[Samba] name resolution: dns name different to windows machine name
Hi, we are running a samba domain controller as master with activated WINS. The machine name of each WinXP client is set manually during the installation initial process. Each client is using the WINS server of our domain controller. The WinXP clients are getting their IP by a global DNS Server, which sets the DNS and reverse DNS entry identically to the windos machine name. Now the administrator of the DNS server is planning to change the DNS and reverse DNS concept in the way that the DNS and reverse DNS entry will be not identically to the windows machine name any more. The WinXP clients will get a generic, randomly set DNS/reverse DNS entry. Does this have any effects for functionality of my samba domain controller and/or the WinXP clients? Thanks, Marcus -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
[Samba] problem connecting DFS-share with winXP - successful with Vista 7
Hi, i have problems connecting to DFS-Share from Client WindowsXP. Same configuration works fine for Windows Vista and 7. On Windows 7 the LMCompatibility Level is 3. the striking point I see in logfile is following 2010/12/20 10:30:17, 1] smbd/service.c:make_connection_snum(1119) 10.184.144.171 (10.184.144.171) signed connect to service applbin initially as user useracc (uid=45110, gid=45110) (pid 20312) [2010/12/20 10:30:17, 3] smbd/sec_ctx.c:set_sec_ctx(324) setting sec ctx (0, 0) - sec_ctx_stack_ndx = 0 [2010/12/20 10:30:17, 2] smbd/reply.c:reply_tcon_and_X(789) Serving applbin as a Dfs root [2010/12/20 10:30:17, 3] smbd/reply.c:reply_tcon_and_X(794) tconX service=APPLBIN [2010/12/20 10:30:17, 3] smbd/process.c:process_smb(1576) Transaction 3 of length 1404 (0 toread) [2010/12/20 10:30:17, 3] smbd/process.c:switch_message(1393) switch message SMBsesssetupX (pid 20312) conn 0x0 [2010/12/20 10:30:17, 3] smbd/sec_ctx.c:set_sec_ctx(324) setting sec ctx (0, 0) - sec_ctx_stack_ndx = 0 [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_sesssetup_and_X(1412) wct=12 flg2=0xc807 [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_sesssetup_and_X_spnego(1175) Doing spnego session setup [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_sesssetup_and_X_spnego(1210) NativeOS=[Windows 2002 Service Pack 3 2600] NativeLanMan=[Windows 2002 5.1] PrimaryDomain=[] [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_spnego_negotiate(802) reply_spnego_negotiate: Got secblob of size 1172 [2010/12/20 10:30:17, 3] libads/authdata.c:decode_pac_data(301) Found account name from PAC: CLIENTXP$ [] [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_spnego_kerberos(356) Ticket name is [client...@domain.local] [2010/12/20 10:30:17, 1] smbd/sesssetup.c:reply_spnego_kerberos(474) Username DOMAIN00\CLIENTXP$ is invalid on this system [2010/12/20 10:30:17, 3] smbd/error.c:error_packet_set(61) error packet at smbd/sesssetup.c(480) cmd=115 (SMBsesssetupX) NT_STATUS_LOGON_FAILURE On Windows Vista the Same Section in logfile shows different account-name: Doing spnego session setup [2010/12/20 09:12:29, 3] smbd/sesssetup.c:reply_sesssetup_and_X_spnego(1210) NativeOS=[] NativeLanMan=[] PrimaryDomain=[] [2010/12/20 09:12:29, 3] smbd/sesssetup.c:reply_spnego_negotiate(802) reply_spnego_negotiate: Got secblob of size 6566 [2010/12/20 09:12:29, 3] libads/authdata.c:decode_pac_data(301) Found account name from PAC: USERACC [Nachname, Vorname] [2010/12/20 09:12:29, 3] smbd/sesssetup.c:reply_spnego_kerberos(356) Ticket name is [user...@domain.local] The difference i see, is that on windows XP the system don't take the username of connecting account, although it is grabbed in start of session-log. [2010/12/20 10:30:17, 3] libads/authdata.c:decode_pac_data(301) Found account name from PAC: USERACC [Nachname, Vorname] [2010/12/20 10:30:17, 3] smbd/sesssetup.c:reply_spnego_kerberos(356) Ticket name is [efro...@domain.local] Has someone an idea, how i can fix this problem. This problem only occurs, if i connect via DFS Master. If i connect directly to the Share, all is successful. [complete Logfile]: http://paste.ubuntu.com/545870/ [config DFS-master]: http://paste.ubuntu.com/545874/ [config client]: http://paste.ubuntu.com/545873/ Regards, Steffen -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Need a little help with Samba 3.5.5 on FreeNAS 7.2.5543
On 20 December 2010 04:16, Steve B stev...@gmail.com wrote: 1. Not sure what you mean by remote the file. I can open and read the file I believe he meant remove. I think he was just wondering if you got a different error with the file missing, which would prove that Audiotron was actually accessing the file. If you got the same symptoms, then it's possible that Audiotron was not actually finding/reading the file at all. in any editor, but the Audiotron performs some sort of read function on the file that tells it there are X number of radio stations defined in the file. The file is basically an XML file. -- Michael Wood esiot...@gmail.com -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] howto? howto switch domain controllers in an ldap directory?
I've got an ldap directory, but no PDC yet. I want to set up a test PDC, and once things work, I want to be able to set a more appropriate system to be teh PDC using teh same login informations (replica of existing authzn databases). where's the documentation of how to do this? what do I need to know about this process (I'm not sure what it's even called ,., in case I wasn't clear above, I want to just be able to flip a switch (in the proverbial sense) and switch PDC on an existing ldap directory without having to purge it or migrate data or anything else so complex. To switch PDCs basically all you need to do is set the ip addreess of the ldap servers in the PDC and set the ldap password so the pdc can connect to the ldap server. smbpasswd -w ldap-secret Remember that samba and ldap are independent. The PDC need not be on the same server as your ldap servers. And I say servers you should always maintain more than 1 ldap server. This is pretty easy to do with ldap syncrepl http://www.bind9.net/manual/openldap/2.3/syncrepl.html You don't forget to configure pam and nss to use ldap on each linux machine that is on your domain. John -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
[Samba] Problem using smb_audit with samba3-3.5
I don't know if I can ask here, but I don't have any answer in official forum[1]. I really need to implement this, please help me. I did all steps of INSTALL documentation and I got this error: [2010/12/10 11:58:33.911661, 0] lib/module.c:59(do_smb_load_module) Error trying to resolve symbol 'init_samba_module' in /usr/lib64/samba/vfs/mysql_audit.so: /usr/lib64/samba/vfs/mysql_audit.so: undefined symbol: init_samba_module [2010/12/10 11:58:33.911810, 0] smbd/vfs.c:167(vfs_init_custom) error probing vfs module 'mysql_audit': NT_STATUS_UNSUCCESSFUL Any ideas? I am really lost :( [1] - http://sourceforge.net/projects/smbdaudit/forums/forum/275569/topic/4008809 Rafael Gomes Consultor em TI LPIC-1 MCSO (71) 8318-0284 IV Encontro Nordestino de Software Livre IV Encontro Potiguar de Software Livre http://ensl.org.br Atenção: Este e-mail pode conter anexos no formato ODF (Open Document Format)/ABNT (extensões odt, ods, odp, odb, odg). Antes de pedir os anexos em outro formato, você pode instalar gratuita e livremente o BrOffice (http://www.broffice.org). -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Samba upgrade HowTo requested
Hi, Volker! On Sat, Dec 18, 2010 at 10:10 AM, Volker Lendecke volker.lende...@sernet.de wrote: On Fri, Dec 17, 2010 at 11:26:12AM +0100, Willy Offermans wrote: 20101026: AFFECTS: users of net/samba35 AUTHOR: Timur Bakeyev ti...@freebsd.org This is the latest stable release of the Samba3 distribution. It has been extended with the experimental support of the NFS4-like ACLs on ZFS partitions, thanks to the sysutils/libsunacl library by Edward Tomasz Napierala(trasz). This support haven't been tested thoroughly, so try it on your own risk. This looks interesting. I just did a portsnap fetch update in my FreeBSD 8.1 box, but I don't find that snippet. Where can I find those patches? There are small patches in the port itself to detect and incorporate libsunacl via configure and build vfs_zfsacl module OOTB. As for the lib itself - it is situated in /usr/ports/sysutils/libsunacl. With regards, Timur. -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] winbind / trust questions and issues
On 12/15/2010 4:19 PM, Eric A. Hall wrote: First issue is that I would like to filter out the local (LABS) users and groups in winbind if possible. I tried using winbind: ignore domains = LABS but this causes winbindd to completely ignore its own domain, which in turn causes it to freak out, dump core, and die I have also tried using require_membership_of=SID for CORP/Domain Users (the remote users I want to allow local access through winbind), but pam_winbind complains about the foreign SID and then allows everybody to login despite the error Anything else I could try? -- Eric A. Hall http://www.eric-a-hall.com/ Network Technology Research Grouphttp://www.ntrg.com/ Internet Core Protocols http://www.oreilly.com/catalog/coreprot/ -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Samba upgrade question
