Wpa/Wpa2 dictionary attack is very slow after the latest update
#1
after the latest updates  (current version: 4.0.1) I saw that the brute force speed is slower than before...

Example.... 
I use hash cat utils to convert .cap to hccpax and devices available are the following:

OpenCL Platform #1: Apple
=========================
* Device #1: Intel(R) Core(TM) i7-3635QM CPU @ 2.40GHz, skipped.
* Device #2: HD Graphics 4000, 384/1536 MB allocatable, 16MCU
* Device #3: GeForce GT 650M, 256/1024 MB allocatable, 2MCU

Now... the devices speeds are very lower than before, example:

now:

Speed.Dev.#2.....:     2345 H/s (5.11ms)
Speed.Dev.#3.....:     4944 H/s (5.10ms)
Speed.Dev.#*.....:     7286 H/s

but before the update (to 4.0.1) speed was about (I don't remember precisely):

Speed.Dev.#1.....:     3500 H/s 

Speed.Dev.#2.....:     5800 H/s
Speed.Dev.#3.....:     16890 H/s

The time for a dictionary attack or brute force was increased so much.
Why now Is slower than before for no apparently reasons?
How can I resolve?

Thank you
#2
current version is 4.1.0

https://github.com/hashcat/hashcat/relea...t-4.1.0.7z
#3
I updated it automatically through brew on osx
#4
so, are you on version 4.1.0 yet?
#5
(02-22-2018, 10:13 PM)undeath Wrote: so, are you on version 4.1.0 yet?

I successfully updated hashcat to 4.1.0 but crack speed is still lower than before

Speed.Dev.#2.....: 2400 H/s (5.69ms) @ Accel:4 Loops:2 Thr:512 Vec:1
Speed.Dev.#3.....: 3919 H/s (8.06ms) @ Accel:16 Loops:4 Thr:1024 Vec:1
Speed.Dev.#*.....: 6319 H/s
#6
You may want to try with -w 3 or -w 4... dunno how the optimized kernels will perform on old cards like those.
Code:
hashcat -m 2500 cap.hccapx dictionary.txt -w 3
or -w 4

Or try adding -O for optimized kernels, I don't recall if optimized kernels are used by default in an actual run, though they are in a benchmark.
#7
(02-22-2018, 10:46 PM)soxrok2212 Wrote: You may want to try with -w 3 or -w 4... dunno how the optimized kernels will perform on old cards like those.
Code:
hashcat -m 2500 cap.hccapx dictionary.txt -w 3
or -w 4

Or try adding -O for optimized kernels, I don't recall if optimized kernels are used by default in an actual run, though they are in a benchmark.

Code:
hashcat --session hashcat --restore -w 4 -O

Speed is the same... It was about 16000 H/s but....

Speed.Dev.#2.....: 3080 H/s (9.86ms) @ Accel:8 Loops:2 Thr:512 Vec:1
Speed.Dev.#3.....: 3906 H/s (7.60ms) @ Accel:16 Loops:4 Thr:1024 Vec:1
Speed.Dev.#*.....: 6985 H/s
#8
EDIT:

Code:
hashcat -m 2500 net.hccapx /dict -w 3 --session different

Speed.Dev.#2.....: 3345 H/s (74.82ms) @ Accel:16 Loops:8 Thr:512 Vec:1
Speed.Dev.#3.....: 4717 H/s (49.27ms) @ Accel:32 Loops:16 Thr:1024 Vec:1
Speed.Dev.#*.....: 8062 H/s

But my computer now lags a little and before not and the speed was 16.000 H/s not 8.000!.... I don't understand... maybe problem is due to latest Osx o.s updates...
#9
(02-22-2018, 11:27 PM)Virgula Wrote: EDIT:

Code:
hashcat -m 2500 net.hccapx /dict -w 3 --session different

Speed.Dev.#2.....:     3345 H/s (74.82ms) @ Accel:16 Loops:8 Thr:512 Vec:1
Speed.Dev.#3.....:     4717 H/s (49.27ms) @ Accel:32 Loops:16 Thr:1024 Vec:1
Speed.Dev.#*.....:     8062 H/s

But my computer now lags a little and before not and the speed was 16.000 H/s not 8.000!.... I don't understand... maybe problem is due to latest Osx o.s updates...

It may be Apple’s OpenCL runtime causing issues. It killed my AMD speeds, not sure about nvidia though. Might be worth your time opening a bug on GitHub.
#10
(02-23-2018, 12:37 AM)soxrok2212 Wrote:
(02-22-2018, 11:27 PM)Virgula Wrote: EDIT:

Code:
hashcat -m 2500 net.hccapx /dict -w 3 --session different

Speed.Dev.#2.....:     3345 H/s (74.82ms) @ Accel:16 Loops:8 Thr:512 Vec:1
Speed.Dev.#3.....:     4717 H/s (49.27ms) @ Accel:32 Loops:16 Thr:1024 Vec:1
Speed.Dev.#*.....:     8062 H/s

But my computer now lags a little and before not and the speed was 16.000 H/s not 8.000!.... I don't understand... maybe problem is due to latest Osx o.s updates...

It may be Apple’s OpenCL runtime causing issues. It killed my AMD speeds, not sure about nvidia though. Might be worth your time opening a bug on GitHub.

I'll do thank you

if it could be usefu,l before to start the brute force hashcat show me this warning:

Code:
/usr/local/Cellar/hashcat/4.1.0/share/hashcat/OpenCL/m02500-optimized.cl: Optimized OpenCL kernel requested but not needed - falling back to pure OpenCL kernel