summaryrefslogtreecommitdiff
path: root/source3
AgeCommit message (Collapse)AuthorFilesLines
2007-10-10r7454: couple of winbindd fixesGerald Carter4-9/+17
* make sure to use our domain as the account name in the net_req_auth2() request when running on a Samba DC * make sure to lookup the correct domain (not default to ours) when getting an async getpwnam() call (This used to be commit c9c3e3c122a6a04847c448d298b6f1adb4f35a60)
2007-10-10r7452: Set sparse flag if needed. Based on code from jpeach@sgi.com.Jeremy Allison1-5/+12
Fixes bug #2774. Jeremy. (This used to be commit 5d366047debed68f36d44d34233ba4670e412d1e)
2007-10-10r7450: fix my bone head mistake with ntlm authentcation and 'map to guest = ↵Gerald Carter2-9/+12
bad uid'; make sure the authentication suceeds (This used to be commit 5de1ffce2f2a0a340f6591939b8f63a3d96a627e)
2007-10-10r7447: Sync up a little between HEAD and 3.0 in preparation for movingJeremy Allison1-5/+30
the AIO code across. Jeremy. (This used to be commit e9604e65f4337a4163398f1f2c71e13da7993ed4)
2007-10-10r7440: * merge registry server changes from trunk (so far) for moreGerald Carter13-19/+142
printmig.exe work * merge the sys_select_signal(char c) change from trunk in order to keeo the winbind code in sync (This used to be commit a112c5570a7f8ddddde1af0fa665f40a6067e8cf)
2007-10-10r7423: Fix the build on sol10Volker Lendecke1-1/+1
(This used to be commit e8fde1bd92e8b19c8b5ce22f92cb048de773be1c)
2007-10-10r7415: * big change -- volker's new async winbindd from trunkGerald Carter51-3157/+7539
(This used to be commit a0ac9a8ffd4af31a0ebc423b4acbb2f043d865b8)
2007-10-10r7412: Now we're not memset'ing ensure we're valgrind clean.Jeremy Allison1-0/+1
Jeremy. (This used to be commit 0d303ab2f30f6047fc3f2f6fc4916e93704959d2)
2007-10-10r7398: commiting abartlet's patch for kerberos authentication when using a ↵Gerald Carter2-6/+9
keytab and security != ads (This used to be commit 3faaa5c3eb3b2057984586e069a47cb210c99140)
2007-10-10r7396: cleaning up vestiges of winbind local accounts codeGerald Carter1-4/+0
(This used to be commit c57c23cf52184171ac588c2919dc4ee4e68f2431)
2007-10-10r7395: * new feature 'map to guest = bad uid' (based on patch fromGerald Carter4-34/+55
aruna.prabakar@hp.com). This re-enables the Samba 2.2 behavior where a user that was successfully authenticated by a remote DC would be mapped to the guest account if there was not existing UNIX account for that user and we could not create one. (This used to be commit b7455fbf81f4e47c087c861f70d492a328730a9b)
2007-10-10r7391: - Added client-support for various lsa_query_trust_dom_info-calls and aGünther Deschner7-18/+1146
rpcclient-tester for some info-levels. Jerry, I tried to adopt to prs_pointer() where possible and to not interfere with your work for usrmgr. - Add "net rpc trustdom vampire"-tool. This allows to retrieve Interdomain Trust(ed)-Relationships from NT4-Servers including cleartext-passwords (still stored in the local secrets.tdb). The net-hook was done in cooperation with Lars Mueller <lmuelle@suse.de>. To vampire trusted domains simply call: net rpc trustdom vampire -S nt4dc -Uadmin%pass Guenther (This used to be commit 512585293963a1737f831af697ea1dc092d63cb0)
2007-10-10r7390: Revision 7371 did not help :-(Volker Lendecke1-4/+0
(This used to be commit a405e7f479cb71be5634b79f2d7847fcf018f516)
2007-10-10r7385: Rewrite the RPC bind parsing functions to follow the spec. I haven't yetJeremy Allison5-145/+157
tested this so I may have screwed this up - however it now follows the DCE spec. valgrinded tests to follow.... Jeremy. (This used to be commit 877e0a61f5821c89149b1403d08675dd7db8039e)
2007-10-10r7374: BUG 1261: remove unusable libbiconv after discussion with Timur ↵Gerald Carter1-11/+0
(FreeBSD Samba maintainer) (This used to be commit 2ead684635a21d737a385d38e0da0b6c53cb1f0d)
2007-10-10r7372: abartet's patch for BUG 2391 (segv caused by free a static pointer)Gerald Carter4-25/+46
(This used to be commit 4cda2bd035276bd090bf0fbd4e3b2eff657a80cb)
2007-10-10r7371: On one build farm box dirtest1 does not return, I think this might be ↵Volker Lendecke1-0/+4
