Andrew Tridgell [Mon, 22 May 2006 06:17:10 +0000 (06:17 +0000)]
r15796: this talloc_reference() was causing a memory leak on every NTVFS
request. The problem is that the talloc_free() in
smbsrv_send_reply_nosign() frees this reference, but doesn't free the
memory iself, thus leaving it attached to the connection context.
Metze, what was the idea behind this reference? If we need it, then we
need a corresponding talloc_free() to go with it.
Simo Sorce [Mon, 22 May 2006 03:55:01 +0000 (03:55 +0000)]
r15795: Try to use the async code by default
It passess all my tests, but I still need to work on a lot of stuff.
Shouldn't impact anybody else work, so I want to commit now and see what happens
Will work to remove the old code from modules and backends soon, and make some
more restyling in ldb internals.
So, if there is something you don't like in this desgin please speak now.
Simo.
Andrew Tridgell [Mon, 22 May 2006 02:07:11 +0000 (02:07 +0000)]
r15794: fixed a problem with DOS status codes - found by kukks (thanks!)
Simo Sorce [Sun, 21 May 2006 23:58:09 +0000 (23:58 +0000)]
r15789: hmm, damn, testing uncovcer all your bugs :-)
Simo Sorce [Sun, 21 May 2006 23:57:19 +0000 (23:57 +0000)]
r15788: Do not crash when no result is returned
Simo Sorce [Sun, 21 May 2006 21:25:31 +0000 (21:25 +0000)]
r15787: NET-API-CREATEUSER was crashing for me because of this uninitialized value...
now it still does not work (auth failure ?!) but at least it does not crash
Simo Sorce [Sun, 21 May 2006 21:24:11 +0000 (21:24 +0000)]
r15786: another fix in rdn_name
Rafal Szczesniak [Sun, 21 May 2006 21:19:34 +0000 (21:19 +0000)]
r15785: Change the test names according to changes made in libnet torture test.
rafal
Rafal Szczesniak [Sun, 21 May 2006 21:18:35 +0000 (21:18 +0000)]
r15784: Redesign of NET-API-RPCCONNECT tests to cover more cases of usage.
Right now, each level of libnet_RpcConnect can be tested with
correct and incorrect credentials (testing against expected nt
status). Various transports can be tested on NET-API-RPCCONN-BIND
test which does testing on explicitly passed binding string.
Tested services include LSA, SAMR, SRVSVC at the moment.
rafal
Simo Sorce [Sun, 21 May 2006 20:21:34 +0000 (20:21 +0000)]
r15783: Fix previous commit, was the wrong way to deal with the problem
Simo Sorce [Sun, 21 May 2006 20:06:01 +0000 (20:06 +0000)]
r15782: More fixes for async cases
Jelmer Vernooij [Sun, 21 May 2006 15:35:33 +0000 (15:35 +0000)]
r15780: Fix indentation of deferred code (fixed #3739)
Jelmer Vernooij [Sun, 21 May 2006 15:02:01 +0000 (15:02 +0000)]
r15779: Remove keepref from atsvc and efs.
Jelmer Vernooij [Sun, 21 May 2006 15:01:47 +0000 (15:01 +0000)]
r15778: Fix bug printing ref pointers.
Jelmer Vernooij [Sun, 21 May 2006 14:59:02 +0000 (14:59 +0000)]
r15777: Fix unresolved symbols in shared library build.
Jelmer Vernooij [Sun, 21 May 2006 12:58:39 +0000 (12:58 +0000)]
r15776: Don't generate ref pointers in Samba4-generated code. There is no point
in having pointers for outgoing data when you can already modify the top-level
element.
This can be overridden (temporarily) by specifying the new "keepref"
attribute. Once we've removed keepref from all IDL files, I'll remove this
attribute as well.
Stefan Metzmacher [Sun, 21 May 2006 12:57:36 +0000 (12:57 +0000)]
r15775: add some privilege related WERROR codes
metze
Stefan Metzmacher [Sun, 21 May 2006 12:56:49 +0000 (12:56 +0000)]
r15774: take care of the SYSTEM_SECURITY flag
metze
Stefan Metzmacher [Sun, 21 May 2006 12:15:04 +0000 (12:15 +0000)]
r15773: - implement srvsvc_NetGetFileSecurity() and srvsvc_NetSetFileSecurity()
- we do this by just creating a ntvfs_context (this doesn't need and smbsrv_* stuff :-)
and then call ntvfs_qpathinfo() and ntvfs_setpathinfo()
metze
Stefan Metzmacher [Sun, 21 May 2006 11:39:50 +0000 (11:39 +0000)]
r15772: pass the messaging context and server_id to the dcerpc server subsystem
metze
Stefan Metzmacher [Sun, 21 May 2006 10:42:29 +0000 (10:42 +0000)]
r15771: this should be a o32s32 blob
metze
Stefan Metzmacher [Sun, 21 May 2006 10:13:49 +0000 (10:13 +0000)]
r15770: when there's a dynamic body, we need to send the first byte even if the
dynamic size if 0
metze
Stefan Metzmacher [Sun, 21 May 2006 10:06:15 +0000 (10:06 +0000)]
r15769: this functions belong to params/ so move it.
metze
Stefan Metzmacher [Sun, 21 May 2006 08:29:56 +0000 (08:29 +0000)]
r15768: make it possible that the caller of the ntvfs_subsystem doesn't
provide handle callbacks
metze
Stefan Metzmacher [Sun, 21 May 2006 08:24:11 +0000 (08:24 +0000)]
r15767: we need to free here
metze
Stefan Metzmacher [Sun, 21 May 2006 07:14:18 +0000 (07:14 +0000)]
r15766: fix parsing of srvsvc_NetGetFileSecurity() and srvsvc_NetSetFileSecurity()
metze
Simo Sorce [Sat, 20 May 2006 19:38:01 +0000 (19:38 +0000)]
r15762: It make no sense for rdn_name to implement modify, it will need to implement rename ...
Simo Sorce [Sat, 20 May 2006 19:37:21 +0000 (19:37 +0000)]
r15761: Fix-as-you-go ...
Testing various async paths and uncovering bugs
Stefan Metzmacher [Sat, 20 May 2006 19:03:06 +0000 (19:03 +0000)]
r15760: - add RAW_READ_SMB2 => generic mapping
- Implement SMB2 Read
metze
Stefan Metzmacher [Sat, 20 May 2006 19:00:53 +0000 (19:00 +0000)]
r15759: samba4 currently only supports read sizes up to with UINT16_MAX
metze
Stefan Metzmacher [Sat, 20 May 2006 18:57:32 +0000 (18:57 +0000)]
r15758: - handle RAW_FLUSH_SMB2 in the posix ntvfs backend
- Implement SMB2 Flush
metze
Stefan Metzmacher [Sat, 20 May 2006 18:56:08 +0000 (18:56 +0000)]
r15757: - add RAW_WRITE_SMB2 => generic mapping
- implement SMB2 Write
metze
Stefan Metzmacher [Sat, 20 May 2006 18:34:59 +0000 (18:34 +0000)]
r15756: handle RAW_OPEN_SMB2
metze
Stefan Metzmacher [Sat, 20 May 2006 17:20:34 +0000 (17:20 +0000)]
r15754: - implement SMB2 Close
- add RAW_CLOSE_SMB2 generic mapping
metze
Stefan Metzmacher [Sat, 20 May 2006 17:06:28 +0000 (17:06 +0000)]
r15753: implement SMB2 Create
metze
Stefan Metzmacher [Sat, 20 May 2006 16:53:44 +0000 (16:53 +0000)]
r15752: - add generic mapping for RAW_OPEN_SMB2
metze
Stefan Metzmacher [Sat, 20 May 2006 16:48:29 +0000 (16:48 +0000)]
r15751: thanks to talloc_get_type() I noticed that I used smbsrv_request in the smb2srv code
metze
Stefan Metzmacher [Sat, 20 May 2006 16:10:22 +0000 (16:10 +0000)]
r15750: don't clear after setting
metze
Stefan Metzmacher [Sat, 20 May 2006 12:54:55 +0000 (12:54 +0000)]
r15749: - pass in a dummy oplock callback
- pass in SMB2 ntvfs_handle callbacks
metze
Stefan Metzmacher [Sat, 20 May 2006 12:25:45 +0000 (12:25 +0000)]
r15748: - implement SMB2 TreeConnect server code
TODO: pass in oplock and handle callbacks, but as we don't do file access
yet, it's no problem to skip them
metze
Stefan Metzmacher [Sat, 20 May 2006 12:15:05 +0000 (12:15 +0000)]
r15747: move some functions to the protocol independend part of the
smb server
metze
Stefan Metzmacher [Sat, 20 May 2006 12:11:46 +0000 (12:11 +0000)]
r15746: - reorder elements of smb2srv_request
- move SMB2 specific elements to the end
metze
Stefan Metzmacher [Sat, 20 May 2006 12:08:27 +0000 (12:08 +0000)]
r15745: - reorder the elements of smbsrv_request
- move the SMB specific elements to the end
metze
Stefan Metzmacher [Sat, 20 May 2006 11:50:10 +0000 (11:50 +0000)]
r15744: convert_string_talloc() handles src_len == 0 as error