On Thu, Dec 16, 2010 at 07:56:55AM -0800, Chernoguz, Inessa wrote: Hi All, I have a problem with upgrade samba on Solaris 10 (x86). The current version is 3.0.25b (distributed with Solaris 10). I am trying to upgrade version to 3.4.5. I compiled samba from sources, has not any problem with make and installation. Now I have 2 different versions and when I am running svcadm enable samba I am receiving old version. The old version installed under: /usr/sfw/sbin, new one under /usr/sbin... # /usr/sfw/sbin/smbd -V Version 3.0.25a # /usr/sbin/smbd -V Version 3.4.5 # ps -ef | grep smbd root 7531 7305 0 17:55:32 pts/1 0:00 grep smbd root 6981 6979 0 15:19:58 ? 0:00 /usr/sfw/sbin/smbd -D root 6979 1 0 15:19:58 ? 0:00 /usr/sfw/sbin/smbd -D Can you, please help me here? Thanks a lot, Inessa As someone already replied, you need to disable the vendor provided Samba first: # svcadm disable samba Then set up SysV init scripts for your new, custom installation of Samba or something via SMF. Alternately, Sunfreeware or Blastwave may have pre-packaged Samba installs that include these scripts. I prefer sticking with the vendor-provided packages when possible (hand built packages don't scale administratively). The latest Sun version of Samba is 3.0.37 for Solaris 10 and I believe they have a refresh due out shortly. Thanks, Ray -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Samba upgrade question
The key advantages of the Solaris provided version is that - someone already did the compilation work - ZFS support is included (this is backported by Sun and not included with 3.0.x source code from samba.) - nsswitch support is also enabled The big problems I had with Sun's version (which is why I also ended up recompiling) - domain trusts don't always work, and the idmap functionality is buggy. So I think 3.0.x is pretty much a dead-end at this point. According to the Oracle forums, Samba 3.4.x is included with the recent Solaris Express release- so I expect it to be included with Solaris 11 (whenever that comes out.) My guess (no evidence to back this up) is that Solaris 10 will stay on samba 3.0.x. You have to make some changes when you change versions and if they pushed that our as an automatic update it could possible break things. With the sunfreeware and blastwave versions of samba- - these might be 32-bit only. - I don't think zfs is included. - nsswitch support may or may not be included.I had problems with this with sunfreeware samba in the past. On 12/20/2010 11:34 AM, Ray Van Dolson wrote: On Thu, Dec 16, 2010 at 07:56:55AM -0800, Chernoguz, Inessa wrote: Hi All, I have a problem with upgrade samba on Solaris 10 (x86). The current version is 3.0.25b (distributed with Solaris 10). I am trying to upgrade version to 3.4.5. I compiled samba from sources, has not any problem with make and installation. Now I have 2 different versions and when I am running svcadm enable samba I am receiving old version. The old version installed under: /usr/sfw/sbin, new one under /usr/sbin... # /usr/sfw/sbin/smbd -V Version 3.0.25a # /usr/sbin/smbd -V Version 3.4.5 # ps -ef | grep smbd root 7531 7305 0 17:55:32 pts/1 0:00 grep smbd root 6981 6979 0 15:19:58 ? 0:00 /usr/sfw/sbin/smbd -D root 6979 1 0 15:19:58 ? 0:00 /usr/sfw/sbin/smbd -D Can you, please help me here? Thanks a lot, Inessa As someone already replied, you need to disable the vendor provided Samba first: # svcadm disable samba Then set up SysV init scripts for your new, custom installation of Samba or something via SMF. Alternately, Sunfreeware or Blastwave may have pre-packaged Samba installs that include these scripts. I prefer sticking with the vendor-provided packages when possible (hand built packages don't scale administratively). The latest Sun version of Samba is 3.0.37 for Solaris 10 and I believe they have a refresh due out shortly. Thanks, Ray -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Keeping Windows ACL's when migrating to SAMBA Server
I've tried it with the /copyall switch and /copy:datsou and still no luck On Dec 20, 2010 6:31 PM, Andrew Masterson andrew.master...@nuvistaenergy.com wrote: -Original Message- From: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] On Behalf Of TAKAHASHI Motonobu Sent: Saturday, December 18, 2010 10:20 PM To: George Cc: samba@lists.samba.org; Jeremy Allison Subject: Re: [Samba] Keeping Windows ACL's when migrating to SAMBA Server 2010/12/19 George greenadm...@gmail.com: Rightnow i got it half working but am having trouble with part of it. If I use robocopy, It copies the files and ACL but not the correct create of modify dates robocopy \\man_fs2\Batteries \\bed-fs1\servers\man_fs2\Batteries /sec /e If I use xxcopy I can copy all the files with the proper create and modify dates but it wont copy the ACL's. xxcopy \\man_fs2\Batteries \\bed-fs1\servers\man_fs2\Batteries /backup George Are you using the /COPYALL switch? http://www.stevelu.com/TechnicalArticles/DevTools/1206.aspx (for gory flag details) I have successfully migrated several servers with terabytes of data from windows to linux by using robocopy (although I don't remember the exact flags at the moment). xcopy is unreliable for that much data IMX. -=Andrew -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Keeping Windows ACL's when migrating to SAMBA Server
-Original Message- From: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] On Behalf Of TAKAHASHI Motonobu Sent: Saturday, December 18, 2010 10:20 PM To: George Cc: samba@lists.samba.org; Jeremy Allison Subject: Re: [Samba] Keeping Windows ACL's when migrating to SAMBA Server 2010/12/19 George greenadm...@gmail.com: Rightnow i got it half working but am having trouble with part of it. If I use robocopy, It copies the files and ACL but not the correct create of modify dates robocopy \\man_fs2\Batteries \\bed-fs1\servers\man_fs2\Batteries /sec /e If I use xxcopy I can copy all the files with the proper create and modify dates but it wont copy the ACL's. xxcopy \\man_fs2\Batteries \\bed-fs1\servers\man_fs2\Batteries /backup George Are you using the /COPYALL switch? http://www.stevelu.com/TechnicalArticles/DevTools/1206.aspx (for gory flag details) I have successfully migrated several servers with terabytes of data from windows to linux by using robocopy (although I don't remember the exact flags at the moment). xcopy is unreliable for that much data IMX. -=Andrew -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] name resolution: dns name different to windows machine name
I am not sure how you would have a DNS server assign random names. But if you aren't going to have the correct entries for the XP machine in DNS, then you probably should not have any DNS entries for those machines.XP machines can register their hostnames directly in DNS (if the DNS server supports it.) In general I don't think workstations machines need to be in DNS at all since (unless you have shared folders or printers.) I general DISABLE dynamic updates in DNS and don't have DNS assignments/static IP's for most XP machines. -Original Message- From: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] On Behalf Of Marcus Sent: Monday, December 20, 2010 5:17 AM To: samba@lists.samba.org Subject: [Samba] name resolution: dns name different to windows machine name Hi, we are running a samba domain controller as master with activated WINS. The machine name of each WinXP client is set manually during the installation initial process. Each client is using the WINS server of our domain controller. The WinXP clients are getting their IP by a global DNS Server, which sets the DNS and reverse DNS entry identically to the windos machine name. Now the administrator of the DNS server is planning to change the DNS and reverse DNS concept in the way that the DNS and reverse DNS entry will be not identically to the windows machine name any more. The WinXP clients will get a generic, randomly set DNS/reverse DNS entry. Does this have any effects for functionality of my samba domain controller and/or the WinXP clients? Thanks, Marcus -- 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
Re: [Samba] Ldap Users only?
Your example is a text-book case of why you would use domains. I am not sure why it would be unacceptable- since you already have a samba machine and, from the user POV, the login process is practically the same. The LDAP component is for the samba backend- you would need an LDAP backend for samba if you have more than one domain controller, and you might want to use LDAP backend for samba if you are already using LDAP for other stuff. But since windows domain model is unacceptable, the only other solution I can think of is the use Kerberos.You can configure XP machines to use Kerberos authentication, and then configure your linux server as a Kerberos server. You would have to look on support.microsoft.com for instructions for the XP side of things.Although I am then not sure how you would configure samba to use Kerberos authentication as well. Maybe if you configure plaintext auth in samba so it can use the OS level authentication -Original Message- From: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] On Behalf Of Christ Schlacta Sent: Sunday, December 19, 2010 10:28 PM To: samba@lists.samba.org Subject: Re: [Samba] Ldap Users only? firstly, I'd like to apologize, somehow we ended up off-list. my setup looks like this: Density: Massive file storage rebirth: Media server (most files come from density, but it has a share for local files, too) faithful: backup server (handles backups from density, rebirth, and other systems) smb shares are used for laptops and desktops (a bunch of laptops and desktops): Joining these systems to the domain is inconvenient, awkward, and not going to happen because we have a small family household, and most of them are non-techie. the trick is, each time we change a password on a laptop or desktop, we have to change the corresponding passwords on all the systems. since each user only uses 1-2 laptops and a desktop (my wife and I have a laptop and netbook each), that's simple. Changing the server passwords requires logging into 3 servers over ssh, typing their old and new passwords, then changing their samba passwords. I want to store all the usernames and passwords for the servers in the ldap directory, so that users can update their passwords once. as I stated above, joining the end user systems to a domain is UNACCEPTABLE, and with one of the servers set to be master browser, I can set all of the systems to join the workgroup aarcane.info, and all the systems show up on the network view in windows 7 (and windows XP, and linux, and mac also.). All I can find is howtos on using samba as an ldap-backed domain controller. thanks, again, tms3 and everyone else for any help. On 12/19/2010 18:20, t...