Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 6 Jul 2013 21:50:04 +0200
From: Katja Malvoni <kmalvoni@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: Parallella: bcrypt

Hi Yaniv,

On Fri, Jul 5, 2013 at 5:59 AM, Yaniv Sapir <yaniv@...pteva.com> wrote:

> Katja,
>
> From a quick glance at the attached *host *program you attached, I
> spotted a couple of things that *may* be a problem:
> [...]


I tried all your suggestions and changed the code but it still doesn't work
and it behaves strange. For example, different runs of program give
different results because data is not transferred correctly. But cores that
return wrong result are different for every run and it looks like it's
random. Also, number of cores with correct result and correct data aren't
always equal. There is no difference between busy wait and sleep(1) - same
random things happen.
I checked *.srec file and nothing is written on addresses I'm using. I'm
not succeeding in finding a mistake and I can't figure out what is causing
this behaviour.
Something like this happens every time I make changes to working
implementation and I've tried lot of things to make it work but usually
every try fails. What can cause this type of behaviour? Am I very wrong in
my approach because this is taking lot of my time and things that are
happening don't make much sense?

Thank you

Katja

Content of type "text/html" skipped

View attachment "e_bcrypt.c" of type "text/x-csrc" (27966 bytes)

View attachment "host.c" of type "text/x-csrc" (5771 bytes)

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.