API tools faq
paste
Bond697

xfr wondercards

Bond697
Dec 9th, 2011
440
0
Never
1
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 4.75 KB | None | 0 0
raw download clone embed print report
  1. Session Start: Wed Dec 07 00:43:49 2011
  2. Session Ident: maxg
  3. [00:43] Session Ident: maxg (KottNet, Mat) ([email protected])
  4. [00:43] <maxg> I managed to send myself a self made wondercard on a retail game
  5. 01[00:44] <Mat> whoa
  6. 01[00:44] <Mat> that's awesome
  7. [00:44] <maxg> yeah, but the problem is that the attack is so complicated I have doubts it would be possible to make a public wondercard distribution system
  8. 01[00:44] <Mat> right
  9. [00:45] <maxg> basically I tried everything
  10. [00:45] <maxg> for instance contraty to what the debug messages in the rom say the comon name is actually checked on the certificate
  11. 01[00:46] <Mat> really
  12. 01[00:46] <Mat> huh.
  13. [00:46] <maxg> I suppose it wasn't in the dev stage because the dev cert has a wildcard
  14. 01[00:46] <Mat> ah, right
  15. [00:46] <maxg> the cert is for *.dev.nintendowifi.net
  16. 01[00:46] <Mat> makes sense
  17. 01[00:46] <Mat> yeah
  18. [00:47] <maxg> ok so that sucks major balls here because I had a valid cert chained on the thawte root CA which is present in the game
  19. [00:47] <maxg> and if the comon name wasn't checked it would have been so easy
  20. 01[00:47] <Mat> :(
  21. [00:48] <maxg> then I tried all sorts of signature/cert alteration but the game does the job properly and actually checks the signature against the hash of the sent certificate
  22. [00:49] <maxg> finally the only thing we're left with is the rng which seems to be the only way to bypass ssl
  23. 01[00:49] <Mat> how so?
  24. [00:49] <maxg> (except action replay but what good is a wondercard distribution server if you have an action replay)
  25. 01[00:49] <Mat> right
  26. [00:50] <maxg> basically what you have to do is stick to a seed and see the ssl session key used
  27. [00:50] <maxg> then redo the protocol on the same state with ssl disabled
  28. [00:50] <maxg> then do some replay with the master to find the write AES key
  29. 01[00:50] <Mat> ohhh
  30. 01[00:50] <Mat> i see what you're saying
  31. 01[00:51] <Mat> right
  32. 01[00:51] <Mat> ok
  33. [00:51] <maxg> and basically you can RNG abuse the game to replay an altered ssl session
  34. [00:51] <maxg> but you see the problem here, it is fully dependant on the seeding which includes the ds parameters
  35. 01[00:51] <Mat> yeah, good luck with that
  36. [00:51] <maxg> that's doable
  37. 01[00:52] <Mat> not really
  38. 01[00:52] <Mat> http://pastebin.com/7ZtSsEcT
  39. 01[00:52] <Mat> if you mean it seeds with the standard seed building function
  40. [00:52] <maxg> yeah
  41. [00:52] <maxg> yeah that's really won't do it will it
  42. 01[00:52] <Mat> not at all
  43. 01[00:52] <Mat> that's it, btw
  44. [00:52] <maxg> nice function by the way
  45. 01[00:52] <Mat> that's the entropy function
  46. 01[00:53] <Mat> that builds the sha digest array
  47. [00:53] <maxg> ok
  48. 01[00:53] <Mat> just saying
  49. [00:53] <maxg> so basically we need a new idea to make this work
  50. 01[00:53] <Mat> somehow managing to manipulate that would be pretty well impossible unless it happens as the game boots
  51. 01[00:53] <Mat> yeah, pretty much
  52. [00:54] <maxg> the problem with a session key attack is that you need a previous session with ram alteration anyway
  53. [00:54] <maxg> so basically useless for a public server
  54. 01[00:55] <Mat> right
  55. [00:56] <maxg> I even tried using CNAME in DNS to force a domain name change for the comon name
  56. [00:56] <maxg> but it didn't work
  57. 01[00:56] <Mat> damn it
  58. [00:57] <maxg> yeah
  59. [00:58] <maxg> basically there's very little to do except try a very low level attack like buffer overflow or a bug in the ssl protocol functions
  60. [00:58] <maxg> but I kinda rage quitted trying to find them in the arm9
  61. 01[00:59] <Mat> what are you looking for in the arm9, exactly?
  62. [00:59] <maxg> the comon network code
  63. 01[00:59] <Mat> you sure it's there?
  64. [01:00] <maxg> yeah without a doubt
  65. 01[01:00] <Mat> you might find this useful:
  66. 01[01:01] <Mat> http://hack.thundaga.com/arm9-white_decompressed_U.idb
  67. 01[01:01] <Mat> pop that open in ida
  68. [01:03] <maxg> but then I'll give gen 4 games a try since there's a chance they might be easier to work with
  69. 01[01:03] <Mat> that's true, actually
  70. [01:03] <maxg> a gen4 wondercard server would still be neat I guess
  71. 01[01:03] <Mat> yeah
  72. [01:04] <maxg> and then we have the trading protocol to research if you're up to it
  73. 01[01:04] <Mat> oh, forgot about that
  74. 01[01:04] <Mat> yeah
  75. [01:04] <maxg> it should be much easier I think
  76. 01[01:04] <Mat> you would think so, yes
  77. [01:27] <maxg> I even tested the nintendo CA public key for the debian openssl seeding bug but it isn't one of the weak keys :(
  78. 01[01:28] <Mat> :(
  79. 01[01:28] <Mat> unfortunate
  80. [01:28] <maxg> that would have been great way to hack the ssl though
  81. 01[01:28] <Mat> right
  82. [01:30] <maxg> anyway
  83. 01[01:30] <Mat> right
  84. 01[01:30] <Mat> later
  85. Session Close: Wed Dec 07 12:31:13 2011
  86.  
  87. Session Start: Wed Dec 07 12:31:13 2011
  88. Session Ident: maxg
  89. Session Close: Thu Dec 08 00:00:00 2011
Advertisement
Comments
  • User was banned
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!