Re: [Samba] Samba4 Multi-Master replication

2012-06-13 Thread Daniel Müller
You  always need to have two Samba4/dns and the entries of both dns in your
clients network settings too.

---
EDV Daniel Müller

Leitung EDV
Tropenklinik Paul-Lechler-Krankenhaus
Paul-Lechler-Str. 24
72076 Tübingen

Tel.: 07071/206-463, Fax: 07071/206-499
eMail: muel...@tropenklinik.de
Internet: www.tropenklinik.de
---
-Ursprüngliche Nachricht-
Von: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] Im
Auftrag von Gémes Géza
Gesendet: Dienstag, 12. Juni 2012 19:19
An: samba@lists.samba.org
Betreff: Re: [Samba] Samba4 Multi-Master replication

On 2012-06-12 12:16, Morten Kramer wrote:
 Hi guys,


 I'm trying to get the Samba4 multi-master replication to work.



 I set up the primary domain controller using this howto (under CentOS
 6.2 x64):


 http://wiki.samba.org/index.php/Samba4/HOWTO



 I installed bind 9.8.3 and enabled encrypted dns updates.




 I set up another VM with the same CentOS version and oriented myself 
 on this howto:

 http://wiki.samba.org/index.php/Samba4/HOWTO/Join_a_domain_as_a_DC

 to join the second DC into the domain. I edited /etc/resolv.conf and 
 set the nameserver to the IP of the primary DC (bind dns server).


 Basic replication seems to work (not doing the rsync for sysvol yet).
 However, when i take the primary DC offline (bind keeps running), I 
 can't use any of the .msc domain admin tools anymore.

 I always get an error message, telling me that there is no RPC server 
 available.
 When i run gpmc.msc i can choose the DC i want to work on and I can 
 see the secondary one, but it will come back with the RPC error.

 I had Wireshark running on one of the Windows7 clients. It seems like 
 it tries to talk to the 2nd DC (DCERPC packets). But i'm not an expert 
 in packet analysis, could somebody give me a hint what to look for here?



 User authentication does still work and Kerberos tickets are generated 
 by the 2nd DC.



 I can find this in the log:


 ../source4/dsdb/kcc/kcc_topology.c:1402: failed to find nCName 
 attribute of object 
 CN=ac7bf69c-9458-4205-acba-6fe172412d1b,CN=Partitions,CN=Configuration
 ,DC=aeriatest2,DC=dc,DC=loc
 ../source4/dsdb/kcc/kcc_topology.c:3158: failed to color vertices:
 NT_STATUS_INTERNAL_DB_CORRUPTION
 ../source4/dsdb/kcc/kcc_topology.c:3415: failed to create connections:
 NT_STATUS_INTERNAL_DB_CORRUPTION
 ...
 Warning: 60 extra bytes in incoming RPC request
 ../source4/rpc_server/drsuapi/dcesrv_drsuapi.c:74: doing DsBind with 
 system_session



 Do i need to configure something extra, so the secondary DC will be 
 able to act as an RPC server?





 Thanks,
 freezer
Hi,

With your setup DNS is the single point of failure, because with the
(default) DLZ setup bind9 is able to serve DNS records only when samba4 is
running on that box.
My recommendation would be to try to set up DNS on the second DC too.

Regards.

Geza

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba3+OpenLDAP - Samba4 implications.

2012-06-13 Thread Alex Ferrara
Hi everyone,

I might be going over old stuff, and if so, I apologise.

I administer a network that uses Samba 3 with an OpenLDAP backend for domain 
logons, printing and file sharing. I am interested in moving to Samba4 for the 
domain control side of things, but the twist is that I have many other things 
relying on OpenLDAP for authentication and configuration, with several custom 
schemas.

Is there a samba4 schema for OpenLDAP or is there a migration path for networks 
like mine?

Alex Ferrara
Director
Receptive IT Solutions




-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba 4 samba-tool dns question

2012-06-13 Thread Trever L. Adams
Hello Everyone,

I am used to messing with bind zones directly. With samba 4, I cannot do
that. Is the following the correct way to add a TXT record for SPF to
the zone using samba-tool?


/usr/local/samba/bin/samba-tool dns add s4server.example.org example.org
example.org TXT v=spf1 mx -all

I am not sure it is because it is showing up with dig as v=spf1 mx
-all where it should be all one string, I believe.

Thank you for any help,
Trever
-- 
Life is a comedy for those who think and a tragedy for those who feel.
-- Unknown



signature.asc
Description: OpenPGP digital signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba4 Multi-Master replication ( slave setup possible ?)

2012-06-13 Thread L . P . H . van Belle
Is it possible to create slave DNS servers of the samba server. 

I would like to setup like this. ( this is also my running setup with samba 3) 

  SLAVEDNS1 \
SAMBA(DNS)  / Clients pc's with dns servers slave1 and slave2 
\ SLAVEDNS2 / 

I my case slavedns(1) is also master of an external lan which is replicated to 
slavedns2. 
This was needed because of security and the way i extract the external dns 
settings, 
dont ask why, i just created a solution which my ICT company could not. 

so the question is, can i do this with samba4 and bind DLZ setup?

Louis


-Oorspronkelijk bericht-
Van: muel...@tropenklinik.de 
[mailto:samba-boun...@lists.samba.org] Namens Daniel Müller
Verzonden: 2012-06-13 09:01
Aan: 'Gémes Géza'; samba@lists.samba.org
Onderwerp: Re: [Samba] Samba4 Multi-Master replication

You  always need to have two Samba4/dns and the entries of 
both dns in your
clients network settings too.

---
EDV Daniel Müller

Leitung EDV
Tropenklinik Paul-Lechler-Krankenhaus
Paul-Lechler-Str. 24
72076 Tübingen

Tel.: 07071/206-463, Fax: 07071/206-499
eMail: muel...@tropenklinik.de
Internet: www.tropenklinik.de
---
-Ursprüngliche Nachricht-
Von: samba-boun...@lists.samba.org 
[mailto:samba-boun...@lists.samba.org] Im
Auftrag von Gémes Géza
Gesendet: Dienstag, 12. Juni 2012 19:19
An: samba@lists.samba.org
Betreff: Re: [Samba] Samba4 Multi-Master replication

On 2012-06-12 12:16, Morten Kramer wrote:
 Hi guys,


 I'm trying to get the Samba4 multi-master replication to work.



 I set up the primary domain controller using this howto (under CentOS
 6.2 x64):


 http://wiki.samba.org/index.php/Samba4/HOWTO



 I installed bind 9.8.3 and enabled encrypted dns updates.




 I set up another VM with the same CentOS version and oriented myself 
 on this howto:

 http://wiki.samba.org/index.php/Samba4/HOWTO/Join_a_domain_as_a_DC

 to join the second DC into the domain. I edited /etc/resolv.conf and 
 set the nameserver to the IP of the primary DC (bind dns server).


 Basic replication seems to work (not doing the rsync for sysvol yet).
 However, when i take the primary DC offline (bind keeps running), I 
 can't use any of the .msc domain admin tools anymore.

 I always get an error message, telling me that there is no 
RPC server 
 available.
 When i run gpmc.msc i can choose the DC i want to work on and I can 
 see the secondary one, but it will come back with the RPC error.

 I had Wireshark running on one of the Windows7 clients. It 
seems like 
 it tries to talk to the 2nd DC (DCERPC packets). But i'm not 
an expert 
 in packet analysis, could somebody give me a hint what to 
look for here?



 User authentication does still work and Kerberos tickets are 
generated 
 by the 2nd DC.



 I can find this in the log:


 ../source4/dsdb/kcc/kcc_topology.c:1402: failed to find nCName 
 attribute of object 
 
CN=ac7bf69c-9458-4205-acba-6fe172412d1b,CN=Partitions,CN=Configuration
 ,DC=aeriatest2,DC=dc,DC=loc
 ../source4/dsdb/kcc/kcc_topology.c:3158: failed to color vertices:
 NT_STATUS_INTERNAL_DB_CORRUPTION
 ../source4/dsdb/kcc/kcc_topology.c:3415: failed to create 
connections:
 NT_STATUS_INTERNAL_DB_CORRUPTION
 ...
 Warning: 60 extra bytes in incoming RPC request
 ../source4/rpc_server/drsuapi/dcesrv_drsuapi.c:74: doing DsBind with 
 system_session



 Do i need to configure something extra, so the secondary DC will be 
 able to act as an RPC server?





 Thanks,
 freezer
Hi,

With your setup DNS is the single point of failure, because with the
(default) DLZ setup bind9 is able to serve DNS records only 
when samba4 is
running on that box.
My recommendation would be to try to set up DNS on the second DC too.

Regards.

Geza

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba



-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Self-Service-Password check_ntlm_password

2012-06-13 Thread Cédric Carlen
Hello all,

I've got a problem to change my password with self-service-password (
http://ltb-project.org/wiki/documentation/self-service-password/latest/start )
who is an web interface. In fact, I can change my password but when I close
the session windows return me an error who said that windows cannot change
my profile blabla and talk about security rights. However, the password is
well changed :/

In the log i've got this :

  check_ntlm_password:  Authentication for user [troll] - [troll] FAILED
with error NT_STATUS_WRONG_PASSWORD

I googled this, but no solution for my problem :/

Has anyone use self-service-password  and anyone can help me ?

-- 
Cédric CARLEN
Élève-ingénieur à TELECOM Lille 1
Promotion FI15
☎ 06.59.42.81.55
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba 4 samba-tool dns question

2012-06-13 Thread Kai Blin
On 2012-06-13 11:21, Trever L. Adams wrote:

Hi Trever,

 /usr/local/samba/bin/samba-tool dns add s4server.example.org example.org
 example.org TXT v=spf1 mx -all
 
 I am not sure it is because it is showing up with dig as v=spf1 mx
 -all where it should be all one string, I believe.

Well, this is how the current code is set up, Amitay added this behavior
to the DLZ backend recently, and I followed along for the internal
server. I'm not aware of how the string tokenization is actually
described in the standard. I'm also not sure if the issue actually is
that samba-tool doesn't preserve the quotes around the string, so the
data is tokenized when added to the AD record.

Amitay, any insights on that one?

Cheers,
Kai

-- 
Kai Blin
Worldforge developer http://www.worldforge.org/
Wine developer http://wiki.winehq.org/KaiBlin
Samba team member http://www.samba.org/samba/team/


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Samba4 for AD using existing LDAP, Kerberos, and Bind Setup.

2012-06-13 Thread David Minard
G'day Dewayne,

Thank you your reply.  I'll submit the question on the technical list.

Sorry for the slowness of my reply.  I've been on leave.

Cheers,
David Minard.
Ph:0247 360 155
Fax:0247 360 770

School of Computing, Engineering, and Mathematics
Building Y - Penrith Campus (Kingswood)
Locked bag 1797
Penrith South DC
NSW 1797

[Sometimes waking up just isn't worth the insult of the day to come.]

On 19/05/2012, at 6:59 AM, Dewayne wrote:

 David, 
 
 I'd echo Gemes comment about posting your question to the
 samba-techni...@lists.samba.org list which would be more appropriate.
 There is some topical discussion going on there regarding content of a
 samba4 Beta release, and your question would be well timed.
 
 I'd suggest that you also consider the samba4 on existing: dhcp, dns  ntp
 infrastructure.  
 
 Good sources of information are at: 
 WhatsNew -
 http://gitweb.samba.org/?p=samba.git;a=blob;f=WHATSNEW.txt;h=8798a875cc7618
 da819e9ecd1db6cb7f25f85a94;hb=edb15ffef29fbb69a4d1dfc862fe8d6a3a027347
 
 Other useful references:
 1. https://wiki.samba.org/index.php/Samba4/HOWTO
 2. https://wiki.samba.org/index.php/Samba4/HOWTO/Join_a_domain_as_a_DC
 3. https://wiki.samba.org/index.php/Samba4/samba3upgrade/HOWTO 
 4. Last updated March 2011
 https://wiki.samba.org/index.php/Samba4_DRS_TODO_List#Support_RODC
 
 Kind regards, Dewayne.
 
