Support Install hashcat 6.2.3

[John1980]

I try to install it many times. But I still the error . Hashcat only run CPU , no load PGU. Please, anyone to help me to fix it.

D:\hashcat\hashcat-6.2.3>hashcat -m 2500 -a 3 -o cracked.txt hash.txt
hashcat (v6.2.3) starting

* Device #3: Unstable OpenCL driver detected!

This OpenCL driver may fail kernel compilation or produce false negatives.
You can use --force to override, but do not report related errors.

nvmlDeviceGetFanSpeed(): Not Supported

CUDA API (CUDA 11.4)
====================
* Device #1: NVIDIA GeForce RTX 2070 Super, 7153/8192 MB, 40MCU

OpenCL API (OpenCL 3.0 CUDA 11.4.94) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #2: NVIDIA GeForce RTX 2070 Super, skipped

OpenCL API (OpenCL 3.0 ) - Platform #2 [Intel(R) Corporation]
=============================================================
* Device #3: Intel(R) UHD Graphics 630, skipped

Minimum password length supported by kernel: 8
Maximum password length supported by kernel: 63

No hashes loaded.

Started: Mon Jul 26 10:09:39 2021
Stopped: Mon Jul 26 10:09:40 2021

===========================================================================
D:\hashcat\hashcat-6.2.3>hashcat -I
hashcat (v6.2.3) starting in backend information mode

* Device #3: Unstable OpenCL driver detected!

CUDA Info:
==========

CUDA.Version.: 11.4

Backend Device ID #1 (Alias: #2)
Name...........: NVIDIA GeForce RTX 2070 Super
Processor(s)...: 40
Clock..........: 1380
Memory.Total...: 8192 MB
Memory.Free....: 7153 MB
PCI.Addr.BDFe..: 0000:01:00.0

OpenCL Info:
============

OpenCL Platform ID #1
Vendor..: NVIDIA Corporation
Name....: NVIDIA CUDA
Version.: OpenCL 3.0 CUDA 11.4.94

Backend Device ID #2 (Alias: #1)
Type...........: GPU
Vendor.ID......: 32
Vendor.........: NVIDIA Corporation
Name...........: NVIDIA GeForce RTX 2070 Super
Version........: OpenCL 3.0 CUDA
Processor(s)...: 40
Clock..........: 1380
Memory.Total...: 8192 MB (limited to 2048 MB allocatable in one block)
Memory.Free....: 7360 MB
OpenCL.Version.: OpenCL C 1.2
Driver.Version.: 471.41
PCI.Addr.BDF...: 01:00.0

OpenCL Platform ID #2
Vendor..: Intel(R) Corporation
Name....: Intel(R) OpenCL HD Graphics
Version.: OpenCL 3.0

Backend Device ID #3
Type...........: GPU
Vendor.ID......: 8
Vendor.........: Intel(R) Corporation
Name...........: Intel(R) UHD Graphics 630
Version........: OpenCL 3.0 NEO
Processor(s)...: 24
Clock..........: 1200
Memory.Total...: 13009 MB (limited to 4095 MB allocatable in one block)
Memory.Free....: 12945 MB
OpenCL.Version.: OpenCL C 3.0
Driver.Version.: 27.20.100.8935

[ZerBea]

Your command line is faulty and the hash mode (2500) is deprecated.

Convert your dump file to new hash mode here:
https://hashcat.net/cap2hashcat/

than run hashcat in mode 22000:
$ hashcat -m 22000 xxxxx_xxxxxxxxxx.hc22000 cracked.txt.gz

where xxxxx_xxxxxxxxxx is the filename you received from the online converter (e.g.: 20702_1627284296.hc22000)
and cracked.txt.gz is a (good basic) wordlist from here (https://wpa-sec.stanev.org):
https://wpa-sec.stanev.org/dict/cracked.txt.gz

[John1980]

I try your command but it's no run

thanks

[Snoopy]

you can use option
-D 2 -d 1

to tell hashcat to just use your Geforce, the combined intel CPU/GPU (device 3) and the shipped drivers are known to be "broken"

[ZerBea]

Please comment the entire hashcat message after running:
$ hashcat -m 22000 xxxxx_xxxxxxxxxx.hc22000 cracked.txt.gz

Take also a look at your first comment and hashcat messages:

Code:

No hashes loaded.

That means there is no hash file hashcat can work on! This is the major issue.
Make sure you have a valid hash file for the chosen hash mode and a valid wordlist inside your working directory.
Or if you run mode -a 3, make sure you add a valid mask to your command line (e.g.: -a 3 ?d?d?d?d?d?d?d?d).
That is not the case in your command line!

Code:

CUDA API (CUDA 11.4)
====================
* Device #1: NVIDIA GeForce RTX 2070 Super, 7153/8192 MB, 40MCU

That means that hashcat will use your GPU running the CUDA API.

Code:

OpenCL API (OpenCL 3.0 CUDA 11.4.94) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #2: NVIDIA GeForce RTX 2070 Super, skipped

OpenCL API (OpenCL 3.0 ) - Platform #2 [Intel(R) Corporation]
=============================================================
* Device #3: Intel(R) UHD Graphics 630, skipped

That means hashcat doesn't use the OpenCL runtime on your GPU and doesn't use
the instable Intel Runtime on the embedded Intel CPU/GPU.

