sfrench/samba-autobuild/.git
20 years agoMention security=ads in introductory paragraph too
Alexander Bokovoy [Mon, 11 Aug 2003 08:05:59 +0000 (08:05 +0000)]
Mention security=ads in introductory paragraph too
(This used to be commit 33a53550df16bf89a8904ee9496896ff790ed32b)

20 years agofix bug #281 by surrounding pdb_getgrgid() with become/unbecome_root()
Gerald Carter [Mon, 11 Aug 2003 07:05:23 +0000 (07:05 +0000)]
fix bug #281 by surrounding pdb_getgrgid() with become/unbecome_root()
(This used to be commit f6a01f51159ccd822c6e764b7243fff375f22747)

20 years agoadding '.' special name to --domain to mean our domain
Gerald Carter [Mon, 11 Aug 2003 06:36:30 +0000 (06:36 +0000)]
adding '.' special name to --domain to mean our domain
(This used to be commit f833d1f15691a418b99f82dbda75b673fb3428f6)

20 years agoadding missing options (mine) to wbinfo
Gerald Carter [Mon, 11 Aug 2003 06:30:22 +0000 (06:30 +0000)]
adding missing options (mine) to wbinfo
(This used to be commit ae6d704946867e92ad6263799f28712442d55617)

20 years agoUpdate upgrading docs
Jelmer Vernooij [Mon, 11 Aug 2003 01:41:56 +0000 (01:41 +0000)]
Update upgrading docs
(This used to be commit cd81eaaf2d2bc407a4f9d037fedaa7eb82abcc65)

20 years agoImprove MySQL library detection, add support for MySQL 4
Jelmer Vernooij [Mon, 11 Aug 2003 00:31:28 +0000 (00:31 +0000)]
Improve MySQL library detection, add support for MySQL 4
(This used to be commit 90cbc4361e86d591e38c80092612edaa3875ea67)

20 years agonot used
Gerald Carter [Sun, 10 Aug 2003 22:11:26 +0000 (22:11 +0000)]
not used
(This used to be commit b3074a893bc0bee11d628907bd83b53f596af0eb)

20 years agoadd --domain=DOMAINNAME to wbinfo
Gerald Carter [Sun, 10 Aug 2003 22:01:11 +0000 (22:01 +0000)]
add --domain=DOMAINNAME to wbinfo

Add support for geting the sequence number, list of users, and list
of groups for a specific domain (assuming on reported back by
wbinfo -m)

wbinfo -u --domain=DOA
(This used to be commit 34fc6e1bf97d514d3b1763a808d08d730191e03b)

20 years agoWe get the server and workgroup list only over port 139. Fall back to that
Volker Lendecke [Sun, 10 Aug 2003 21:43:28 +0000 (21:43 +0000)]
We get the server and workgroup list only over port 139. Fall back to that
and if NetBIOS over TCP/IP is disabled, give an error message.

Fixes Bug #284

Volker
(This used to be commit a2902a6cf8d05569d1c690436f69b4a4cedac400)

20 years agoFix style
Jelmer Vernooij [Sun, 10 Aug 2003 20:41:27 +0000 (20:41 +0000)]
Fix style
(This used to be commit ba60a11b62a093565459efca1879f2b028724af1)

20 years agoStore the server domain from the ntlmssp challenge in the client struct
Volker Lendecke [Sun, 10 Aug 2003 20:18:05 +0000 (20:18 +0000)]
Store the server domain from the ntlmssp challenge in the client struct
to be able to ask a LMB for the servers in its workgroup. Against
W2k this only works on port 139....

Volker
(This used to be commit 62b04d7776852098dd768268500f36c3a362f688)

20 years agofix MS-DFS (my bug) stupid return code error that cose me more time
Gerald Carter [Sun, 10 Aug 2003 06:45:04 +0000 (06:45 +0000)]
fix MS-DFS (my bug)  stupid return code error that cose me more time
that I would like to admit.
(This used to be commit 082b1cf06fd3b144b9713045672028f16e20aff4)

20 years agofix for BUG #267 (problem with supplementary groups).
Gerald Carter [Sat, 9 Aug 2003 23:12:35 +0000 (23:12 +0000)]
fix for BUG #267 (problem with supplementary groups).
Use winbindd to get the group list if possible since we already
know it from netsamlogon_cache.tdb. More effecient than letting
libc call getgrent() to get seconary groups.