 
 -- 
 This message has been scanned for viruses and
 dangerous content by MailScanner, and is
 believed to be clean.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Samba4 Multi-Master replication

2012-06-13 Thread Gémes Géza
On 2012-06-13 17:10, steve wrote:
 On 12/06/12 19:19, Gémes Géza wrote:
 On 2012-06-12 12:16, Morten Kramer wrote:
 Hi guys,


 I'm trying to get the Samba4 multi-master replication to work.


 With your setup DNS is the single point of failure, because with the
 (default) DLZ setup bind9 is able to serve DNS records only when samba4
 is running on that box.
 My recommendation would be to try to set up DNS on the second DC too.


 Hi
 Would both DC's and every client have both IP's in their resolv.conf
 (or whatever windoze calls it)?
 Cheers,
 Steve
Short answer: Yes
Longer answer: The easiest is to do that via dhcp

Regards

Geza
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] how to automount a kerberos cifs share

2012-06-13 Thread steve

Hi

I have an automount map:
* -fstype=cifs,sec=krb5 ://server/share/

It works fine, but only if Administrator has tickets. I can't do that on 
every client!


Is there any way I can store the Administrator key in a keytab and use 
that? Or any other solution?


Cheers,
Steve

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] how to automount a kerberos cifs share

2012-06-13 Thread Gaiseric Vandal
How about if you use NFS v4 with kerberos instead of CIFS?



On 06/13/12 14:58, steve wrote:
 Hi

 I have an automount map:
 * -fstype=cifs,sec=krb5 ://server/share/

 It works fine, but only if Administrator has tickets. I can't do that
 on every client!

 Is there any way I can store the Administrator key in a keytab and use
 that? Or any other solution?

 Cheers,
 Steve


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] how to automount a kerberos cifs share

2012-06-13 Thread steve

On 13/06/12 21:10, Gaiseric Vandal wrote:

How about if you use NFS v4 with kerberos instead of CIFS?



On 06/13/12 14:58, steve wrote:

Hi

I have an automount map:
* -fstype=cifs,sec=krb5 ://server/share/

It works fine, but only if Administrator has tickets. I can't do that
on every client!

Is there any way I can store the Administrator key in a keytab and use
that? Or any other solution?

Cheers,
Steve




Hi Gaiseric
Yes, that would be perfect as we are using kerberized nfs3 for 
everything else.


The problem with nfs4 is that you can't have group rw shares and also 
there is no document locking between libreoffice and m$office:-(


This particular share _has_ to be cifs.
Thanks,
Steve

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Samba 64 bit compilation

2012-06-13 Thread Gaiseric Vandal
Which platform?

If on Solaris 10 sparc, GCC (either from Sun or sunfreeware.com) should
be 64-bit by default.

GCC from Sunfreeware for Solaris 10 x86 will compile 32-bit by default.

For Solaris, you are better off using Sun Studio and Dmake.  Actually,
you are better off just using the compiled version from Oracle/Sun.




On 06/13/12 02:08, prabu.muru...@emc.com wrote:
 Hi,

 Have tried to compile samba 64 bit. By default it is compiling 32 bit. Google 
 doesn't give much info about it.
 Please share your experience on 64bit  samba.

 Thanks,
 Prabu


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] how to automount a kerberos cifs share

2012-06-13 Thread Gaiseric Vandal
On 06/13/12 17:08, steve wrote:
 On 13/06/12 21:10, Gaiseric Vandal wrote:
 How about if you use NFS v4 with kerberos instead of CIFS?



 On 06/13/12 14:58, steve wrote:
 Hi

 I have an automount map:
 * -fstype=cifs,sec=krb5 ://server/share/

 It works fine, but only if Administrator has tickets. I can't do that
 on every client!

 Is there any way I can store the Administrator key in a keytab and use
 that? Or any other solution?

 Cheers,
 Steve


 Hi Gaiseric
 Yes, that would be perfect as we are using kerberized nfs3 for
 everything else.

 The problem with nfs4 is that you can't have group rw shares and also
 there is no document locking between libreoffice and m$office:-(

 This particular share _has_ to be cifs.
 Thanks,
 Steve

What OS are you running?  My experience is that Solaris backported
kerberos to nfs v3 but that linux requires nfs v4 for kerberos.NFS
talks to GSS  which in turn talks to Kerberos.autofs runs as root so
with nfs  you would add creds to the local keytab for root  to make that
work.  No   I take it autofs on linux works with more than just NFS.





-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Two attempts required to join domain

2012-06-13 Thread Bill Arlofski
Hi Everyone.

I have run across an issue that is driving me crazy. This is a new deployment
of Samba v3.6.5 with openldap v2.4.30 and smbldap-tools v0.9.8


When trying to join the domain, on the first attempt the machine account is
properly created in the correct ou - e.g. ou=Computers,dc=domain,dc=local

But the failed to join domain pop-up with reason of The user name could not
be found is displayed (which really means the machine name was not found in
LDAP)  and of course the machine is not yet a domain member.

However, a 2nd attempt to join the domain with the same credentials,
immediately after the failure results in a Welcome to the X domain and the
machine is now a domain member.


Setting the openldap slapd loglevel to 416 to show the queries during this
process reveals the following:

On 1st join attempt Samba searches the whole directory from dc=domain,dc=local
with a scope of 2 (sub) for uid=MyMachine, objectClass=sambaSamAccount.

It of course does not find it, so the smbldap-useradd script is called and the
machine account is properly added to ou=Computers.

Then Samba immediately searches _ONLY_ ou=People,dc=domain,dc=local for the
newly created machine account and of course does not find it. And the failed
to join domain pop-up is displayed on the WinXP machine.

On the second join attempt, Samba _ONLY_ searches
ou=Computers,dc=domain,dc=local, which is where it SHOULD search for machines
as defined everywhere in my configs and it finds the machine and the machine
successfully joins the domain.

If I set all configs - samba, smbldap etc to be such that computers are in the
People organizational unit, then joining the domain works on the first try,
every time.

Also, if I un-join the domain, but leave the machine account in LDAP in
ou=Computers and then re-join the domain, this always works on first try too
since Samba's initial scope 2 sub search of the directory starting at the
top will find the machine account under ou=Computers.

Can someone offer guidance as to why during the new machine creation process
(joining a domain) Samba does not look for the machine in the defined machines
ou but always in the People ou?

Thank you in advance for any help on this!

--
Bill Arlofski
Reverse Polarity, LLC
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Gentoo Linux-installed Samba 4 alpha 21 getting python error

2012-06-13 Thread John Dekowski
I installed the Samba 4 alpha 21 using the Samba 4 alpha 20 ebuild as a
template:

When I try to run the samba-tool command, I keep getting

*Code:*File /usr/bin/samba-tool, line 38
except SystemExit, e:
 ^
SyntaxError: invalid syntax
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba4 Multi-Master replication

2012-06-13 Thread Morten Kramer

Hi guys,


I'm trying to get the Samba4 multi-master replication to work.



I set up the primary domain controller using this howto (under CentOS 
6.2 x64):



http://wiki.samba.org/index.php/Samba4/HOWTO



I installed bind 9.8.3 and enabled encrypted dns updates.




I set up another VM with the same CentOS version and oriented myself on 
this howto:


http://wiki.samba.org/index.php/Samba4/HOWTO/Join_a_domain_as_a_DC

to join the second DC into the domain. I edited /etc/resolv.conf and set 
the nameserver to the IP of the primary DC (bind dns server).



Basic replication seems to work (not doing the rsync for sysvol yet). 
However, when i take the primary DC offline, I can't use any of the .msc 
domain admin tools anymore.


I always get an error message, telling me that there is no RPC server 
available.
When i run gpmc.msc i can choose the DC i want to work on and I can see 
the secondary one, but it will come back with the RPC error.



User authentication does still work and Kerberos tickets are generated 
by the 2nd DC.




I can find this in the log:


../source4/dsdb/kcc/kcc_topology.c:1402: failed to find nCName attribute 
of object 
CN=ac7bf69c-9458-4205-acba-6fe172412d1b,CN=Partitions,CN=Configuration,DC=aeriatest2,DC=dc,DC=loc
../source4/dsdb/kcc/kcc_topology.c:3158: failed to color vertices: 
NT_STATUS_INTERNAL_DB_CORRUPTION
../source4/dsdb/kcc/kcc_topology.c:3415: failed to create connections: 
NT_STATUS_INTERNAL_DB_CORRUPTION

...
Warning: 60 extra bytes in incoming RPC request
../source4/rpc_server/drsuapi/dcesrv_drsuapi.c:74: doing DsBind with 
system_session




Do i need to configure something extra, so the secondary DC will be able 
to act as an RPC server?






Thanks,
freezer








--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Creating a package to upgrade several solaris servers

2012-06-13 Thread Kincho .
Hi,

I have to upgrade several servers with different versions of Solaris (9 
10) and different installations and I thought that creating a package will
help me save some time, but I am facing more issues this way than
installing manually.
I understand that I have to run ./configure first from source3 folder, then
make and last the makepkg.sh script, right? I have also changed the install
path to match the install directory. configure  make seem to work without
errors after a couple of fixes, but makepkg.sh is throwing me some errors
(I don't have them here but I can post them later)

My question is, did anyone create a successful package for Solaris with
latest version? can anyone give me a little bit of guidance, please?

Thanks in advance.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba4 Multi-Master replication

2012-06-13 Thread Morten Kramer

Hi guys,


I'm trying to get the Samba4 multi-master replication to work.



I set up the primary domain controller using this howto (under CentOS 
6.2 x64):



http://wiki.samba.org/index.php/Samba4/HOWTO



I installed bind 9.8.3 and enabled encrypted dns updates.




I set up another VM with the same CentOS version and oriented myself on 
this howto:


http://wiki.samba.org/index.php/Samba4/HOWTO/Join_a_domain_as_a_DC

to join the second DC into the domain. I edited /etc/resolv.conf and set 
the nameserver to the IP of the primary DC (bind dns server).



Basic replication seems to work (not doing the rsync for sysvol yet). 
However, when i take the primary DC offline (bind keeps running), I 
can't use any of the .msc domain admin tools anymore.


I always get an error message, telling me that there is no RPC server 
available.
When i run gpmc.msc i can choose the DC i want to work on and I can see 
the secondary one, but it will come back with the RPC error.


I had Wireshark running on one of the Windows7 clients. It seems like it 
tries to talk to the 2nd DC (DCERPC packets). But i'm not an expert in 
packet analysis, could somebody give me a hint what to look for here?




User authentication does still work and Kerberos tickets are generated 
by the 2nd DC.




I can find this in the log:


../source4/dsdb/kcc/kcc_topology.c:1402: failed to find nCName attribute 
of object 
CN=ac7bf69c-9458-4205-acba-6fe172412d1b,CN=Partitions,CN=Configuration,DC=aeriatest2,DC=dc,DC=loc
../source4/dsdb/kcc/kcc_topology.c:3158: failed to color vertices: 
NT_STATUS_INTERNAL_DB_CORRUPTION
../source4/dsdb/kcc/kcc_topology.c:3415: failed to create connections: 
NT_STATUS_INTERNAL_DB_CORRUPTION

...
Warning: 60 extra bytes in incoming RPC request
../source4/rpc_server/drsuapi/dcesrv_drsuapi.c:74: doing DsBind with 
system_session




Do i need to configure something extra, so the secondary DC will be able 
to act as an RPC server?






Thanks,
freezer








--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba4 (S4) [homes] special share availability

2012-06-13 Thread OBFC - Olivier BILHAUT
I mates,

We'd like to know if the beta version of S4 with the s3fs file system
now allow to use the [homes] special share. Are we now allowed to use
it in smb.conf, and does it works like in S3, as the home share of the
user is converted by it's username?

Sorry for my approximate english.
Cheers. Thank you for your work. Good luck.

O.Bilhaut
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Gentoo Linux-installed Samba 4 alpha 21 getting python error

2012-06-13 Thread Amitay Isaacs
On Fri, Jun 8, 2012 at 6:19 AM, John Dekowski jdekow...@gmail.com wrote:
 I installed the Samba 4 alpha 21 using the Samba 4 alpha 20 ebuild as a
 template:

 When I try to run the samba-tool command, I keep getting

 *Code:*File /usr/bin/samba-tool, line 38
    except SystemExit, e:
                     ^
 SyntaxError: invalid syntax

What version of python are you using? Samba requires python 2.x and not 3.x.

