muhmadkhalrd / reaver-wps

Automatically exported from code.google.com/p/reaver-wps
0 stars 0 forks source link

Rev. 52 barely starts #66

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1. start program
2.
3.

What is the expected output? What do you see instead?
normal program flow

reaver associates (than sometimes checks 1 pin) thereafter or directly at the 
check of the first pin stops after Warning "Last message ..." and the Warning 
"Out of order...)
prgramm did run properly in v 48

What version of the product are you using? On what operating system?
v 52 - bt5

Please provide any additional information below.

Original issue reported on code.google.com by wernerli...@gmail.com on 3 Jan 2012 at 9:00

GoogleCodeExporter commented 8 years ago
Same issue, previous versions worked fine.

Reaver still runs, But allot more:

[!] WARNING: Last message not processed properly, reverting state to previous 
message
[!] WARNING: Out of order packet received, re-trasmitting last message

And sometimes when these messages appear reaver will stop until restarted.

Original comment by Sca...@gmail.com on 3 Jan 2012 at 9:11

GoogleCodeExporter commented 8 years ago
All revisions are working for me right now. Can you try r49 and r51 and see if 
the issues persist with those revisions? 

Original comment by cheff...@tacnetsol.com on 3 Jan 2012 at 9:20

GoogleCodeExporter commented 8 years ago
Attempting on Rev 51 now.

[+] Trying pin 49048629
[+] 7.84% complete @ 2012-01-03 16:54:00 (3 seconds/attempt)
[+] Trying pin 66658627
[+] Trying pin 94238624
[!] WARNING: Last message not processed properly, reverting state to previous 
message
[!] WARNING: Out of order packet received, re-trasmitting last message
[+] Trying pin 94238624
[!] WARNING: Last message not processed properly, reverting state to previous 
message
[!] WARNING: Out of order packet received, re-trasmitting last message

And then locked up, waited about 5 minutes.

Will now get r49

Original comment by Sca...@gmail.com on 3 Jan 2012 at 9:55

GoogleCodeExporter commented 8 years ago
Rev 49:

[+] Trying pin 45498626
[+] 8.15% complete @ 2012-01-03 16:59:26 (3 seconds/attempt)
[+] Trying pin 36108626
[+] Trying pin 06928629
[+] Trying pin 85828629
[+] Trying pin 19678627
[!] WARNING: Last message not processed properly, reverting state to previous 
message
[!] WARNING: Out of order packet received, re-trasmitting last message

And locked.

Will now try r48

Original comment by Sca...@gmail.com on 3 Jan 2012 at 10:02

GoogleCodeExporter commented 8 years ago
I feed from the read-only version and do not know how to get now to the older 
versions - but I just see that Sca...@gmail.com is doing the checks.

Original comment by wernerli...@gmail.com on 3 Jan 2012 at 10:03

GoogleCodeExporter commented 8 years ago
You can get an older revision like this:
svn checkout http://reaver-wps.googlecode.com/svn/trunk/ reaver-wps-read-only 
-r [number here]

I can't always get it to lock up sometimes it runs for 30 minutes without a 
hitch.
r48 did give me 2 pages full of "out of order" messages twice. But hasn't 
locked up as of yet.

Original comment by Sca...@gmail.com on 3 Jan 2012 at 10:17

GoogleCodeExporter commented 8 years ago
thanks for the info on how to get to the older versions. 
I can confirm that v. 48 did run for me some hours without any lock up till I 
killed it and during this time the program overcame many 
"
!] WARNING: Last message not processed properly, reverting state to previous 
message
[!] WARNING: Out of order packet received, re-trasmitting last message"

messages.
I will try now to go now back to v. 48 

Original comment by wernerli...@gmail.com on 3 Jan 2012 at 10:22

GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago
I am back at v.48 and it works. Just want to correct my statement regarding 
OS-system.
I act. use bt4 on this issue.

Original comment by wernerli...@gmail.com on 3 Jan 2012 at 10:38

GoogleCodeExporter commented 8 years ago
It indeed seems that rev 48, doesn't lock up.

I have also noticed with rev 52, a high powered connection (-30/-40) doesn't 
lockup.
But -60 does lock up pretty quickly.

Original comment by Sca...@gmail.com on 3 Jan 2012 at 10:46

GoogleCodeExporter commented 8 years ago
wl & Scalee, thanks a lot for the feedback. I suspect this has to do with the 
M2D message changes in r49; I've updated the code, see if r53 fixes this for 
you.

Original comment by cheff...@tacnetsol.com on 4 Jan 2012 at 1:51

GoogleCodeExporter commented 8 years ago
@At the office will check, when i get home.

Original comment by Sca...@gmail.com on 4 Jan 2012 at 8:20

GoogleCodeExporter commented 8 years ago
I am on v.54 now and the program no longer locks up - however I see a reduced 
rate in pin-checking on the same AP. v.48 -about 80sec/attempt - now 180 
sec/attempt and I do not think that the AP has caused this.
Besides that I no longer receive a message at program start that the previous 
sessions are included (something in this regard - forgot the act. wording).

Original comment by wernerli...@gmail.com on 4 Jan 2012 at 9:46

GoogleCodeExporter commented 8 years ago
wl, if you are attacking an AP that you have previously attacked, you should 
get a prompt asking if you want to restore the previous session. The session 
files are stored in /etc/reaver and are named <BSSID>.wpc. Double check to see 
that you are attacking the same BSSID and that the wpc file exists.

Original comment by cheff...@tacnetsol.com on 4 Jan 2012 at 1:00

GoogleCodeExporter commented 8 years ago
wl, can you also provide a pcap of the attack? 80 sec/attempt and 180 
sec/attempt are both very slow...

Original comment by cheff...@tacnetsol.com on 4 Jan 2012 at 1:34

GoogleCodeExporter commented 8 years ago
I fixed (with your help) my problem to restore a previuos session. So that is 
not really an issue but was a mistake by me!

I changed in the meantime with v.54 to a different adapter (ralink 2573 - 
rt73usb) which has addition. an extra attenna so I get a better signal - 
result: every 3 sec. or so I get a pin checked until after about 2 min. when 
the AP locks out for about 5 min. I consider this a perfect result by the 
program.
Therefore I believe now (as Scalee also hinted to) that our problems with v. 49 
till 52 are signal-strengh related. Unfortunat. I cannot test at the moment 
v.49 against the rt73usb adapter and I also have not access to the pcap, 
because I now use a different computer. I should also mention that I now test 
under Vmware workst.8 + bt5r1.

Original comment by wernerli...@gmail.com on 4 Jan 2012 at 2:31

GoogleCodeExporter commented 8 years ago
Thanks wl! I'll mark this as complete.

Original comment by cheff...@tacnetsol.com on 4 Jan 2012 at 2:43