Tested by Ken Cross.
(This used to be commit 3c537c906f29a08e75895c8c8e3ed5c5abaaa940)

20 years agoupdating smb.conf manpage to reflect changes in ./configure options
Gerald Carter [Sat, 9 Aug 2003 23:01:35 +0000 (23:01 +0000)]
updating smb.conf manpage to reflect changes in ./configure options
(This used to be commit ca29fdb9c28b7a61ce35eaea8864a8cd647b45cc)

20 years agofix 2 bugs:
Gerald Carter [Fri, 8 Aug 2003 23:53:13 +0000 (23:53 +0000)]
fix 2 bugs:

  1)  don't ask trusted DC's for a list of trusted domains.  This causes
      us to treat non-transitive ones as if they were transitive.  Not
      needed anyways

  2)  Fix dc lookup bug where we would always try to use DNS to resolve
      the DC's for a domain (even if it was a trusted NT4 domain).
(This used to be commit 4d3acce5066d3adf53ee8fbaa627c42523b3cbc3)

20 years agoFormat tidyup.
Jeremy Allison [Fri, 8 Aug 2003 23:09:09 +0000 (23:09 +0000)]
Format tidyup.
Jeremy.
(This used to be commit 049e77d636c5abd0fdd8840c3c4c465708354ed7)

20 years agoTurn on client ntlmv2 by default.
Jeremy Allison [Fri, 8 Aug 2003 22:19:09 +0000 (22:19 +0000)]
Turn on client ntlmv2 by default.
Jeremy.
(This used to be commit 729b468f7e0e5522dfdede481947826851842483)

20 years agoRPC fix from Ronan Waide <waider@waider.ie>. Tested with rpcecho.
Jeremy Allison [Fri, 8 Aug 2003 17:08:35 +0000 (17:08 +0000)]
RPC fix from Ronan Waide <waider@waider.ie>. Tested with rpcecho.
Jeremy.
(This used to be commit 68590b9e2266cf76b46a68cca0acaa47733811fe)

20 years agoneed to be able to connect to a domain member as a local account; don't always map...
Gerald Carter [Fri, 8 Aug 2003 05:11:11 +0000 (05:11 +0000)]
need to be able to connect to a domain member as a local account; don't always map to the domain name
(This used to be commit 20b6e64da2669e5dfc7265cae331ec9c89aa02dc)

20 years agoneed to make sure that the connection struct mem_ctx is initialized and destroyed
Gerald Carter [Fri, 8 Aug 2003 05:10:12 +0000 (05:10 +0000)]
need to make sure that the connection struct mem_ctx is initialized and destroyed
(This used to be commit e0c061b94d4cb181afcd70c5d6398e0c2f446449)

20 years agoAdded by request of "Stefan (metze) Metzmacher" <metze@metzemix.de>.
Jeremy Allison [Fri, 8 Aug 2003 00:53:46 +0000 (00:53 +0000)]
Added by request of "Stefan (metze) Metzmacher" <metze@metzemix.de>.
Jeremy.
(This used to be commit 80e14b176526a3038ea03214fc3dd0fd00fef922)

20 years agoTest modules for shadow copy by "Stefan (metze) Metzmacher" <metze@metzemix.de>.
Jeremy Allison [Thu, 7 Aug 2003 21:49:01 +0000 (21:49 +0000)]
Test modules for shadow copy by "Stefan (metze) Metzmacher" <metze@metzemix.de>.
Jeremy.
(This used to be commit 3f8c77b116b96b01fd333c4b45d84666ef108fe9)

20 years agoShadow copy API - Original work by "Ken Cross" <kcross@nssolutions.com>, adapted
Jeremy Allison [Thu, 7 Aug 2003 21:47:46 +0000 (21:47 +0000)]
Shadow copy API - Original work by "Ken Cross" <kcross@nssolutions.com>, adapted
into a patch by "Stefan (metze) Metzmacher" <metze@metzemix.de>.
Jeremy.
(This used to be commit ce5c91d35dabc5ff6fb3df2b259ed186d6a7e0da)

20 years agoAn oplock break reply from the client causes the sequence number to be
Jeremy Allison [Thu, 7 Aug 2003 05:36:08 +0000 (05:36 +0000)]
An oplock break reply from the client causes the sequence number to be
updated by 2 if there is no open reply outstanding, else by one....
Yes - this makes no sense....
Jeremy.
(This used to be commit b43ce1ff6109f6422a621329ceb713b42df40040)

20 years agoadding generic docs for defining VFS module options
Gerald Carter [Thu, 7 Aug 2003 05:00:51 +0000 (05:00 +0000)]
adding generic docs for defining VFS module options
(This used to be commit 6168d3da670f7bb75e34b65a91c76728576f8607)

20 years agoTurns out I had my packet sequences wrong for oplock break code.
Jeremy Allison [Thu, 7 Aug 2003 02:59:52 +0000 (02:59 +0000)]
Turns out I had my packet sequences wrong for oplock break code.
I was storing the mid of the oplock break - I should have been
storing the mid from the open. There are thus 2 types of deferred
packet sequence returns - ones that increment the sequence number
(returns from oplock causing opens) and ones that don't (change notify
returns etc). Running with signing forced on does lead to some
interesting tests :-).
Jeremy.
(This used to be commit 85907f02cec566502d9e4adabbd414020a26064d)

20 years agoCosmetic fix from waider@waider.ie.
Jeremy Allison [Thu, 7 Aug 2003 01:04:57 +0000 (01:04 +0000)]
Cosmetic fix from waider@waider.ie.
Jeremy.
(This used to be commit cb326c2dbff1fad87d5c72df4a004d5a42d17472)