Also I recommend to read the wiki to get some basic knowledge:
https://hashcat.net/wiki/doku.php?id=fre..._knowledge

[ZerBea]

Additional we can test if we get a benchmark result for your GPU:
For the new hash mode:
$ hashcat -m 22000 --benchmark

or for the deprecated hash mode:
$ hashcat -m 2500 --benchmark

Output should look like this on hash mode 22000:

Code:

$ hashcat -m 22000 --benchmark
hashcat (v6.2.3-56-gf3f6cfadb) starting in benchmark mode

Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.

CUDA API (CUDA 11.4)
====================
* Device #1: NVIDIA GeForce GTX 1080 Ti, 10737/11175 MB, 28MCU

OpenCL API (OpenCL 3.0 CUDA 11.4.94) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #2: NVIDIA GeForce GTX 1080 Ti, skipped

Benchmark relevant options:
===========================
* --optimized-kernel-enable

Hashmode: 22000 - WPA-PBKDF2-PMKID+EAPOL (Iterations: 4095)

Speed.#1.........:   622.5 kH/s (89.85ms) @ Accel:64 Loops:128 Thr:1024 Vec:1

Started: Mon Jul 26 14:21:58 2021
Stopped: Mon Jul 26 14:22:32 2021

and on hash mode 2500:

Code:

$ hashcat -m 2500 --benchmark
hashcat (v6.2.3-56-gf3f6cfadb) starting in benchmark mode

Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.

CUDA API (CUDA 11.4)
====================
* Device #1: NVIDIA GeForce GTX 1080 Ti, 10737/11175 MB, 28MCU

OpenCL API (OpenCL 3.0 CUDA 11.4.94) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #2: NVIDIA GeForce GTX 1080 Ti, skipped

Benchmark relevant options:
===========================
* --optimized-kernel-enable

Hashmode: 2500 - WPA-EAPOL-PBKDF2 (Iterations: 4095)

Speed.#1.........:   586.5 kH/s (89.26ms) @ Accel:8 Loops:1024 Thr:1024 Vec:1

Started: Mon Jul 26 14:22:39 2021
Stopped: Mon Jul 26 14:23:06 2021

As you can see, I'm running CUDA API and OpenCL API is skipped.

BTW:
The commands to recover the PSK are similar on the WiFi hash modes:
on the new hash mode WPA-PBKDF2-PMKID+EAPOL:
$ hashcat -m 22000 hashfile.hc22000 wordlist

and on the deprecated hash mode WPA-EAPOL-PBKDF2:
$ hashcat -m 2500 hashfile.hccapx wordlist

as well as on the deprecated hash mode WPA-PMKID-PBKDF2:
$ hashcat -m 16800 hashfile.hc16800 wordlist

[Jacek]

Hey buddy ZERABEA update your graphics card drivers and CUDA because your results are HIGHLY understated, I have the same graphics card as you GTX1080TI on default settings, manufacturer MSI, check in which PCX port you have the card
My results for the windows10 comparison:

PS C: \ Hashcat> ./hashcat.exe -m 2500 -b
hashcat (v6.2.3) starting in benchmark mode
CUDA API (CUDA 11.4)
====================
* Device # 1: NVIDIA GeForce GTX 1080 Ti, 9326/11264 MB, 28MCU
Hashmode: 2500 - WPA-EAPOL-PBKDF2 (Iterations: 4095)

Speed. # 1 .........: 682.2 kH / s (83.14ms) @ Accel: 8 Loops: 1024 Thr: 1024 Vec: 1
=============================
Hashmode: 22000 - WPA-PBKDF2-PMKID + EAPOL (Iterations: 4095)

Speed. # 1 .........: 678.4 kH / s (83.39ms) @ Accel: 8 Loops: 1024 Thr: 1024 Vec: 1

Regards Jacek

[ZerBea]

Thanks for sharing your results.
My GPU is clocked down, because I'm a coder/developer (not a cracker) and I don't need full GPU speed.
In addition to that, my word lists are very small (mostly generated from hcxtools sources). That can't utilize full GPU workload, eg.:

Code:

Hardware.Mon.#1..: Temp: 74c Fan: 81% Util: 68% Core:1295MHz Mem:5005MHz Bus:16

Right now, I'm running a task (GTX1080TI) to analyze some hashes to improve hcxtools:
Session..........: hashcat
Status...........: Running
Hash.Name........: WPA-PBKDF2-PMKID+EAPOL
...
Speed.#1.........: 84540 H/s (94.49ms) @ Accel:64 Loops:256 Thr:512 Vec:1
...

Will say, no need for me to take care about "best" hashcat benchmark results.

To answer your question:
GPU is connected to the first PCI slot and the system is a little bit oversized for my purpose:

Code:

Motherboard is:
MSI X370

CPU is:
AMD Ryzen 7 1700 Eight-Core Processor

GPU is:
MSI GTX 1080 Ti AERO

RAM is:
16GB

Driver is latest:
$ pacman -Q | grep nvidia
nvidia 470.63.01-3
nvidia-settings 470.63.01-1
nvidia-utils 470.63.01-1
opencl-nvidia 470.63.01-1

CUDA is latest:
$ pacman -Q | grep cuda
cuda 11.4.1-1