> Full_Name: Matus Honek
> Version: 2.4.46
> OS: Fedora 28
> URL: ftp://ftp.openldap.org/incoming/Matus-Honek-180731.sh
> Submission from: (NULL) (18.104.22.168)
> Since commit d2d534b, slapadd does not remove spaces from the 'manager'
> attribute value any more. Reverting the commit makes the slapadd work as before.
> Please, see attached script for a reproducer. If this is an expected behaviour
> feel free to close this ticket.
slapadd's job is to load data exactly as provided. Primarily, loading data that
was dumped by slapcat. As such, it's not intended to do any cleanup of the input
data; the input is expected to be clean already.
Closing this ITS.
-- Howard Chu
CTO, Symas Corp. http://www.symas.com
Director, Highland Sun http://highlandsun.com/hyc/
Chief Architect, OpenLDAP http://www.openldap.org/project/
Full_Name: Matus Honek
OS: Fedora 28
Submission from: (NULL) (22.214.171.124)
Since commit d2d534b, slapadd does not remove spaces from the 'manager'
attribute value any more. Reverting the commit makes the slapadd work as before.
Please, see attached script for a reproducer. If this is an expected behaviour
feel free to close this ticket.
On Mon, Jul 30, 2018 at 01:44:31AM +0000, manu(a)netbsd.org wrote:
>@@ -1312,6 +1312,14 @@ if test $ol_enable_lmpasswd != no; then
> AC_DEFINE(SLAPD_LMHASH, 1, [define to support LAN Manager passwords])
Copypasta from the paragraph below? Looks off topic for this hunk.
Full_Name: Emmanuel Dreyfus
Submission from: (NULL) (126.96.36.199)
contrib/slapd-modules/smbk5pwd uses hdb_generate_key_set_password() from
Heimdal, which was shortly turned from a 5 arguments function to a 7 arguments
function before the prototype change was rolled back to address API
Unfortunately, the 7 arguments hdb_generate_key_set_password() made it into
released NetBSD 8.0, causing a build break in contrib/slapd-modules/smbk5pwd.
The attached patch adds a configure test for 7 arguments prototype so that
contrib/slapd-modules/smbk5pwd build again on NetBSD 8.0, and other OS that
would include the 7 arguments hdb_generate_key_set_password().
I really wonder why function rwm_attrs() is called with
stripEntryDN = 1. A comment indicates the front-end generates 'entryDN'.
BTW: The database uses back-mdb. I did not test whether it behaves
differently with back-hdb yet.
Submission from: (NULL) (188.8.131.52)
Enabling slapo-rwm for a database makes operational attribute 'entryDN'
invisible (tested with rootdn).
It's sufficient to add this line to the database section:
IMO this is a serious bug.