20 years agoPatch from waider@waider.ie to print out Port Type.
Jeremy Allison [Thu, 7 Aug 2003 00:55:35 +0000 (00:55 +0000)]
Patch from waider@waider.ie to print out Port Type.
Jeremy.
(This used to be commit 8516baf58d333a54bcbe1c1a08eee499b3dd2636)

20 years agoEnsure smbclient obeys -s smb.conf option.
Jeremy Allison [Wed, 6 Aug 2003 22:45:46 +0000 (22:45 +0000)]
Ensure smbclient obeys -s smb.conf option.
Jeremy.
(This used to be commit 7950f7d6b2c5e8247bce1e96ddd1a9ec1045f7f9)

20 years agoReversed replacement. Oops.
Jeremy Allison [Wed, 6 Aug 2003 21:31:11 +0000 (21:31 +0000)]
Reversed replacement. Oops.
Jeremy.
(This used to be commit 520bee397d946766cef8ab9f0d7d89064106f510)

20 years agoAnal formatting tidyup :-).
Jeremy Allison [Wed, 6 Aug 2003 20:01:31 +0000 (20:01 +0000)]
Anal formatting tidyup :-).
Jeremy.
(This used to be commit 9c0197471366ad1acf922e02bb694769863054a6)

20 years agoGet rid of MAXPATHLEN, move to standard PATH_MAX.
Jeremy Allison [Wed, 6 Aug 2003 19:30:42 +0000 (19:30 +0000)]
Get rid of MAXPATHLEN, move to standard PATH_MAX.
Jeremy.
(This used to be commit 455ed2d51d86f39ce0fa6e6abca31a5425d2ea17)

20 years agofix bug #208; have to get the gid of the user's primary group for %G
Gerald Carter [Wed, 6 Aug 2003 19:16:17 +0000 (19:16 +0000)]
fix bug #208; have to get the gid of the user's primary group for %G
(This used to be commit 575483a1efe18a90055490117ba6894512ae568a)

20 years agorework winbindd_accountdb_init() to reduce error messages in the log
Gerald Carter [Wed, 6 Aug 2003 18:01:39 +0000 (18:01 +0000)]
rework winbindd_accountdb_init() to reduce error messages in the log
(This used to be commit d98a68e0ebaf2fbd360b826f5df472bc3f94285b)

20 years agoWhen doing 'net groupmap add', default to algorithmic mapping for the rid.
Volker Lendecke [Wed, 6 Aug 2003 09:24:11 +0000 (09:24 +0000)]
When doing 'net groupmap add', default to algorithmic mapping for the rid.

Volker
(This used to be commit 7ce94d39add6e056e3b1deea21bf0438ba61e4cc)

20 years agoSpelling.
Tim Potter [Wed, 6 Aug 2003 01:14:51 +0000 (01:14 +0000)]
Spelling.
(This used to be commit 7d009ebf66c82b254828bac267102eb6e6a4a75e)

20 years agooops; fix typo. Noticed by gcc warning
Gerald Carter [Wed, 6 Aug 2003 00:06:37 +0000 (00:06 +0000)]
oops; fix typo.  Noticed by gcc warning
(This used to be commit 4c36ef65e5101899f730adaeacf754f5f3647d89)

20 years agofix bug #245; local_lookupsid() needed to make a getpwuid() call to get the username...
Gerald Carter [Tue, 5 Aug 2003 23:24:14 +0000 (23:24 +0000)]
fix bug #245; local_lookupsid() needed to make a getpwuid() call to get the username instead of making up unix_user.##
(This used to be commit b947fc3eed464d7a64914f3965964d29be031614)

20 years agoChanges to make gss-spnego ntlmssp client work against W2k AD.
Volker Lendecke [Mon, 4 Aug 2003 13:10:43 +0000 (13:10 +0000)]
Changes to make gss-spnego ntlmssp client work against W2k AD.

Now I know where the mechListMIC changes came from: Ethereal ;-)

Volker
(This used to be commit 4e9eed1273035d09ac3b427b9711327ba8c6ebfc)

20 years agoFix unused variable warning.
Tim Potter [Mon, 4 Aug 2003 06:16:03 +0000 (06:16 +0000)]
Fix unused variable warning.
(This used to be commit 73d02e3a2b0f9e84ab6d8685e4ad6a03ef9249b2)

20 years agoMemory leak fix for create_rpc_bind_req()
Tim Potter [Mon, 4 Aug 2003 02:51:30 +0000 (02:51 +0000)]
Memory leak fix for create_rpc_bind_req()
(This used to be commit 4d26feabd75d5b298276b0c5880b9765507bb6ae)

20 years agoMore patches from Brett:
Tim Potter [Mon, 4 Aug 2003 00:50:00 +0000 (00:50 +0000)]
More patches from Brett:

  - remove 'if(mem_ctx)' tests prior to 'talloc_destroy' call to make
consistent with other modules; 'talloc_destroy' already test for NULL
anyway.