Amitay.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Samba 4 samba-tool dns question

2012-06-13 Thread Amitay Isaacs
On Wed, Jun 13, 2012 at 8:45 PM, Kai Blin k...@samba.org wrote:
 On 2012-06-13 11:21, Trever L. Adams wrote:

 Hi Trever,

 /usr/local/samba/bin/samba-tool dns add s4server.example.org example.org
 example.org TXT v=spf1 mx -all

 I am not sure it is because it is showing up with dig as v=spf1 mx
 -all where it should be all one string, I believe.

 Well, this is how the current code is set up, Amitay added this behavior
 to the DLZ backend recently, and I followed along for the internal
 server. I'm not aware of how the string tokenization is actually
 described in the standard. I'm also not sure if the issue actually is
 that samba-tool doesn't preserve the quotes around the string, so the
 data is tokenized when added to the AD record.

 Amitay, any insights on that one?

 Cheers,
 Kai

Hi Trever/Kai,

The tokenization changes were added to support multiple strings in txt
record. If you want to use a single string with spaces in it as a txt
record, you have to quote it twice.

  samba-tool dns add s4server example.org example.org TXT 'v=spf1 mx -all'

If you are using nsupdate to add the record, then make sure to quote
it as follows:

  server s4server
  update add example.org 3600 IN TXT v=spf1 mx -all
  send

That should get you the desired result.

Amitay.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] Samba 4 samba-tool dns question

2012-06-13 Thread Trever L. Adams
Thank you kindly!
-- 
Sent from my Android phone with K-9 Mail. Please excuse my brevity.

Amitay Isaacs ami...@samba.org wrote:

On Wed, Jun 13, 2012 at 8:45 PM, Kai Blin k...@samba.org wrote:
 On 2012-06-13 11:21, Trever L. Adams wrote:

 Hi Trever,

 /usr/local/samba/bin/samba-tool dns add s4server.example.org example.org
 example.org TXT v=spf1 mx -all

 I am not sure it is because it is showing up with dig as v=spf1 mx
 -all where it should be all one string, I believe.

 Well, this is how the current code is set up, Amitay added this behavior
 to the DLZ backend recently, and I followed along for the internal
 server. I'm not aware of how the string tokenization is actually
 described in the standard. I'm also not sure if the issue actually is
 that samba-tool doesn't preserve the quotes around the string, so the
 data is tokenized when added to the AD record.

 Amitay, any insights on that one?

 Cheers,
 Kai

Hi Trever/Kai,

The tokenization changes were added to support multiple strings in txt
record. If you want to use a single string with spaces in it as a txt
record, you have to quote it twice.

samba-tool dns add s4server example.org example.org TXT 'v=spf1 mx -all'

If you are using nsupdate to add the record, then make sure to quote
it as follows:

server s4server
update add example.org 3600 IN TXT v=spf1 mx -all
send

That should get you the desired result.

Amitay.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[SCM] Samba Shared Repository - branch master updated

2012-06-13 Thread Stefan Metzmacher
The branch, master has been updated
   via  96ada4d s4:ntvfs: add '_fn' suffix to all ntvfs_ops function 
pointers
  from  855d23b s3: Use talloc_tos() in more places in dbwrap_util

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master


- Log -
commit 96ada4d87b6d8618919c614765f2ec1d706f4f4a
Author: Stefan Metzmacher me...@samba.org
Date:   Wed Jun 13 09:11:42 2012 +0200

s4:ntvfs: add '_fn' suffix to all ntvfs_ops function pointers

This hopefully fixes the build on systems where _LARGE_FILES
triggers defines of syscalls e.g. '#define lseek lseek64'
on AIX.

metze

Autobuild-User(master): Stefan Metzmacher me...@samba.org
Autobuild-Date(master): Wed Jun 13 11:03:15 CEST 2012 on sn-devel-104

---

Summary of changes:
 source4/ntvfs/cifs/vfs_cifs.c |   64 +++---
 source4/ntvfs/cifs_posix_cli/vfs_cifs_posix.c |   60 +++---
 source4/ntvfs/ipc/vfs_ipc.c   |   60 +++---
 source4/ntvfs/nbench/vfs_nbench.c |   62 +++---
 source4/ntvfs/ntvfs.h |   68 
 source4/ntvfs/ntvfs_generic.c |   54 +++---
 source4/ntvfs/ntvfs_interface.c   |  256 
 source4/ntvfs/posix/vfs_posix.c   |   62 +++---
 source4/ntvfs/print/vfs_print.c   |8 +-
 source4/ntvfs/simple/vfs_simple.c |   60 +++---
 source4/ntvfs/smb2/vfs_smb2.c |   62 +++---
 source4/ntvfs/unixuid/vfs_unixuid.c   |   62 +++---
 12 files changed, 439 insertions(+), 439 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source4/ntvfs/cifs/vfs_cifs.c b/source4/ntvfs/cifs/vfs_cifs.c
index fb7a485..552f664 100644
--- a/source4/ntvfs/cifs/vfs_cifs.c
+++ b/source4/ntvfs/cifs/vfs_cifs.c
@@ -1191,38 +1191,38 @@ NTSTATUS ntvfs_cifs_init(void)
ops.type = NTVFS_DISK;

/* fill in all the operations */
-   ops.connect = cvfs_connect;
-   ops.disconnect = cvfs_disconnect;
-   ops.unlink = cvfs_unlink;
-   ops.chkpath = cvfs_chkpath;
-   ops.qpathinfo = cvfs_qpathinfo;
-   ops.setpathinfo = cvfs_setpathinfo;
-   ops.open = cvfs_open;
-   ops.mkdir = cvfs_mkdir;
-   ops.rmdir = cvfs_rmdir;
-   ops.rename = cvfs_rename;
-   ops.copy = cvfs_copy;
-   ops.ioctl = cvfs_ioctl;
-   ops.read = cvfs_read;
-   ops.write = cvfs_write;
-   ops.seek = cvfs_seek;
-   ops.flush = cvfs_flush; 
-   ops.close = cvfs_close;
-   ops.exit = cvfs_exit;
-   ops.lock = cvfs_lock;
-   ops.setfileinfo = cvfs_setfileinfo;
-   ops.qfileinfo = cvfs_qfileinfo;
-   ops.fsinfo = cvfs_fsinfo;
-   ops.lpq = cvfs_lpq;
-   ops.search_first = cvfs_search_first;
-   ops.search_next = cvfs_search_next;
-   ops.search_close = cvfs_search_close;
-   ops.trans = cvfs_trans;
-   ops.logoff = cvfs_logoff;
-   ops.async_setup = cvfs_async_setup;
-   ops.cancel = cvfs_cancel;
-   ops.notify = cvfs_notify;
-   ops.trans2 = cvfs_trans2;
+   ops.connect_fn = cvfs_connect;
+   ops.disconnect_fn = cvfs_disconnect;
+   ops.unlink_fn = cvfs_unlink;
+   ops.chkpath_fn = cvfs_chkpath;
+   ops.qpathinfo_fn = cvfs_qpathinfo;
+   ops.setpathinfo_fn = cvfs_setpathinfo;
+   ops.open_fn = cvfs_open;
+   ops.mkdir_fn = cvfs_mkdir;
+   ops.rmdir_fn = cvfs_rmdir;
+   ops.rename_fn = cvfs_rename;
+   ops.copy_fn = cvfs_copy;
+   ops.ioctl_fn = cvfs_ioctl;
+   ops.read_fn = cvfs_read;
+   ops.write_fn = cvfs_write;
+   ops.seek_fn = cvfs_seek;
+   ops.flush_fn = cvfs_flush;
+   ops.close_fn = cvfs_close;
+   ops.exit_fn = cvfs_exit;
+   ops.lock_fn = cvfs_lock;
+   ops.setfileinfo_fn = cvfs_setfileinfo;
+   ops.qfileinfo_fn = cvfs_qfileinfo;
+   ops.fsinfo_fn = cvfs_fsinfo;
+   ops.lpq_fn = cvfs_lpq;
+   ops.search_first_fn = cvfs_search_first;
+   ops.search_next_fn = cvfs_search_next;
+   ops.search_close_fn = cvfs_search_close;
+   ops.trans_fn = cvfs_trans;
+   ops.logoff_fn = cvfs_logoff;
+   ops.async_setup_fn = cvfs_async_setup;
+   ops.cancel_fn = cvfs_cancel;
+   ops.notify_fn = cvfs_notify;
+   ops.trans2_fn = cvfs_trans2;
 
/* register ourselves with the NTVFS subsystem. We register
   under the name 'cifs'. */
diff --git a/source4/ntvfs/cifs_posix_cli/vfs_cifs_posix.c 
b/source4/ntvfs/cifs_posix_cli/vfs_cifs_posix.c
index 949b6db..8c5a53b 100644
--- a/source4/ntvfs/cifs_posix_cli/vfs_cifs_posix.c
+++ b/source4/ntvfs/cifs_posix_cli/vfs_cifs_posix.c
@@ -1061,36 +1061,36 @@ NTSTATUS ntvfs_cifs_posix_init(void)
ZERO_STRUCT(ops);
 
/* fill in all the operations */
-   ops.connect = cifspsx_connect;
-   ops.disconnect = cifspsx_disconnect;
-   

[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  ec7a5f2 Forward port of Richard Sharpe's 
realrichardsha...@gmail.com fix for bug #8970 - Possible memory leaks in the 
samba master process.
  from  7ca2654 s3: fix compile of krb5 locator on Solaris

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit ec7a5f2593b67797fc3924611d3b0b05b807d0bf
Author: Jeremy Allison j...@samba.org
Date:   Thu May 31 16:25:52 2012 -0700

Forward port of Richard Sharpe's realrichardsha...@gmail.com fix for bug 
#8970 - Possible memory leaks in the samba master process.

---

Summary of changes:
 source3/include/proto.h  |6 +++---
 source3/nmbd/nmbd.c  |3 ++-
 source3/param/loadparm.c |   12 
 source3/printing/load.c  |7 ++-
 source3/smbd/server_reload.c |1 +
 source3/winbindd/winbindd.c  |3 ++-
 6 files changed, 22 insertions(+), 10 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/include/proto.h b/source3/include/proto.h
index 28b58b2..e22fc9c 100644
--- a/source3/include/proto.h
+++ b/source3/include/proto.h
@@ -1338,9 +1338,9 @@ NTSTATUS change_trust_account_password( const char 
*domain, const char *remote_m
 /* The following definitions come from param/loadparm.c  */
 
 char *lp_smb_ports(void);
-char *lp_dos_charset(void);
-char *lp_unix_charset(void);
-char *lp_display_charset(void);
+const char *lp_dos_charset(void);
+const char *lp_unix_charset(void);
+const char *lp_display_charset(void);
 char *lp_logfile(void);
 char *lp_configfile(void);
 char *lp_smb_passwd_file(void);
diff --git a/source3/nmbd/nmbd.c b/source3/nmbd/nmbd.c
index 986c575..0172e08 100644
--- a/source3/nmbd/nmbd.c
+++ b/source3/nmbd/nmbd.c
@@ -365,11 +365,12 @@ static bool reload_nmbd_services(bool test)
set_remote_machine_name(nmbd, False);
 
if ( lp_loaded() ) {
-   const char *fname = lp_configfile();
+   char *fname = lp_configfile();
if (file_exist(fname)  
!strcsequal(fname,get_dyn_CONFIGFILE())) {
set_dyn_CONFIGFILE(fname);
test = False;
}
+   TALLOC_FREE(fname);
}
 
if ( test  !lp_file_list_changed() )
diff --git a/source3/param/loadparm.c b/source3/param/loadparm.c
index 6ad2452..dd63339 100644
--- a/source3/param/loadparm.c
+++ b/source3/param/loadparm.c
@@ -5575,9 +5575,9 @@ static char *lp_string(const char *s)
  char fn_name(const struct share_params *p) {return(LP_SNUM_OK(p-service)? 
ServicePtrs[(p-service)]-val : sDefault.val);}
 
