Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug
Bug#: 134010
Alias:
Product:
Component:
Status: RESOLVED
Resolution: FIXED
Assigned To: Gentoo Security <security@gentoo.org>
Hardware:
OS:
Version:
Priority:
Severity:
Reporter: Dax <gentoomail@gmail.com>
Add CC:
CC:
Remove selected CCs
URL:
Summary:
Status Whiteboard:
Keywords:
Flags: Requestee:
 
 
gentoomail: ()

Filename Description Type Creator Created Size Actions
openldap-2.3.24.ebuild.patch openldap-2.3.24.ebuild.patch with new stuff removed. patch Tuan Van (RETIRED) 2006-06-02 11:25 0000 5.21 KB Details | Diff
Create a New Attachment (proposed patch, testcase, etc.) View All

Bug 134010 depends on: Show dependency tree
Bug 134010 blocks:

Additional Comments: (this is where you put emerge --info)


Not eligible to see or edit group visibility for this bug.






View Bug Activity   |   Format For Printing   |   XML   |   Clone This Bug


Description:   Opened: 2006-05-22 02:16 0000
Hi All,

Date:           19 May 2006
SecWatch ID:            1014107
Vendor URL:             http://www.openldap.org/
Original Advisory:             
http://www.openldap.org/software/release/changes.html
http://www.openldap...tic=1sortbydate=0f=h

Description:
A weakness with unknown impact has been reported in OpenLDAP.

The weakness is caused due to a boundary error in slurpd within the handling of
the status file. This can be exploited to cause a stack-based buffer overflow
via an overly long hostname read from the status file.


Affected:
OpenLDAP version 2.3.21. Prior versions may also be affected.


Solution:
The vulnerability has been fixed in version 2.3.22 or later, available:
http://www.openldap.org/software/download/


Credits:
Reported by vendor.

rgds
Daxomatic

------- Comment #1 From Andrea Barisani (RETIRED) 2006-05-23 02:04:38 0000 -------
Seems very oscure to me and I don't think (not fully checked though) you can
inject arbitrary hostnames in slurpd status file, so it's dependent on DNS and
your configuration.

So it looks far from being critical to me.

------- Comment #2 From Dax 2006-05-23 04:41:07 0000 -------
Hi Lcars
Is the new version ready to be unmasked? If so, could you unmask it for stable?

rgds
Daxomatic

------- Comment #3 From Markus Ullmann 2006-05-23 15:30:15 0000 -------
The new version has an issue with new-style config directory, mentioned in bug
#133898 which upstream already knows about

So really non-easy decision here...

------- Comment #4 From Stefan Cornelius (RETIRED) 2006-05-23 15:50:08 0000 -------
http://tinyurl.com/s34lu

This seems to be the patch - could somebody do a revbump of the old-config
style version with this?

------- Comment #5 From Thierry Carrez (RETIRED) 2006-05-30 11:26:38 0000 -------
ldap team, please bump with provided patch or comment

------- Comment #6 From Markus Ullmann 2006-05-30 14:06:13 0000 -------
As .24 was released some hours ago, I bumped right now

------- Comment #7 From Sune Kloppenborg Jeppesen 2006-06-01 09:39:51 0000 -------
Arches please test and mark stable and sorry for the delay.

------- Comment #8 From Thomas Cort (RETIRED) 2006-06-01 11:32:50 0000 -------
(In reply to comment #7)
> Arches please test and mark stable and sorry for the delay.

What version are we supposed to stable? Seems like the unaffected versions are
still masked.

# Markus Ullmann <jokey@gentoo.org (21 May 2006)
# OpenLDAP serious config problem, see bug #133898
>=net-nds/openldap-2.3.23

------- Comment #9 From Markus Ullmann 2006-06-01 12:21:07 0000 -------
Readjusted the package mask so that all versions below 2.3.24 are masked.
2.3.24 is the candidate for stable

------- Comment #10 From Thomas Cort (RETIRED) 2006-06-01 13:50:13 0000 -------
alpha done.

------- Comment #11 From Markus Ullmann 2006-06-01 14:06:20 0000 -------
*** Bug 130975 has been marked as a duplicate of this bug. ***

------- Comment #12 From Jason Wever (RETIRED) 2006-06-01 16:30:16 0000 -------
If mit-krb5 is used to satisfy the virtual/kerberos use flag dependency for
kerberos, openldap will fail to build as mit-krb5 does not provide both
kadm5/admin.h or hdb.h headers.

------- Comment #13 From Jakub Moc (RETIRED) 2006-06-01 17:42:20 0000 -------
(In reply to comment #9)
> Readjusted the package mask so that all versions below 2.3.24 are masked.
> 2.3.24 is the candidate for stable

# masking older versions due to security bug #134010 and bug #133898
<net-nds/openldap-2.3.24

Well uh... you've killed all stable openldap, not exactly a good thing,
considering that noone besides alpha keyworded the fixed version.