  - initialize PyObject* result pointers to NULL in function
declarations; enables removal of redundant NULL assignments.

  - use local scope TALLOC_CTX in lsa_lookup_names to prevent unbounded
memory growth during python policy object lifetime.

  - change context name string used in lsa_lookup_sids from
'lsa_open_policy' to 'lsa_lookup_sids' (cut'npaste oversight from
previous patch)

  - change docstring to match module name (apparently another cut'npaste
situation)
(This used to be commit 98cac74c67c454f636df5c1ea2baa029c482e402)

20 years agoFix memory leak in py_smb_set_setdesc()
Tim Potter [Mon, 4 Aug 2003 00:48:49 +0000 (00:48 +0000)]
Fix memory leak in py_smb_set_setdesc()

Consistency fixups in py_smb_query_secdesc()

Thanks to Brett A. Funderburg for these patches.
(This used to be commit 6c33189e50ae3704275198e324fcdb7facd209f7)

20 years agoFix up #defines around utmp_host and utmp_name. Noticed by Cord.Hockemeyer@uni-graz.at
Jeremy Allison [Sun, 3 Aug 2003 18:50:00 +0000 (18:50 +0000)]
Fix up #defines around utmp_host and utmp_name. Noticed by Cord.Hockemeyer@uni-graz.at
Jeremy.
(This used to be commit 07c5ecb945c1189fd6e8628f5e989b90dd15163b)

20 years agoOutput message saying "signed connect" instead of just connect when signing
Jeremy Allison [Sun, 3 Aug 2003 07:20:05 +0000 (07:20 +0000)]
Output message saying "signed connect" instead of just connect when signing
is active.
Jeremy.
(This used to be commit c6674fa62865b64aa788a1903db118e4d773bcae)

20 years agoFix oplock break detection code on incoming oplock break responses. This
Jeremy Allison [Sun, 3 Aug 2003 07:12:46 +0000 (07:12 +0000)]
Fix oplock break detection code on incoming oplock break responses. This
fixes signing for oplocks.
Jeremy.
(This used to be commit 69c56ee8bce122839a8fec4e59198f84b0757166)

20 years agoRemoved duplicated file, no longer in use. Note: profiles.1.xml is still
John Terpstra [Sun, 3 Aug 2003 05:10:57 +0000 (05:10 +0000)]
Removed duplicated file, no longer in use. Note: profiles.1.xml is still
present and relevant.
(This used to be commit 7b24759c270ffbc914aa0a45c8b7facbf6217c69)

20 years agoChanged "winbind uid/gid" to "idmap uid/gid"
John Terpstra [Sun, 3 Aug 2003 05:09:19 +0000 (05:09 +0000)]
Changed "winbind uid/gid" to "idmap uid/gid"
(This used to be commit d205dd963bb306935d3c02c94684fa1143436158)

20 years agomake sure to initialize the backend methods when enumerating sequence numbers; report...
Gerald Carter [Sat, 2 Aug 2003 18:15:33 +0000 (18:15 +0000)]
make sure to initialize the backend methods when enumerating sequence numbers; reported by Ken Cross
(This used to be commit 10c7a1af67e556c17d4b3495934a2dad19728d77)

20 years agoEnsure we don't leak any sign records on cancel of pending requests.
Jeremy Allison [Sat, 2 Aug 2003 08:48:01 +0000 (08:48 +0000)]
Ensure we don't leak any sign records on cancel of pending requests.
Jeremy.
(This used to be commit 9a8ffc239c0f1aada713de7e9e007066738d8874)

20 years agoOnly look for mid sign records on incoming packets for oplock break replies.
Jeremy Allison [Sat, 2 Aug 2003 08:38:34 +0000 (08:38 +0000)]
Only look for mid sign records on incoming packets for oplock break replies.
Otherwise we find spurious mid sign records on reply_ntcancel calls (they cancel
by mid). That took a *lot* of tracking down. I still need to remove the mid
records from the sign state on reply_ntcancel to avoid leaking memory....
Jeremy.
(This used to be commit 270bf20fe3e226ab5cfc689bd20ed4c22b2fa7e6)

20 years agoMore fixes for client and server side signing. Ensure sequence numbers
Jeremy Allison [Sat, 2 Aug 2003 07:07:38 +0000 (07:07 +0000)]
More fixes for client and server side signing. Ensure sequence numbers
are updated correctly on returning an error for server trans streams.
Ensure we turn off client trans streams on error.
Jeremy.
(This used to be commit 3a789cb7f01115c37404e5a696de363287cb0e5f)

20 years agoLeave the packet sequence checkers enabled whilst I track down a smbclient -> smbd
Jeremy Allison [Sat, 2 Aug 2003 03:12:39 +0000 (03:12 +0000)]
Leave the packet sequence checkers enabled whilst I track down a smbclient -> smbd
sequence number problem.
Jeremy.
(This used to be commit 844898dbd8e99837ef1621aa73024714aa819ce4)