 FN_GLOBAL_STRING(lp_smb_ports, Globals.smb_ports)
-FN_GLOBAL_STRING(lp_dos_charset, Globals.dos_charset)
-FN_GLOBAL_STRING(lp_unix_charset, Globals.unix_charset)
-FN_GLOBAL_STRING(lp_display_charset, Globals.display_charset)
+FN_GLOBAL_CONST_STRING(lp_dos_charset, Globals.dos_charset)
+FN_GLOBAL_CONST_STRING(lp_unix_charset, Globals.unix_charset)
+FN_GLOBAL_CONST_STRING(lp_display_charset, Globals.display_charset)
 FN_GLOBAL_STRING(lp_logfile, Globals.szLogFile)
 FN_GLOBAL_STRING(lp_configfile, Globals.szConfigFile)
 FN_GLOBAL_STRING(lp_smb_passwd_file, Globals.szSMBPasswdFile)
@@ -9647,7 +9647,11 @@ static bool lp_load_ex(const char *pszFname,
}
}
 
-   lp_add_auto_services(lp_auto_services());
+   {
+   char *serv = lp_auto_services();
+   lp_add_auto_services(serv);
+   TALLOC_FREE(serv);
+   }
 
if (add_ipc) {
/* When 'restrict anonymous = 2' guest connections to ipc$
diff --git a/source3/printing/load.c b/source3/printing/load.c
index 5acc258..829c3e3 100644
--- a/source3/printing/load.c
+++ b/source3/printing/load.c
@@ -30,6 +30,7 @@ static void add_auto_printers(void)
int pnum = lp_servicenumber(PRINTERS_NAME);
char *str;
char *saveptr;
+   char *auto_serv = NULL;
 
if (pnum  0)
if (process_registry_service(PRINTERS_NAME))
@@ -38,8 +39,12 @@ static void add_auto_printers(void)
if (pnum  0)
return;
 
-   if ((str = SMB_STRDUP(lp_auto_services())) == NULL)
+   auto_serv = lp_auto_services();
+   str = SMB_STRDUP(auto_serv);
+   TALLOC_FREE(auto_serv);
+   if (str == NULL) {
return;
+   }
 
for (p = strtok_r(str, LIST_SEP, saveptr); p;
 p = strtok_r(NULL, LIST_SEP, saveptr)) {
diff --git a/source3/smbd/server_reload.c b/source3/smbd/server_reload.c
index 6e0ab39..fd6dc1a 100644
--- a/source3/smbd/server_reload.c
+++ b/source3/smbd/server_reload.c
@@ -127,6 +127,7 @@ bool reload_services(struct messaging_context *msg_ctx, int 
smb_sock,
set_dyn_CONFIGFILE(fname);
test = False;
}
+   TALLOC_FREE(fname);
}
 
 

[SCM] Samba Shared Repository - branch v3-5-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-5-test has been updated
   via  d853bc0 Fix bug #8970 - Possible memory leaks in the samba master 
process.
  from  0529cf9 Fix bug #8882 - Broken processing of %U with vfs_full_audit 
when force user is set.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-5-test


- Log -
commit d853bc02c32a4c1172bf5f3f64c75db5ecc5ccca
Author: Richard Sharpe realrichardsha...@gmail.com
Date:   Thu May 31 15:43:14 2012 -0700

Fix bug #8970 - Possible memory leaks in the samba master process.

Signed-off-by: Jeremy Allison j...@samba.org

---

Summary of changes:
 source3/include/proto.h |6 +++---
 source3/lib/debug.c |   13 -
 source3/nmbd/nmbd.c |3 ++-
 source3/param/loadparm.c|   12 
 source3/smbd/server.c   |1 +
 source3/winbindd/winbindd.c |3 ++-
 6 files changed, 24 insertions(+), 14 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/include/proto.h b/source3/include/proto.h
index 579fc1b..559a34e 100644
--- a/source3/include/proto.h
+++ b/source3/include/proto.h
@@ -3918,9 +3918,9 @@ void expire_workgroups_and_servers(time_t t);
 /* The following definitions come from param/loadparm.c  */
 
 char *lp_smb_ports(void);
-char *lp_dos_charset(void);
-char *lp_unix_charset(void);
-char *lp_display_charset(void);
+const char *lp_dos_charset(void);
+const char *lp_unix_charset(void);
+const char *lp_display_charset(void);
 char *lp_logfile(void);
 char *lp_configfile(void);
 char *lp_smb_passwd_file(void);
diff --git a/source3/lib/debug.c b/source3/lib/debug.c
index 80b8310..05e9eee 100644
--- a/source3/lib/debug.c
+++ b/source3/lib/debug.c
@@ -657,9 +657,11 @@ bool reopen_logs( void )
SAFE_FREE(fname);
fname = SMB_STRDUP(logfname);
if (!fname) {
+   TALLOC_FREE(logfname);
return false;
}
}
+   TALLOC_FREE(logfname);
}
 
debugf = fname;
@@ -1028,6 +1030,8 @@ bool dbghdrclass(int level, int cls, const char 
*location, const char *func)
 */
if( lp_timestamp_logs() || lp_debug_prefix_timestamp() || 
!(lp_loaded()) ) {
char header_str[200];
+   char *curtime = current_timestring(talloc_tos(),
+   lp_debug_hires_timestamp());
 
header_str[0] = '\0';
 
@@ -1050,19 +1054,18 @@ bool dbghdrclass(int level, int cls, const char 
*location, const char *func)
 , class=%s,
 default_classname_table[cls]);
}
-  
+
/* Print it all out at once to prevent split syslog output. */
if( lp_debug_prefix_timestamp() ) {
(void)Debug1( [%s, %2d%s] ,
-   current_timestring(talloc_tos(),
-  lp_debug_hires_timestamp()),
+   curtime,
level, header_str);
} else {
(void)Debug1( [%s, %2d%s] %s(%s)\n,
-   current_timestring(talloc_tos(),
-  lp_debug_hires_timestamp()),
+   curtime,
level, header_str, location, func );
}
+   TALLOC_FREE(curtime);
}
 
errno = old_errno;
diff --git a/source3/nmbd/nmbd.c b/source3/nmbd/nmbd.c
index 48e6d93..2a7b28d 100644
--- a/source3/nmbd/nmbd.c
+++ b/source3/nmbd/nmbd.c
@@ -366,11 +366,12 @@ static bool reload_nmbd_services(bool test)
set_remote_machine_name(nmbd, False);
 
if ( lp_loaded() ) {
-   const char *fname = lp_configfile();
+   char *fname = lp_configfile();
if (file_exist(fname)  
!strcsequal(fname,get_dyn_CONFIGFILE())) {
set_dyn_CONFIGFILE(fname);
test = False;
}
+   TALLOC_FREE(fname);
}
 
if ( test  !lp_file_list_changed() )
diff --git a/source3/param/loadparm.c b/source3/param/loadparm.c
index 28ffc08..8c1cf09 100644
--- a/source3/param/loadparm.c
+++ b/source3/param/loadparm.c
@@ -5318,9 +5318,9 @@ static char *lp_string(const char *s)
  char fn_name(const struct share_params *p) {return(LP_SNUM_OK(p-service)? 
ServicePtrs[(p-service)]-val : sDefault.val);}
 
 FN_GLOBAL_STRING(lp_smb_ports, Globals.smb_ports)
-FN_GLOBAL_STRING(lp_dos_charset, Globals.dos_charset)
-FN_GLOBAL_STRING(lp_unix_charset, Globals.unix_charset)
-FN_GLOBAL_STRING(lp_display_charset, Globals.display_charset)
+FN_GLOBAL_CONST_STRING(lp_dos_charset, Globals.dos_charset)
+FN_GLOBAL_CONST_STRING(lp_unix_charset, Globals.unix_charset)

[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  445f314 s3: Fix uninitialized memory read in talloc_free()
  from  ec7a5f2 Forward port of Richard Sharpe's 
realrichardsha...@gmail.com fix for bug #8970 - Possible memory leaks in the 
samba master process.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 445f314614e4e514a70ff5f1fbbfedb4d3ab0aac
Author: Volker Lendecke v...@samba.org
Date:   Tue May 29 15:31:49 2012 +0200

s3: Fix uninitialized memory read in talloc_free()

Thanks to laurent gaffie laurent.gaf...@gmail.com for reporting
this issue!

---

Summary of changes:
 source3/libsmb/clispnego.c |3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/libsmb/clispnego.c b/source3/libsmb/clispnego.c
index 4581ce4..0a907ba 100644
--- a/source3/libsmb/clispnego.c
+++ b/source3/libsmb/clispnego.c
@@ -127,6 +127,9 @@ bool spnego_parse_negTokenInit(TALLOC_CTX *ctx,
asn1_start_tag(data,ASN1_SEQUENCE(0));
for (i=0; asn1_tag_remaining(data)  0  i  ASN1_MAX_OIDS-1; i++) {
asn1_read_OID(data,ctx, OIDs[i]);
+   if (data-has_error) {
+   break;
+   }
}
OIDs[i] = NULL;
asn1_end_tag(data);


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-5-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-5-test has been updated
   via  bc4a2c1 s3: Fix uninitialized memory read in talloc_free()
  from  d853bc0 Fix bug #8970 - Possible memory leaks in the samba master 
process.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-5-test


- Log -
commit bc4a2c143b531f9362acb8f3d6e099cbac070840
Author: Volker Lendecke v...@samba.org
Date:   Mon Jun 4 12:22:21 2012 -0700

s3: Fix uninitialized memory read in talloc_free()

Thanks to laurent gaffie laurent.gaf...@gmail.com for reporting
this issue!

---

Summary of changes:
 source3/libsmb/clispnego.c |3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/libsmb/clispnego.c b/source3/libsmb/clispnego.c
index 36d21d0..3322529 100644
--- a/source3/libsmb/clispnego.c
+++ b/source3/libsmb/clispnego.c
@@ -161,6 +161,9 @@ bool spnego_parse_negTokenInit(DATA_BLOB blob,
for (i=0; asn1_tag_remaining(data)  0  i  ASN1_MAX_OIDS-1; i++) {
const char *oid_str = NULL;
asn1_read_OID(data,talloc_autofree_context(),oid_str);
+   if (data-has_error) {
+   break;
+   }
OIDs[i] = CONST_DISCARD(char *, oid_str);
}
OIDs[i] = NULL;


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  8b3b1aa We are triggering the cleanup_timeout_fn() too often, on 
exiting when an smbd is idle.
   via  777ac04 Stop spamming the logs with Could not remove pid XX from 
serverid.tdb messages and initiating the cleanup function on every process deat
  from  445f314 s3: Fix uninitialized memory read in talloc_free()

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 8b3b1aa4ff76d7f03285e3fa87f30f3068a7ea5d
Author: Jeremy Allison j...@samba.org
Date:   Fri Jun 1 12:28:33 2012 -0700

We are triggering the cleanup_timeout_fn() too often, on exiting when an 
smbd is idle.

Calls to exit_server_cleanly() should be treated as a clean shutdown,
and not trigger the master smbd to call cleanup_timeout_fn.

The last 2 patches address bug #8971 (cleanup_timeout_fn() is called too 
often,
on exiting when an smbd is idle.)

commit 777ac04a99467594805a03635b04011c495ff7e7
Author: Jeremy Allison j...@samba.org
Date:   Fri Jun 1 12:02:04 2012 -0700

Stop spamming the logs with Could not remove pid XX from serverid.tdb 
messages and initiating the cleanup function on every process deat

We now have many sub-processes from smbd that don't serve SMB1/SMB2 
requests and
don't register themselves in the serverid.tdb. Only initiate the cleanup
from processes that were explicitly in the child list.