@tms3.com wrote: On Sunday 19/12/2010 at 5:54 pm, Christ Schlacta wrote: actually, it's because we have a few samba servers here, it's just a home, Honestly, I have a lab/cloud at home. I can't for the life of me even contemplate running them with out full Samba/LDAP domain mode...well, I've converted over to Samba4 since August, but it is by far the easiest way to manage things. but different machines use samba for different reasons. as such, it's a pain in the butt to have to change passwords on all systems. I'm aware that we'd need to have matching unix accounts, and the plan is to use 10K+ UIDs for samba users to make logging into the shell a simple matter as well. What's bugging me at the moment, is that in workgroup mode, it uses the HOSTNAME HOSTNAME is a NETBios name, and it is based on that whole host of protocols/services. It has NOTHING to do with DNS names, or machine names. for the domain name.. but there are several different hosts.. can I just use the workgroup name and have it work? will it be smart enough to say I'm not in domain mode, so the domain doesn't matter, or will I need to add a user for each host, thus mitigating any benefit? It would perhaps be better if you laid out what you have and are doing. I'm having a tough time understanding what you are doing/having issues with. On 12/19/2010 17:37, t...@tms3.com wrote: On Sunday 19/12/2010 at 5:02 pm, Christ Schlacta wrote: X-SpamDetect-Info: - Start ASpam results --- X-SpamDetect-Info: This message may be spam. This message BODY has been altered to show you the spam information X-SpamDetect: ***: 3.8 sd=3.8 [96]12%-6.0(Accept Orbs) [212]87%5.6(!46,60) [129]44%-0.0(from_return_nomatch) [27]46%-0.0(X-LangGuess:English) X-SpamDetect-Info: - End ASpam results - how to do only users from ldap? Same way as domain mode...I'm assuming for workstation users to have access to smb shares You don't need to add machines to the domain, though why you wouldn't want to I don't know. Then you need to have pam
Re: [Samba] name resolution: dns name different to windows machinename
--- Original message --- Subject: Re: [Samba] name resolution: dns name different to windows machinename From: Gaiseric Vandal gaiseric.van...@gmail.com To: samba@lists.samba.org Date: Monday, 20/12/2010 5:15 PM I am not sure how you would have a DNS server assign random names. But if you aren't going to have the correct entries for the XP machine in DNS, then you probably should not have any DNS entries for those machines.XP machines can register their hostnames directly in DNS (if the DNS server supports it.) In general I don't think workstations machines need to be in DNS at all since (unless you have shared folders or printers.) It's certainly easier when a help desk request comes in to ask the user for the machine name label, then try and get the machines IP addy. rdp://acct1 rdp://lab7 rdp://eng21 rdp://admin5 soo But yeah, WinXP and up have no problem doing this. dhcp can be a big help as well. I general DISABLE dynamic updates in DNS and don't have DNS assignments/static IP's for most XP machines. -Original Message- From: samba-boun...@lists.samba.org [mailto:samba-boun...@lists.samba.org] On Behalf Of Marcus Sent: Monday, December 20, 2010 5:17 AM To: samba@lists.samba.org Subject: [Samba] name resolution: dns name different to windows machine name Hi, we are running a samba domain controller as master with activated WINS. The machine name of each WinXP client is set manually during the installation initial process. Each client is using the WINS server of our domain controller. The WinXP clients are getting their IP by a global DNS Server, which sets the DNS and reverse DNS entry identically to the windos machine name. Now the administrator of the DNS server is planning to change the DNS and reverse DNS concept in the way that the DNS and reverse DNS entry will be not identically to the windows machine name any more. The WinXP clients will get a generic, randomly set DNS/reverse DNS entry. Does this have any effects for functionality of my samba domain controller and/or the WinXP clients? Thanks, Marcus -- 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
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 33a57db s4-pydsdb: Use local memory context in
py_dsdb_get_oid_from_attid()
from e1d3de3 dnspython: Merge in new upstream.
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 33a57db3a542c364eb82499e5765a2caef00af33
Author: Kamen Mazdrashki kame...@samba.org
Date: Mon Dec 20 11:39:26 2010 +0200
s4-pydsdb: Use local memory context in py_dsdb_get_oid_from_attid()
This version reverts changes from commit
b974966cc2b4d0b5b0d83206070b5f7c5c6495d1
and is what Matthieu Patou had commited in
d784ecec555a3d9737e6f4b3894f27904d2b833c
with added reference to the schema cache.
I think referencing schema here is the right thing to be done
as thus we garantee that schema cache will stay in memory
for the time our function is executed
Autobuild-User: Kamen Mazdrashki kame...@samba.org
Autobuild-Date: Mon Dec 20 12:01:53 CET 2010 on sn-devel-104
---
Summary of changes:
source4/dsdb/pydsdb.c | 20
1 files changed, 16 insertions(+), 4 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source4/dsdb/pydsdb.c b/source4/dsdb/pydsdb.c
index af03504..1a51b86 100644
--- a/source4/dsdb/pydsdb.c
+++ b/source4/dsdb/pydsdb.c
@@ -249,25 +249,37 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject
*self, PyObject *args)
const char *oid;
PyObject *ret;
WERROR status;
+ TALLOC_CTX *mem_ctx;
if (!PyArg_ParseTuple(args, Oi, py_ldb, attid))
return NULL;
PyErr_LDB_OR_RAISE(py_ldb, ldb);
- schema = dsdb_get_schema(ldb, NULL);
+ mem_ctx = talloc_new(NULL);
+ if (!mem_ctx) {
+ PyErr_NoMemory();
+ return NULL;
+ }
+ schema = dsdb_get_schema(ldb, mem_ctx);
if (!schema) {
PyErr_SetString(PyExc_RuntimeError, Failed to find a schema
from ldb \n);
+ talloc_free(mem_ctx);
return NULL;
}
status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid,
- NULL, oid);
- PyErr_WERROR_IS_ERR_RAISE(status);
+ mem_ctx, oid);
+ if (!W_ERROR_IS_OK(status)) {
+ PyErr_SetWERROR(status);
+ talloc_free(mem_ctx);
+ return NULL;
+ }
ret = PyString_FromString(oid);
- talloc_free(discard_const_p(char, oid));
+
+ talloc_free(mem_ctx);
return ret;
}
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 9dce43a s4-smbtorture: run the winreg tests also against an added
driver.
from 33a57db s4-pydsdb: Use local memory context in
py_dsdb_get_oid_from_attid()
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 9dce43a809eae4b44bf1e2b301844a0753c037e9
Author: Günther Deschner g...@samba.org
Date: Fri Dec 17 16:34:47 2010 +0100
s4-smbtorture: run the winreg tests also against an added driver.
Guenther
Autobuild-User: Günther Deschner g...@samba.org
Autobuild-Date: Mon Dec 20 14:31:43 CET 2010 on sn-devel-104
---
Summary of changes:
source4/torture/rpc/spoolss.c | 19 +++
1 files changed, 19 insertions(+), 0 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source4/torture/rpc/spoolss.c b/source4/torture/rpc/spoolss.c
index 114b3e1..630694d 100644
--- a/source4/torture/rpc/spoolss.c
+++ b/source4/torture/rpc/spoolss.c
@@ -8628,6 +8628,25 @@ static bool test_PrinterDriver_args(struct
torture_context *tctx,
return ret;
}
+ {
+ struct dcerpc_pipe *p2;
+ struct policy_handle hive_handle;
+ struct dcerpc_binding_handle *b2;
+
+ torture_assert_ntstatus_ok(tctx,
+ torture_rpc_connection(tctx, p2, ndr_table_winreg),
+ could not open winreg pipe);
+ b2 = p2-binding_handle;
+
+ torture_assert(tctx, test_winreg_OpenHKLM(tctx, b2,
hive_handle), );
+
+ ret = test_GetDriverInfo_winreg(tctx, b, NULL, NULL,
r-driver_name, r-architecture, r-version, b2, hive_handle, server_name);
+
+ test_winreg_CloseKey(tctx, b2, hive_handle);
+
+ talloc_free(p2);
+ }
+
if (ex) {
return test_DeletePrinterDriverEx(tctx, b, server_name,
r-driver_name, r-architecture, delete_flags, r-version);
} else {
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via bfc4fe4 s3: Remove unused retry from cli_start_connection
via d096de5 s3: Remove unused retry from cli_full_connection
via 6601906 s3: Always retry the DC connection in auth_domain
via 09cccee s3: Remove some unused variables
from 9dce43a s4-smbtorture: run the winreg tests also against an added
driver.
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit bfc4fe401259085f9e17c75a7ec171be0bc5f35f
Author: Volker Lendecke v...@samba.org
Date: Mon Dec 20 16:43:39 2010 +0100
s3: Remove unused retry from cli_start_connection
Autobuild-User: Volker Lendecke vlen...@samba.org
Autobuild-Date: Mon Dec 20 17:58:33 CET 2010 on sn-devel-104
commit d096de56b16c50c7cc22df08895dc29567ee15d7
Author: Volker Lendecke v...@samba.org
Date: Mon Dec 20 16:37:23 2010 +0100
s3: Remove unused retry from cli_full_connection
commit 660190632e4f8ace65f4944f4686ea42444056c0
Author: Volker Lendecke v...@samba.org
Date: Mon Dec 20 16:22:02 2010 +0100
s3: Always retry the DC connection in auth_domain
The only condition that cli_full_connection marks as non-retryable is the
basic
name lookup and TCP connect. To me this is pretty fishy. For example if the
negprot fails, this is supposed to be more retryable than a NetBIOS name
lookup
failure? I'd rather think the opposite is true.