20 years agoAdd the same signing code to the server. Ensure we use identical session
Jeremy Allison [Sat, 2 Aug 2003 03:06:07 +0000 (03:06 +0000)]
Add the same signing code to the server. Ensure we use identical session
numbers and MIDs when in trans/trans2/nttrans code.
Jeremy.
(This used to be commit 901544b29b4d815709b3dbad3012f1d2c419d904)

20 years agoCorrect fix (removed the earlier band-aid) for what I thought was a signing
Jeremy Allison [Sat, 2 Aug 2003 00:29:45 +0000 (00:29 +0000)]
Correct fix (removed the earlier band-aid) for what I thought was a signing
bug with w2k. Turns out that when we're doing a trans/trans2/nttrans call
the MID and send_sequence_number and reply_sequence_number must remain constant.
This was something we got very wrong in earlier versions of Samba. I can now
get a directory listing from WINNT\SYSTEM32 with the older earlier parameters
for clilist.c
This still needs to be fixed for the server side of Samba, client appears to
be working happily now (I'm doing a signed smbtar download of an entire W2K3
image to test this :-).
Jeremy.
(This used to be commit 2093a3130d4087d0659b497eebd580e7a66e5aa3)

20 years agoFix the option processing for smbtar. Does no one check this !
Jeremy Allison [Fri, 1 Aug 2003 21:09:10 +0000 (21:09 +0000)]
Fix the option processing for smbtar. Does no one check this !
Jeremy.
(This used to be commit c4af1ec73584656fb9d2b372461da4748e1643f1)

20 years agoadd tests for IRIX attr functions
Herb Lewis [Fri, 1 Aug 2003 19:45:12 +0000 (19:45 +0000)]
add tests for IRIX attr functions
(This used to be commit ea57a57db036de6245ff44cd953dae3b482b4744)

20 years agoUpdate my copyrights according to my agreement with IBM
Jim McDonough [Fri, 1 Aug 2003 15:30:44 +0000 (15:30 +0000)]
Update my copyrights according to my agreement with IBM
(This used to be commit a2bd8f0bfa12f2a1e33c96bc9dabcc0e2171700d)

20 years agoUpdate my copyrights according to my agreement with IBM
Jim McDonough [Fri, 1 Aug 2003 15:21:20 +0000 (15:21 +0000)]
Update my copyrights according to my agreement with IBM
(This used to be commit c9b209be2b17c2e4677cc30b46b1074f48878f43)

20 years agoFix copyright statements for various pieces of Anthony Liguori's work.
Jim McDonough [Fri, 1 Aug 2003 14:47:39 +0000 (14:47 +0000)]
Fix copyright statements for various pieces of Anthony Liguori's work.
(This used to be commit 15d2bc47854df75f8b2644ccbc887d0357d9cd27)

20 years agofix cut-n-paste error found by abartlet
Gerald Carter [Fri, 1 Aug 2003 13:28:13 +0000 (13:28 +0000)]
fix cut-n-paste error found by abartlet
(This used to be commit 042492b862c4f2389086275dd8a0e044d7432537)

20 years agoAdd ntlmssp client support to ntlm_auth. Find the corresponding cyrus sasl
Volker Lendecke [Fri, 1 Aug 2003 07:59:23 +0000 (07:59 +0000)]
Add ntlmssp client support to ntlm_auth. Find the corresponding cyrus sasl
module under http://samba.sernet.de/cyrus-gss-spnego.diff

Volker
(This used to be commit a82f6a00969f7ea377626c28ec05ace04f8135a9)

20 years agoFix a memory leak. I did not check all the calls to winbindd_request, but
Volker Lendecke [Fri, 1 Aug 2003 07:46:42 +0000 (07:46 +0000)]
Fix a memory leak. I did not check all the calls to winbindd_request, but
we might leak the extra_data somewhere else as well.

Volker
(This used to be commit 5d379345fa06f4253f67b40cb8127b70072db561)

20 years agolocking.c now refers to map_nt_error_from_unix, so link it in with
Volker Lendecke [Fri, 1 Aug 2003 07:45:02 +0000 (07:45 +0000)]
locking.c now refers to map_nt_error_from_unix, so link it in with
smbstatus and smbcontrol

Volker
(This used to be commit c2c901e06b9723e33cccdb592c609072b4376c10)

20 years agoFinish reformatting.
Jeremy Allison [Fri, 1 Aug 2003 06:29:16 +0000 (06:29 +0000)]
Finish reformatting.
Jeremy.
(This used to be commit 5ef75fa36620a87e8fd8567b854ea0770d8125ea)

20 years agoFinal fix for the bug tridge found. Only push locks onto a blocking lock
Jeremy Allison [Fri, 1 Aug 2003 06:10:30 +0000 (06:10 +0000)]
Final fix for the bug tridge found. Only push locks onto a blocking lock
queue if the posix lock failed with EACCES or EAGAIN (this means another
lock conflicts). Else return an error and don't queue the request.
Jeremy.
(This used to be commit 43fbc18fdc184bf29c15186c16bc99fb208de963)

