[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: shadow passwd reserved field
- From: Ben Collins <bcollins debian org>
- To: pam-list redhat com
- Cc: "Dublin Dale, Terror of the Sunol Grade" <rodmur piratehaven org>
- Subject: Re: shadow passwd reserved field
- Date: Sun, 3 Sep 2000 08:12:32 -0400
On Sat, Sep 02, 2000 at 09:12:18PM -0700, Andrew Morgan wrote:
> I don't know. Perhaps someone else can enlighten us?
>
> [Do you use pam_pwdb or pam_unix?]
>
> Cheers
>
> Andrew
>
> "Dublin Dale, Terror of the Sunol Grade" wrote:
> >
> > Hi,
> >
> > A short question for you hopefully. In shadow passwd files there is a
> > reserved filed that PAM appears to use (but I don't know how since it
> > doesn't appear that sp_flag in the spwd struct is apparently never used).
> > For example an entry like:
> >
> > wilma:xQBvJxB368gfg:11202:0:99999:7:-1:-1:134527444
This is from shadow.h:
unsigned long int sp_flag; /* Reserved. */
And here is the code from putspent.c:
if (p->sp_flag != ~0ul
&& fprintf (stream, "%ld", p->sp_flag) < 0)
++errors;
IMO, whatever code is using putspnam(), needs to set this field to ~0ul, so
it doesn't get written (note, I checked this in glibc 2.1.3 and 2.1.93, both
the same). Neither pam_unix now pam_pwdb calls putspnam(). I don't have
the pwdb sources, so I haven't checked that, but I'de be willing to bet it
has some similar internal call that it uses.
Of couse it would be nice if this were documented. I don't find a
reference to putspnam() in glibc's info pages.
--
-----------=======-=-======-=========-----------=====------------=-=------
/ Ben Collins -- ...on that fantastic voyage... -- Debian GNU/Linux \
` bcollins@debian.org -- bcollins@openldap.org -- bcollins@linux.com '
`---=========------=======-------------=-=-----=-===-======-------=--=---'
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
[]