Jeremy, this is code from 2002, 389a16d9d533. If you have any comments from
back then, let me know :-)
Volker
commit 09cccee27302044dd2fa3e94c60bf7854c4d2e45
Author: Volker Lendecke v...@samba.org
Date: Mon Dec 20 15:58:55 2010 +0100
s3: Remove some unused variables
---
Summary of changes:
libgpo/gpo_fetch.c |2 +-
source3/auth/auth_domain.c | 12
source3/client/smbspool.c |2 +-
source3/include/proto.h |6 ++
source3/libnet/libnet_join.c|6 +++---
source3/libsmb/cliconnect.c | 19 ---
source3/libsmb/libsmb_server.c |2 +-
source3/libsmb/trusts_util.c|5 ++---
source3/rpc_server/srv_spoolss_nt.c |2 +-
source3/rpcclient/cmd_spoolss.c |2 +-
source3/rpcclient/rpcclient.c |3 +--
source3/torture/locktest2.c |2 +-
source3/torture/torture.c |8 +++-
source3/utils/net_ads.c |2 +-
source3/utils/net_util.c|6 +++---
source3/utils/netlookup.c |3 +--
source3/utils/smbcacls.c|3 +--
source3/utils/smbcquotas.c |3 +--
18 files changed, 32 insertions(+), 56 deletions(-)
Changeset truncated at 500 lines:
diff --git a/libgpo/gpo_fetch.c b/libgpo/gpo_fetch.c
index ad6c2a8..00f9b5c 100644
--- a/libgpo/gpo_fetch.c
+++ b/libgpo/gpo_fetch.c
@@ -150,7 +150,7 @@ static NTSTATUS gpo_connect_server(ADS_STRUCT *ads, struct
loadparm_context *lp_
ads-auth.password,
CLI_FULL_CONNECTION_USE_KERBEROS |
CLI_FULL_CONNECTION_FALLBACK_AFTER_KERBEROS,
- Undefined, NULL);
+ Undefined);
if (!NT_STATUS_IS_OK(result)) {
DEBUG(10,(check_refresh_gpo:
failed to connect: %s\n,
diff --git a/source3/auth/auth_domain.c b/source3/auth/auth_domain.c
index cac482c..0f541cd 100644
--- a/source3/auth/auth_domain.c
+++ b/source3/auth/auth_domain.c
@@ -113,8 +113,7 @@ static NTSTATUS connect_to_domain_password_server(struct
cli_state **cli,
const char *domain,
const char *dc_name,
struct sockaddr_storage *dc_ss,
- struct rpc_pipe_client
**pipe_ret,
- bool *retry)
+ struct rpc_pipe_client
**pipe_ret)
{
NTSTATUS result;
struct rpc_pipe_client *netlogon_pipe = NULL;
@@ -143,9 +142,8 @@ static NTSTATUS connect_to_domain_password_server(struct
cli_state **cli,
}
/* Attempt connection */
- *retry = True;
result = cli_full_connection(cli, global_myname(), dc_name, dc_ss, 0,
- IPC$, IPC, , , , 0, Undefined, retry);
+ IPC$, IPC, , , , 0, Undefined);
if (!NT_STATUS_IS_OK(result)) {
/* map to something more useful */
@@ -267,7 +265,6 @@ static NTSTATUS domain_client_validate(TALLOC_CTX *mem_ctx,
struct rpc_pipe_client *netlogon_pipe = NULL;
NTSTATUS nt_status = NT_STATUS_NO_LOGON_SERVERS;
int i;
- bool retry = True;
/*
* At this point,
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 139e2cb s3-printing: fix printer_list_traverse()
from bfc4fe4 s3: Remove unused retry from cli_start_connection
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 139e2cbb13ca230b65ad4b1f5ff91a795d8cca02
Author: David Disseldorp dd...@suse.de
Date: Mon Dec 20 16:08:02 2010 +0100
s3-printing: fix printer_list_traverse()
The tdb traverse function returns the number of elements traversed, or
less than zero on error, printer_list_traverse() is incorrectly checking
for non-zero return.
Autobuild-User: Andreas Schneider a...@samba.org
Autobuild-Date: Mon Dec 20 18:44:41 CET 2010 on sn-devel-104
---
Summary of changes:
source3/printing/printer_list.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/printing/printer_list.c b/source3/printing/printer_list.c
index edc398b..f3f00f0 100644
--- a/source3/printing/printer_list.c
+++ b/source3/printing/printer_list.c
@@ -293,7 +293,7 @@ static NTSTATUS printer_list_traverse(printer_list_trv_fn_t
*fn,
}
ret = db-traverse(db, fn, private_data);
- if (ret != 0) {
+ if (ret 0) {
return NT_STATUS_UNSUCCESSFUL;
}
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch v3-6-test updated
The branch, v3-6-test has been updated
via eb03383 From metze's work on sparse attributes.
FILE_ATTRIBUTE_SPARSE is valid on get but not on set. (cherry picked from
commit e9f9e803f03be20e5f573484be7b5e7351472786)
via 22d9791 s3-printing: fix printer_list_traverse()
from f1c3c4b s3: Fix bug 7066 -- wbcAuthenticateEx gives unix times
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test
- Log -
commit eb03383e8a51f489f30412f7be23aace4a36c24c
Author: Jeremy Allison j...@samba.org
Date: Mon Dec 20 10:23:27 2010 -0800
From metze's work on sparse attributes. FILE_ATTRIBUTE_SPARSE is valid on
get but not on set.
(cherry picked from commit e9f9e803f03be20e5f573484be7b5e7351472786)
commit 22d979142941ac4135b20c937156a83d8439fb06
Author: David Disseldorp dd...@suse.de
Date: Mon Dec 20 16:08:02 2010 +0100
s3-printing: fix printer_list_traverse()
The tdb traverse function returns the number of elements traversed, or
less than zero on error, printer_list_traverse() is incorrectly checking
for non-zero return.
Autobuild-User: Andreas Schneider a...@samba.org
Autobuild-Date: Mon Dec 20 18:44:41 CET 2010 on sn-devel-104
(cherry picked from commit 139e2cbb13ca230b65ad4b1f5ff91a795d8cca02)
---
Summary of changes:
source3/include/smb.h |3 +--
source3/printing/printer_list.c |2 +-
source3/smbd/dosmode.c |3 ++-
3 files changed, 4 insertions(+), 4 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/include/smb.h b/source3/include/smb.h
index b87caf2..2b397cc 100644
--- a/source3/include/smb.h
+++ b/source3/include/smb.h
@@ -1211,8 +1211,7 @@ struct bitmap {
FILE_ATTRIBUTE_HIDDEN|\
FILE_ATTRIBUTE_SYSTEM|\
FILE_ATTRIBUTE_DIRECTORY|\
- FILE_ATTRIBUTE_ARCHIVE|\
- FILE_ATTRIBUTE_SPARSE)
+ FILE_ATTRIBUTE_ARCHIVE)
/* Flags - combined with attributes. */
#define FILE_FLAG_WRITE_THROUGH0x8000L
diff --git a/source3/printing/printer_list.c b/source3/printing/printer_list.c
index edc398b..f3f00f0 100644
--- a/source3/printing/printer_list.c
+++ b/source3/printing/printer_list.c
@@ -293,7 +293,7 @@ static NTSTATUS printer_list_traverse(printer_list_trv_fn_t
*fn,
}
ret = db-traverse(db, fn, private_data);
- if (ret != 0) {
+ if (ret 0) {
return NT_STATUS_UNSUCCESSFUL;
}
diff --git a/source3/smbd/dosmode.c b/source3/smbd/dosmode.c
index 628f887..2c6dcd0 100644
--- a/source3/smbd/dosmode.c
+++ b/source3/smbd/dosmode.c
@@ -318,7 +318,8 @@ static bool get_ea_dos_attribute(connection_struct *conn,
if (S_ISDIR(smb_fname-st.st_ex_mode)) {
dosattr |= aDIR;
}
- *pattr = (uint32)(dosattr SAMBA_ATTRIBUTES_MASK);
+ /* FILE_ATTRIBUTE_SPARSE is valid on get but not on set. */
+ *pattr = (uint32)(dosattr
(SAMBA_ATTRIBUTES_MASK|FILE_ATTRIBUTE_SPARSE));
DEBUG(8,(get_ea_dos_attribute returning (0x%x), dosattr));
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via e7707d5 From metze's work on sparse attributes.
FILE_ATTRIBUTE_SPARSE is valid on get but not on set.
from 139e2cb s3-printing: fix printer_list_traverse()
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit e7707d5abd7365c89405963c7cac0c0382d9179e
Author: Jeremy Allison j...@samba.org
Date: Mon Dec 20 10:23:27 2010 -0800
From metze's work on sparse attributes. FILE_ATTRIBUTE_SPARSE is valid on
get but not on set.
Autobuild-User: Jeremy Allison j...@samba.org
Autobuild-Date: Mon Dec 20 20:11:22 CET 2010 on sn-devel-104
---
Summary of changes:
source3/include/smb.h |3 +--
source3/smbd/dosmode.c |3 ++-
2 files changed, 3 insertions(+), 3 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/include/smb.h b/source3/include/smb.h
index b87caf2..2b397cc 100644
--- a/source3/include/smb.h
+++ b/source3/include/smb.h
@@ -1211,8 +1211,7 @@ struct bitmap {
FILE_ATTRIBUTE_HIDDEN|\
FILE_ATTRIBUTE_SYSTEM|\
FILE_ATTRIBUTE_DIRECTORY|\
- FILE_ATTRIBUTE_ARCHIVE|\
- FILE_ATTRIBUTE_SPARSE)
+ FILE_ATTRIBUTE_ARCHIVE)
/* Flags - combined with attributes. */
#define FILE_FLAG_WRITE_THROUGH0x8000L
diff --git a/source3/smbd/dosmode.c b/source3/smbd/dosmode.c
index 628f887..2c6dcd0 100644
--- a/source3/smbd/dosmode.c
+++ b/source3/smbd/dosmode.c
@@ -318,7 +318,8 @@ static bool get_ea_dos_attribute(connection_struct *conn,
if (S_ISDIR(smb_fname-st.st_ex_mode)) {
dosattr |= aDIR;
}
- *pattr = (uint32)(dosattr SAMBA_ATTRIBUTES_MASK);
+ /* FILE_ATTRIBUTE_SPARSE is valid on get but not on set. */
+ *pattr = (uint32)(dosattr
(SAMBA_ATTRIBUTES_MASK|FILE_ATTRIBUTE_SPARSE));
DEBUG(8,(get_ea_dos_attribute returning (0x%x), dosattr));
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch v3-6-test updated
The branch, v3-6-test has been updated
via 7dff32f CREATE in a compound CREATE/NOTIFY sequence was being
passed through set_operation_credits() twice (ultimately perhaps because of bug
7331 involving this compound sequence and the need to be ready for any incoming
CANCEL of the NOTIFY). This had the server thinking it had granted more credit
than it actually had, which lead to zero-credits being granted in interim
NOTIFY responses. (cherry picked from commit
0a9b65262c76fdad8331fbc580dbe578a403407b)
from eb03383 From metze's work on sparse attributes.