20 years agoReformat lots of clitar code as I hate the style so much :-).
Jeremy Allison [Fri, 1 Aug 2003 01:03:05 +0000 (01:03 +0000)]
Reformat lots of clitar code as I hate the style so much :-).
Jeremy.
(This used to be commit 77f2a91549c8b61e74e2088faef2b61612803aed)

20 years agoReformat clitar option processing - getting ready to fix it for popt...
Jeremy Allison [Fri, 1 Aug 2003 00:41:57 +0000 (00:41 +0000)]
Reformat clitar option processing - getting ready to fix it for popt...
Jeremy.
(This used to be commit 94b30e7b4a4f71d3aa2fefee60b9ea8ead2acccd)

20 years agoCVAL_NC() doesn't need the (unsigned) fix and breaks the IRIX build
Andrew Tridgell [Thu, 31 Jul 2003 23:22:21 +0000 (23:22 +0000)]
CVAL_NC() doesn't need the (unsigned) fix and breaks the IRIX build

Thanks to Herb for pointing this out!
(This used to be commit 87ede8d310db10d92b4ff57e67d3b53cbb7697fb)

20 years agoAdded a note inspired by andrew@cis.uoguelph.ca to explain when this
Jeremy Allison [Thu, 31 Jul 2003 21:47:22 +0000 (21:47 +0000)]
Added a note inspired by andrew@cis.uoguelph.ca to explain when this
parameter gets run.
Jeremy.
(This used to be commit c5e11fd01003fa2d85c91ec81afd2cb2498472f5)

20 years agoonly honor the first OID in the sessetup snego negotiate. Deviates
Gerald Carter [Thu, 31 Jul 2003 19:01:22 +0000 (19:01 +0000)]
only honor the first OID in the sessetup snego negotiate.  Deviates
from RFC but I'm smelling a client bug here.

/* only look at the first OID for determining the mechToken --
   accoirding to RFC2478, we should choose the one we want
   and renegotiate, but i smell a client bug here..

   Problem observed when connecting to a member (samba box)
   of an AD domain as a user in a Samba domain.  Samba member
   server sent back krb5/mskrb5/ntlmssp as mechtypes, but the
   client (2ksp3) replied with ntlmssp/mskrb5/krb5 and an
   NTLMSSP mechtoken.                 --jerry              */
(This used to be commit 731420b03dbc15977822f74047e931dc62284fc0)

20 years agoReturn proper error when it is impossible to change quota flags
Alexander Bokovoy [Thu, 31 Jul 2003 17:08:38 +0000 (17:08 +0000)]
Return proper error when it is impossible to change quota flags
(This used to be commit 44bae9da9e03a1df738367f8d86adb0f315da41f)

20 years agoFix off-by-one found by valgrind.
Volker Lendecke [Thu, 31 Jul 2003 15:53:59 +0000 (15:53 +0000)]
Fix off-by-one found by valgrind.

Volker
(This used to be commit bc39c9b57fa6258674e1ee44b3446f25bf63661e)

20 years agospnego.c has function definitions. Prototype them.
Volker Lendecke [Thu, 31 Jul 2003 15:53:26 +0000 (15:53 +0000)]
spnego.c has function definitions. Prototype them.

Anybody familiar with Makefile.in could you please look at this?
This is probably the wrong way to fix this.

Volker
(This used to be commit 9a04750dea1b614604e6f9a0f69a251391994a71)

20 years agoFixes for memory leaks in gss spnego handling by aliguori.
Volker Lendecke [Thu, 31 Jul 2003 10:24:10 +0000 (10:24 +0000)]
Fixes for memory leaks in gss spnego handling by aliguori.

Volker
(This used to be commit 946695242f0a6db0e8c712f9bd97ac3cb99c4119)

20 years agoThis fixes an error I must have made when playing with spnego.c found
Volker Lendecke [Thu, 31 Jul 2003 10:23:13 +0000 (10:23 +0000)]
This fixes an error I must have made when playing with spnego.c found
by aliguori: NegTokenInit.mechListMIC is an Octet String.

Second: add a free_spnego_data function.

Both thanks to aliguori.

Volker
(This used to be commit 6c252440fba33eb69827d5515a95fbb3e8e9a653)

20 years agoApply some const
Volker Lendecke [Thu, 31 Jul 2003 10:21:13 +0000 (10:21 +0000)]
Apply some const
(This used to be commit 692ff44ba39cd24dbc906e0319bc51c8be9cc267)

20 years agomake sure the domain sid is set when enumerating trusted domains
Gerald Carter [Thu, 31 Jul 2003 06:37:37 +0000 (06:37 +0000)]
make sure the domain sid is set when enumerating trusted domains
(we don't always get it back)
(This used to be commit 6e8f23d1b1dd5122691cd9363a5c572fcab182af)

20 years agoworking on transtive trusts issue:
Gerald Carter [Thu, 31 Jul 2003 05:43:47 +0000 (05:43 +0000)]
working on transtive trusts issue:

  * use DsEnumerateDomainTrusts() instead of LDAP search.
    wbinfo -m now lists all trusted downlevel domains and
    all domains in the forest.