------- Comment #14 From Jakub Moc (RETIRED) 2006-06-01 17:43:09 0000 -------
*** Bug 135216 has been marked as a duplicate of this bug. ***

------- Comment #15 From solar 2006-06-01 17:55:44 0000 -------
(In reply to comment #9)
> Readjusted the package mask so that all versions below 2.3.24 are masked.
> 2.3.24 is the candidate for stable

Jokey,
Please never ever do that again. You started to cause a world of pain for alot 
of people. We first have the arches test and decide if it can be marked stable.
Then p.masking if needed.

Carlo,
Thanks for reverting that but please next time find the bug which caused 
the breakage and comment on it.

------- Comment #16 From Tobias Scherbaum 2006-06-01 21:59:00 0000 -------
ppc stable

------- Comment #17 From Markus Rothe 2006-06-02 05:55:47 0000 -------
stable on ppc64

------- Comment #18 From Wes 2006-06-02 09:58:57 0000 -------
(In reply to comment #12)
> If mit-krb5 is used to satisfy the virtual/kerberos use flag dependency for
> kerberos, openldap will fail to build as mit-krb5 does not provide both
> kadm5/admin.h or hdb.h headers.
> 

Will/Can it be made to work with mit-krb5 or do I have to switch to heimdal?

------- Comment #19 From Tuan Van (RETIRED) 2006-06-02 10:38:19 0000 -------
a diff berween a last ebuild with the most arch keywork (openldap-2.2.28-r3)
and this one, we can see that there are a lot more modules in contrib/ comes
with openldap-2.3.24.ebuild . IMHO, it's a bad practice to introduce new
"features" in an ebuild requires secutiry stable keyword which causes problem
as seen with smbk5pwd module.

------- Comment #20 From Tuan Van (RETIRED) 2006-06-02 11:25:38 0000 -------
Created an attachment (id=88202) [details]
openldap-2.3.24.ebuild.patch with new stuff removed.

dsaschema, smbk5pwd, addrdnvalues are new stuff added in 2.3.24. I propose we
remove them for now.

~/cvs/gentoo-x86/net-nds/openldap $ grep -l dsaschema *.ebuild
openldap-2.3.24.ebuild
~/cvs/gentoo-x86/net-nds/openldap $ grep -l smbk5pwd *.ebuild
openldap-2.3.24.ebuild
~/cvs/gentoo-x86/net-nds/openldap $ grep -l addrdnvalues *.ebuild
openldap-2.3.24.ebuild

------- Comment #21 From Markus Ullmann 2006-06-02 11:48:31 0000 -------
Okay, a short "what went wrong here" story to clarify things a bit.

First let me say sorry for the wrong package mask. I just had 2.3 branch in
mind when setting that mask.

Then in 2.3.21-r1 I started testing the contrib overlays as described in bug
#116045 but I decided to let it not hit the tree as I didn't have enough time
to test.
After 2.3.23 was out I prepared things to go live as in the meantime I was fine
with it. Then the slurpd bug came in and just thought "okay, not that many
changes (some overlays) so let it go.

Afterwards at least I do know better now.

Right now I prepare an 2.3.24-r1 ebuild without all the extra overlays that can
go stable then (already talked to arches who stabled already to make sure they
help here) and the new overlays (with what I learned from now) will go into an
-r3 then which should be at best ~ keyworded or just  stay hardmasked for
further development.

All in all not the best one would expect, I admit that but now try to make best
out of it and get this crap sorted.

------- Comment #22 From Markus Ullmann 2006-06-02 17:32:12 0000 -------
Okay, candidate for stable is now 2.3.24-r1

------- Comment #23 From Pieter Van den Abeele 2006-06-02 23:48:38 0000 -------
-r1 works fine here on ppc with mit-krb5, but r2 still fails. 

------- Comment #24 From René Nussbaumer 2006-06-04 04:12:08 0000 -------
Stable on hppa

------- Comment #25 From Thomas Cort (RETIRED) 2006-06-04 18:07:01 0000 -------
amd64 stable.

------- Comment #26 From Jason Wever (RETIRED) 2006-06-05 16:36:13 0000 -------
SPARC stable

------- Comment #27 From Joshua Jackson 2006-06-11 13:17:40 0000 -------
x86 done sorry about the delay

------- Comment #28 From Markus Rothe 2006-06-11 23:32:16 0000 -------
stable on ppc64

we are not CC'd...

------- Comment #29 From Sune Kloppenborg Jeppesen 2006-06-15 09:01:47 0000 -------
GLSA 200606-17

arm, ia64, mips and s390 don't forget to mark stable to benifit from the GLSA.

------- Comment #30 From Markus Ullmann 2006-06-15 17:33:26 0000 -------
Arm done

------- Comment #31 From Joshua Kinard 2006-09-04 22:12:18 0000 -------
2.3.24 stable on mips.

Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug