Jeremy Allison [Tue, 27 Jul 2010 07:23:37 +0000 (00:23 -0700)]
Second part of fix for bug 7578 - 'net idmap restore' fails to set HWM, causing duplicates.
Jeremy.
Karolin Seeger [Tue, 27 Jul 2010 07:02:36 +0000 (09:02 +0200)]
WHATSNEW: Update changes.
Thanks to Holger for providing!
Karolin
Justin Maggard [Tue, 27 Jul 2010 06:42:04 +0000 (08:42 +0200)]
s3: Fix bug 7578
Uninitialized variable read in _wbint_SetHWM
Volker Lendecke [Fri, 23 Jul 2010 18:08:58 +0000 (20:08 +0200)]
s3: Early start of an async nbench
The current nbench implementations have the problem that they fork a child
per simulated client. With hundreds or thousands of clients this can put
quite some load on the client. This test (when finished) will read the
standard dbench client.txt and run completely async from within one
process.
Volker
Volker Lendecke [Mon, 26 Jul 2010 08:35:15 +0000 (10:35 +0200)]
s3: Convert cli_qpathinfo_streams to cli_qpathinfo_send
Volker Lendecke [Mon, 26 Jul 2010 08:08:38 +0000 (10:08 +0200)]
s3: Factor out parse_streams_blob
Volker Lendecke [Mon, 26 Jul 2010 07:41:25 +0000 (09:41 +0200)]
s3: Convert cli_qpathinfo_basic to cli_qpathinfo_send
Volker Lendecke [Mon, 26 Jul 2010 07:27:11 +0000 (09:27 +0200)]
s3: Convert cli_qpathinfo2 to cli_qpathinfo_send
Volker Lendecke [Mon, 26 Jul 2010 07:21:17 +0000 (09:21 +0200)]
s3: Use error status instead of cli_errstr
Volker Lendecke [Mon, 26 Jul 2010 20:57:47 +0000 (22:57 +0200)]
s3: Callers of cli_qpathinfo_recv might ignore the output
Volker Lendecke [Fri, 23 Jul 2010 10:22:23 +0000 (12:22 +0200)]
s3: Log the correct core path on Linux if core_pattern is set
Contributions from Bill Hymas, IBM
Volker Lendecke [Mon, 26 Jul 2010 13:04:52 +0000 (15:04 +0200)]
s3: fix a typo
Volker Lendecke [Mon, 26 Jul 2010 07:11:40 +0000 (09:11 +0200)]
s3: cli_qpathinfo2 expects at least 68 bytes
Volker Lendecke [Mon, 26 Jul 2010 06:34:35 +0000 (08:34 +0200)]
s3: Convert cli_qpathinfo1 to cli_qpathinfo
Volker Lendecke [Mon, 26 Jul 2010 05:49:58 +0000 (07:49 +0200)]
s3: Fix a structure mess-up
I wonder why the compiler did not complain -- maybe because
the structs have the same data members? No clue.
Volker Lendecke [Mon, 26 Jul 2010 05:47:39 +0000 (07:47 +0200)]
s3: Remove some unused struct members
Andrew Tridgell [Mon, 26 Jul 2010 01:15:14 +0000 (11:15 +1000)]
lib: added samba-util.pc creation
Thanks to Brad Hards for this patch
Volker Lendecke [Sun, 25 Jul 2010 19:32:13 +0000 (21:32 +0200)]
s3: Convert cli_get_ea_list_path to cli_qpathinfo_send
Volker Lendecke [Sun, 25 Jul 2010 17:42:22 +0000 (19:42 +0200)]
s3: Factor out parse_ea_blob
Volker Lendecke [Sun, 25 Jul 2010 15:49:20 +0000 (17:49 +0200)]
s3: Convert cli_posix_stat to cli_qpathinfo_send
Volker Lendecke [Sun, 25 Jul 2010 15:46:01 +0000 (17:46 +0200)]
s3: Convert cli_posix_getfacl to cli_qpathinfo_send
Volker Lendecke [Sun, 25 Jul 2010 15:37:16 +0000 (17:37 +0200)]
s3: Convert cli_posix_readlink to cli_qpathinfo_send
Volker Lendecke [Sun, 25 Jul 2010 15:26:43 +0000 (17:26 +0200)]
s3: Add async cli_qpathinfo
Volker Lendecke [Sun, 25 Jul 2010 11:19:34 +0000 (13:19 +0200)]
s3: cli_qpathinfo->cli_qpathinfo1
Volker Lendecke [Sun, 25 Jul 2010 17:12:28 +0000 (19:12 +0200)]
s3: Fix cli_posix_stat
nlink seems to be defined as 8 bytes, not 4
Jeremy, please check!
Volker Lendecke [Sun, 25 Jul 2010 17:11:30 +0000 (19:11 +0200)]
s3: use dump_data in query FILE_BASIC_UNIX_INFO
Jeremy, please check! To me this is a lot more readable than the decimal
listing of the bytes.
Volker Lendecke [Sat, 24 Jul 2010 11:17:15 +0000 (13:17 +0200)]
s3: Fix a typo (missing space)
Jeremy Allison [Fri, 23 Jul 2010 17:54:46 +0000 (10:54 -0700)]
Fix bug 7583 - Smbclient fails to kerberos connect to a Alfresco JLAN CIFS Server
Correctly calculate the gssapi channel binding checkum.
Jeremy
Signed off by: simo <idra@samba.org>
Björn Jacke [Tue, 20 Jul 2010 17:48:39 +0000 (19:48 +0200)]
s3: not use as-needed by default and auto use if enable-developer is active
as discussed on samba-technical, we don't want people out there to use it
unwittingly by default. However by enabling it when enable-developer is used we
make sure that we don't break the feature.
Simo Sorce [Sat, 17 Jul 2010 20:34:33 +0000 (16:34 -0400)]
s3-rpc_server: Cleanup the way we free pipe auth data
Simo Sorce [Mon, 19 Jul 2010 19:43:22 +0000 (15:43 -0400)]
cleaup: Reformat cli_get_session_key()
merge cli_get_session_key
Simo Sorce [Sat, 17 Jul 2010 20:01:07 +0000 (16:01 -0400)]
s3-dcerpc: Fix potential memleak
Use a free function for schannel data too for now.
Simo Sorce [Tue, 20 Jul 2010 17:56:25 +0000 (13:56 -0400)]
s3-dcerpc: Properly initialize output buffers on nak/fault
Simo Sorce [Wed, 21 Jul 2010 00:00:12 +0000 (20:00 -0400)]
s3-libsmb: Use data_blob_talloc to get krb5 ticket and session keys
Simo Sorce [Tue, 20 Jul 2010 23:45:00 +0000 (19:45 -0400)]
misc: cleanup get_krb5_smb_session_key()
Simo Sorce [Tue, 20 Jul 2010 23:41:19 +0000 (19:41 -0400)]
misc: cleanup cli_krb5_get_ticket()
Jeremy Allison [Tue, 20 Jul 2010 23:17:58 +0000 (16:17 -0700)]
Add approriate TALLOC_CTX's thoughout the spnego code. No more implicit NULL contexts.
Jeremy.
Jeremy Allison [Tue, 20 Jul 2010 22:12:07 +0000 (15:12 -0700)]
Fix warning on 64-bit machines.
Jeremy.
Jeremy Allison [Tue, 20 Jul 2010 21:59:31 +0000 (14:59 -0700)]
Fix one more data_blob -> data_blob_talloc. Move away from implicit NULL context tallocs.
Jeremy.
Jeremy Allison [Tue, 20 Jul 2010 20:35:43 +0000 (13:35 -0700)]
Add TALLOC_CTX argument to spnego_parse_negTokenInit, reduce
use of malloc, and data_blob().
Jeremy.
Volker Lendecke [Tue, 20 Jul 2010 19:54:45 +0000 (21:54 +0200)]
s3: Fix two uninitialized variables
Jeremy Allison [Tue, 20 Jul 2010 18:14:49 +0000 (11:14 -0700)]
Rename spnego_gen_negTokenTarg() -> spnego_gen_krb5_negTokenInit()
as this correctly describes what this function does.
Jeremy.
Jeremy Allison [Tue, 20 Jul 2010 18:04:19 +0000 (11:04 -0700)]
Remove gen_negTokenTarg(), as it's not actually creating a TokenTarg frame, but a TokenInit one.
Move to using spnego_gen_negTokenInit() instead.