Thnigs to do:

  o Look at Krb5 connection trusted domains
  o make sure to initial the trusted domain cache as soon
    as possible
(This used to be commit 0ab00ccaedf204b39c86a9e1c2fcac5f15d0e033)

20 years agoWhoops - this is probably better shell syntax.
Tim Potter [Thu, 31 Jul 2003 04:28:43 +0000 (04:28 +0000)]
Whoops - this is probably better shell syntax.
(This used to be commit 4644085b682c806796f9acf2ed73de9dd9872738)

20 years agoTurn on automatic winbindd support for FreeBSD and see what the compile farm
Tim Potter [Thu, 31 Jul 2003 04:27:41 +0000 (04:27 +0000)]
Turn on automatic winbindd support for FreeBSD and see what the compile farm
thinks of it.
(This used to be commit e6210295e4daa5f1768616b4d8390a173c5c67ae)

20 years agoThis is a critical bug fix for a data corruption bug. If you
Andrew Tridgell [Thu, 31 Jul 2003 04:01:32 +0000 (04:01 +0000)]
This is a critical bug fix for a data corruption bug. If you
maintain another tree then please apply!

On non-X86 machines out byte-order macros fails for one particular
value. If you asked for IVAL() of 0xFFFFFFFF and assigned it to a 64
bit quantity then you got a 63 bit number 0x7FFFFFFFFFFFFFFF rather
than the expected 0xFFFFFFFF. This is due to some rather bizarre and
obscure sign extension rules to do with unsigned chars and arithmetic
operators (basically if you | together two unsigned chars you get a
signed result!)

This affected a byte range lock using the large lockingX format and a
lock of offset 0 and length 0xFFFFFFFF. Microsoft Excel does one of
these locks when opening a .csv file. If the platform you run on does
not then handle locks of length 0x7FFFFFFFFFFFFFFF then the posix lock
fails and the client is given a lockingX failure. This causes the .csv
file to be trunated!!
(This used to be commit 886661c3777dbfd4fa431746c8a5f48674a12b8e)

20 years agoWrap calls to change_oem_password() in become_root()/unbecome_root() pairs
Jeremy Allison [Thu, 31 Jul 2003 01:33:44 +0000 (01:33 +0000)]
Wrap calls to change_oem_password() in become_root()/unbecome_root() pairs
to allow UNIX password change scripts to work correctly. This is safe as
the old password has been checked as correct before invoking this.
Jeremy.
(This used to be commit 1734d43eb55561d46a6ffb5d806afedfd3746f9f)

20 years agoTurn the 'doing_signing' variable on - fix bug where it was only being set
Jeremy Allison [Thu, 31 Jul 2003 00:30:01 +0000 (00:30 +0000)]
Turn the 'doing_signing' variable on - fix bug where it was only being set
on when signing was mandatory.
Jeremy.
(This used to be commit 7c58673a103195435ca75ebb2684880d1f7242d3)

20 years agoAdd a command line option (-S on|off|required) to enable signing on client
Jeremy Allison [Wed, 30 Jul 2003 23:49:29 +0000 (23:49 +0000)]
Add a command line option (-S on|off|required) to enable signing on client
connections. Overrides smb.conf parameter if set.
Jeremy.
(This used to be commit 879309671df6b530e0bff69559422a417da4a307)

20 years agoSave us from possibly uninitialised variable (caught by gcc).
Jeremy Allison [Wed, 30 Jul 2003 23:33:56 +0000 (23:33 +0000)]
Save us from possibly uninitialised variable (caught by gcc).
Jeremy.
(This used to be commit f3f29665bd2c396c4756cd23f603ac768fea66fd)

20 years agoFix bug we discovered in W2K client signing on secondary trans2 packets.
Jeremy Allison [Wed, 30 Jul 2003 19:00:52 +0000 (19:00 +0000)]
Fix bug we discovered in W2K client signing on secondary trans2 packets.
Use W2K parameters. tpot please re-test smbclient with your problem
directory.
Jeremy.
(This used to be commit 677d3a3c4ca0b67148e5e56fa876773a067679bd)

20 years agoEliminate valgrind error when client gets bad sig on list. Some reformatting.
Jeremy Allison [Wed, 30 Jul 2003 18:57:37 +0000 (18:57 +0000)]
Eliminate valgrind error when client gets bad sig on list. Some reformatting.
Jeremy.
(This used to be commit b8f6b836468b3a0ae75977dc65cae8400f74734c)

20 years agoadd a few more tidy ups. Now onto winbindd
Gerald Carter [Wed, 30 Jul 2003 17:37:46 +0000 (17:37 +0000)]
add a few more tidy ups.  Now onto winbindd
(This used to be commit f8abdd23e1d4aed56c263c3228e702b191af4c64)

