Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 30 Sep 2013 11:17:07 +0100
From: Dániel Bali <>
Subject: Re: Daniel's weekly report #15

Hello Lukas,

2013/9/29 Lukas Odzioba <>

> 2013/9/24 Dániel Bali <>:
> > 1. I was working on the CRC32 format.
> >   - The OpenCL implementation computes the correct CRC32 values
> >   - There is still a bug where the results are compared.
> Sorry for delay. The problem in comparision is in pure opencl crc32
> implementation or in "gcnasm'ed" version?

It's in the pure OpenCL version, so the error is either from the
KernelAnalyzer's generated ISA or from the patching.
My guess is that it's the former.

> > 2. I wrote a short list of steps that might be needed to run GCN code
> inside
> > JtR
> >   - This is on the top of the wiki page now
> >
> >
> At first glance it looks ok, but later we should try to make it
> easier/automatic to be redistributable.

Yes, this should definitely be done by some sort of script. However, I
think I should get it working properly first.

> > By now the patched ELF doesn't freeze the GPU, but it runs with errors. I
> > get the following error:
> > only)]... Warning: binary() returned misaligned pointer
> > FAILED (cmp_all(1))
> >
> > Here is the two hex values that are compared in cmp_all:
> > 7ada0bcc, 6e65706f
> Does this refer to idea from previous mail (below)?

Sorry, I don't understand. What do you mean here?

2013/9/27 Dániel Bali <>:
> > This would mean setting up a minimal host-side environment that passes
> the correct buffers and compares the output to the required results.
> If not have you tried that first?

This is the next thing I will do, preferably tonight.


Content of type "text/html" skipped

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.