hashcat OpenCL error
#1
Hello...

Long time I didn't use this tool. Today I tried it again and got surprised by the errors I got.

I'm using Debian Stretch and hashcat v3.30 from Debian repositories. I have an ATI/AMD HD7970M running open source drivers and all that stuff.

I have just generated an MD5 hash to try it out and when I run the following command:

Code:
hashcat -m 0 -a 3 63c385429a290dd34a872ee6be7a559d ?l?u?d

It returned the following problems:

Code:
hashcat -m 0 -a 3 63c385429a290dd34a872ee6be7a559d ?l?u?d
hashcat (v3.30) starting...

OpenCL Platform #1: Mesa
========================
* Device #1: AMD PITCAIRN (DRM 2.49.0 / 4.9.0-7-amd64, LLVM 3.9.1), 1433/2048 MB allocatable, 20MCU

Hashes: 1 digests; 1 unique digests, 1 unique salts
Bitmaps: 16 bits, 65536 entries, 0x0000ffff mask, 262144 bytes, 5/13 rotates

Applicable Optimizers:
* Zero-Byte
* Precompute-Init
* Precompute-Merkle-Demgard
* Meet-In-The-Middle
* Early-Skip
* Not-Salted
* Not-Iterated
* Single-Hash
* Single-Salt
* Brute-Force
* Raw-Hash

Watchdog: Hardware Monitoring Interface not found on your system
Watchdog: Temperature abort trigger disabled
Watchdog: Temperature retain trigger disabled

* Device #1: build_opts '-I /usr/share/hashcat/OpenCL -D VENDOR_ID=16 -D CUDA_ARCH=0 -D VECT_SIZE=4 -D DEVICE_TYPE=4 -D DGST_R0=0 -D DGST_R1=3 -D DGST_R2=2 -D DGST_R3=1 -D DGST_ELEM=4 -D KERN_TYPE=0 -D _unroll -cl-std=CL1.2'
* Device #1: Kernel m00000_a3.3e988fe8.kernel not found in cache! Building may take a while...
clBuildProgram(): CL_BUILD_PROGRAM_FAILURE

input.cl:34:1: error: OpenCL version 1.1 does not support the 'static' storage class specifier
input.cl:204:1: error: OpenCL version 1.1 does not support the 'static' storage class specifier

* Device #1: Kernel /usr/share/hashcat/OpenCL/m00000_a3.cl build failure. Proceeding without this device.

Started: Sat Aug 11 13:45:30 2018
Stopped: Sat Aug 11 13:45:33 2018

or
https://paste.debian.net/1037532/

Is there a way of fixing this? Workaround?
Reply
#2
1. use the latest version of hashcat
2. don't use mesa. use rocm if it supports your card or try the official fglrx driver
Reply
#3
(08-11-2018, 04:25 PM)undeath Wrote: 1. use the latest version of hashcat
2. don't use mesa. use rocm if it supports your card or try the official fglrx driver

I have tried to install rocm-dkms but I got an whole bunch of problems. I tried to install latest version of kernel 4.17 but also run into problems. I did an whole sort of things but nothing worked. I just gave up because I ran into a situation where I got a lot of unmet dependencies and got sick of trying to work it around! I also tried to install amdgpu pro 18.20 and more trouble came even following all instructions from AMD site.
Just couldn't make this work. At least without the help of a more expert guy than me. And to be honest I'm no expert, I had the help of what I consider an expert guy in Linux, just not expert in hashcat and all it's dependencies and mostly depressing, all AMD cards dependencies and f*ck ups with Linux support!
If anyone is willing to try to make it work on my laptop, I can give you access to it via ssh connection or screen or so, and give it a try! I would love to make this VGA card get some heavy load!

Edited;
And I used latest version of hashcat (v4.2.1-4-g188a9568). I tried first the repository version and later I tried the github version.

Edited 1;
And apparently rocm-dkms is having compatibility issues with Linux Kernel 4.17 and I have no knowledge to work things around!
Reply
#4
Is there a way to be noticed when ROCm is available for the next kernel release? I've seen someone saying that next compatible kernel version with ROCm will be 4.19.

How can I know when will it be compatible with the next kernel version?
Reply