Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Mon, 20 May 2013 08:37:22 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: latest bleeding jumbo error

On 20 May, 2013, at 0:09 , magnum <john.magnum@...hmail.com> wrote:
> On 19 May, 2013, at 20:46 , Erik Winkler <ewinkler@...ls.com> wrote:
>> Bleeding jumbo fails for raw-sha256-ng on MacOS 10.8.3 with OpenMP enabled (target macosx-x86-64):
>> 
>> ../run/john -test -format:raw-sha256-ng
>> Will run 2 OpenMP threads
>> Benchmarking: raw-sha256-ng, Raw SHA-256 [128/128 SSE2 instrinsics 4x]... (2xOMP) FAILED (get_hash[0](4))
>> 
>> 
>> The test passes with OpenMP disabled (target macosx-x86-64):
>> 
>> ../run/john -test -format:raw-sha256-ng
>> Benchmarking: raw-sha256-ng, Raw SHA-256 [128/128 SSE2 instrinsics 4x]... DONE
> 
> I can reproduce it. It's not specific to OSX. It is specific to OMP-build (even running one thread) and also to vanilla SSE2:
> 
> Benchmarking: raw-sha256-ng, Raw SHA-256 [128/128 SSE4.1 instrinsics 4x]... (8xOMP) DONE
> Raw:	19333K c/s real, 4750K c/s virtual
> 
> Benchmarking: raw-sha256-ng, Raw SHA-256 [128/128 XOP instrinsics 4x]... (8xOMP) DONE
> Raw:    16891K c/s real, 2114K c/s virtual
> 
> Benchmarking: raw-sha256-ng, Raw SHA-256 [128/128 SSE2 instrinsics 4x]... (8xOMP) FAILED (get_hash[0](4))

This is fixed now, it was a trivial bug. Thanks for reporting.

magnum

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