FILE_ATTRIBUTE_SPARSE is valid on get but not on set. (cherry picked from
commit e9f9e803f03be20e5f573484be7b5e7351472786)
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test
- Log -
commit 7dff32fb7263cfdce015ffd66ea9997685edf817
Author: Ken Harris ken.har...@mathworks.com
Date: Mon Dec 20 10:44:48 2010 -0800
CREATE in a compound CREATE/NOTIFY sequence was being passed through
set_operation_credits()
twice (ultimately perhaps because of bug 7331 involving this compound
sequence and the need
to be ready for any incoming CANCEL of the NOTIFY). This had the server
thinking it had
granted more credit than it actually had, which lead to zero-credits being
granted in interim
NOTIFY responses.
(cherry picked from commit 0a9b65262c76fdad8331fbc580dbe578a403407b)
---
Summary of changes:
source3/smbd/smb2_server.c | 34 +-
1 files changed, 17 insertions(+), 17 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/smbd/smb2_server.c b/source3/smbd/smb2_server.c
index 924e41f..025f403 100644
--- a/source3/smbd/smb2_server.c
+++ b/source3/smbd/smb2_server.c
@@ -1443,23 +1443,6 @@ static NTSTATUS smbd_smb2_request_reply(struct
smbd_smb2_request *req)
req-subreq = NULL;
- smb2_setup_nbt_length(req-out.vector, req-out.vector_count);
-
- /* Set credit for this operation (zero credits if this
- is a final reply for an async operation). */
- smb2_set_operation_credit(req-sconn,
- req-async ? NULL : req-in.vector[i],
- req-out.vector[i]);
-
- if (req-do_signing) {
- NTSTATUS status;
- status = smb2_signing_sign_pdu(req-session-session_key,
- req-out.vector[i], 3);
- if (!NT_STATUS_IS_OK(status)) {
- return status;
- }
- }
-
req-current_idx += 3;
if (req-current_idx req-out.vector_count) {
@@ -1482,6 +1465,23 @@ static NTSTATUS smbd_smb2_request_reply(struct
smbd_smb2_request *req)
return NT_STATUS_OK;
}
+ smb2_setup_nbt_length(req-out.vector, req-out.vector_count);
+
+ /* Set credit for this operation (zero credits if this
+ is a final reply for an async operation). */
+ smb2_set_operation_credit(req-sconn,
+ req-async ? NULL : req-in.vector[i],
+ req-out.vector[i]);
+
+ if (req-do_signing) {
+ NTSTATUS status;
+ status = smb2_signing_sign_pdu(req-session-session_key,
+ req-out.vector[i], 3);
+ if (!NT_STATUS_IS_OK(status)) {
+ return status;
+ }
+ }
+
if (DEBUGLEVEL = 10) {
dbgtext(smbd_smb2_request_reply: sending...\n);
print_req_vectors(req);
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 09aea03 CREATE in a compound CREATE/NOTIFY sequence was being
passed through set_operation_credits() twice (ultimately perhaps because of bug
7331 involving this compound sequence and the need to be ready for any incoming
CANCEL of the NOTIFY). This had the server thinking it had granted more credit
than it actually had, which lead to zero-credits being granted in interim
NOTIFY responses.
from e7707d5 From metze's work on sparse attributes.
FILE_ATTRIBUTE_SPARSE is valid on get but not on set.
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 09aea038139f8717d38f0fdae6be9cf46bd86b15
Author: Ken Harris ken.har...@mathworks.com
Date: Mon Dec 20 10:44:48 2010 -0800
CREATE in a compound CREATE/NOTIFY sequence was being passed through
set_operation_credits()
twice (ultimately perhaps because of bug 7331 involving this compound
sequence and the need
to be ready for any incoming CANCEL of the NOTIFY). This had the server
thinking it had
granted more credit than it actually had, which lead to zero-credits being
granted in interim
NOTIFY responses.
Autobuild-User: Jeremy Allison j...@samba.org
Autobuild-Date: Mon Dec 20 20:59:55 CET 2010 on sn-devel-104
---
Summary of changes:
source3/smbd/smb2_server.c | 34 +-
1 files changed, 17 insertions(+), 17 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/smbd/smb2_server.c b/source3/smbd/smb2_server.c
index 924e41f..025f403 100644
--- a/source3/smbd/smb2_server.c
+++ b/source3/smbd/smb2_server.c
@@ -1443,23 +1443,6 @@ static NTSTATUS smbd_smb2_request_reply(struct
smbd_smb2_request *req)
req-subreq = NULL;
- smb2_setup_nbt_length(req-out.vector, req-out.vector_count);
-
- /* Set credit for this operation (zero credits if this
- is a final reply for an async operation). */
- smb2_set_operation_credit(req-sconn,
- req-async ? NULL : req-in.vector[i],
- req-out.vector[i]);
-
- if (req-do_signing) {
- NTSTATUS status;
- status = smb2_signing_sign_pdu(req-session-session_key,
- req-out.vector[i], 3);
- if (!NT_STATUS_IS_OK(status)) {
- return status;
- }
- }
-
req-current_idx += 3;
if (req-current_idx req-out.vector_count) {
@@ -1482,6 +1465,23 @@ static NTSTATUS smbd_smb2_request_reply(struct
smbd_smb2_request *req)
return NT_STATUS_OK;
}
+ smb2_setup_nbt_length(req-out.vector, req-out.vector_count);
+
+ /* Set credit for this operation (zero credits if this
+ is a final reply for an async operation). */
+ smb2_set_operation_credit(req-sconn,
+ req-async ? NULL : req-in.vector[i],
+ req-out.vector[i]);
+
+ if (req-do_signing) {
+ NTSTATUS status;
+ status = smb2_signing_sign_pdu(req-session-session_key,
+ req-out.vector[i], 3);
+ if (!NT_STATUS_IS_OK(status)) {
+ return status;
+ }
+ }
+
if (DEBUGLEVEL = 10) {
dbgtext(smbd_smb2_request_reply: sending...\n);
print_req_vectors(req);
--
Samba Shared Repository
[SCM] Samba Website Repository - branch master updated
The branch, master has been updated via d2e2b00 Update addresses for Conservancy from 932dd98 Updated entries for PrimaStasys. http://gitweb.samba.org/?p=samba-web.git;a=shortlog;h=master - Log - commit d2e2b007ca9fa7c103ba24d0a0a75702831c2912 Author: Jim McDonough j...@samba.org Date: Mon Dec 20 15:25:14 2010 -0500 Update addresses for Conservancy --- Summary of changes: donations.html |8 1 files changed, 4 insertions(+), 4 deletions(-) Changeset truncated at 500 lines: diff --git a/donations.html b/donations.html index bc26511..bdeb9dd 100755 --- a/donations.html +++ b/donations.html @@ -32,9 +32,9 @@ src=https://www.paypal.com/en_US/i/btn/x-click-but21.gif; border=0 name=subm pre Samba Team - c/o The Software Freedom Conservancy - 1995 BROADWAY FL 17 - NEW YORK NY 10023-5882 + c/o Software Freedom Conservancy, Inc. + 137 Montague St Ste 380 + Brooklyn, NY 11201-3548 /pre br @@ -47,7 +47,7 @@ the USA, so donations in the USA may be tax-deductible./p pIf you would like to make a larger corporate donation then we would certainly like to discuss that. Please send a email to a -href=mailto:conserva...@softwarefreedom.org;conserva...@softwarefreedom.org/a +href=mailto:donat...@sfconservancy.org;donat...@sfconservancy.org/a or talk to any Samba Team member./p h3Why do we need money?/h3 -- Samba Website Repository
[SCM] Samba Shared Repository - branch v3-6-test updated
The branch, v3-6-test has been updated via 132c57d Added call out to a Linux-compatible fallocate() when we need to extend a file allocation extent without changing end-of-file size. (cherry picked from commit 00d2d16262909fde2c144a504d7d554767b7fd45) from 7dff32f CREATE in a compound CREATE/NOTIFY sequence was being passed through set_operation_credits() twice (ultimately perhaps because of bug 7331 involving this compound sequence and the need to be ready for any incoming CANCEL of the NOTIFY). This had the server thinking it had granted more credit than it actually had, which lead to zero-credits being granted in interim NOTIFY responses. (cherry picked from commit 0a9b65262c76fdad8331fbc580dbe578a403407b) http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test - Log - commit 132c57df21571ad0400e3d26e5e0a0db3a9d9fd7 Author: Jeremy Allison j...