I haven't quite figured out everything that's going on, but this patch
definitely shows when it occurs in test036:
Index: result.c
===================================================================
RCS file: /repo/OpenLDAP/pkg/ldap/libraries/libldap/result.c,v
retrieving revision 1.140
diff -u -r1.140 result.c
--- result.c 3 Dec 2006 23:11:12 -0000 1.140
+++ result.c 12 Dec 2006 22:53:25 -0000
@@ -914,6 +914,10 @@
}
if ( lr != NULL ) {
+ if ( tag == LDAP_RES_SEARCH_RESULT ) {
+ Debug( LDAP_DEBUG_TRACE, "request done: ld %p
msgid %ld but mis-handled!\n",
+ (void *)ld, (long) id, 0 );
+ }
ldap_return_request( ld, lr, 0 );
lr = NULL;
}
The response is actually returned to the caller properly, but we never
decide to "mark request completed." Setting a breakpoint here will let
you poke around at it. v3ref is set to V3REF_TOAPP at this point. This
should mean that the result included a referral and it will be returned
to the caller, which means that we can dispose of the request, but we
never do.
--
-- Howard Chu
Chief Architect, Symas Corp. http://www.symas.com
Director, Highland Sun http://highlandsun.com/hyc
OpenLDAP Core Team http://www.openldap.org/project/
kuenne(a)rentec.com wrote:
> Full_Name: Karsten Kuenne
> Version: 2.3.30
> OS: Solaris 10
> URL: ftp://ftp.openldap.org/incoming/
> Submission from: (NULL) (216.223.241.248)
>
>
> The following loop in translucent.c at around line 297 fails if the first mod in
> the modlist is a delete and eventually crashes slapd:
This is now fixed in HEAD, thanks for the report.
--
-- Howard Chu
Chief Architect, Symas Corp. http://www.symas.com
Director, Highland Sun http://highlandsun.com/hyc
OpenLDAP Core Team http://www.openldap.org/project/
Full_Name: Karsten Kuenne
Version: 2.3.30
OS: Solaris 10
URL: ftp://ftp.openldap.org/incoming/
Submission from: (NULL) (216.223.241.248)
The following loop in translucent.c at around line 297 fails if the first mod in
the modlist is a delete and eventually crashes slapd:
....
Debug(LDAP_DEBUG_TRACE,
"=> translucent_modify: silently dropping delete: %s\n",
m->sml_desc->ad_cname.bv_val, 0, 0);
==> for(mm = op->orm_modlist; mm->sml_next != m; mm = mm->sml_next);
mm->sml_next = m->sml_next;
m->sml_next = NULL;
slap_mods_free(m, 1);
....
I tried to come up with a fix but so far I always got a "Invalid modification"
because I apparently got something wrong so I worked around it on the client
side for now.
<quote who="Howard Chu">
> ghenry(a)suretecsystems.com wrote:
>> <quote who="hyc(a)symas.com">
>>> ghenry(a)suretecsystems.com wrote:
>>>> Full_Name: Gavin Henry
>>>> Version: 2.3.30
>>>> OS: FC6
>>>> URL:
>>>> http://www.suretecsystems.com/our_docs/admin-guide-monitoring.patch
>>>> Submission from: (NULL) (212.159.59.85)
>>>>
>>>>
>>>> Hi all,
>>>>
>>>> Monitoring section updated. Please review and provide feedback.
>>> You use the term "subentries" in several places where I'm pretty sure
>>> you mean "child entries"; "subentry" has a special meaning in
>>> X.500/LDAP
>>> that isn't relevant here.
>>
>> Sure, that's what I meant really ;-)
>>
>> I'll update it and send another patch. Or is there an easier way?
>
> Your original patch is already in CVS now. Just send a diff against the
> CVS version.
New version with TOC set to 3:
http://www.suretecsystems.com/our_docs/admin_guide.patch
Thanks.
--
Kind Regards,
Gavin Henry.
Managing Director.
T +44 (0) 1224 279484
M +44 (0) 7930 323266
F +44 (0) 1224 824887
E ghenry(a)suretecsystems.com
Open Source. Open Solutions(tm).
http://www.suretecsystems.com/
At 12:59 AM 12/12/2006, lukeh(a)padl.com wrote:
>Full_Name: Luke Howard
>Version: HEAD
>OS:
>URL: http://www.padl.com/~lukeh/openldap-HEAD-nto.diff
>Submission from: (NULL) (203.13.32.80)
>
>
>A customer had requested we add an option to ldapsearch to allow a network
>timeout to be specified on the command line.