---

Summary of changes:
 source3/smbd/server.c  |   35 +++
 source3/smbd/server_exit.c |   15 ---
 2 files changed, 23 insertions(+), 27 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/smbd/server.c b/source3/smbd/server.c
index 8cda180..9a8cdc0 100644
--- a/source3/smbd/server.c
+++ b/source3/smbd/server.c
@@ -283,6 +283,25 @@ static void remove_child_pid(pid_t pid, bool 
unclean_shutdown)
static struct timed_event *cleanup_te;
struct server_id child_id;
 
+   child_id = procid_self(); /* Just initialize pid and potentially vnn */
+   child_id.pid = pid;
+
+   for (child = children; child != NULL; child = child-next) {
+   if (child-pid == pid) {
+   struct child_pid *tmp = child;
+   DLIST_REMOVE(children, child);
+   SAFE_FREE(tmp);
+   num_children -= 1;
+   break;
+   }
+   }
+
+   if (child == NULL) {
+   /* not all forked child processes are added to the children 
list */
+   DEBUG(2, (Could not find child %d -- ignoring\n, (int)pid));
+   return;
+   }
+
if (unclean_shutdown) {
/* a child terminated uncleanly so tickle all
   processes to see if they can grab any of the
@@ -301,26 +320,10 @@ static void remove_child_pid(pid_t pid, bool 
unclean_shutdown)
}
}
 
-   child_id = procid_self(); /* Just initialize pid and potentially vnn */
-   child_id.pid = pid;
-
if (!serverid_deregister(child_id)) {
DEBUG(1, (Could not remove pid %d from serverid.tdb\n,
  (int)pid));
}
-
-   for (child = children; child != NULL; child = child-next) {
-   if (child-pid == pid) {
-   struct child_pid *tmp = child;
-   DLIST_REMOVE(children, child);
-   SAFE_FREE(tmp);
-   num_children -= 1;
-   return;
-   }
-   }
-
-   /* not all forked child processes are added to the children list */
-   DEBUG(1, (Could not find child %d -- ignoring\n, (int)pid));
 }
 
 /
diff --git a/source3/smbd/server_exit.c b/source3/smbd/server_exit.c
index 4c71d8f..fc77dee 100644
--- a/source3/smbd/server_exit.c
+++ b/source3/smbd/server_exit.c
@@ -83,7 +83,6 @@ static void exit_server_common(enum server_exit_reason how,
 static void exit_server_common(enum server_exit_reason how,
const char *const reason)
 {
-   bool had_open_conn = false;
struct smbd_server_connection *sconn = smbd_server_conn;
 
if (!exit_firsttime)
@@ -101,7 +100,7 @@ static void exit_server_common(enum server_exit_reason how,
bool found = false;
files_forall(sconn, log_writeable_file_fn, found);
}
-   had_open_conn = conn_close_all(sconn);
+   (void)conn_close_all(sconn);
invalidate_all_vuids(sconn);
}
 
@@ -175,6 +174,8 @@ static void exit_server_common(enum server_exit_reason how,
 
dump_core();
 
+   /* Notreached. */
+   exit(1);
} else {

[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  185c205 s3-librpc-crypto: avoid crash with MIT krb5 1.10.0 in 
gss_get_name_attribute()
  from  8b3b1aa We are triggering the cleanup_timeout_fn() too often, on 
exiting when an smbd is idle.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 185c2054fd22de0ab07a762a279a7ef0db5d802c
Author: Alexander Bokovoy a...@samba.org
Date:   Thu Jun 7 18:24:38 2012 +0300

s3-librpc-crypto: avoid crash with MIT krb5 1.10.0 in 
gss_get_name_attribute()

gss_get_name_attribute() can return unintialized pac_display_buffer
and later gss_release_buffer() will crash on attempting to release it.

The fix on MIT krb5 side is in 1.10.1, reported in both Debian and MIT 
upstream:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658514
http://krbdev.mit.edu/rt/Ticket/Display.html?user=guestpass=guestid=7087

We need to initialize variables before using gss_get_name_attribute()

Fix bug #8988 (avoid crash with MIT krb5 1.10.0 in 
gss_get_name_attribute()).

---

Summary of changes:
 source3/librpc/crypto/gse.c |   20 ++--
 1 files changed, 18 insertions(+), 2 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/librpc/crypto/gse.c b/source3/librpc/crypto/gse.c
index 335dc1c..02fb0f6 100644
--- a/source3/librpc/crypto/gse.c
+++ b/source3/librpc/crypto/gse.c
@@ -688,8 +688,24 @@ NTSTATUS gse_get_pac_blob(struct gse_context *gse_ctx,
  TALLOC_CTX *mem_ctx, DATA_BLOB *pac_blob)
 {
OM_uint32 gss_min, gss_maj;
-   gss_buffer_desc pac_buffer;
-   gss_buffer_desc pac_display_buffer;
+/*
+ * gss_get_name_attribute() in MIT krb5 1.10.0 can return unintialized 
pac_display_buffer
+ * and later gss_release_buffer() will crash on attempting to release it.
+ *
+ * So always initialize the buffer descriptors.
+ *
+ * See following links for more details:
+ * http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658514
+ * http://krbdev.mit.edu/rt/Ticket/Display.html?user=guestpass=guestid=7087
+ */
+   gss_buffer_desc pac_buffer = {
+   .value = NULL,
+   .length = 0
+   };
+   gss_buffer_desc pac_display_buffer = {
+   .value = NULL,
+   .length = 0
+   };
gss_buffer_desc pac_name = {
.value = discard_const_p(char, urn:mspac:),
.length = sizeof(urn:mspac:) - 1


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  87d75c6 Fix bug #8311 - Winzip occasionally can not read files out 
of an open winzip dialog.
  from  185c205 s3-librpc-crypto: avoid crash with MIT krb5 1.10.0 in 
gss_get_name_attribute()

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 87d75c675ccd1cd455afb67e149ce1e835ad6b21
Author: Stefan Metzmacher me...@samba.org
Date:   Mon Jun 4 15:59:35 2012 -0700

Fix bug #8311 - Winzip occasionally can not read files out of an open 
winzip dialog.

Backport of the changes in master without the VFS change. Move
all processing into smb_fsctl().

---

Summary of changes:
 source3/smbd/nttrans.c|  392 -
 source3/smbd/proto.h  |9 +
 source3/smbd/smb2_ioctl.c |  138 +++--
 3 files changed, 242 insertions(+), 297 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/smbd/nttrans.c b/source3/smbd/nttrans.c
index de508eb..429250e 100644
--- a/source3/smbd/nttrans.c
+++ b/source3/smbd/nttrans.c
@@ -2112,44 +2112,29 @@ static void 
call_nt_transact_set_security_desc(connection_struct *conn,
return;
 }
 
-/
- Reply to NT IOCTL
-/
-
-static void call_nt_transact_ioctl(connection_struct *conn,
-  struct smb_request *req,
-  uint16 **ppsetup, uint32 setup_count,
-  char **ppparams, uint32 parameter_count,
-  char **ppdata, uint32 data_count,
-  uint32 max_data_count)
-{
-   uint32 function;
-   uint16 fidnum;
-   files_struct *fsp;
-   uint8 isFSctl;
-   uint8 compfilter;
-   char *pdata = *ppdata;
-
-   if (setup_count != 8) {
-   DEBUG(3,(call_nt_transact_ioctl: invalid setup count %d\n, 
setup_count));
-   reply_nterror(req, NT_STATUS_NOT_SUPPORTED);
-   return;
-   }
-
-   function = IVAL(*ppsetup, 0);
-   fidnum = SVAL(*ppsetup, 4);
-   isFSctl = CVAL(*ppsetup, 6);
-   compfilter = CVAL(*ppsetup, 7);
+/*
+ * Implement the default fsctl operation.
+ */
 
-   DEBUG(10,(call_nt_transact_ioctl: function[0x%08X] FID[0x%04X] 
isFSctl[0x%02X] compfilter[0x%02X]\n, 
-function, fidnum, isFSctl, compfilter));
+static bool vfswrap_logged_ioctl_message = false;
 
-   fsp=file_fsp(req, fidnum);
-   /* this check is done in each implemented function case for now
-  because I don't want to break anything... --metze
-   FSP_BELONGS_CONN(fsp,conn);*/
-
-   SMB_PERFCOUNT_SET_IOCTL(req-pcd, function);
+/*
+ * In 3.6 we do not have a SMB_VFS_FSCTL() function
+ * it's just faked to make it more look like
+ * master (4.0)
+ */
+NTSTATUS smb_fsctl(struct files_struct *fsp,
+  TALLOC_CTX *ctx,
+  uint32_t function,
+  uint16_t req_flags,  /* Needed for UNICODE ... */
+  const uint8_t *_in_data,
+  uint32_t in_len,
+  uint8_t **_out_data,
+  uint32_t max_out_len,
+  uint32_t *out_len)
+{
+   const char *in_data = (const char *)_in_data;
+   char **out_data = (char **)_out_data;
 
switch (function) {
case FSCTL_SET_SPARSE:
@@ -2157,79 +2142,61 @@ static void call_nt_transact_ioctl(connection_struct 
*conn,
bool set_sparse = true;
NTSTATUS status;
 
-   if (data_count = 1  pdata[0] == 0) {
+   if (in_len = 1  in_data[0] == 0) {
set_sparse = false;
}
 
-   DEBUG(10,(FSCTL_SET_SPARSE: called on FID[0x%04X]set[%u]\n,
-fidnum, set_sparse));
+   status = file_set_sparse(fsp-conn, fsp, set_sparse);
 
-   if (!check_fsp_open(conn, req, fsp)) {
-   return;
-   }
-
-   status = file_set_sparse(conn, fsp, set_sparse);
-   if (!NT_STATUS_IS_OK(status)) {
-   DEBUG(9,(FSCTL_SET_SPARSE: fname[%s] set[%u] - %s\n,
-smb_fname_str_dbg(fsp-fsp_name), set_sparse, 
nt_errstr(status)));
-   reply_nterror(req, status);
-   return;
-   }
+   DEBUG(NT_STATUS_IS_OK(status) ? 10 : 9,
+ (FSCTL_SET_SPARSE: fname[%s] set[%u] - %s\n,
+  smb_fname_str_dbg(fsp-fsp_name), set_sparse,
+  nt_errstr(status)));
 
-   DEBUG(10,(FSCTL_SET_SPARSE: fname[%s] 

[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  4cf3fb8 s3-winbindd: call dump_core_setup after command line option 
has been parsed
  from  87d75c6 Fix bug #8311 - Winzip occasionally can not read files out 
of an open winzip dialog.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 4cf3fb815610c6f0939f8b142296cd836faac7e6
Author: Matthieu Patou m...@matws.net
Date:   Fri Jun 1 15:33:04 2012 -0700

s3-winbindd: call dump_core_setup after command line option has been parsed

Without this fix in some situations winbindd can't coredump.
Such cases append when samba is compiled in a custom prefix (ie.
/home/build/mat/prod/1/) in this case get_dyn_LOGFILEBASE or 
basename(lp_logfile)
before the configuration file and the command line is parsed will be 
something like /home/build/mat/prod/1/var
which might not exists on the host where you run it (where it's most
probably more normal directories).
Specifying --log-basename didn't help as dump_core_setup is called before 
the command line and
the config file is read so it didn't help getting a correct value in 
dump_core_setup.
We fix this issue by calling dump_core_setup() also after the command
line has been read and also after the configfile has been parsed so that
the final location for the coredump is coherent with the final logile
location.

Fix bug #8975 (winbindd can't coredump).

---

Summary of changes:
 source3/winbindd/winbindd.c |   14 ++
 1 files changed, 14 insertions(+), 0 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/winbindd/winbindd.c b/source3/winbindd/winbindd.c
index f80949d..15aafeb 100644
--- a/source3/winbindd/winbindd.c
+++ b/source3/winbindd/winbindd.c
@@ -1303,6 +1303,15 @@ int main(int argc, char **argv, char **envp)
}
}
 
+   /* We call dump_core_setup one more time because the command line can
+* set the log file or the log-basename and this will influence where
+* cores are stored. Without this call get_dyn_LOGFILEBASE will be
+* the default value derived from build's prefix. For EOM this value
+* is often not related to the path where winbindd is actually run
+* in production.
+*/
+   dump_core_setup(winbindd);
+
if (is_daemon  interactive) {
d_fprintf(stderr,\nERROR: 
  Option -i|--interactive is not allowed together with 
-D|--daemon\n\n);
@@ -1341,6 +1350,11 @@ int main(int argc, char **argv, char **envp)
DEBUG(0, (error opening config file\n));
exit(1);
}
+   /* After parsing the configuration file we setup the core path one more 
time
+* as the log file might have been set in the configuration and cores's
+* path is by default basename(lp_logfile()).
+*/
+   dump_core_setup(winbindd);
 
/* Initialise messaging system */
 


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-5-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-5-test has been updated
   via  529 s3-winbindd: call dump_core_setup after command line option 
has been parsed
  from  bc4a2c1 s3: Fix uninitialized memory read in talloc_free()

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-5-test


- Log -
commit 52922fb373ad23a0ce9034bf9630cdb17765
Author: Matthieu Patou m...@matws.net
Date:   Fri Jun 1 15:33:04 2012 -0700

s3-winbindd: call dump_core_setup after command line option has been parsed

Without this fix in some situations winbindd can't coredump.
Such cases append when samba is compiled in a custom prefix (ie.
/home/build/mat/prod/1/) in this case get_dyn_LOGFILEBASE or 
basename(lp_logfile)
before the configuration file and the command line is parsed will be 
something like /home/build/mat/prod/1/var
which might not exists on the host where you run it (where it's most
probably more normal directories).
Specifying --log-basename didn't help as dump_core_setup is called before 
the command line and
the config file is read so it didn't help getting a correct value in 
dump_core_setup.
We fix this issue by calling dump_core_setup() also after the command
line has been read and also after the configfile has been parsed so that
the final location for the coredump is coherent with the final logile
location.

