Andrew Tridgell [Thu, 17 May 2007 08:47:04 +0000 (08:47 +0000)]
r22969: fix some more places where we could end up with more than one event
context. We now have an event context on the torture_context, and we
can also get one from the cli_credentials structure
Andrew Tridgell [Thu, 17 May 2007 07:52:33 +0000 (07:52 +0000)]
r22968: andrew bartlett pointed out that the parent process could add a fd to
the epoll context which would then appear in the children. To fix this
we need to check for pid changes in more places. Luckily on platforms
where we have epoll(), getpid() is very very cheap.
Andrew Bartlett [Thu, 17 May 2007 05:46:45 +0000 (05:46 +0000)]
r22967: Move to the TCP packet interface for the krb5_send_to_kdc plugin.
This replaces a lump of hand-crafted code with the generic packet
system used in the rest of Samba4.
(I started this while chasing down the epoll bug, which turned out to
be seperate)
Andrew Bartlett [Thu, 17 May 2007 05:44:51 +0000 (05:44 +0000)]
r22966: Make sure to return LOGON_FAILURE if the user's kerberos password is
incorrect.
Andrew Bartlett
Andrew Tridgell [Thu, 17 May 2007 03:42:28 +0000 (03:42 +0000)]
r22965: changed RAW-LOCK-BENCH to remove the scheduling uncertainty. We should
always get balanced locking now
Andrew Tridgell [Thu, 17 May 2007 02:42:14 +0000 (02:42 +0000)]
r22964: log an error on epoll_create failing on reopen
Andrew Tridgell [Thu, 17 May 2007 02:38:07 +0000 (02:38 +0000)]
r22963: fixed the epoll/fork interaction in the epoll and aio backends
Andrew Tridgell [Thu, 17 May 2007 02:22:29 +0000 (02:22 +0000)]
r22962: show progress separately for each client in RAW-BENCH-LOCK, this is
much more useful for seeing why we get unbalanced locking
Andrew Tridgell [Thu, 17 May 2007 02:21:07 +0000 (02:21 +0000)]
r22961: use EVENT_FD_AUTOCLOSE and SOCKET_FLAG_NOCLOSE to fix up some hairy
problems with order of socket closing in krb5
Andrew Tridgell [Thu, 17 May 2007 02:19:28 +0000 (02:19 +0000)]
r22960: added a SOCKET_FLAG_NOCLOSE to allow us to tell the socket layer that
we will handle the close of the socket
Andrew Tridgell [Thu, 17 May 2007 02:18:29 +0000 (02:18 +0000)]
r22959: cope with a rather interesting interaction between epoll() and
fork(). See
http://junkcode.samba.org/ftp/unpacked/junkcode/epoll_fork.c for why
this is needed
Ronnie Sahlberg [Thu, 17 May 2007 01:17:43 +0000 (01:17 +0000)]
r22958: For SRVSVC/NetFileClose only Administrator is allowed to close open files.
If a normal user tries to close a file that exists, even that users own files the server
responds with this error on w2k
if the file does not exist, the server instead responds with WERR_BADFILE
James Peach [Wed, 16 May 2007 19:23:08 +0000 (19:23 +0000)]
r22952: Make sure we look at extra_cflags when building custom targets.
Stefan Metzmacher [Wed, 16 May 2007 14:52:54 +0000 (14:52 +0000)]
r22944: fix bug #4618:
rename private -> private_data
metze
Michael Adam [Wed, 16 May 2007 11:15:16 +0000 (11:15 +0000)]
r22931: Fix logic in detection of the need to replace dlopen and friends.
Originally, dlfcn.o was only added to LIBREPLACEOBJ if dlopen
was found in libdl but header dlfcn.h was not appropriate.
Michael
Andrew Bartlett [Wed, 16 May 2007 00:17:45 +0000 (00:17 +0000)]
r22921: This index saves another 7 seconds off a 'make quicktest', and is a common search operator.