>
>A patch against HEAD is attached.
>
>I am aware that you can specify NETWORK_TIMEOUT in ldap.conf, and that free
>command line options are precious.
Given this, -e nettimeout=X might be better choice here.
Or -o nettimeout=X.
Full_Name: Pierangelo Masarati
Version: HEAD,re23
OS: irrelevant
URL: ftp://ftp.openldap.org/incoming/
Submission from: (NULL) (87.28.220.33)
When chasing a referral, if the DN in the referral differs from the one in the
original request slapo-chain(5) disregards it and uses the original request DN.
This behavior disagrees with that of the client library.
p.
Full_Name: Luke Howard
Version: HEAD
OS:
URL: http://www.padl.com/~lukeh/openldap-HEAD-nto.diff
Submission from: (NULL) (203.13.32.80)
A customer had requested we add an option to ldapsearch to allow a network
timeout to be specified on the command line.
A patch against HEAD is attached.
I am aware that you can specify NETWORK_TIMEOUT in ldap.conf, and that free
command line options are precious.
<quote who="Howard Chu">
> ghenry(a)suretecsystems.com wrote:
>> <quote who="hyc(a)symas.com">
>>> ghenry(a)suretecsystems.com wrote:
>>>> Full_Name: Gavin Henry
>>>> Version: 2.3.30
>>>> OS: FC6
>>>> URL:
>>>> http://www.suretecsystems.com/our_docs/admin-guide-monitoring.patch
>>>> Submission from: (NULL) (212.159.59.85)
>>>>
>>>>
>>>> Hi all,
>>>>
>>>> Monitoring section updated. Please review and provide feedback.
>>> You use the term "subentries" in several places where I'm pretty sure
>>> you mean "child entries"; "subentry" has a special meaning in
>>> X.500/LDAP
>>> that isn't relevant here.
>>
>> Sure, that's what I meant really ;-)
>>
>> I'll update it and send another patch. Or is there an easier way?
>
> Your original patch is already in CVS now. Just send a diff against the
> CVS version.
Thanks.
>
> --
> -- Howard Chu
> Chief Architect, Symas Corp. http://www.symas.com
> Director, Highland Sun http://highlandsun.com/hyc
> OpenLDAP Core Team http://www.openldap.org/project/
>
ghenry(a)suretecsystems.com wrote:
> <quote who="hyc(a)symas.com">
>> ghenry(a)suretecsystems.com wrote:
>>> Full_Name: Gavin Henry
>>> Version: 2.3.30
>>> OS: FC6
>>> URL: http://www.suretecsystems.com/our_docs/admin-guide-monitoring.patch
>>> Submission from: (NULL) (212.159.59.85)
>>>
>>>
>>> Hi all,
>>>
>>> Monitoring section updated. Please review and provide feedback.
>> You use the term "subentries" in several places where I'm pretty sure
>> you mean "child entries"; "subentry" has a special meaning in X.500/LDAP
>> that isn't relevant here.
>
> Sure, that's what I meant really ;-)
>
> I'll update it and send another patch. Or is there an easier way?
Your original patch is already in CVS now. Just send a diff against the
CVS version.
--
-- Howard Chu
Chief Architect, Symas Corp. http://www.symas.com
Director, Highland Sun http://highlandsun.com/hyc
OpenLDAP Core Team http://www.openldap.org/project/
<quote who="hyc(a)symas.com">
> ghenry(a)suretecsystems.com wrote:
>> Full_Name: Gavin Henry
>> Version: 2.3.30
>> OS: FC6
>> URL: http://www.suretecsystems.com/our_docs/admin-guide-monitoring.patch
>> Submission from: (NULL) (212.159.59.85)
>>
>>
>> Hi all,
>>
>> Monitoring section updated. Please review and provide feedback.
>
> You use the term "subentries" in several places where I'm pretty sure
> you mean "child entries"; "subentry" has a special meaning in X.500/LDAP
> that isn't relevant here.
Sure, that's what I meant really ;-)
I'll update it and send another patch. Or is there an easier way?
Thanks.
>
> --
> -- Howard Chu
> Chief Architect, Symas Corp. http://www.symas.com
> Director, Highland Sun http://highlandsun.com/hyc
> OpenLDAP Core Team http://www.openldap.org/project/
>
>
>