Fix bug #8975 (winbindd can't coredump).
(cherry picked from commit 4cf3fb815610c6f0939f8b142296cd836faac7e6)

---

Summary of changes:
 source3/winbindd/winbindd.c |   14 ++
 1 files changed, 14 insertions(+), 0 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/winbindd/winbindd.c b/source3/winbindd/winbindd.c
index 0550da8..ca5a53b 100644
--- a/source3/winbindd/winbindd.c
+++ b/source3/winbindd/winbindd.c
@@ -1073,6 +1073,15 @@ int main(int argc, char **argv, char **envp)
}
}
 
+   /* We call dump_core_setup one more time because the command line can
+* set the log file or the log-basename and this will influence where
+* cores are stored. Without this call get_dyn_LOGFILEBASE will be
+* the default value derived from build's prefix. For EOM this value
+* is often not related to the path where winbindd is actually run
+* in production.
+*/
+   dump_core_setup(winbindd);
+
if (is_daemon  interactive) {
d_fprintf(stderr,\nERROR: 
  Option -i|--interactive is not allowed together with 
-D|--daemon\n\n);
@@ -1107,6 +1116,11 @@ int main(int argc, char **argv, char **envp)
DEBUG(0, (error opening config file\n));
exit(1);
}
+   /* After parsing the configuration file we setup the core path one more 
time
+* as the log file might have been set in the configuration and cores's
+* path is by default basename(lp_logfile()).
+*/
+   dump_core_setup(winbindd);
 
/* Initialise messaging system */
 


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  bb750d7 Fix bug #8972 - Directory group write permission bit is set 
if unix extensions are enabled
  from  4cf3fb8 s3-winbindd: call dump_core_setup after command line option 
has been parsed

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit bb750d7232bd266c06a14ac3ea577aeecfb81b14
Author: Jeremy Allison j...@samba.org
Date:   Wed Jun 13 10:48:32 2012 -0700

Fix bug #8972 - Directory group write permission bit is set if unix 
extensions are enabled

We can't manipulate file_attributes if it's a posix call.

---

Summary of changes:
 source3/smbd/open.c |6 --
 1 files changed, 4 insertions(+), 2 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/smbd/open.c b/source3/smbd/open.c
index 202643f..72b7d8e 100644
--- a/source3/smbd/open.c
+++ b/source3/smbd/open.c
@@ -2583,8 +2583,10 @@ static NTSTATUS open_directory(connection_struct *conn,
 
SMB_ASSERT(!is_ntfs_stream_smb_fname(smb_dname));
 
-   /* Ensure we have a directory attribute. */
-   file_attributes |= FILE_ATTRIBUTE_DIRECTORY;
+   if (!(file_attributes  FILE_FLAG_POSIX_SEMANTICS)) {
+   /* Ensure we have a directory attribute. */
+   file_attributes |= FILE_ATTRIBUTE_DIRECTORY;
+   }
 
DEBUG(5,(open_directory: opening directory %s, access_mask = 0x%x, 
 share_access = 0x%x create_options = 0x%x, 


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-5-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-5-test has been updated
   via  fe7d9d8 Fix bug #8972 - Directory group write permission bit is set 
if unix extensions are enabled
  from  529 s3-winbindd: call dump_core_setup after command line option 
has been parsed

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-5-test


- Log -
commit fe7d9d85102613346a1365929f4545e43f412ab8
Author: Jeremy Allison j...@samba.org
Date:   Wed Jun 13 10:48:32 2012 -0700

Fix bug #8972 - Directory group write permission bit is set if unix 
extensions are enabled

We can't manipulate file_attributes if it's a posix call.
(cherry picked from commit bb750d7232bd266c06a14ac3ea577aeecfb81b14)

---

Summary of changes:
 source3/smbd/open.c |6 --
 1 files changed, 4 insertions(+), 2 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/smbd/open.c b/source3/smbd/open.c
index ded07a1..dfa45ef 100644
--- a/source3/smbd/open.c
+++ b/source3/smbd/open.c
@@ -2296,8 +2296,10 @@ static NTSTATUS open_directory(connection_struct *conn,
 
SMB_ASSERT(!is_ntfs_stream_smb_fname(smb_dname));
 
-   /* Ensure we have a directory attribute. */
-   file_attributes |= FILE_ATTRIBUTE_DIRECTORY;
+   if (!(file_attributes  FILE_FLAG_POSIX_SEMANTICS)) {
+   /* Ensure we have a directory attribute. */
+   file_attributes |= FILE_ATTRIBUTE_DIRECTORY;
+   }
 
DEBUG(5,(open_directory: opening directory %s, access_mask = 0x%x, 
 share_access = 0x%x create_options = 0x%x, 


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-5-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-5-test has been updated
   via  c615d8e Fix bug #8994 - winbind normalize names.
  from  fe7d9d8 Fix bug #8972 - Directory group write permission bit is set 
if unix extensions are enabled

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-5-test


- Log -
commit c615d8e8e037996a9dd8d5a1982cf49d7c19a831
Author: Jeremy Allison j...@samba.org
Date:   Thu Jul 29 13:47:27 2010 -0700

Fix bug #8994 - winbind normalize names.

We should be using the winbindd separator in this case, not hardcoding a \\ 
value.

Jeremy.
(cherry picked from commit b7f029016a6a3fb98652c65f27ae80ad78048396)

Signed-off-by: Andreas Schneider a...@samba.org

---

Summary of changes:
 source3/winbindd/winbindd_pam.c |4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/winbindd/winbindd_pam.c b/source3/winbindd/winbindd_pam.c
index b0b8e40..c8910d6 100644
--- a/source3/winbindd/winbindd_pam.c
+++ b/source3/winbindd/winbindd_pam.c
@@ -1558,7 +1558,9 @@ enum winbindd_result winbindd_dual_pam_auth(struct 
winbindd_domain *domain,
parse_domain_user(mapped_user, name_domain, name_user);
 
if ( mapped_user != state-request-data.auth.user ) {
-   fstr_sprintf( domain_user, %s\\%s, name_domain, name_user );
+   fstr_sprintf( domain_user, %s%c%s, name_domain,
+   *lp_winbind_separator(),
+   name_user );
safe_strcpy( state-request-data.auth.user, domain_user,
 sizeof(state-request-data.auth.user)-1 );
}


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  2041ef3 s3: remove dependency on automake for make everything
  from  bb750d7 Fix bug #8972 - Directory group write permission bit is set 
if unix extensions are enabled

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 2041ef3df94842822102d371029cbb14a7f2c024
Author: Björn Jacke b...@sernet.de
Date:   Tue Jun 5 15:37:00 2012 +0200

s3: remove dependency on automake for make everything

the dependency was introduced by 737a1c9b96a4ba8d8688f4dba1df6d931e10e64d

We now call auto* in examples/VFS from within the main autogen.sh.

This fixes bug #8978.

Autobuild-User: Björn Jacke b...@sernet.de
Autobuild-Date: Tue Jun  5 17:32:22 CEST 2012 on sn-devel-104
(cherry picked from commit 5cc86fd560568202bef069eb89f5906f20050085)

---

Summary of changes:
 source3/Makefile.in |1 -
 source3/autogen.sh  |9 +
 2 files changed, 9 insertions(+), 1 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/Makefile.in b/source3/Makefile.in
index 2c28e2b..1a7ad8a 100644
--- a/source3/Makefile.in
+++ b/source3/Makefile.in
@@ -3557,7 +3557,6 @@ bin/ndrdump4: $(BINARY_PREREQS)
 vfs_examples:
( \
cd ../examples/VFS  \
-   ./autogen.sh  \
./configure  \
make clean  \
make \
diff --git a/source3/autogen.sh b/source3/autogen.sh
index d9f6030..52b5fa6 100755
--- a/source3/autogen.sh
+++ b/source3/autogen.sh
@@ -75,6 +75,15 @@ $AUTOCONF $IPATHS || exit 1
 
 rm -rf autom4te*.cache
 
+( cd ../examples/VFS || exit 1
+  echo $0: running $AUTOHEADER in ../examples/VFS/
+  $AUTOHEADER || exit 1
+  echo $0: running $AUTOCONF in ../examples/VFS/
+  $AUTOCONF || exit 1
+  rm -rf autom4te*.cache
+) || exit 1
+
+
 if gcc -E tests/preproc-dummy.c -o /dev/null ;
 then
 PIDL_OUTPUTDIR=librpc/gen_ndr CPP=gcc -E PIDL=../pidl/pidl \


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  e5e8668 Complete fix for bug #8811 - sd_has_inheritable_components 
segfaults on an SD that se_access_check accepts.
  from  2041ef3 s3: remove dependency on automake for make everything

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit e5e86688bdcb91817954fb23ecf10a3b6ddd3933
Author: Jeremy Allison j...@samba.org
Date:   Fri Mar 30 11:58:31 2012 -0700

Complete fix for bug #8811 - sd_has_inheritable_components segfaults on an 
SD that se_access_check accepts.

---

Summary of changes:
 source3/lib/secdesc.c|4 
 source3/modules/vfs_acl_common.c |   25 -
 source3/smbd/file_access.c   |5 -
 3 files changed, 28 insertions(+), 6 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/lib/secdesc.c b/source3/lib/secdesc.c
index 001eccb..007e097 100644
--- a/source3/lib/secdesc.c
+++ b/source3/lib/secdesc.c
@@ -534,6 +534,10 @@ bool sd_has_inheritable_components(const struct 
security_descriptor *parent_ctr,
unsigned int i;
const struct security_acl *the_acl = parent_ctr-dacl;
 
+   if (the_acl == NULL) {
+   return false;
+   }
+
for (i = 0; i  the_acl-num_aces; i++) {
const struct security_ace *ace = the_acl-aces[i];
 
diff --git a/source3/modules/vfs_acl_common.c b/source3/modules/vfs_acl_common.c
index 56da3af..a537011 100644
--- a/source3/modules/vfs_acl_common.c
+++ b/source3/modules/vfs_acl_common.c
@@ -166,7 +166,7 @@ static NTSTATUS create_acl_blob(const struct 
security_descriptor *psd,
  CREATOR_OWNER/CREATOR_GROUP/WORLD.
 ***/
 
-static void add_directory_inheritable_components(vfs_handle_struct *handle,
+static NTSTATUS add_directory_inheritable_components(vfs_handle_struct *handle,
 const char *name,
SMB_STRUCT_STAT *psbuf,
struct security_descriptor *psd)
@@ -184,7 +184,7 @@ static void 
add_directory_inheritable_components(vfs_handle_struct *handle,
num_aces + 3);
 
if (new_ace_list == NULL) {
-   return;
+   return NT_STATUS_NO_MEMORY;
}
 
/* Fake a quick smb_filename. */
@@ -236,8 +236,19 @@ static void 
add_directory_inheritable_components(vfs_handle_struct *handle,
SEC_ACE_FLAG_CONTAINER_INHERIT|
SEC_ACE_FLAG_OBJECT_INHERIT|
SEC_ACE_FLAG_INHERIT_ONLY);
-   psd-dacl-aces = new_ace_list;
-   psd-dacl-num_aces += 3;
+   if (psd-dacl) {
+   psd-dacl-aces = new_ace_list;
+   psd-dacl-num_aces += 3;
+   } else {
+   psd-dacl = make_sec_acl(talloc_tos(),
+   NT4_ACL_REVISION,
+   3,
+   new_ace_list);
+   if (psd-dacl == NULL) {
+   return NT_STATUS_NO_MEMORY;
+   }
+   }
+   return NT_STATUS_OK;
 }
 