Andrew Bartlett
Stefan Metzmacher [Tue, 15 May 2007 18:23:58 +0000 (18:23 +0000)]
r22913: try to work arround the unaligned memory access bug on Tru64
metze
Stefan Metzmacher [Tue, 15 May 2007 11:00:20 +0000 (11:00 +0000)]
r22896: we pass SMB2-LOCK now
metze
Stefan Metzmacher [Tue, 15 May 2007 10:50:25 +0000 (10:50 +0000)]
r22894: don't send a 2nd request before the last one returned (at least with
STATUS_PENDING) otherwise longhorn beta3 just drops the connection
metze
Andrew Tridgell [Tue, 15 May 2007 08:01:37 +0000 (08:01 +0000)]
r22887: and servicePrincipalName ....
Volker Lendecke [Tue, 15 May 2007 08:01:30 +0000 (08:01 +0000)]
r22886: Workaround to make RAW-NOTIFY a bit more reliable against Samba3. It
might take a bit for events to propagate from one to another connection,
in particular with a process per connection.
Andrew Tridgell [Tue, 15 May 2007 08:00:49 +0000 (08:00 +0000)]
r22885: now we use kernberos more, we need to index on userPrincipalName
Andrew Bartlett [Tue, 15 May 2007 07:16:04 +0000 (07:16 +0000)]
r22884: Be consistant with the case of these constants.
Andrew Bartlett
Andrew Bartlett [Tue, 15 May 2007 07:15:23 +0000 (07:15 +0000)]
r22883: Indentation.
Andrew Bartlett [Tue, 15 May 2007 06:51:18 +0000 (06:51 +0000)]
r22882: It seems entirly reasonable to follow metze's suggestion and check for
a valid file handle first.
Andrew Bartlett
Andrew Tridgell [Tue, 15 May 2007 05:42:16 +0000 (05:42 +0000)]
r22881: show number of connected clients
Andrew Bartlett [Tue, 15 May 2007 03:49:21 +0000 (03:49 +0000)]
r22877: Remove stray 'l'
Andrew Tridgell [Tue, 15 May 2007 03:00:58 +0000 (03:00 +0000)]
r22876: - try to reconnect once per second, not continously
- patch from ronnie to fix the lock offset on reconnect
Andrew Bartlett [Tue, 15 May 2007 01:57:21 +0000 (01:57 +0000)]
r22875: We want to skip this test, it will fail unless run against IPC$ (which the pattern does not).
We need a way to list this against IPC$, but not have this skip entry
override it.
Andrew Bartlett
Andrew Bartlett [Tue, 15 May 2007 01:21:20 +0000 (01:21 +0000)]
r22874: Expand the RPC-QFILEINFO-IPC test, and add a server implementation to match.
This should help with Vista joins.
Andrew Bartlett
Andrew Bartlett [Tue, 15 May 2007 00:10:18 +0000 (00:10 +0000)]
r22873: Make the RAW-QFILEINFO-IPC test pass against Win2k3.
Next step is working on Samba4's IPC ntvfs module.
Andrew Bartlett
Rafal Szczesniak [Mon, 14 May 2007 23:24:50 +0000 (23:24 +0000)]
r22871: restore testing against host name instead of ip address
rafal
Stefan Metzmacher [Mon, 14 May 2007 18:02:49 +0000 (18:02 +0000)]
r22866: handle incoming chained smb2 requests in our server code to let
the windows explorer in longhorn beta3 work.