@samba.org Date: Mon Dec 20 16:53:16 2010 -0800 Added call out to a Linux-compatible fallocate() when we need to extend a file allocation extent without changing end-of-file size. (cherry picked from commit 00d2d16262909fde2c144a504d7d554767b7fd45) --- Summary of changes: source3/configure.in | 35 +++ source3/include/proto.h |1 + source3/lib/system.c | 35 +++ source3/modules/vfs_default.c |5 +++-- source3/smbd/vfs.c| 13 - 5 files changed, 86 insertions(+), 3 deletions(-) Changeset truncated at 500 lines: diff --git a/source3/configure.in b/source3/configure.in index 5f81a19..c772f58 100644 --- a/source3/configure.in +++ b/source3/configure.in @@ -741,6 +741,7 @@ AC_CHECK_HEADERS(sys/syslog.h syslog.h) AC_CHECK_HEADERS(langinfo.h locale.h) AC_CHECK_HEADERS(xfs/libxfs.h) AC_CHECK_HEADERS(netgroup.h) +AC_CHECK_HEADERS(linux/falloc.h) AC_CHECK_HEADERS(rpcsvc/yp_prot.h,,,[[ #if HAVE_RPC_RPC_H @@ -1095,6 +1096,7 @@ AC_CHECK_FUNCS(sigprocmask sigblock sigaction sigset innetgr setnetgrent getnetg AC_CHECK_FUNCS(initgroups select poll rdchk getgrnam getgrent pathconf) AC_CHECK_FUNCS(setpriv setgidx setuidx setgroups sysconf stat64 fstat64) AC_CHECK_FUNCS(lstat64 fopen64 atexit grantpt lseek64 ftruncate64 posix_fallocate posix_fallocate64) +AC_CHECK_FUNCS(fallocate fallocate64) AC_CHECK_FUNCS(fseek64 fseeko64 ftell64 ftello64 setluid getpwanam) AC_CHECK_FUNCS(opendir64 readdir64 seekdir64 telldir64 rewinddir64 closedir64) AC_CHECK_FUNCS(getpwent_r) @@ -2563,6 +2565,39 @@ fi fi # end utmp details +AC_CACHE_CHECK([for linux fallocate],samba_cv_HAVE_LINUX_FALLOCATE,[ +AC_TRY_COMPILE([ +#if defined(HAVE_UNISTD_H) +#include unistd.h +#endif +#include sys/types.h +#define _GNU_SOURCE +#include fcntl.h +#if defined(HAVE_LINUX_FALLOC_H) +#include linux/falloc.h +#endif], +[int ret = fallocate(0, FALLOC_FL_KEEP_SIZE, 0, 10);], +samba_cv_HAVE_LINUX_FALLOCATE=yes,samba_cv_HAVE_LINUX_FALLOCATE=no)]) +if test x$samba_cv_HAVE_LINUX_FALLOCATE = xyes test x$ac_cv_func_fallocate = xyes; then +AC_DEFINE(HAVE_LINUX_FALLOCATE,1,[Whether the Linux 'fallocate' function is available]) +fi + +AC_CACHE_CHECK([for linux fallocate64],samba_cv_HAVE_LINUX_FALLOCATE64,[ +AC_TRY_COMPILE([ +#if defined(HAVE_UNISTD_H) +#include unistd.h +#endif +#include sys/types.h +#define _GNU_SOURCE +#include fcntl.h +#if defined(HAVE_LINUX_FALLOC_H) +#include linux/falloc.h +#endif], +[int ret = fallocate64(0, FALLOC_FL_KEEP_SIZE, 0, 10);], +samba_cv_HAVE_LINUX_FALLOCATE64=yes,samba_cv_HAVE_LINUX_FALLOCATE64=no)]) +if test x$samba_cv_HAVE_LINUX_FALLOCATE64 = xyes test x$ac_cv_func_fallocate64 = xyes; then +AC_DEFINE(HAVE_LINUX_FALLOCATE64,1,[Whether the Linux 'fallocate64' function is available]) +fi ICONV_LOOK_DIRS=/usr /usr/local /sw /opt AC_ARG_WITH(libiconv, diff --git a/source3/include/proto.h b/source3/include/proto.h index e26f43e..8b2ba95 100644 --- a/source3/include/proto.h +++ b/source3/include/proto.h @@ -888,6 +888,7 @@ int sys_lstat(const char *fname,SMB_STRUCT_STAT *sbuf, bool fake_dir_create_times); int sys_ftruncate(int fd, SMB_OFF_T offset); int sys_posix_fallocate(int fd, SMB_OFF_T offset, SMB_OFF_T len); +int sys_fallocate(int fd, enum vfs_fallocate_mode mode, SMB_OFF_T offset, SMB_OFF_T len); SMB_OFF_T sys_lseek(int fd, SMB_OFF_T offset, int whence); int sys_fseek(FILE *fp, SMB_OFF_T offset, int whence); SMB_OFF_T sys_ftell(FILE *fp); diff --git a/source3/lib/system.c b/source3/lib/system.c index 1c00ad8..d5b833c 100644 --- a/source3/lib/system.c +++ b/source3/lib/system.c @@ -695,6 +695,41 @@ int sys_posix_fallocate(int fd, SMB_OFF_T offset, SMB_OFF_T len) } /*** + An fallocate() function that matches the semantics of the Linux one.
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated via 8998f4b Added call out to a Linux-compatible fallocate() when we need to extend a file allocation extent without changing end-of-file size. from 09aea03 CREATE in a compound CREATE/NOTIFY sequence was being passed through set_operation_credits() twice (ultimately perhaps because of bug 7331 involving this compound sequence and the need to be ready for any incoming CANCEL of the NOTIFY). This had the server thinking it had granted more credit than it actually had, which lead to zero-credits being granted in interim NOTIFY responses. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 8998f4b01310e4b45e75d8d5f3260b5ba5c1cdf9 Author: Jeremy Allison j...@samba.org Date: Mon Dec 20 16:53:16 2010 -0800 Added call out to a Linux-compatible fallocate() when we need to extend a file allocation extent without changing end-of-file size. Autobuild-User: Jeremy Allison j...@samba.org Autobuild-Date: Tue Dec 21 02:41:24 CET 2010 on sn-devel-104 --- Summary of changes: source3/configure.in | 35 +++ source3/include/proto.h |1 + source3/lib/system.c | 35 +++ source3/modules/vfs_default.c |5 +++-- source3/smbd/vfs.c| 13 - 5 files changed, 86 insertions(+), 3 deletions(-) Changeset truncated at 500 lines: diff --git a/source3/configure.in b/source3/configure.in index b43d0b3..ed99b17 100644 --- a/source3/configure.in +++ b/source3/configure.in @@ -741,6 +741,7 @@ AC_CHECK_HEADERS(sys/syslog.h syslog.h) AC_CHECK_HEADERS(langinfo.h locale.h) AC_CHECK_HEADERS(xfs/libxfs.h) AC_CHECK_HEADERS(netgroup.h) +AC_CHECK_HEADERS(linux/falloc.h) AC_CHECK_HEADERS(rpcsvc/yp_prot.h,,,[[ #if HAVE_RPC_RPC_H @@ -1095,6 +1096,7 @@ AC_CHECK_FUNCS(sigprocmask sigblock sigaction sigset innetgr setnetgrent getnetg AC_CHECK_FUNCS(initgroups select poll rdchk getgrnam getgrent pathconf) AC_CHECK_FUNCS(setpriv setgidx setuidx setgroups sysconf stat64 fstat64) AC_CHECK_FUNCS(lstat64 fopen64 atexit grantpt lseek64 ftruncate64 posix_fallocate posix_fallocate64) +AC_CHECK_FUNCS(fallocate fallocate64) AC_CHECK_FUNCS(fseek64 fseeko64 ftell64 ftello64 setluid getpwanam) AC_CHECK_FUNCS(opendir64 readdir64 seekdir64 telldir64 rewinddir64 closedir64) AC_CHECK_FUNCS(getpwent_r) @@ -2563,6 +2565,39 @@ fi fi # end utmp details +AC_CACHE_CHECK([for linux fallocate],samba_cv_HAVE_LINUX_FALLOCATE,[ +AC_TRY_COMPILE([ +#if defined(HAVE_UNISTD_H) +#include unistd.h +#endif +#include sys/types.h +#define _GNU_SOURCE +#include fcntl.h +#if defined(HAVE_LINUX_FALLOC_H) +#include linux/falloc.h +#endif], +[int ret = fallocate(0, FALLOC_FL_KEEP_SIZE, 0, 10);], +samba_cv_HAVE_LINUX_FALLOCATE=yes,samba_cv_HAVE_LINUX_FALLOCATE=no)]) +if test x$samba_cv_HAVE_LINUX_FALLOCATE = xyes test x$ac_cv_func_fallocate = xyes; then +AC_DEFINE(HAVE_LINUX_FALLOCATE,1,[Whether the Linux 'fallocate' function is available]) +fi + +AC_CACHE_CHECK([for linux fallocate64],samba_cv_HAVE_LINUX_FALLOCATE64,[ +AC_TRY_COMPILE([ +#if defined(HAVE_UNISTD_H) +#include unistd.h +#endif +#include sys/types.h +#define _GNU_SOURCE +#include fcntl.h +#if defined(HAVE_LINUX_FALLOC_H) +#include linux/falloc.h +#endif], +[int ret = fallocate64(0, FALLOC_FL_KEEP_SIZE, 0, 10);], +samba_cv_HAVE_LINUX_FALLOCATE64=yes,samba_cv_HAVE_LINUX_FALLOCATE64=no)]) +if test x$samba_cv_HAVE_LINUX_FALLOCATE64 = xyes test x$ac_cv_func_fallocate64 = xyes; then +AC_DEFINE(HAVE_LINUX_FALLOCATE64,1,[Whether the Linux 'fallocate64' function is available]) +fi ICONV_LOOK_DIRS=/usr /usr/local /sw /opt AC_ARG_WITH(libiconv, diff --git a/source3/include/proto.h b/source3/include/proto.h index dabb315..566b3f3 100644 --- a/source3/include/proto.h +++ b/source3/include/proto.h @@ -888,6 +888,7 @@ int sys_lstat(const char *fname,SMB_STRUCT_STAT *sbuf, bool fake_dir_create_times); int sys_ftruncate(int fd, SMB_OFF_T offset); int sys_posix_fallocate(int fd, SMB_OFF_T offset, SMB_OFF_T len); +int sys_fallocate(int fd, enum vfs_fallocate_mode mode, SMB_OFF_T offset, SMB_OFF_T len); SMB_OFF_T sys_lseek(int fd, SMB_OFF_T offset, int whence); int sys_fseek(FILE *fp, SMB_OFF_T offset, int whence); SMB_OFF_T sys_ftell(FILE *fp); diff --git a/source3/lib/system.c b/source3/lib/system.c index 02322b7..4cf6a29 100644 --- a/source3/lib/system.c +++ b/source3/lib/system.c @@ -695,6 +695,41 @@ int sys_posix_fallocate(int fd, SMB_OFF_T offset, SMB_OFF_T len) } /*** + An fallocate() function that matches the semantics of the Linux one. +/ + +#ifdef HAVE_LINUX_FALLOC_H +#include linux/falloc.h
[SCM] Samba Shared Repository - branch v3-6-test updated
The branch, v3-6-test has been updated
via a2956e7 Keep track of the sparse status of an open file handle.
