Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 8 Aug 2015 12:53:59 +0800
From: Lei Zhang <zhanglei.april@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: JtR on ARM (NEON)


> On Jul 31, 2015, at 4:35 PM, Solar Designer <solar@...nwall.com> wrote:
> 
>> (OpenMP is disabled in this test. PBKDF2-HMAC-SHA512 failed somehow, so I chose sha512crypt here.)
> 
> You'll need to investigate why PBKDF2-HMAC-SHA512 fails.  This might
> provide a clue as to why sha512crypt became slower.

I see some weird stuff happening in pbkdf2-hmac-sha512. So far every formats that use SIMDSHA512body() work correctly, except for pbkdf2-hmac-sha512. 

I tried to print out the content of crypt_out[] after each invocation of crypt_all(), and compared it with another instance of JtR running on my laptop(AVX). The first few invocations of crypt_all() on ARM gave the correct result, but at some point (the 10th invocation) it started to misbehave and failed the self-test.

More strangely, when I ran it in GDB, the output is different from a normal run; even each run in gdb gave different results. Do you have any idea what might be happening here?


Lei

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.