Return to BSD News archive
Newsgroups: comp.os.386bsd.bugs
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!agate!howland.reston.ans.net!news.sprintlink.net!hookup!swrinde!news.dell.com!tadpole.com!uunet!zib-berlin.de!news.th-darmstadt.de!News.Uni-Marburg.DE!news.belwue.de!delos.stgt.sub.org!nadia.s.bawue.de!luva!migieger
From: migieger@luva.lb.bawue.de (Michael Giegerich)
Subject: Re: Bug in libcrypt.a on NetBSD 1.0?
Organization: private FreeBSD site
Message-ID: <CyrLJp.4Dr@luva.lb.bawue.de>
References: <HILDEKCA.94Nov4095109@shorty.eng.umd.edu>
Date: Fri, 4 Nov 1994 22:38:13 GMT
Lines: 39
In article <HILDEKCA.94Nov4095109@shorty.eng.umd.edu>,
Kevin Hildebrand <hildekca@eng.umd.edu> wrote:
>It seems that crypt() doesn't work correctly with non-alphanumeric
>salts.
>
>main()
>{
> printf ("%s\n", crypt ("dork", "B$"));
>}
>
>The result from NetBSD: B$PCRvVbyopZc
>The result from others: B$nCWoDwQadXI
I tried this with FreeBSD-1.1.5.1: it gives the same results as NetBSD.
Then I built a library with the crypt.c from the original NET/2 release:
same results as with NetBSD or FreeBSD.
So nothing seems wrong to me...
>
>However, if you use an alphanumeric salt:
>
>main()
>{
> printf ("%s\n", crypt ("dork", "B4"));
>}
>
>The result from NetBSD: B4.H1Menh1vrc
>The result from others: B4.H1Menh1vrc
Same for FreeBSD.
-Michael
--
------------------------------+------------------------------
Michael Giegerich | migieger@luva.lb.bawue.de
------------------------------+------------------------------