Jeremy
Jeremy Allison [Tue, 20 Jul 2010 16:51:48 +0000 (09:51 -0700)]
Fixes pointed out by <david.kondrad@legrand.us>. Free memory if not being
returned to caller. Remove unneeded asn1_tag_remaining() calls.
Jeremy.
Jeremy Allison [Tue, 20 Jul 2010 16:34:32 +0000 (09:34 -0700)]
Fix warning - no return value for a non-void fn.
Jeremy.
Björn Jacke [Tue, 20 Jul 2010 16:03:13 +0000 (18:03 +0200)]
s3: also run the test prog from the explicit external lib bug check
on Linux linking already stops with a broken ld, on FreeBSD8 linking works but
the binary is broken, so we must check both.
this should be the final fix for bug #7561
Simo Sorce [Mon, 19 Jul 2010 22:57:12 +0000 (18:57 -0400)]
dcerpc: Clarify what the arguments of dcerpc_pull_auth_trailer() actually represent.
Björn Jacke [Tue, 20 Jul 2010 11:39:59 +0000 (13:39 +0200)]
s3: check for explicit external lib bug in ld's as-needed code
Second part of fix for bug #7561
Based on test provided by Olaf Flebbe.
Björn Jacke [Tue, 20 Jul 2010 11:35:03 +0000 (13:35 +0200)]
s3: add configure swtich to manually disable as-needed
first part fix for bug #7561
based on earlier patch from metze
Björn Jacke [Tue, 20 Jul 2010 09:13:51 +0000 (11:13 +0200)]
Revert "s3: Don't use as-needed"
This reverts commit
b9835a1f9d8b81e4121a01a190dffaefe67cbe5c.
Simo Sorce [Sat, 17 Jul 2010 20:58:50 +0000 (16:58 -0400)]
s3-dcerpc: Unifiy cli_pipe_auth_data and pipe_auth_data
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Mon, 19 Jul 2010 17:39:01 +0000 (13:39 -0400)]
s3-auth: Add auth_ntlmssp wrapper for client side
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Mon, 19 Jul 2010 17:36:33 +0000 (13:36 -0400)]
s3-auth: Move auth_ntlmssp wrappers in their own file
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Fri, 16 Jul 2010 22:33:38 +0000 (18:33 -0400)]
s3-auth: Refactor and rename auth_ntlmssp_server_info()
Rename it to auth_ntlmssp_steal_server_info() to make it clear that
the server_info struct is stolen from the auth_ntlmssp_state structure.
Use talloc_move instead of manual steal&clear
Add comments to explain what is going on.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Jeremy Allison [Tue, 20 Jul 2010 00:14:26 +0000 (17:14 -0700)]
Remove gen_negTokenInit() - change all callers to spnego_gen_negTokenInit().
We now have one function to do this in all calling code. More rationalization
to follow.
Jeremy.
Jeremy Allison [Mon, 19 Jul 2010 23:45:16 +0000 (16:45 -0700)]
Move the addition of the 16 byte guid out of spnego_gen_negTokenInit() and
into negprot_spnego() where it belongs (it's not an SPNEGO operation).
Add a TALLOC_CTX for callers of negprot_spnego(). Closer to unifying all
the gen_negTokenXXX calls.
Jeremy.
Jeremy Allison [Mon, 19 Jul 2010 22:41:45 +0000 (15:41 -0700)]
Remove parse_negTokenTarg(), as it's actually incorrect. We're processing
negTokenInit's here. Use common code in spnego_parse_negTokenInit().
Jeremy.
Jeremy Allison [Mon, 19 Jul 2010 21:21:01 +0000 (14:21 -0700)]
Fix [Bug 7577] SPNEGO auth fails when contacting Win7 system using Microsoft Live Sign-in Assistant
Based on code from <david.kondrad@legrand.us>. Cope with every NegTokenInit ::= SEQUENCE value.
Jeremy.
Volker Lendecke [Mon, 19 Jul 2010 14:15:01 +0000 (16:15 +0200)]
s3: Avoid pwrite calls for the 1-byte writes with zero content
When a write cache is empty and we write one 0, there is no point in actually
writing into the write cache, which would trigger a potentially superfluous
write cache flush. We can rely on posix semantics to give us 0's when
subsequent read calls want to read from file areas not written yet.
