Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 07 Apr 2015 18:42:40 +0200
From: magnum <>
Subject: Re: bechmark versus self-test

On 2015-04-06 00:48, Claudio André wrote:
> magnum, all
> JtR should allow the format to set which ciphers in test[] should be
> used to benchmark. Reason:
> - work only with passwords of the same size/cost/whatever.
> A POC can be seen here:

Yes, we are about to do the following format struct change unless
someone has better ideas.

diff --git a/src/formats.h b/src/formats.h
index 53b4c9b..0ca2a94 100644
--- a/src/formats.h
+++ b/src/formats.h
@@ -20,7 +20,7 @@
  * in case of any problem with the new additions
  * (tunable cost parameters)
-#define FMT_MAIN_VERSION 13    /* change if structure fmt_main changes */
+#define FMT_MAIN_VERSION 14    /* change if structure fmt_main changes */

  * fmt_main is declared for real further down this file, but we refer
to it in
@@ -102,6 +102,9 @@ struct db_salt;
 struct fmt_tests {
        char *ciphertext, *plaintext;
        char *fields[10];
+       int skip_on_bench;
+       char *mask_plain;
+       char *mask;


The skip_on_bench can be used to (de-)select test vectors used for
benchmarking (as opposed to self-test, which will ignore this field).

The mask_plain and mask can be used for testing/benchmarking GPU-side
mask. For example, if the plaintext is "magnum", mask_plain could be
"num" and the mask could be "?l?l?l?w".

Like with the *fields[10] we'll leave most formats unchanged, leaving
the new items at NULL and 0 - which is perfect for our needs.

Any objections?


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.