but it's valid in this case
metze
Stefan Metzmacher [Sat, 20 May 2006 11:03:51 +0000 (11:03 +0000)]
r15743: use union smb_tcon
metze
Stefan Metzmacher [Sat, 20 May 2006 10:59:26 +0000 (10:59 +0000)]
r15742: use union smb_sessetup
metze
Stefan Metzmacher [Sat, 20 May 2006 10:46:38 +0000 (10:46 +0000)]
r15741: move smb2 request structures into the main smb request structs
as new levels
metze
Stefan Metzmacher [Sat, 20 May 2006 10:40:10 +0000 (10:40 +0000)]
r15740: add TODO, that we should check if the server supports
large offsets, before sending large offset requests
metze
Stefan Metzmacher [Sat, 20 May 2006 10:31:23 +0000 (10:31 +0000)]
r15739: return False when the function is called with an invalid level
metze
Stefan Metzmacher [Sat, 20 May 2006 09:54:10 +0000 (09:54 +0000)]
r15738: make the code more readable
metze
Stefan Metzmacher [Sat, 20 May 2006 09:38:59 +0000 (09:38 +0000)]
r15737: fix a typos and comment
metze
Stefan Metzmacher [Sat, 20 May 2006 08:15:22 +0000 (08:15 +0000)]
r15734: This is a major change to the NTVFS subsystem:
- to use a struct ntvfs_handle instead of a uint16_t fnum.
(to make it independend from the frontend protocol)
- the allocation of handles now is provided by the frontend
(smbsrv_*) via callbacks and not by each backend module
- this also makes sure that file handles are only passed
to the ntvfs subsystem when the tcon and session matches,
so modules can rely on this and need to check this.
- this allows multiple modules in the ntvfs module chain
to allocate file handles. This can be used for virtual
files like "\\$Extend\\$Quota:$Q:$INDEX_ALLOCATION"...
- also this will make SMB2 with 128 bit file handles possible
metze
Andrew Tridgell [Sat, 20 May 2006 03:08:44 +0000 (03:08 +0000)]
r15731: module init functions should return NTSTATUS, not void
Rafal Szczesniak [Fri, 19 May 2006 21:28:28 +0000 (21:28 +0000)]
r15726: Comments and a couple of minor formattings in the code.
rafal
Simo Sorce [Fri, 19 May 2006 21:12:26 +0000 (21:12 +0000)]
r15725: First shot at making password_hash async
The async path is not yet enabled by default so it should make no harm
Jim McDonough [Fri, 19 May 2006 18:37:35 +0000 (18:37 +0000)]
r15719: Fix build on systems (AIX) that don't have vsyslog or strcasestr, with
--enable-developer on. syslog() and toupper() required more includes.
Someone more familiar with samba4 builds should verify this, please.
Stefan Metzmacher [Fri, 19 May 2006 15:10:39 +0000 (15:10 +0000)]
r15718: - split the SMBflush with the 0xFFFF wildcard fnum into a different level
metze
Stefan Metzmacher [Fri, 19 May 2006 14:32:42 +0000 (14:32 +0000)]
r15716: disable the cifs_posix ntvfs backend as it will not be updated
with coming ntvfs subsystem changes
metze
Stefan Metzmacher [Fri, 19 May 2006 14:25:15 +0000 (14:25 +0000)]
r15715: add my copyright
metze
Stefan Metzmacher [Fri, 19 May 2006 14:15:21 +0000 (14:15 +0000)]
r15714: fix async handling of ntioctl calls and push the DATA_BLOB into the reply
metze
Stefan Metzmacher [Fri, 19 May 2006 14:10:14 +0000 (14:10 +0000)]
r15713: - initialize ntioctl.in.max_data and ntioctl.in.blob
- use define instead of hex value
metze
Stefan Metzmacher [Fri, 19 May 2006 13:58:06 +0000 (13:58 +0000)]
r15711: - make use of nttrans_setup_reply()
metze
Stefan Metzmacher [Fri, 19 May 2006 13:44:39 +0000 (13:44 +0000)]
r15710: - we no longer need to set the nttrans out fields to zero
in the backends
- some whitespace cleanups
metze
Stefan Metzmacher [Fri, 19 May 2006 13:37:43 +0000 (13:37 +0000)]
r15709: - return an error is better than ignoring a non supported request
- zero nttrans output fields before calling the backends
metze
Stefan Metzmacher [Fri, 19 May 2006 13:32:17 +0000 (13:32 +0000)]
r15708: setup count is 8bit and param_count and data_count are 32bit
metze
Volker Lendecke [Fri, 19 May 2006 07:47:28 +0000 (07:47 +0000)]
r15706: This is a very specific test that helps me in making sure I don't screw up too
badly in getting rid of set_saved_ntstatus in Samba3. If this is not the
appropriate place to put it, please tell me.