Allows bypass of strict allocation on sparse files. Files opened as POSIX opens
are always sparse. (cherry picked from commit
3db2614ccd1948792cc403f0302c7516319461b2)
from 132c57d Added call out to a Linux-compatible fallocate() when we
need to extend a file allocation extent without changing end-of-file size.
(cherry picked from commit 00d2d16262909fde2c144a504d7d554767b7fd45)
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=v3-6-test
- Log -
commit a2956e7d2b66a828d7355841bcb3edfe188a12d3
Author: Jeremy Allison j...@samba.org
Date: Mon Dec 20 17:58:33 2010 -0800
Keep track of the sparse status of an open file handle. Allows bypass of
strict allocation on sparse files. Files opened as POSIX opens are always
sparse.
(cherry picked from commit 3db2614ccd1948792cc403f0302c7516319461b2)
---
Summary of changes:
source3/include/smb.h |1 +
source3/modules/vfs_default.c |2 +-
source3/smbd/dosmode.c|2 ++
source3/smbd/fileio.c |3 ++-
source3/smbd/open.c |9 +
5 files changed, 15 insertions(+), 2 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/include/smb.h b/source3/include/smb.h
index 2b397cc..8d12fb9 100644
--- a/source3/include/smb.h
+++ b/source3/include/smb.h
@@ -344,6 +344,7 @@ typedef struct files_struct {
bool initial_delete_on_close; /* Only set at NTCreateX if file was
created. */
bool delete_on_close;
bool posix_open;
+ bool is_sparse;
struct smb_filename *fsp_name;
struct vfs_fsp_data *vfs_extension;
diff --git a/source3/modules/vfs_default.c b/source3/modules/vfs_default.c
index 54f38c3..9cca349 100644
--- a/source3/modules/vfs_default.c
+++ b/source3/modules/vfs_default.c
@@ -895,7 +895,7 @@ static int vfswrap_ftruncate(vfs_handle_struct *handle,
files_struct *fsp, SMB_O
START_PROFILE(syscall_ftruncate);
- if (lp_strict_allocate(SNUM(fsp-conn))) {
+ if (lp_strict_allocate(SNUM(fsp-conn)) !fsp-is_sparse) {
result = strict_allocate_ftruncate(handle, fsp, len);
END_PROFILE(syscall_ftruncate);
return result;
diff --git a/source3/smbd/dosmode.c b/source3/smbd/dosmode.c
index 2c6dcd0..cf95348 100644
--- a/source3/smbd/dosmode.c
+++ b/source3/smbd/dosmode.c
@@ -911,6 +911,8 @@ NTSTATUS file_set_sparse(connection_struct *conn,
FILE_NOTIFY_CHANGE_ATTRIBUTES,
fsp-fsp_name-base_name);
+ fsp-is_sparse = sparse;
+
return NT_STATUS_OK;
}
diff --git a/source3/smbd/fileio.c b/source3/smbd/fileio.c
index aec6554..da40013 100644
--- a/source3/smbd/fileio.c
+++ b/source3/smbd/fileio.c
@@ -128,7 +128,8 @@ static ssize_t real_write_file(struct smb_request *req,
ret = vfs_write_data(req, fsp, data, n);
} else {
fsp-fh-pos = pos;
- if (pos lp_strict_allocate(SNUM(fsp-conn))) {
+ if (pos lp_strict_allocate(SNUM(fsp-conn)
+ !fsp-is_sparse)) {
if (vfs_fill_sparse(fsp, pos) == -1) {
return -1;
}
diff --git a/source3/smbd/open.c b/source3/smbd/open.c
index 80756d6..32a08b5 100644
--- a/source3/smbd/open.c
+++ b/source3/smbd/open.c
@@ -2221,6 +2221,15 @@ static NTSTATUS open_file_ntcreate(connection_struct
*conn,
}
}
+ /* Determine sparse flag. */
+ if (posix_open) {
+ /* POSIX opens are sparse by default. */
+ fsp-is_sparse = true;
+ } else {
+ fsp-is_sparse = (file_existed
+ (existing_dos_attributes FILE_ATTRIBUTE_SPARSE));
+ }
+
/*
* Take care of inherited ACLs on created files - if default ACL not
* selected.
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 4820c97 dns: fixed the padding for dnsp_name fields in LDAP
via 049a16c dns: auto-calculate the wDataLength field in DNS records
via 8c04657 s4-dns: fixed a crash bug in dlz_bind9 code
via b9a2852 dnsp: fixed parsing of dns_name structures
from 8998f4b Added call out to a Linux-compatible fallocate() when we
need to extend a file allocation extent without changing end-of-file size.
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 4820c97e9ea00e73f3188f9834a03913ed74df80
Author: Andrew Tridgell tri...@samba.org
Date: Tue Dec 21 11:59:54 2010 +1100
dns: fixed the padding for dnsp_name fields in LDAP
all names are NUL terminated, but may have additional padding as well
Autobuild-User: Andrew Tridgell tri...@samba.org
Autobuild-Date: Tue Dec 21 03:26:26 CET 2010 on sn-devel-104
commit 049a16c8ef475f6b327292d231022fd4c5aaddf1
Author: Andrew Tridgell tri...@samba.org
Date: Tue Dec 21 11:59:05 2010 +1100
dns: auto-calculate the wDataLength field in DNS records
we need this for creating new records
commit 8c04657600cd6702dbfc66744fe2268c59ebea0b
Author: Andrew Tridgell tri...@samba.org
Date: Tue Dec 21 11:57:50 2010 +1100
s4-dns: fixed a crash bug in dlz_bind9 code
we need to keep el_ctx for the next part of the loop
commit b9a2852fdd68a0691ff567557824be44bb08b27a
Author: Andrew Tridgell tri...@samba.org
Date: Wed Dec 15 23:52:32 2010 +1100
dnsp: fixed parsing of dns_name structures
its not a pad byte, its a trailing zero
---
Summary of changes:
librpc/idl/dnsp.idl|4 ++--
librpc/ndr/ndr_dnsp.c | 25 -
source4/dns_server/dlz_bind9.c |2 --
3 files changed, 22 insertions(+), 9 deletions(-)
Changeset truncated at 500 lines:
diff --git a/librpc/idl/dnsp.idl b/librpc/idl/dnsp.idl
index 905e420..eed0c47 100644
--- a/librpc/idl/dnsp.idl
+++ b/librpc/idl/dnsp.idl
@@ -92,7 +92,7 @@ interface dnsp
dnsp_name nameTarget;
} dnsp_srv;
- typedef [nodiscriminant] union {
+ typedef [nodiscriminant,gensize] union {
[case(DNS_TYPE_A)] [flag(NDR_BIG_ENDIAN)] ipv4address ipv4;
[case(DNS_TYPE_NS)] dnsp_name ns;
[case(DNS_TYPE_CNAME)] dnsp_name cname;
@@ -109,7 +109,7 @@ interface dnsp
/* this is the format for the dnsRecord attribute in the DNS
partitions in AD */
typedef [public] struct {
- uint16 wDataLength;
+ [value(ndr_size_dnsRecordData(data,wType,ndr-flags))] uint16
wDataLength;
dns_record_type wType;
uint32 dwFlags;
uint32 dwSerial;
diff --git a/librpc/ndr/ndr_dnsp.c b/librpc/ndr/ndr_dnsp.c
index 256638a..ae78425 100644
--- a/librpc/ndr/ndr_dnsp.c
+++ b/librpc/ndr/ndr_dnsp.c
@@ -36,14 +36,16 @@ _PUBLIC_ void ndr_print_dnsp_name(struct ndr_print *ndr,
const char *name,
*/
_PUBLIC_ enum ndr_err_code ndr_pull_dnsp_name(struct ndr_pull *ndr, int
ndr_flags, const char **name)
{
- uint8_t len, count;
+ uint8_t len, count, termination;
int i;
- uint32_t total_len;
+ uint32_t total_len, raw_offset;
char *ret;
NDR_CHECK(ndr_pull_uint8(ndr, ndr_flags, len));
NDR_CHECK(ndr_pull_uint8(ndr, ndr_flags, count));
+ raw_offset = ndr-offset;
+
ret = talloc_strdup(ndr-current_mem_ctx, );
if (!ret) {
return ndr_pull_error(ndr, NDR_ERR_ALLOC, Failed to pull
dnsp);
@@ -68,19 +70,32 @@ _PUBLIC_ enum ndr_err_code ndr_pull_dnsp_name(struct
ndr_pull *ndr, int ndr_flag
ret[newlen-1] = 0;
total_len = newlen;
}
+ NDR_CHECK(ndr_pull_uint8(ndr, ndr_flags, termination));
+ if (termination != 0) {
+ return ndr_pull_error(ndr, NDR_ERR_ALLOC, Failed to pull dnsp
- not NUL terminated);
+ }
+ if (ndr-offset raw_offset + len) {
+ return ndr_pull_error(ndr, NDR_ERR_ALLOC, Failed to pull dnsp
- overrun by %u bytes,
+ ndr-offset - (raw_offset + len));
+ }
+ /* there could be additional pad bytes */
+ while (ndr-offset raw_offset + len) {
+ uint8_t pad;
+ NDR_CHECK(ndr_pull_uint8(ndr, ndr_flags, pad));
+ }
(*name) = ret;
- NDR_PULL_ALIGN(ndr, 2);
return NDR_ERR_SUCCESS;
}
enum ndr_err_code ndr_push_dnsp_name(struct ndr_push *ndr, int ndr_flags,
const char *name)
{
int count, total_len, i;
+
/* count the dots */
for (count=i=0; name[i]; i++) {
if (name[i] == '.') count++;
}
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated
via 0a5f4f5 Keep track of the sparse status of an open file handle.
