Device change?

[Pand4]

Do I need to choose device 2 instead? I looked at Nvidia's documentation and it seem I have the correct driver version and the correct cuda version.

Code:

➤ hashcat -b
hashcat (v6.2.6) 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 12.4)
====================
* Device #1: NVIDIA GeForce RTX 4090, 22776/24165 MB, 128MCU

OpenCL API (OpenCL 3.0 CUDA 12.4.131) - Platform #1 [NVIDIA Corporation]
========================================================================
* Device #2: NVIDIA GeForce RTX 4090, skipped

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

-------------------
* Hash-Mode 0 (MD5)
-------------------

cuLinkAddData(): the provided PTX was compiled with an unsupported toolchain.

* Device #1: Kernel /usr/share/hashcat/OpenCL/shared.cl link failed. Error Log:

ptxas application ptx input, line 9; fatal  : Unsupported .version 8.5; current version is '8.4'



* Device #1: Kernel /usr/share/hashcat/OpenCL/shared.cl build failed.

[ZerBea]

The installed driver doesn't match to the installed CUDA toolkit.
More information is here:
https://docs.nvidia.com/cuda/cuda-toolki...index.html
Please take a look at "Table 3 CUDA Toolkit and Corresponding Driver Versions".

[penguinkeeper]

That error only happens with mis-matched driver/CUDA versions so you may have to backdate CUDA or ideally, update your driver

[Pand4]

Code:

➤ nvidia-smi
Sun Jun 23 04:09:14 2024     
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.90.07              Driver Version: 550.90.07      CUDA Version: 12.4    |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf          Pwr:Usage/Cap |          Memory-Usage | GPU-Util  Compute M. |
|                                        |                        |              MIG M. |
|=========================================+========================+======================|
|  0  NVIDIA GeForce RTX 4090        Off |  00000000:01:00.0  On |                  Off |
|  0%  29C    P8              6W /  450W |    845MiB /  24564MiB |      1%      Default |
|                                        |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+                                                   

I wish that were the case. According to table 3, for Linux x86_64 I need
>= Driver Version 550.54.15 for CUDA 12.4

My Driver Version is 550.90.07. I'll see if I can upgrade driver to 555 and CUDA 12.5. Thank you

[ZerBea]

Do not trust nvida-smi!

I can reproduce this problem (if I force a driver/cuda mis-match):

Code:

$ pacman -Q
nvidia 550.90.07-3
nvidia-lts 1:550.90.07-3
nvidia-settings 550.78-1
nvidia-utils 550.90.07-3
opencl-nvidia 550.90.07-3
cuda 12.5.0-1

Code:

$ nvidia-smi
Sun Jun 23 20:01:35 2024      
| NVIDIA-SMI 550.90.07              Driver Version: 550.90.07      CUDA Version: 12.4

The installed version is definitely not 12.4 (as reported by nvidia-smi).


And I end up into the same problem:

Code:

$ hashcat -m 22000 --benchmark
hashcat (v6.2.6-848-gc1a10518f) starting in benchmark mode
...
The device #2 specifically listed was skipped because it is an alias of device #1
CUDA API (CUDA 12.4)
====================
* Device #1: NVIDIA GeForce GTX 1080 Ti, 11027/11165 MB, 28MCU

OpenCL API (OpenCL 3.0 CUDA 12.4.131) - Platform #1 [NVIDIA Corporation]
========================================================================
* Device #2: NVIDIA GeForce GTX 1080 Ti, skipped
...
-------------------------------------------------------------
* Hash-Mode 22000 (WPA-PBKDF2-PMKID+EAPOL) [Iterations: 4095]
-------------------------------------------------------------

cuLinkAddData(): the provided PTX was compiled with an unsupported toolchain.

* Device #1: Kernel /usr/share/hashcat/OpenCL/shared.cl link failed. Error Log:

ptxas application ptx input, line 9; fatal   : Unsupported .version 8.5; current version is '8.4'

After I fixed it (going back to cuda 12.4.1):

Code:

$ pacman -Q
nvidia 550.90.07-3
nvidia-lts 1:550.90.07-3
nvidia-settings 550.78-1
nvidia-utils 550.90.07-3
opencl-nvidia 550.90.07-3
cuda 12.4.1-4

Code:

$ nvidia-smi
Sun Jun 23 20:14:51 2024      
NVIDIA-SMI 550.90.07              Driver Version: 550.90.07      CUDA Version: 12.4

everything is fine:

Code:

$ hashcat -m22000 --benchmark
hashcat (v6.2.6-848-gc1a10518f) starting in benchmark mode
...
Started: Sun Jun 23 20:11:51 2024
Stopped: Sun Jun 23 20:11:54 2024

BTW:
If you're running Arch Linux - this problem has been reported here:
https://gitlab.archlinux.org/archlinux/p...-/issues/7

[Pand4]

Solved. Thank you