Volker
Jelmer Vernooij [Thu, 18 May 2006 21:28:54 +0000 (21:28 +0000)]
r15702: Fix typo
Stefan Metzmacher [Thu, 18 May 2006 12:34:37 +0000 (12:34 +0000)]
r15694: for the cifs backend it's desired that we pull and repush
the packets with our parsing code for the most stuff.
So make cifs:maptrans2=yes the default.
metze
Stefan Metzmacher [Thu, 18 May 2006 12:32:23 +0000 (12:32 +0000)]
r15693: fix the logic for the "cifs:map_trans2" parameter
we need to return NOT_IMPLEMENTED when mapping is desired
metze
Stefan Metzmacher [Thu, 18 May 2006 12:17:56 +0000 (12:17 +0000)]
r15690: - disable the lock cancel test against samba3
- disable the LOCK_NOT_GRANTED vs. FILE_LOCK_CONFLICT test against samba3
metze
Stefan Metzmacher [Thu, 18 May 2006 11:21:17 +0000 (11:21 +0000)]
r15689: dummy commit
metze
Stefan Metzmacher [Thu, 18 May 2006 11:15:18 +0000 (11:15 +0000)]
r15688: windows sends 4 as max_setup count on NT IOCTL
metze
Stefan Metzmacher [Thu, 18 May 2006 09:29:08 +0000 (09:29 +0000)]
r15685: let samba3 pass RAW-READ and ignore one bug
this demonstrates how the target:samba3=yes option
passed in samba3's make test can be used in samba4's smbtorture
metze
Stefan Metzmacher [Thu, 18 May 2006 09:22:57 +0000 (09:22 +0000)]
r15684: fix option passing
metze
Rafal Szczesniak [Wed, 17 May 2006 22:21:24 +0000 (22:21 +0000)]
r15667: Make sure the rpc pipe pointers are zeroed during
initialisation of libnet_context.
This fixes a valgrind warning.
rafal
Rafal Szczesniak [Wed, 17 May 2006 22:07:32 +0000 (22:07 +0000)]
r15666: Fix a silly typo.
rafal
Rafal Szczesniak [Wed, 17 May 2006 21:45:00 +0000 (21:45 +0000)]
r15665: ejsnet test should be better now, after changes in delete
user routine.
rafal
Rafal Szczesniak [Wed, 17 May 2006 21:44:06 +0000 (21:44 +0000)]
r15664: Add NET-API-DELETEUSER test for libnet_DeleteUser function.
Only the sync version right now.
rafal
Rafal Szczesniak [Wed, 17 May 2006 21:41:59 +0000 (21:41 +0000)]
r15663: Turn libnet_DeleteUser into async function and fix subtle bug probably
causing ejsnet test to segfault. Also, cleanup a bit and add monitor
fn pointer to internal user delete libnet function.
Time for some comments now.
rafal
Stefan Metzmacher [Wed, 17 May 2006 15:16:46 +0000 (15:16 +0000)]
r15661: add NT_STATUS_OBJECTID_NOT_FOUND
metze
Jelmer Vernooij [Wed, 17 May 2006 13:58:53 +0000 (13:58 +0000)]
r15659: Implement opening a key by name (significant better performance
in some situations). Patch by Wilco Baan Hofman.