Jeremy, please check!
Simo Sorce [Mon, 19 Jul 2010 17:53:34 +0000 (13:53 -0400)]
s3-auth: Remove unused variable.
It was a spurious remnant after a rebase.
Simo Sorce [Mon, 19 Jul 2010 17:48:31 +0000 (13:48 -0400)]
Merge branch 'master' of ssh://git.samba.org/data/git/samba
Andrew Tridgell [Mon, 19 Jul 2010 04:26:20 +0000 (14:26 +1000)]
waf: make the error msg when gen_ndr directory is missing clearer
the next time someone removes gen_ndr/README the fix should be more
obvious.
Andrew Tridgell [Mon, 19 Jul 2010 04:20:33 +0000 (14:20 +1000)]
waf-idl: put the gen_ndr/README file for source3 back
This was removed by
100d37fc4624690423f6a932709b3f9046d05c64, probably
accidentially.
Andreas Schneider [Fri, 16 Jul 2010 12:52:42 +0000 (14:52 +0200)]
s3-rpc_server: Added callbacks for init and shutdown of a rpc service.
This adds two callback function for each rpc service. One is for
initialisation and the other for shutdown. rpc_<service>_unregister()
needs to be called to execute the shutdown function.
Andreas Schneider [Thu, 15 Jul 2010 10:47:19 +0000 (12:47 +0200)]
s3-rpc_server: Don't register the same rpc commands twice.
Volker Lendecke [Mon, 19 Jul 2010 10:47:09 +0000 (12:47 +0200)]
s3-docs: Add more verbose description of "log writeable files on exit".
Volker Lendecke [Mon, 19 Jul 2010 10:35:58 +0000 (12:35 +0200)]
s3-docs: Add more verbose description of "ctdb locktime warn threshold".
Volker Lendecke [Mon, 19 Jul 2010 10:22:05 +0000 (12:22 +0200)]
s3-docs: Add more verbose description of "username map cache time".
Karolin Seeger [Mon, 19 Jul 2010 09:54:05 +0000 (11:54 +0200)]
s3-docs: Remove older documentation of smb2 parameters.
Jeremy, these parameters were already documented. My fault.
Sorry for the noise!
Karolin
Kamen Mazdrashki [Fri, 16 Jul 2010 11:19:07 +0000 (14:19 +0300)]
s4: Remove trailing whitespaces
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:18:49 +0000 (14:18 +0300)]
ldb: Mark _DEPRECATED_ ldb_msg_diff() and ldb_msg_canonicalize() functions
They are not quite safe to use (requires caller to steal
resulting message in own context) and may lead to holding
memory for too long.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:18:14 +0000 (14:18 +0300)]
s4-dsdb: use ldb_msg_normalize() in ldb_msg_difference()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:16:38 +0000 (14:16 +0300)]
s4-test: Use ldb_msg_normalize() in sqlite3 backend
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:13:20 +0000 (14:13 +0300)]
s4-test: Use ldb_msg_normalize() in torture/rpc/dssync.c test
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:03:53 +0000 (14:03 +0300)]
s4-dsdb: use ldb_msg_normalize() in ldbadd-process_file()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:01:49 +0000 (14:01 +0300)]
s4-dsdb: use ldb_msg_normalize() in source4/lib/ldb/common/ldb.c
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:59:40 +0000 (13:59 +0300)]
s4-dsdb: use ldb_msg_normalize() in source4/dsdb/schema/schema_set.c
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:55:42 +0000 (13:55 +0300)]
s4-ldb: Add ldb_msg_normalize() to accept a memory context from client
Previos implementation from ldb_msg_canonicalize()
was moved into this function and now ldb_msg_canonicalize()
is based on ldb_msg_normalize()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:47:41 +0000 (13:47 +0300)]
s4-ldb: Use _ldb_msg_add_el() in ldb_msg_add()
Previous implementation was 'leaking' attribute name
string, that is allocated by ldb_msg_add_empty()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:46:05 +0000 (13:46 +0300)]
s4-ldb: Use _ldb_msg_add_el() in ldb_msg_add_empty()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:44:13 +0000 (13:44 +0300)]
s4-ldb: Add separate function to add empty element into ldb_msg
It just adds another element, nothing more.
