00010246
function: sysman
00465c37 49 dec ecx
00465c38 eb02 jmp sysman+0x65c3c (00465c3c)
00465c3a 33c9 xor ecx,ecx
00465c3c 8d542428 lea edx,[esp+0x28]
00465c40 52 push edx
00465c41 51 push ecx
00465c42 8d4c2418 lea ecx,[esp+0x18]
00465c46 e8d5c0fcff call sysman+0x31d20 (00431d20)
00465c4b 33c0 xor eax,eax
00465c4d 8d4c2418 lea ecx,[esp+0x18]
FAULT ->00465c51 8a441eff mov al,[esi+ebx-0x1] ds:0023:003a3000=??
00465c55 50 push eax
00465c56 6864074900 push 0x490764
00465c5b 51 push ecx
00465c5c e8cfd0fcff call sysman+0x32d30 (00432d30)
00465c61 8d542424 lea edx,[esp+0x24]
00465c65 68689f4800 push 0x489f68
00465c6a 8d44242c lea eax,[esp+0x2c]
00465c6e 52 push edx
00465c6f 50 push eax
00465c70 e83bc3fcff call sysman+0x31fb0 (00431fb0)
*----> Stack Back Trace <----*
ChildEBP RetAddr Args to Child
00000000 00000000 00000000 00000000 00000000 sysman+0x65c51
*----> Raw Stack Dump <----*
000000000854ff30 58 01 55 08 75 07 c8 09 - 00 00 00 00 18 6d c7 01 X.U.u........m..
000000000854ff40 fc 4b e4 09 04 bd 47 00 - 04 bd 47 00 fc 0c c9 01 .K....G...G.....
000000000854ff50 ac ca ae 09 64 5f c4 01 - 20 37 37 30 32 34 3a 20 ....d_.. 77024:
000000000854ff60 00 b3 42 00 a8 ff 54 08 - 90 a6 47 00 02 00 00 00 ..B...T...G.....
000000000854ff70 8b c5 42 00 b8 ff 54 08 - 2e 03 39 00 28 99 cb 01 ..B...T...9.(...
000000000854ff80 ff ff ff ff 00 00 00 00 - 00 00 00 00 20 1e cb 01 ............ ...
000000000854ff90 a6 f7 ba 77 06 00 00 00 - c9 f7 ba 77 e1 6b d9 09 ...w.......w.k..
000000000854ffa0 06 00 00 00 1f 00 00 00 - 68 00 55 08 c1 a0 47 00 ........h.U...G.
000000000854ffb0 00 00 00 00 58 c4 42 00 - c9 a5 ca 09 d1 fb 38 0a ....X.B.......8.
000000000854ffc0 27 00 00 00 e1 6b d9 09 - ef f2 41 00 c9 a5 ca 09 '....k....A.....
000000000854ffd0 01 59 cc 01 38 00 55 08 - ec 00 55 08 00 00 00 00 .Y..8.U...U.....
000000000854ffe0 e0 00 55 08 ff ff ff ff - 89 00 00 00 01 00 01 01 ..U.............
000000000854fff0 c8 ff 54 08 b8 ff 54 08 - 77 00 55 08 29 a5 ca 09 ..T...T.w.U.)...
0000000008550000 51 00 00 00 5f 00 00 00 - 00 9f 82 7c 61 36 ca 01 Q..._......|a6..
0000000008550010 25 00 00 00 3f 00 00 00 - 00 ce bb 77 91 b7 c7 01 %...?......w....
0000000008550020 19 00 00 00 1f 00 00 00 - 00 ff ff ff d9 28 cc 01 .............(..
0000000008550030 0b 00 00 00 1f 00 00 00 - 00 00 55 08 d1 fb 38 0a ..........U...8.
0000000008550040 27 00 00 00 3f 00 00 00 - 00 20 ba 77 00 00 00 00 '...?.... .w....
0000000008550050 00 00 00 00 00 00 00 00 - 20 b7 c7 01 00 00 00 00 ........ .......
0000000008550060 00 00 00 00 00 00 00 00 - b4 00 55 08 1b 90 47 00 ..........U...G.`
To summarize:
You get a register dump here:
eax=00000000 ebx=00390320 ecx=0854ff48 edx=09e44bfc esi=00012ce1 edi=0854ff61
eip=00465c51 esp=0854ff30 ebp=00000000 iopl=0 nv up ei pl zr na po nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00010246
The eip indicates the instruction that failed:
FAULT ->00465c51 8a441eff mov al,[esi+ebx-0x1] ds:0023:003a3000=??
The stuff at the end is the address that failed to read, which is the "usual" data segment of 23, and address 3A3000, whcih is composed of esi and ebx minus 1: 390320+12ce1-1. To me, that looks like an index gone bad - 3a3000 would be the first address of a new "page" in memory, so that's why it's failing at that point. 77025 bytes into an array is quite a long way, but it is of course possible that it's something else that is wrong.