So I slimmed my app way way down and made it multithreaded, here is what its
doing right now.
Time
KeysProc
dTime
dKeys
kps
KeysLeft
ETA Seconds
ETA Minutes
ETA Hours
ETA Days
Num Threads
Thread 0 Counter
9/27/2007 12:01:31
33
0:00:11
15018
1365.273
4294967263
3145867.635
52431.12724
873.8521207
36.41050503
9/27/2007 12:01:42
15051
0:00:59
82201
1393.237
4294952245
3082714.107
51378.56845
856.3094741
35.67956142
4
14
9/27/2007 12:02:41
97252
0:01:04
90292
1410.813
4294870044
3044252.894
50737.54823
845.6258038
35.23440849
4
3158
9/27/2007 12:03:45
187544
0:04:54
417813
1421.133
4294779752
3022082.242
50368.03737
839.4672896
34.97780373
4
21070
9/27/2007 12:08:39
605357
0:04:17
362125
1409.047
4294361939
3047707.331
50795.12218
846.5853697
35.2743904
4
41084
9/27/2007 12:12:56
967482
0:05:59
505800
1408.914
4293999814
3047738.107
50795.63512
846.5939186
35.27474661
4
136086
9/27/2007 12:18:55
1473282
0:03:17
278268
1412.528
4293494014
3039581.703
50659.69506
844.3282509
35.18034379
4
220830
9/27/2007 12:22:12
1751550
0:03:13
271989
1409.269
4293215746
3046412.312
50773.53853
846.2256422
35.25940176
4
338699
9/27/2007 12:25:25
2023539
0:17:54
1523321
1418.362
4292943757
3026690.761
50444.84601
840.7474335
35.03114306
4
462360
9/27/2007 12:43:19
3546860
0:09:59
824726
1376.838
4291420436
3116866.499
51947.77499
865.7962498
36.07484374
4
818799
9/27/2007 12:53:18
4371586
4
1010076
Keysproc is the total number of keys processed, dTime and dKeys are 'delta'
values from subtracting the previous row , kps is keys per second. Keys
left. 2^32 hah. And the eta. It an excel spread sheet I just copy and paste
the values into from the app.
Im searching osos for the string "MTOOL399". Im only searching the first
0x800 of the data area (>0x800). Ive based this assumption on the rsrc.fw
file that has mtool399 written in it at 0x0e03. I guess this is the ipod's
partition / filestructure marker. Im using 32-bit rc4 with the small chunk
of data so its faster. I started the decryption at 0x800. Hmm reading the
crypto synth I see that MTOOL399 is probably not in the data portion. So new
approach. Im going to get the first 20 instructions. 20x4 bytes. And verify
that the opcode portion of the instruction bitmap is a valid opcode. This
method will probably return many false keys. Im going to do some more
coding. I expect the performance (kps) to increase because im only rc4
decoding the first 80 bytes not 2048 now. Ill update soon.
Once I get the app able to actually be functional enough to find the correct
key would people be willing to split up the keyspace so we could get the
result sooner. If my core 2 duo 1.83ghz can do the entire osos in 36 days.
get a few more people on it and boom maybe a week or so. But we all need to
agree on what a 'correct' key would be. Wasting all the cpu time on a
decyption scheme that is flawed is a big waste. Well im going to hunt down a
set of arm opcodes. Later. -- Jeremy
_______________________________________________
Linux4nano-dev mailing list
[email protected]
https://mail.gna.org/listinfo/linux4nano-dev
http://www.linux4nano.org