Stefan Metzmacher [Wed, 17 May 2006 09:52:14 +0000 (09:52 +0000)]
r15656: for NT IOCTL's we need to control the max_data field for some calls
metze
Jelmer Vernooij [Tue, 16 May 2006 23:51:01 +0000 (23:51 +0000)]
r15653: Remove idl_types.h include where possible. Remove
types from .h file that are now in pidls' aliases list.
Jelmer Vernooij [Tue, 16 May 2006 23:49:31 +0000 (23:49 +0000)]
r15652: Fix aliases mechanism (fixes #3710)
Jelmer Vernooij [Tue, 16 May 2006 22:25:13 +0000 (22:25 +0000)]
r15651: Add 'MANUAL' command in Ethereal conformance files. This allows overriding
individual functions for specific element levels without losing the other
functions for an element or the hf/ett fields allocated.
Fixes #3738.
Jelmer Vernooij [Tue, 16 May 2006 22:21:44 +0000 (22:21 +0000)]
r15650: Fix perl warning when generating IDL for elements with no properties
Jelmer Vernooij [Tue, 16 May 2006 19:54:31 +0000 (19:54 +0000)]
r15648: Update MANIFEST to install Samba3 code generators.
Stefan Metzmacher [Tue, 16 May 2006 16:59:08 +0000 (16:59 +0000)]
r15643: add some FSCTL_ codes
metze
Stefan Metzmacher [Tue, 16 May 2006 16:57:56 +0000 (16:57 +0000)]
r15642: add an '--always-create' option to mkproto.pl
with this the proto headers will be always regenerated,
even if the content hasn't changed
you can use currently by
#>make MK_PROTO_ALWAYS_CREATE=yes'
or
#>export MK_PROTO_ALWAYS_CREATE=yes
#>make
metze
Stefan Metzmacher [Tue, 16 May 2006 16:51:34 +0000 (16:51 +0000)]
r15641: we know have source/version.h
metze
Stefan Metzmacher [Tue, 16 May 2006 16:50:50 +0000 (16:50 +0000)]
r15640: - NT IOCTL calls also have an 'in' data_blob
- fix the receive code of NT IOCTL's we only need the NTTRANS data payload
not the SMB payload here
metze
Stefan Metzmacher [Tue, 16 May 2006 16:43:34 +0000 (16:43 +0000)]
r15639: fix warnings
metze
Stefan Metzmacher [Tue, 16 May 2006 16:28:36 +0000 (16:28 +0000)]
r15638: add a usefull macro to get a pointer the smb_handle union
of smb_open, as it's not nicely alligned for all levels,
If someone has an idea for a better solution where we can access
it via op->generic.out.file.* please let me know:-)
metze
Stefan Metzmacher [Tue, 16 May 2006 00:37:13 +0000 (00:37 +0000)]
r15629: fix some warnings
metze
Stefan Metzmacher [Tue, 16 May 2006 00:35:57 +0000 (00:35 +0000)]
r15628: fix compiler warning
metze
Rafal Szczesniak [Mon, 15 May 2006 22:35:17 +0000 (22:35 +0000)]
r15627: Disabling test of ejsnet till the time I find the reason why
it's segfaulting. It's late so I just don't want to keep the
build farm all in red PANIC messages.
rafal
Rafal Szczesniak [Mon, 15 May 2006 21:50:53 +0000 (21:50 +0000)]
r15626: Modify the tests to fit them in current changes in libnet
functions.
rafal
Rafal Szczesniak [Mon, 15 May 2006 21:49:27 +0000 (21:49 +0000)]
r15625: Partial commit of my current work. It makes libnet api functions
a bit more smart and more aware of what libnet_context can offer.
The context is a help when some of the arguments are not passed
(programmer counts on using sensible defaults) and stores some of
results so that similar subsequent calls don't need to reopen some
of policy handles, pipes, etc. again. It also helps to hide some
of details the library user don't really want to know much about.
Also, change domain open function to be part of public api, as
it is going to be used in ejsnet interface.
Note, this is work in progress. Comments are welcome.
rafal
Stefan Metzmacher [Mon, 15 May 2006 14:38:04 +0000 (14:38 +0000)]
r15624: add some useful helper functions
metze
Stefan Metzmacher [Mon, 15 May 2006 13:12:01 +0000 (13:12 +0000)]
r15622: pass 'target:samba4=yes' to samba4's smbtorture when running samba4's make test
we can use this in samba4's smbtorture to disable tests sections which doesn't pass
against samba4
metze