 /***
@@ -393,10 +404,14 @@ static NTSTATUS get_nt_acl_internal(vfs_handle_struct 
*handle,
if (is_directory 
!sd_has_inheritable_components(psd,
true)) {
-   add_directory_inheritable_components(handle,
+   status = add_directory_inheritable_components(
+   handle,
name,
psbuf,
psd);
+   if (!NT_STATUS_IS_OK(status)) {
+   return status;
+   }
}
/* The underlying POSIX module always sets
   the ~SEC_DESC_DACL_PROTECTED bit, as ACLs
diff --git a/source3/smbd/file_access.c b/source3/smbd/file_access.c
index 9f95d68..bd65a70 100644
--- a/source3/smbd/file_access.c
+++ b/source3/smbd/file_access.c
@@ -261,7 +261,10 @@ bool directory_has_default_acl(connection_struct *conn, 
const char *fname)
NTSTATUS status = SMB_VFS_GET_NT_ACL(conn, fname,
SECINFO_DACL, secdesc);
 
-   if (!NT_STATUS_IS_OK(status) || secdesc == NULL) {
+   if (!NT_STATUS_IS_OK(status) ||
+   secdesc == NULL ||
+   secdesc-dacl == NULL) {
+   TALLOC_FREE(secdesc);
  

[SCM] Samba Shared Repository - branch v3-6-test updated

2012-06-13 Thread Karolin Seeger
The branch, v3-6-test has been updated
   via  6330936 Fix bug #8922.
  from  e5e8668 Complete fix for bug #8811 - sd_has_inheritable_components 
segfaults on an SD that se_access_check accepts.

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test


- Log -
commit 633093685734e44afd7a0de1d58a0ffe905ae13a
Author: Jeremy Allison j...@samba.org
Date:   Fri May 11 03:27:01 2012 -0700

Fix bug #8922.

Looking at the do_list status returns was historically
ignored, allowing tar to continue after NT_STATUS_ACCESS_DENIED.
Return to this state.

Autobuild-User: Jeremy Allison j...@samba.org
Autobuild-Date: Fri May 11 14:55:53 CEST 2012 on sn-devel-104

---

Summary of changes:
 source3/client/clitar.c |2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/client/clitar.c b/source3/client/clitar.c
index 3c08734..b658688 100644
--- a/source3/client/clitar.c
+++ b/source3/client/clitar.c
@@ -907,7 +907,7 @@ strlen(finfo-name)=%d\nname=%s,cur_dir=%s\n,
return NT_STATUS_NO_MEMORY;
}
DEBUG(5, (Doing list with mtar_mask: %s\n, mtar_mask));
-   status = do_list(mtar_mask, attribute, do_tar, False, True);
+   do_list(mtar_mask, attribute, do_tar, False, True);
client_set_cur_dir(saved_curdir);
TALLOC_FREE(saved_curdir);
TALLOC_FREE(new_cd);


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch master updated

2012-06-13 Thread Björn Jacke
The branch, master has been updated
   via  43c56dc s3: fix build without ads support
   via  e2747fc replace: fix unused variable warning
   via  dba3b2e doc: fix typo uniq → unique
  from  96ada4d s4:ntvfs: add '_fn' suffix to all ntvfs_ops function 
pointers

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master


- Log -
commit 43c56dc4255a7a6cbd176e6ae66a7652c6d72d2c
Author: Björn Jacke b...@sernet.de
Date:   Wed Jun 13 19:28:06 2012 +0200

s3: fix build without ads support

when we have no ads support we don't have the ads_get_sid_token symbol used 
in
this unused code :-)

Autobuild-User(master): Björn Jacke b...@sernet.de
Autobuild-Date(master): Wed Jun 13 21:20:15 CEST 2012 on sn-devel-104

commit e2747fc62c2acbab143c4971469e0a4fc36d8789
Author: Björn Jacke b...@sernet.de
Date:   Wed Jun 13 18:55:56 2012 +0200

replace: fix unused variable warning

found by the IRIX compiler

commit dba3b2e0b62d385599057a6e6b5d604e1e70394c
Author: Björn Jacke b...@sernet.de
Date:   Tue Jun 12 11:35:29 2012 +0200

doc: fix typo uniq → unique

---

Summary of changes:
 docs-xml/smbdotconf/winbind/idmapconfig.xml |2 +-
 lib/replace/strptime.c  |9 +
 libgpo/gpo_util.c   |7 ---
 3 files changed, 10 insertions(+), 8 deletions(-)


Changeset truncated at 500 lines:

diff --git a/docs-xml/smbdotconf/winbind/idmapconfig.xml 
b/docs-xml/smbdotconf/winbind/idmapconfig.xml
index 265fa12..b9aec46 100644
--- a/docs-xml/smbdotconf/winbind/idmapconfig.xml
+++ b/docs-xml/smbdotconf/winbind/idmapconfig.xml
@@ -83,7 +83,7 @@
Defines the available matching uid and gid range for which the
backend is authoritative. For allocating backends, this also
defines the start and the end of the range for allocating
-   new uniq IDs.
+   new unique IDs.
/para
para
winbind uses this parameter to find the backend that is
diff --git a/lib/replace/strptime.c b/lib/replace/strptime.c
index 0e40f75..181fd12 100644
--- a/lib/replace/strptime.c
+++ b/lib/replace/strptime.c
@@ -251,7 +251,6 @@ strptime_internal (rp, fmt, tm, decided, era_cnt)
  enum locale_status *decided;
  int era_cnt;
 {
-  const char *rp_backup;
   int cnt;
   size_t val;
   int have_I, is_pm;
@@ -261,15 +260,17 @@ strptime_internal (rp, fmt, tm, decided, era_cnt)
   int have_yday;
   int have_mon, have_mday;
 #ifdef _NL_CURRENT
+  const char *rp_backup;
   size_t num_eras;
-#endif
   struct era_entry *era;
 
+  era = NULL;
+#endif
+
   have_I = is_pm = 0;
   century = -1;
   want_century = 0;
   want_era = 0;
-  era = NULL;
 
   have_wday = want_xday = have_yday = have_mon = have_mday = 0;
 
@@ -297,10 +298,10 @@ strptime_internal (rp, fmt, tm, decided, era_cnt)
 #ifndef _NL_CURRENT
   /* We need this for handling the `E' modifier.  */
 start_over:
-#endif
 
   /* Make back up of current processing pointer.  */
   rp_backup = rp;
+#endif
 
   switch (*fmt++)
{
diff --git a/libgpo/gpo_util.c b/libgpo/gpo_util.c
index ca529f8..cfb4512 100644
--- a/libgpo/gpo_util.c
+++ b/libgpo/gpo_util.c
@@ -838,13 +838,11 @@ ADS_STATUS gp_get_machine_token(ADS_STRUCT *ads,
const char *dn,
struct security_token **token)
 {
+#ifdef HAVE_ADS
struct security_token *ad_token = NULL;
ADS_STATUS status;
NTSTATUS ntstatus;
 
-#ifndef HAVE_ADS
-   return ADS_ERROR_NT(NT_STATUS_NOT_SUPPORTED);
-#endif
status = ads_get_sid_token(ads, mem_ctx, dn, ad_token);
if (!ADS_ERR_OK(status)) {
return status;
@@ -855,4 +853,7 @@ ADS_STATUS gp_get_machine_token(ADS_STRUCT *ads,
return ADS_ERROR_NT(ntstatus);
}
return ADS_SUCCESS;
+#else
+   return ADS_ERROR_NT(NT_STATUS_NOT_SUPPORTED);
+#endif
 }


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch master updated

2012-06-13 Thread Björn Jacke
The branch, master has been updated
   via  6edb239 replace: fix unused variable warning
   via  6a3b3fa Revert replace: fix unused variable warning
  from  43c56dc s3: fix build without ads support

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master


- Log -
commit 6edb239f8ed724abaa594fbbac8c2eb6e485209e
Author: Björn Jacke b...@sernet.de
Date:   Wed Jun 13 22:02:53 2012 +0200

replace: fix unused variable warning

e2747fc62c2acbab143c4971469e0a4fc36d8789 fixed ...

Autobuild-User(master): Björn Jacke b...@sernet.de
Autobuild-Date(master): Wed Jun 13 23:57:58 CEST 2012 on sn-devel-104

commit 6a3b3fa3b079072d6a4bc399e3e410a9bd09b42b
Author: Björn Jacke b...@sernet.de
Date:   Wed Jun 13 21:55:42 2012 +0200

Revert replace: fix unused variable warning

This reverts commit e2747fc62c2acbab143c4971469e0a4fc36d8789.

one line slipped into a wrong ifndef ...

---

Summary of changes:
 lib/replace/strptime.c |2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)


Changeset truncated at 500 lines:

diff --git a/lib/replace/strptime.c b/lib/replace/strptime.c
index 181fd12..20e5d8c 100644
--- a/lib/replace/strptime.c
+++ b/lib/replace/strptime.c
@@ -298,7 +298,9 @@ strptime_internal (rp, fmt, tm, decided, era_cnt)
 #ifndef _NL_CURRENT
   /* We need this for handling the `E' modifier.  */
 start_over:
+#endif
 
+#ifdef _NL_CURRENT
   /* Make back up of current processing pointer.  */
   rp_backup = rp;
 #endif


-- 
Samba Shared Repository


[SCM] Samba Shared Repository - branch master updated

2012-06-13 Thread Jeremy Allison
The branch, master has been updated
   via  7e63e22 s3: Fix a comment
  from  6edb239 replace: fix unused variable warning

http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master


- Log -
commit 7e63e2230c25f0cbd3be7dcbbc29f0a26f9a5e90
Author: Volker Lendecke v...@samba.org
Date:   Wed Jun 13 11:11:39 2012 +0200

s3: Fix a comment

The fd count is implicit

Signed-off-by: Jeremy Allison j...@samba.org

Autobuild-User(master): Jeremy Allison j...@samba.org
Autobuild-Date(master): Thu Jun 14 01:53:17 CEST 2012 on sn-devel-104

---

Summary of changes:
 source3/locking/posix.c |   28 +---
 1 files changed, 13 insertions(+), 15 deletions(-)


Changeset truncated at 500 lines:

diff --git a/source3/locking/posix.c b/source3/locking/posix.c
index 2a274f9..02d9b6d 100644
--- a/source3/locking/posix.c
+++ b/source3/locking/posix.c
@@ -392,12 +392,10 @@ bool posix_locking_end(void)
 /
 
 /
- The records in posix_pending_close_tdb are composed of an array of ints
- keyed by dev/ino pair.
- The first int is a reference count of the number of outstanding locks on
- all open fd's on this dev/ino pair. Any subsequent ints are the fd's that
- were open on this dev/ino pair that should have been closed, but can't as
- the lock ref count is non zero.
+ The records in posix_pending_close_db are composed of an array of
+ ints keyed by dev/ino pair. Those ints are the fd's that were open on
+ this dev/ino pair that should have been closed, but can't as the lock
+ ref count is non zero.
 /
 
 /
@@ -568,7 +566,8 @@ static void delete_windows_lock_ref_count(files_struct *fsp)
 static void add_fd_to_close_entry(files_struct *fsp)
 {
struct db_record *rec;
-   uint8_t *new_data;
+   int *fds;
+   size_t num_fds;
NTSTATUS status;
TDB_DATA value;
 
@@ -579,19 +578,18 @@ static void add_fd_to_close_entry(files_struct *fsp)
SMB_ASSERT(rec != NULL);
 
value = dbwrap_record_get_value(rec);
+   SMB_ASSERT((value.dsize % sizeof(int)) == 0);
 
-   new_data = talloc_array(rec, uint8_t,
-   value.dsize + sizeof(fsp-fh-fd));
+   num_fds = value.dsize / sizeof(int);
+   fds = talloc_array(rec, int, num_fds+1);
 