due to it being too slow. The build farm re-runs that with debuglevel 10, and this ends up with a 400MB logfile which is not synced. Lets see if with this change the machine pops up on build.samba.org.... Volker (This used to be commit b4a6ace00dbf1a6b6200704197df7ecd460412e9)
2007-10-10r7344: Correctly return if malloc fails.Volker Lendecke1-0/+1
Jeremy, you might want to check this. Thanks, Volker (This used to be commit 560f9d37353f6b886996144215f99e06f2513007)
2007-10-10r7339: only check for dfs proxy referrals when the server supports dfsGerald Carter1-1/+1
(This used to be commit 9e93244fa5bd06b142a980d4090fa8f95afc983a)
2007-10-10r7335: patch from Joerg Sonnenberger <joerg@britannica.bec.de> to fix ↵Gerald Carter1-4/+4
dragonfly detection (This used to be commit d80ae98d9e3f872c8064ce8316c76bd6de50a848)
2007-10-10r7318: Unify HEAD and 3.0 socket functions. Add HEAD functions here ↵Jeremy Allison1-0/+93
#ifdef'ed out. Correct branch this time ! Jeremy. (This used to be commit 052b4f8e2849b4897176e64cc29b6a5f686d946e)
2007-10-10r7245: bug fixes in libsmbclient, setting time attributesDerrell Lipman1-2/+11
(This used to be commit bf33902c059b7f222f89673ce07c8f929860a85f)
2007-10-10r7243: Don't look at gencache.tdb for the trusted domains if winbind is around.Volker Lendecke3-5/+51
Volker (This used to be commit 94acb93f57b963bf137c6ddd644a147f4d0b5175)
2007-10-10r7217: Only allow schannel connections if a successful Auth2 has been doneVolker Lendecke2-0/+9
before. Things tested: Domain join and subsequent interactive and network logon to NT4, W2kSP and XPSP2 workstations and a NT4 domain trusting us. Right now I've got problems with my W2k3 domain trusts. So this needs testing, although I'm really confident that this does not break. Volker (This used to be commit c25b4afda2b657b73a6215d3ff36461a36496ba3)
2007-10-10r7210: Fix my own mistakes up, sorry.Jeremy Allison3-5/+5
Jeremy. (This used to be commit 53c3a954ee0e1c9dc61950f1a9d0a654de9382c6)
2007-10-10r7202: lock mtab when updating it during umount.cifs, also delete only one ↵Steve French1-17/+63
matching entry at a time (This used to be commit 200db0790a7380e9a68fde391fd09c82da51f52e)
2007-10-10r7200: Don't use memset, use SET_STAT_INVALID (has the same effect).Jeremy Allison6-11/+12
Jeremy. (This used to be commit 0b6f87d5e14da461bd2b1c3a4e6f47a69d2cd1c4)
2007-10-10r7197: fix build failure when running 'make torture' withoutGerald Carter1-4/+4
first running 'make all' first (This used to be commit 8a8408a34089afb08f037b2652f4803399c21d8e)
2007-10-10r7191: Squeezing out more unnecessary memset's gets us up 20% over older ↵Jeremy Allison1-28/+45
code :-). Getting medieval on our ass about memset.... Jeremy. (This used to be commit d0c804bc353afa1712de3885071712af1daffb11)
2007-10-10r7190: I *love* valgrind/cachegrind.....Jeremy Allison1-10/+9
By removing unneeded memsets in qfilepathinfo I just improved our netbench performance by *********15%**********. Check it out :-). Jeremy. (This used to be commit c20a7b10b6a82db6df349fa5e6f2fcf3eeefec58)
2007-10-10r7175: fix incorrect commentDerrell Lipman1-1/+4
(This used to be commit 6f951c863ea4671a62075c6ac963f198fbf322b1)
2007-10-10r7172: This is the proper fix for setting file times from libsmbclient. We nowDerrell Lipman4-86/+244
try setpathinfo, and if that doesn't work (e.g. on win98), revert to the previous slower method. (This used to be commit 6c05812bd90b0db69d974ee2758721dc2974a507)
2007-10-10r7168: Updating file times from libsmbclient was not working for win98. ↵Derrell Lipman2-37/+68
Although the function that was being used to set attributes is a core protocol function (SMBsetatr = 0x09), it does not appear to work on win98. As a temporary measure, when file times are to be set, this version opens the file and uses SMBsetattrE = 0x22 instead. (The other advantage of this function over the original one is that it supports setting access time as well as modification time.) The next step, the proper solution if it can be made to work, is to write functions that use TRANS2_SET_PATH_INFO instead. (This used to be commit bab0bf7f4f9d2a4b6fcee4429094349302bbeb33)
2007-10-10r7157: Ensure we abort a directory listing if we see the sameJeremy Allison1-0/+7
