clEnqueueWriteBuffer() -4 error when using more than 2 rule files

[tibit]

When using  oclhashcat I get this "clEnqueueWriteBuffer() -4" error when I use more than two rule files in the same command.

With 2 rules everything is ok. With 3 rule files, I get this error.

The rule files only have short rules with maximum 4 arguments for each line?

Any idea or solution??

thank you

tibit

[epixoip]

What OS? What device(s)? For all algorithms or just one particular one?

[tibit]

Windows 7 ultimate 64-Bit
Processor : Intel(R) Core(TM) i7-4960X CPU © 3.60GHz 3.60 GHz
RAM : 32GB
GPU : R9 295x2
Catalyst Driver : 15.7

Problem hard to define.
Small rulefiles (2kb) seem to work fine unless the hashfile is really big (500mb) and the rules demand more or less from the system. Typical rulefiles with a lot of insertion or overwrite rules give more trouble than simple pre-and append rules.
Combination of 2 big rule files (+500kb) often give trouble with even very small hashfile (150kb)

I tried several combinations of different rule files and hashloads and couldn't find any consistend explanation.

beside the former "clEnqueueWriteBuffer() -4", I also get an "clEnqueueBuffer() -61" and in my command there is a "kernel1 can't be loaded in cache".
Am I asking too much from my build??

The same problem occurs with other fast hash modules (didn't check on slow ones).

Thanks for helping here

tibit

[epixoip]

Have you tried a newer driver and/or have you tested with latest oclHashcat beta?

[tibit]

OK,

I was a little reluctant about updating the new driver because this Crimson edition gets a lot of negative reviews.
But I will definitely try your advice.

thanks

[epixoip]

So test with latest oclHashcat beta first, and if that doesn't help then try upgrading the driver.

[atom]

> Combination of 2 big rule files (+500kb)

Not wondering

[ZaLdY]

Is there a way to use more than 1 rule in hashcat gui 0.45b1 under the Commands Tab?

[atom]

No, but they are pure text files. you can simply copy the content of all the rules you want to run at once into a single new file.