20 years agoadd support for DsEnumerateDomainTrusted for enumerating all the
Gerald Carter [Wed, 30 Jul 2003 17:29:00 +0000 (17:29 +0000)]
add support for DsEnumerateDomainTrusted for enumerating all the
trusted domains in a forest.
(This used to be commit c691c7f7d9afb8af542dc83cf934df1dfd38ef17)

20 years agoDon't revert something until you've seen if volker has already fixed it :-).
Jeremy Allison [Wed, 30 Jul 2003 16:34:14 +0000 (16:34 +0000)]
Don't revert something until you've seen if volker has already fixed it :-).
Jeremy.
(This used to be commit d57d9b99dfc2a83d91b952a11eb1ae4cf37e6467)

20 years agoComment out mutex until I get dependencies sorted out...
Jeremy Allison [Wed, 30 Jul 2003 15:30:36 +0000 (15:30 +0000)]
Comment out mutex until I get dependencies sorted out...
Jeremy
(This used to be commit 382d1732ca8e4d0909eb5a95f7327213913da37b)

20 years agobin/net needs server_mutex as kerberos_verify now uses it.
Volker Lendecke [Wed, 30 Jul 2003 08:05:34 +0000 (08:05 +0000)]
bin/net needs server_mutex as kerberos_verify now uses it.

Volker
(This used to be commit ee09db891a9a45158e12fcd2a85092d64870fac4)

20 years agoPut mutex around access of replay cache for krb5 tickets. krb5 replay cache
Jeremy Allison [Tue, 29 Jul 2003 21:32:36 +0000 (21:32 +0000)]
Put mutex around access of replay cache for krb5 tickets. krb5 replay cache
is not multi-process safe.
Jeremy.
(This used to be commit 9e0534a1b69bbd4f21b4925337cbab127d060fc6)

20 years agosplit replace into replace and replace1 to allow setenv to be used by
Herb Lewis [Tue, 29 Jul 2003 20:11:18 +0000 (20:11 +0000)]
split replace into replace and replace1 to allow setenv to be used by
nsswitch modules. Add required libraries to get rid of undefined
functions for libns_winbind.so and libns_wins.so
(This used to be commit cec5bf5f99ca98b0f5ee84636fc18f0cb3079683)

20 years agoFix bug #226. Stop unmangle of name into a wildcard name from deleting more
Jeremy Allison [Tue, 29 Jul 2003 19:16:59 +0000 (19:16 +0000)]
Fix bug #226. Stop unmangle of name into a wildcard name from deleting more
than was intended.
Jeremy.
(This used to be commit e2742e0d897a35820a7d8f184292c32a4c3952e3)

20 years agoAdd NT quotas support. Users allowed now to manage quotas on systems with sysquotas...
Alexander Bokovoy [Tue, 29 Jul 2003 18:07:13 +0000 (18:07 +0000)]
Add NT quotas support. Users allowed now to manage quotas on systems with sysquotas interface detected (Linux at least) using native Windows tools. Also move default quota support for NT quotas to VFS module default_quota. Code by Metze
(This used to be commit e856a96c2c42c39843e5e1a3a6b0d538e7179900)

20 years agoFinish tridge's patch as referenced here :
Jeremy Allison [Tue, 29 Jul 2003 17:34:20 +0000 (17:34 +0000)]
Finish tridge's patch as referenced here :

make sure we don't allow the creation of directories containing
wildcard characters. I've only put this in mkdir at the moment, but I
suspect this will apply to all places that can create new filenames.

We need to allow the opening of existing filenames that contain
wildcards, but not allow the creation of new ones.

Jeremy.
(This used to be commit 7f111e545d198faa5fa89f6d360db0d5c32a8bd7)

20 years agoTypo on my part. I typed KRB5_KDB_BAD_ENCTYPE when I meant to type KRB5_BAD_ENCTYPE.
Jeremy Allison [Tue, 29 Jul 2003 17:03:51 +0000 (17:03 +0000)]
Typo on my part. I typed KRB5_KDB_BAD_ENCTYPE when I meant to type KRB5_BAD_ENCTYPE.
Heimdal has the latter, not the former.
Jeremy.
(This used to be commit e8425df77c2e917c819592d93833a164ee3b5338)

20 years agoFix the build on Heimdal. KRB5_KDB_BAD_ENCTYPE doesn't exist on Heimdal, and
Jim McDonough [Tue, 29 Jul 2003 16:34:31 +0000 (16:34 +0000)]
Fix the build on Heimdal.  KRB5_KDB_BAD_ENCTYPE doesn't exist on Heimdal, and
it's a different rc than KRB5_BAD_ENCTYPE (which exists on both MIT and
Heimdal).  This will just make the debug show up at level 3 always.

Jeremy, you may want to revisit this, but it's probably not worth the hassle.
(This used to be commit 4ff322ccf9c8485bcfe67e658d48f190f03547b0)