Caller is responsible to fill-in the added element and
determine how to handle data allocation contexts.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:41:57 +0000 (13:41 +0300)]
s4-ldb: Write more explanatory comment for ldb_msg_add()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:40:50 +0000 (13:40 +0300)]
s4-pyldb: Use ldb_msg_difference() in py_ldb_msg_diff()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:38:09 +0000 (13:38 +0300)]
s4-test: Use ldb_msg_difference() in torture/rpc/dssync.c test
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:35:07 +0000 (13:35 +0300)]
s4-tools: use ldb_msg_difference() in ldbedit - modify_record()
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:27:30 +0000 (14:27 +0300)]
s4-dsdb/schema/schema_set.c: fix trailing spaces and comments spelling
Few comments split on several lines also...
(Sorry Metze, I know you hate reviewing "and this, and that"
type of patches, but those are just cosmetics)
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 10:26:45 +0000 (13:26 +0300)]
s4-dsdb: use ldb_msg_difference() in source4/dsdb/schema/schema_set.c
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Kamen Mazdrashki [Fri, 16 Jul 2010 11:23:24 +0000 (14:23 +0300)]
s4-ldb: Implement ldb_msg_difference() function to accept a memory context from client
Old implementation from ldb_msg_diff() was moved into
this this function but with changed interface
so that a memory context may be passed.
ldb_msg_diff() function is now based on ldb_msg_difference(),
which fixes a hidden leak - internal ldb_msg object
(returned from ldb_msg_canonicalize) wasn't freed
and stays attached to ldb_context for the connection lifetime.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Fri, 16 Jul 2010 23:44:22 +0000 (19:44 -0400)]
s3-auth: Use talloc hierarchies to properly free auth_ntlmssp_state contexts
Turn auth_ntlmssp_end into a destructor and attach it to auth_ntlmssp_state.
Remote auth_ntlmssp_end and use TALLOC_FREE in the callers.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Sat, 17 Jul 2010 14:22:04 +0000 (10:22 -0400)]
s3-auth: auth_make ntlmssp_state the parent context
There is no need for a separate mem_ctx member.
Also make the ntlmssp_state a children of auth_ntlmssp_state
Also cleanup auth_ntlmssp_end to free only what is not automatically freed
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Fri, 16 Jul 2010 22:23:55 +0000 (18:23 -0400)]
s3-auth: Simplify how we free the auth_context
Turn the freeing function into a destructor and attach it to the
auth_context.
Make all callers TALLOC_FREE() the auth_context instead of calling
the free function.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Fri, 16 Jul 2010 21:47:21 +0000 (17:47 -0400)]
s3-auth: Cleanup and readability fixes
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Fri, 16 Jul 2010 21:30:14 +0000 (17:30 -0400)]
s3-ntlmssp: Remove ntlmssp_end and let the talloc hierarchy handle it.
All the members are children of ntlmssp_state anyway.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Simo Sorce [Mon, 19 Jul 2010 00:04:42 +0000 (20:04 -0400)]
Merge branch 'master' of ssh://git.samba.org/data/git/samba
Volker Lendecke [Sun, 18 Jul 2010 15:12:30 +0000 (17:12 +0200)]
s3: Actually use the usecs in aio_fork_suspend
Jeremy, please check!
Volker Lendecke [Sun, 18 Jul 2010 15:12:11 +0000 (17:12 +0200)]
s3: Fix an uninitialized variable
Volker Lendecke [Sun, 18 Jul 2010 12:40:43 +0000 (14:40 +0200)]
s3: Work better without the aio sighandler
Refuse async I/O if we can't set up the signal handler
Volker Lendecke [Sun, 18 Jul 2010 11:39:51 +0000 (13:39 +0200)]
s3: Fix some nonempty blank lines
Volker Lendecke [Sat, 17 Jul 2010 21:16:26 +0000 (23:16 +0200)]
s3: Remove a direct use of procid_self()
Andreas Schneider [Sun, 18 Jul 2010 13:04:20 +0000 (15:04 +0200)]
s3-rpc_client: Fixed a segfault in rpccli_samr_chng_pswd_auth_crap().
This fixes the WINBIND-WBCLIENT test. The test set
old_lm_hash_enc_blob.length to 0 and we don't check the length here. So
the memcpy segfaulted.