john-dev - RE: mscash broken for ages?

Follow @Openwall on Twitter for new release announcements and other news

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <007301cbdeb9$a8f537b0$fadfa710$@net>
Date: Wed, 9 Mar 2011 18:25:29 -0600
From: "jfoug" <jfoug@....net>
To: <john-dev@...ts.openwall.com>
Subject: RE: mscash broken for ages?

Here is the 'prepare' function I have for mscash.  But without all the other john infrastructure changes, this function is 'meaningless'

static char *prepare(char *split_fields[10], struct fmt_main *pFmt)
{
//	if ((db->format == &fmt_mscash) || (db->format == &fmt_mscash2)) {
//		piece = (char *) mem_alloc(strlen(login) + strlen(ciphertext) + 4);
//		sprintf(piece, "M$%s#%s", login, ciphertext);
	char *cp;
	if (!strncmp(split_fields[1], "M$", 2))
		return split_fields[1];
	cp = mem_alloc(strlen(split_fields[0]) + strlen(split_fields[1]) + 4);
	sprintf (cp, "M$%s#%s", split_fields[0], split_fields[1]);
	if (valid(cp, pFmt))
	{
		char *cipher = str_alloc_copy(cp);
		MEM_FREE(cp);
		return cipher;
	}
	MEM_FREE(cp);
	return split_fields[1];
}

Hopefully the above code will get 'through' without problems.  In this 'prepare', I simply return the field[1] if it starts with "M$" which is pretty much the same as the fix you posted was doing.  

I just need to get with Solar, to coordinate getting a couple of larger 'structure' changes out. I personally think they should be put either into john proper, or put into the jumbo.  Changes like this (a new method) into the fmt_methods structure (and I have 2 changes like this), impact pretty much every single format.  Putting a patch out as a stand alone against a jumbo, ends up causing all patches to be very 'order dependant'.  

Jim.

>-----Original Message-----
>From: Bucsay Balázs [mailto:earthquake@...on.hu]
>Hi!
>
>The problem is with the jumbo patch.

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.