name twice between packets. Jeremy. (This used to be commit f9063b383ed2c53841ac27691b6a593b80c20b12)
2007-10-10r7151: Fix for bug #2698. If a unicode to unix charset conversion fails (due ↵Jeremy Allison1-10/+40
to buggy iconv?) we can be left with a filename that doesn't exist on the remote machine. If we then do a findnext with this file the server gets confused and restarts from the beginning of the directory, causing directory listing loops. Fix this by keeping a copy of the "raw" filename data and length and using this as the argument to findnext. This won't fix the incorrect iconv conversion into the finfo struct but at least it ensures that directory listings always terminate. Tested against NTFS and FAT directories. Jeremy. (This used to be commit 848940d5a91b310e58d0631ead293418ea4186f0)
2007-10-10r7148: Fix #2736: winbind race condition with detecting idle clientsJim McDonough1-6/+12
winbind idle connection closing logic is getting invoked under high loads for clients which may already have commands in the pipe. This race condition causes clients to fail with NSS_STATUS_UNAVAIL sometimes. We now retry several times hoping (still not guaranteed, though) it will work. (This used to be commit 05c04cfd2526b8b9a82916b5dffc18bf27c3f198)
2007-10-10r7145: reuse the WINBIND_REQUIRED_MEMBERSHIP constant.Günther Deschner1-12/+14
This is just cosmetic but prevents people from thinking that the pam_winbind "require_membership_of"-option is not yet implemented :) Guenther (This used to be commit ef80a49a858d7d81d427c7bac71fdac4fc0d1bd6)
2007-10-10r7143: removing unused fileGerald Carter1-0/+0
(This used to be commit 8c3886515683735c7449e409f8aa2ee9e6ac14fb)
2007-10-10r7142: removing 'ldap filter' smb.conf optionGerald Carter2-8/+4
(This used to be commit e2f8eeb4e26e8f7bec4d9dc820b86a7b548f82da)
2007-10-10r7140: removing aparser directory since it is obselete in light of pidlGerald Carter56-4009/+0
(This used to be commit 38cad678b8058d5ac99b41c0be07f4c58a720cfc)
2007-10-10r7139: trying to reduce the number of diffs between trunk and 3.0; changing ↵Gerald Carter39-4911/+1327
version to 3.0.20pre1 (This used to be commit 9727d05241574042dd3aa8844ae5c701d22e2da1)
2007-10-10r7130: remove 'winbind enable local accounts' code from the 3.0 treeGerald Carter12-1928/+6
(This used to be commit 318c3db4cb1c85be40b2f812f781bcf5f1da5c19)
2007-10-10r7126: fixing paranoid malloc checker failuresGerald Carter3-6/+5
(This used to be commit b01026674fddb4179a7f002c13f5e341eaaa0a1c)
2007-10-10r7041: compile fixes....long day I guessGerald Carter1-2/+3
(This used to be commit ec3ef5ddbe12fa6ebe8f58979625c671d181c519)
2007-10-10r7039: Ensure we always call read_socket_with_timeout() whenJeremy Allison1-1/+6
timeout > 0. Better solution after much helpful input from derrell@samba.org. We may eventually change the read_socket_with_timeout() interface to count down the timeout value. Jeremy. (This used to be commit 09c9a62aa53ed26e59ed57e577614d02a796c492)
2007-10-10r7038: * upgrade version of nt*tdb files. Have to fixGerald Carter1-10/+145
some issues in the printer security descriptors. Ensure that each printer sd has an oaner and group SID (BUILTIN\Administrators) and that we utilize more than the generic bits assigned in <= 3.0.14a. (This used to be commit c72182c1e20225a655376fd23915ac6053b94633)
2007-10-10r7034: Revert change in util_sock.c from derrell until this isJeremy Allison1-58/+13
discussed more on samba-technical. Jeremy. (This used to be commit 7ad6fdc1ef4754fc25d81ab27fd5ebb40e04a6cf)
2007-10-10r7033: Call a spade a spade :-).Jeremy Allison2-91/+85
Jeremy. (This used to be commit f5027f6370bd085a8d4c1a221881eebb0e4ffbab)
2007-10-10r7031: Added encrypt/decrypt function for LSA secrets and trustedJeremy Allison2-1/+30
domain passwords on the wire. Jeremy. (This used to be commit f82dcac25faf7876655cb1839846cc5e01e4add7)
2007-10-10r7025: 1 if not all data is available at the time we go to read a packet, retryDerrell Lipman2-23/+75
the read using a timeout to ensure that all data for the packet is received. 2 some minor changes to meet coding standards 3 eliminate some compiler warnings (This used to be commit 7b4d4f6109d815ec70c65564435d7d9bd22f66d9)