Allows bypass of strict allocation on sparse files. Files opened as POSIX opens
are always sparse.
from 4820c97 dns: fixed the padding for dnsp_name fields in LDAP
http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master
- Log -
commit 0a5f4f523fe3dcb90033ee53c838ad6030f608b4
Author: Jeremy Allison j...@samba.org
Date: Mon Dec 20 17:58:33 2010 -0800
Keep track of the sparse status of an open file handle. Allows bypass of
strict allocation on sparse files. Files opened as POSIX opens are always
sparse.
Autobuild-User: Jeremy Allison j...@samba.org
Autobuild-Date: Tue Dec 21 04:12:22 CET 2010 on sn-devel-104
---
Summary of changes:
source3/include/smb.h |1 +
source3/modules/vfs_default.c |2 +-
source3/smbd/dosmode.c|2 ++
source3/smbd/fileio.c |3 ++-
source3/smbd/open.c |9 +
5 files changed, 15 insertions(+), 2 deletions(-)
Changeset truncated at 500 lines:
diff --git a/source3/include/smb.h b/source3/include/smb.h
index 2b397cc..8d12fb9 100644
--- a/source3/include/smb.h
+++ b/source3/include/smb.h
@@ -344,6 +344,7 @@ typedef struct files_struct {
bool initial_delete_on_close; /* Only set at NTCreateX if file was
created. */
bool delete_on_close;
bool posix_open;
+ bool is_sparse;
struct smb_filename *fsp_name;
struct vfs_fsp_data *vfs_extension;
diff --git a/source3/modules/vfs_default.c b/source3/modules/vfs_default.c
index 54f38c3..9cca349 100644
--- a/source3/modules/vfs_default.c
+++ b/source3/modules/vfs_default.c
@@ -895,7 +895,7 @@ static int vfswrap_ftruncate(vfs_handle_struct *handle,
files_struct *fsp, SMB_O
START_PROFILE(syscall_ftruncate);
- if (lp_strict_allocate(SNUM(fsp-conn))) {
+ if (lp_strict_allocate(SNUM(fsp-conn)) !fsp-is_sparse) {
result = strict_allocate_ftruncate(handle, fsp, len);
END_PROFILE(syscall_ftruncate);
return result;
diff --git a/source3/smbd/dosmode.c b/source3/smbd/dosmode.c
index 2c6dcd0..cf95348 100644
--- a/source3/smbd/dosmode.c
+++ b/source3/smbd/dosmode.c
@@ -911,6 +911,8 @@ NTSTATUS file_set_sparse(connection_struct *conn,
FILE_NOTIFY_CHANGE_ATTRIBUTES,
fsp-fsp_name-base_name);
+ fsp-is_sparse = sparse;
+
return NT_STATUS_OK;
}
diff --git a/source3/smbd/fileio.c b/source3/smbd/fileio.c
index aec6554..da40013 100644
--- a/source3/smbd/fileio.c
+++ b/source3/smbd/fileio.c
@@ -128,7 +128,8 @@ static ssize_t real_write_file(struct smb_request *req,
ret = vfs_write_data(req, fsp, data, n);
} else {
fsp-fh-pos = pos;
- if (pos lp_strict_allocate(SNUM(fsp-conn))) {
+ if (pos lp_strict_allocate(SNUM(fsp-conn)
+ !fsp-is_sparse)) {
if (vfs_fill_sparse(fsp, pos) == -1) {
return -1;
}
diff --git a/source3/smbd/open.c b/source3/smbd/open.c
index 80756d6..32a08b5 100644
--- a/source3/smbd/open.c
+++ b/source3/smbd/open.c
@@ -2221,6 +2221,15 @@ static NTSTATUS open_file_ntcreate(connection_struct
*conn,
}
}
+ /* Determine sparse flag. */
+ if (posix_open) {
+ /* POSIX opens are sparse by default. */
+ fsp-is_sparse = true;
+ } else {
+ fsp-is_sparse = (file_existed
+ (existing_dos_attributes FILE_ATTRIBUTE_SPARSE));
+ }
+
/*
* Take care of inherited ACLs on created files - if default ACL not
* selected.
--
Samba Shared Repository
[SCM] Samba Shared Repository - branch master updated
The branch, master has been updated via 446f8a1 s4-auth Ensure that we always copy across domain groups via 6f7423c s4-auth Remove duplicate copies of session_info creation code via 1961d7a s4-auth rework session_info handling not to require an auth context via 94a59b7 s4-auth Remove event context from privilage database handling via becaa18 s4-auth Remove obsolete comment via 912faf1 s4:dsdb/schema/schema_* - adaptions needed for removed const on OIDs via ef618f5 s4:lib/ldb-samba/ldif_handlers.c - adaption needed for removed const on OIDs via 13fa674 s3/s4:auth SPNEGO - adaptions for the removed const from OIDs via 464b8fa librpc/ndr/ndr_*.c - remove const from OIDs via 3b591ca lib/util/asn1.c - remove the const specifier from OID via 32bae10 s4:dsdb/common/util.c - remove unused variable ndr_err via 89522ea s4:auth/gensec/spnego.c - remove unused variable principal from 0a5f4f5 Keep track of the sparse status of an open file handle. Allows bypass of strict allocation on sparse files. Files opened as POSIX opens are always sparse. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 446f8a163cfdcfb2c4bb2b8b8adc720bf96c05a5 Author: Andrew Bartlett abart...@samba.org Date: Tue Dec 21 14:08:34 2010 +1100 s4-auth Ensure that we always copy across domain groups Even if we can't calculate the local groups (because we don't have a local SAM to do it with) we still need to include the domain groups in the session_info token. Andrew Bartlett Autobuild-User: Andrew Bartlett abart...@samba.org Autobuild-Date: Tue Dec 21 05:56:22 CET 2010 on sn-devel-104 commit 6f7423c7f1cc3a4596a955a90f315ffbf1025c3b Author: Andrew Bartlett abart...@samba.org Date: Tue Dec 21 11:43:04 2010 +1100 s4-auth Remove duplicate copies of session_info creation code We now just do or do not call into LDB based on some flags. This means there may be some more link time dependencies, but we seem to deal with those better now. Andrew Bartlett commit 1961d7a4119200b8a4ad7b0207e0cdcf2e10d3f8 Author: Andrew Bartlett abart...@samba.org Date: Tue Dec 21 10:19:53 2010 +1100 s4-auth rework session_info handling not to require an auth context This reverts a previous move to have this based around the auth subsystem, which just spread auth deps all over unrelated code. Andrew Bartlett commit 94a59b781ccc5a552a9141484740255977db4637 Author: Andrew Bartlett abart...@samba.org Date: Mon Dec 20 21:26:35 2010 +1100 s4-auth Remove event context from privilage database handling These local TDB operations can quite safely be handled in a new/nested event context, rather than using the main event context. Andrew Bartlett commit becaa18a46f4ee14d8617c22e78da463fda823b2 Author: Andrew Bartlett abart...@samba.org Date: Mon Dec 20 16:07:40 2010 +1100 s4-auth Remove obsolete comment The code that this referred to went away in September with 7dbfeb0dc040889244a1110940af2d070f823374 Andrew Bartlett commit 912faf1b080736a4855e3ea99b49f0ccbf76af33 Author: Matthias Dieter Wallnöfer m...@samba.org Date: Sun Dec 19 16:37:04 2010 +0100 s4:dsdb/schema/schema_* - adaptions needed for removed const on OIDs commit ef618f577870f28b8efd472a87d4dcf094d67d29 Author: Matthias Dieter Wallnöfer m...@samba.org Date: Sun Dec 19 16:36:16 2010 +0100 s4:lib/ldb-samba/ldif_handlers.c - adaption needed for removed const on OIDs commit 13fa6743d86ef6e51b3243cf3045242850358b43 Author: Matthias Dieter Wallnöfer m...@samba.org Date: Wed Dec 15 17:17:09 2010 +0100 s3/s4:auth SPNEGO - adaptions for the removed const from OIDs This is needed in order to suppress warnings. commit 464b8fa3173de7cc801195a28b84786cb1c63833 Author: Matthias Dieter Wallnöfer m...@samba.org Date: Wed Dec 15 17:21:04 2010 +0100 librpc/ndr/ndr_*.c - remove const from OIDs commit 3b591caed00790c5d21b8774c7af87357c329d1c Author: Matthias Dieter Wallnöfer m...@samba.org Date: Wed Dec 15 17:02:49 2010 +0100 lib/util/asn1.c - remove the const specifier from OID There is no reason to have it const since it's an allocated thing. commit 32bae1051272e805e290e49dffb25e4a635ffdfb Author: Matthias Dieter Wallnöfer m...@samba.org Date: Sun Dec 19 16:30:10 2010 +0100 s4:dsdb/common/util.c - remove unused variable ndr_err commit 89522ea5b17dea91608ca49c58bfe7d802e0638e Author: Matthias Dieter Wallnöfer m...@samba.org Date: Wed Dec 15 17:43:45 2010 +0100 s4:auth/gensec/spnego.c - remove unused variable principal --- Summary of changes: lib/util/asn1.c | 17 ++-- lib/util/asn1.h
