Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 23 Jun 2012 12:16:58 +0200
From: Frank Dittrich <>
Subject: Re: segfault with ./john --test --format=md5ns caused
 by commit aab646b42afe93ca909ba1a99bc0e2a1b027a79d

On 06/23/2012 12:02 PM, Frank Dittrich wrote:
> Hi all,
> looks like this one is related to a recent change in core john

I worked around that segfault by repeating the first test case:

diff --git a/src/NS_fmt_plug.c b/src/NS_fmt_plug.c
index 864395d..3f3849b 100644
--- a/src/NS_fmt_plug.c
+++ b/src/NS_fmt_plug.c
@@ -58,6 +58,7 @@
 static struct fmt_tests tests[] = {
        {"admin$nMjFM0rdC9iOc+xIFsGEm3LtAeGZhn", "password"},
        {"a$nMf9FkrCIgHGccRAxsBAwxBtDtPHfn", "netscreen"},
+       {"admin$nMjFM0rdC9iOc+xIFsGEm3LtAeGZhn", "password"},

I could be wrong, but the commit that looks like it could have caused
this regression is
Use first two salts when benchmarking instead of top one.

And indeed, reverting that change fixes the problem for me.
But I think that patch was in there for a reason, so I temporarily just
implemented the above workaround for --format=md5ns instead.


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.