-   SMB_ASSERT(new_data != NULL);
+   SMB_ASSERT(fds != NULL);
 
-   memcpy(new_data, value.dptr, value.dsize);
-   memcpy(new_data + value.dsize,
-  fsp-fh-fd, sizeof(fsp-fh-fd));
+   memcpy(fds, value.dptr, value.dsize);
+   fds[num_fds] = fsp-fh-fd;
 
status = dbwrap_record_store(
-   rec, make_tdb_data(new_data,
-  value.dsize + sizeof(fsp-fh-fd)), 0);
+   rec, make_tdb_data((uint8_t *)fds, talloc_get_size(fds)), 0);
 
SMB_ASSERT(NT_STATUS_IS_OK(status));
 


-- 
Samba Shared Repository


[SCM] CTDB repository - branch 1.13 updated - ctdb-1.43-13-g0cd522f

2012-06-13 Thread Ronnie Sahlberg
The branch, 1.13 has been updated
   via  0cd522f854bb788317e15e5f9a562bdb5abcfb17 (commit)
  from  ba94a0a9ccad7c1de0939e74f0163ae41102 (commit)

http://gitweb.samba.org/?p=ctdb.git;a=shortlog;h=1.13


- Log -
commit 0cd522f854bb788317e15e5f9a562bdb5abcfb17
Author: Ronnie Sahlberg ronniesahlb...@gmail.com
Date:   Wed Jun 13 16:17:18 2012 +1000

STATISTICS: Add tracking of the 10 hottest keys per database measured in 
hopcount

and add mechanisms to dump it using the ctdb dbstatistics command

---

Summary of changes:
 include/ctdb_protocol.h   |   14 
 libctdb/control.c |   40 ---
 server/ctdb_call.c|   50 -
 server/ctdb_ltdb_server.c |   38 -
 tools/ctdb.c  |9 
 5 files changed, 144 insertions(+), 7 deletions(-)


Changeset truncated at 500 lines:

diff --git a/include/ctdb_protocol.h b/include/ctdb_protocol.h
index 5c787ff..33187c7 100644
--- a/include/ctdb_protocol.h
+++ b/include/ctdb_protocol.h
@@ -614,6 +614,7 @@ struct ctdb_traverse_start_ext {
   ctdb statistics information
  */
 #define MAX_COUNT_BUCKETS 16
+#define MAX_HOT_KEYS  10
 
 struct ctdb_statistics {
uint32_t num_clients;
@@ -680,10 +681,23 @@ struct ctdb_statistics_wire {
 /*
  * db statistics
  */
+struct ctdb_db_hot_key {
+   uint32_t count;
+   TDB_DATA key;
+};
 struct ctdb_db_statistics {
uint32_t db_ro_delegations;
uint32_t db_ro_revokes;
uint32_t hop_count_bucket[MAX_COUNT_BUCKETS];
+   uint32_t num_hot_keys;
+   struct ctdb_db_hot_key hot_keys[MAX_HOT_KEYS];
+};
+struct ctdb_db_statistics_wire {
+   uint32_t db_ro_delegations;
+   uint32_t db_ro_revokes;
+   uint32_t hop_count_bucket[MAX_COUNT_BUCKETS];
+   uint32_t num_hot_keys;
+   char hot_keys[1];
 };
 
 /*
diff --git a/libctdb/control.c b/libctdb/control.c
index b4c54cd..f927e08 100644
--- a/libctdb/control.c
+++ b/libctdb/control.c
@@ -120,6 +120,9 @@ bool ctdb_getdbstat_recv(struct ctdb_connection *ctdb,
 {
struct ctdb_reply_control *reply;
struct ctdb_db_statistics *s;
+   struct ctdb_db_statistics_wire *wire;
+   int i;
+   char *ptr;
 
reply = unpack_reply_control(req, CTDB_CONTROL_GET_DB_STATISTICS);
if (!reply) {
@@ -129,16 +132,36 @@ bool ctdb_getdbstat_recv(struct ctdb_connection *ctdb,
DEBUG(ctdb, LOG_ERR, ctdb_getpnn_recv: status -1);
return false;
}
-   if (reply-datalen != sizeof(struct ctdb_db_statistics)) {
-   DEBUG(ctdb, LOG_ERR, ctdb_getdbstat_recv: returned data is %d 
bytes but should be %d, reply-datalen, (int)sizeof(struct 
ctdb_db_statistics));
+   if (reply-datalen  offsetof(struct ctdb_db_statistics_wire, 
hot_keys)) {
+   DEBUG(ctdb, LOG_ERR, ctdb_getdbstat_recv: returned data is %d 
bytes but should be = %d, reply-datalen, (int)sizeof(struct 
ctdb_db_statistics));
return false;
}
 
-   s = malloc(sizeof(struct ctdb_db_statistics));
+   wire = reply-data;
+
+   s = malloc(offsetof(struct ctdb_db_statistics, hot_keys) + 
sizeof(struct ctdb_db_hot_key) * wire-num_hot_keys);
if (!s) {
return false;
}
-   memcpy(s, reply-data, sizeof(struct ctdb_db_statistics));
+   s-db_ro_delegations = wire-db_ro_delegations;
+   s-db_ro_revokes = wire-db_ro_revokes;
+   for (i = 0; i  MAX_COUNT_BUCKETS; i++) {
+   s-hop_count_bucket[i] = wire-hop_count_bucket[i];
+   }
+   s-num_hot_keys  = wire-num_hot_keys;
+   ptr = wire-hot_keys[0];
+   for (i = 0; i  wire-num_hot_keys; i++) {
+   s-hot_keys[i].count = *(uint32_t *)ptr;
+   ptr += 4;
+
+   s-hot_keys[i].key.dsize = *(uint32_t *)ptr;
+   ptr += 4;
+
+   s-hot_keys[i].key.dptr = malloc(s-hot_keys[i].key.dsize);
+   memcpy(s-hot_keys[i].key.dptr, ptr, s-hot_keys[i].key.dsize);
+   ptr += s-hot_keys[i].key.dsize;
+   }
+
*stat = s;
 
return true;
@@ -158,9 +181,18 @@ struct ctdb_request *ctdb_getdbstat_send(struct 
ctdb_connection *ctdb,
 
 void ctdb_free_dbstat(struct ctdb_db_statistics *stat)
 {
+   int i;
+
if (stat == NULL) {
return;
}
+
+   for (i = 0; i  stat-num_hot_keys; i++) {
+   if (stat-hot_keys[i].key.dptr != NULL) {
+   free(stat-hot_keys[i].key.dptr);
+   }
+   }
+
free(stat);
 }
 
diff --git a/server/ctdb_call.c b/server/ctdb_call.c
index fe7e947..56cb5e8 100644
--- a/server/ctdb_call.c
+++ b/server/ctdb_call.c
@@ -667,6 +667,54 @@ ctdb_defer_pinned_down_request(struct ctdb_context *ctdb, 

[SCM] CTDB repository - branch master updated - ctdb-1.13-202-g8307c70

2012-06-13 Thread Ronnie Sahlberg
The branch, master has been updated
   via  8307c70ed98996b430c470e9641a09fdeeb81bd8 (commit)
   via  98e1b46adba11b9549b5c5976e1f561fe732fa6e (commit)
   via  0dc204988eadff214dd149a756d756ab6e96e410 (commit)
   via  7ebc00dc6a89043a971a720e7c21baf5f2a0233d (commit)
   via  cb2bbe93628c1ab932c2e1ad6e2ec199a98f74c6 (commit)
   via  88040778aace229d724de1ba7556aded12e22f86 (commit)
   via  e0c9200c05b1f7a04e002f505ebb5ba9340c0ca1 (commit)
   via  6559106b8b853920f325f2dba532f4008e931fa3 (commit)
  from  1a6a011c772f7d302d114d7c8a151fa7820ec85f (commit)

http://gitweb.samba.org/?p=ctdb.git;a=shortlog;h=master


- Log -
commit 8307c70ed98996b430c470e9641a09fdeeb81bd8
Author: Ronnie Sahlberg ronniesahlb...@gmail.com
Date:   Wed Jun 13 16:17:18 2012 +1000

STATISTICS: Add tracking of the 10 hottest keys per database measured in 
hopcount

and add mechanisms to dump it using the ctdb dbstatistics command

commit 98e1b46adba11b9549b5c5976e1f561fe732fa6e
Author: Martin Schwenke mar...@meltin.net
Date:   Thu Jun 7 15:08:15 2012 +1000

Reimplement logging of long running events

Reimplement 5aba53e6adcfcd7edbdac9e30aa5fcba176aca00 using tevent
trace points.

Signed-off-by: Martin Schwenke mar...@meltin.net

commit 0dc204988eadff214dd149a756d756ab6e96e410
Author: Stefan Metzmacher me...@samba.org
Date:   Fri Jun 8 12:50:21 2012 +0200

tevent: change version to 0.9.16

This adds tevent_*_trace_*() and tevent_context_init_ops()

metze

Autobuild-User(master): Stefan Metzmacher me...@samba.org
Autobuild-Date(master): Fri Jun  8 20:47:41 CEST 2012 on sn-devel-104

commit 7ebc00dc6a89043a971a720e7c21baf5f2a0233d
Author: Stefan Metzmacher me...@samba.org
Date:   Fri May 11 15:19:55 2012 +0200

tevent: expose tevent_context_init_ops

This can be used to implement wrapper backends,
while passing a private pointer to the backens init function
via ev-additional_data.

metze

commit cb2bbe93628c1ab932c2e1ad6e2ec199a98f74c6
Author: Martin Schwenke mar...@meltin.net
Date:   Tue Jun 5 16:00:07 2012 +1000

lib/tevent: Add trace point callback

Set/get a single callback function to be invoked at various trace
points.  Define before wait and after wait trace points - more
trace points can be added later if required.

CTDB wants this to log long waits and events.

Pair-programmed-with: Amitay Isaacs ami...@gmail.com
Signed-off-by: Martin Schwenke mar...@meltin.net
Signed-off-by: Stefan Metzmacher me...@samba.org

commit 88040778aace229d724de1ba7556aded12e22f86
Author: Martin Schwenke mar...@meltin.net
Date:   Thu Jun 7 14:20:13 2012 +1000

Revert TEVENT: Add back tracking of long runnig  events to the local copy 
of tevent library

This reverts commit 5aba53e6adcfcd7edbdac9e30aa5fcba176aca00.

Do this using new tevent trace point callback.

commit e0c9200c05b1f7a04e002f505ebb5ba9340c0ca1
Author: Martin Schwenke mar...@meltin.net
Date:   Thu Jun 7 12:26:02 2012 +1000

lib/tevent: In poll_event_context, add a pointer back to the tevent_context

This makes it consistent with the other backends.

Signed-off-by: Martin Schwenke mar...@meltin.net
Signed-off-by: Stefan Metzmacher me...@samba.org

commit 6559106b8b853920f325f2dba532f4008e931fa3
Author: Stefan Metzmacher me...@samba.org
Date:   Mon May 14 11:48:00 2012 +0200

lib/tevent/testsuite: no longer use 'compat' symbols

metze

---

Summary of changes:
 include/ctdb_protocol.h|   14 +
 .../ABI/{tevent-0.9.15.sigs = tevent-0.9.16.sigs} |4 ++
 lib/tevent/testsuite.c |   24 
 lib/tevent/tevent.c|8 ++-
 lib/tevent/tevent.h|   57 
 lib/tevent/tevent_debug.c  |   24 
 lib/tevent/tevent_epoll.c  |6 +--
 lib/tevent/tevent_internal.h   |   10 +++-
 lib/tevent/tevent_poll.c   |6 ++
 lib/tevent/tevent_select.c |6 +--
 lib/tevent/tevent_standard.c   |6 +--
 lib/tevent/tevent_util.c   |   52 --
 libctdb/control.c  |   40 --
 server/ctdb_call.c |   50 +-
 server/ctdb_daemon.c   |   46 
 server/ctdb_ltdb_server.c  |   38 -
 tools/ctdb.c   |9 +++
 17 files changed, 312 insertions(+), 88 deletions(-)
 copy lib/tevent/ABI/{tevent-0.9.15.sigs = tevent-0.9.16.sigs} (94%)


Changeset