metze
Stefan Metzmacher [Mon, 14 May 2007 17:58:28 +0000 (17:58 +0000)]
r22865: handle pending locks in smb2
metze
Stefan Metzmacher [Mon, 14 May 2007 17:56:00 +0000 (17:56 +0000)]
r22864: in SMB2 mode we need to cancel an existing lock with a conflicting lock
if they're on the same handle and the same range
metze
Stefan Metzmacher [Mon, 14 May 2007 17:51:19 +0000 (17:51 +0000)]
r22863: in SMB2 mode we always need to return NT_STATUS_LOCK_NOT_GRANTED
metze
Stefan Metzmacher [Mon, 14 May 2007 17:08:18 +0000 (17:08 +0000)]
r22858: - let SMB2-LOCK-VALID-REQUEST pass against longhorn beta3
- add modify the SMB2-LOCK-BLOCK-WRITE test to also test reading
and name in SMB2-LOCK-RW-EXCLUSIV
- add SMB2-LOCK-NONE and SMB2-LOCK-SHARED
metze
Stefan Metzmacher [Mon, 14 May 2007 13:47:03 +0000 (13:47 +0000)]
r22849: map smb2 lock to the generic level
metze
Stefan Metzmacher [Mon, 14 May 2007 11:50:51 +0000 (11:50 +0000)]
r22843: pass smb2 lock requests to the ntvfs layer
metze
Stefan Metzmacher [Mon, 14 May 2007 11:45:39 +0000 (11:45 +0000)]
r22842: fix directory listing of SMB2 and include directories and hidden files
metze
Andrew Bartlett [Mon, 14 May 2007 05:53:26 +0000 (05:53 +0000)]
r22838: Add in an explority test for what QFSINFO operations are valid on IPC$
It seems most opertaions are supported, but that the values are not
consistant. I think these are very much hand-hacked stubs, which is
what they will be on Samba4 too.
I'll need to add some more 'don't fail for...' hacks before this
passes against Win2k3.
Andrew Bartlett
Andrew Tridgell [Mon, 14 May 2007 03:49:42 +0000 (03:49 +0000)]
r22837: make RAW-BENCH-LOCK reconnect asynchronously when one of its servers
goges away
Andrew Tridgell [Mon, 14 May 2007 01:05:09 +0000 (01:05 +0000)]
r22836: started adding auto-reconnect logic to lockbench. This needs to be
made async.
Andrew Tridgell [Mon, 14 May 2007 01:04:35 +0000 (01:04 +0000)]
r22835: it's useful to see how much time is remaining in this test (so we can
tell if failures are due to the cleanup code)
Andrew Tridgell [Mon, 14 May 2007 01:04:01 +0000 (01:04 +0000)]
r22834: fixed a memory leak in the torture_open_connection() code, and removed
the duplicate handling of the unclist. This now exposes a function
torture_get_conn_index() which can be used by tests to work out which
host to talk to using the unclist. I will be using that as part of a
set of tests for clustered Samba that need to do auto-reconnect (to
allow testing of cluster node failure)
Andrew Tridgell [Mon, 14 May 2007 01:01:05 +0000 (01:01 +0000)]
r22833: use EVENT_FD_AUTOCLOSE in our event test suite
Andrew Tridgell [Mon, 14 May 2007 01:00:06 +0000 (01:00 +0000)]
r22832: merged the latest tdb changes from ctdb to Samba4
Andrew Tridgell [Mon, 14 May 2007 00:58:24 +0000 (00:58 +0000)]
r22831: take advantage of EVENT_FD_AUTOCLOSE in the inotify code
Andrew Tridgell [Mon, 14 May 2007 00:57:48 +0000 (00:57 +0000)]
r22830: merged the latest lib/events updates from ctdb to Samba4. This
includes a new EVENT_FD_AUTOCLOSE flag that prevents race conditions
where code using fd events might close a fd before releasing the
struct fd_event. That causes headaches for epoll.
Andrew Tridgell [Mon, 14 May 2007 00:31:04 +0000 (00:31 +0000)]
r22829: system/select.h needs to bring in epoll.h for epoll usage in
lib/events
Jelmer Vernooij [Sun, 13 May 2007 15:53:59 +0000 (15:53 +0000)]
r22827: Change license to LGPL (discussed with tridge).
Stefan Metzmacher [Sun, 13 May 2007 12:59:42 +0000 (12:59 +0000)]
r22825: let longhorn beta3 work with samba4
if we return a zero assoc_group_id longhorn beta3 stops
after getting the bind_ack.
metze
Stefan Metzmacher [Sun, 13 May 2007 12:52:13 +0000 (12:52 +0000)]
r22824: - add a simple test which shows how a lock rejects
a write on a different file handle.
SMB2-LOCK-BLOCK-WRITE
- make it possible to run each SMB2-LOCK-* test on its own
metze
Andrew Tridgell [Sat, 12 May 2007 07:44:09 +0000 (07:44 +0000)]
r22813: improve error messages in RAW-NOTIFY test
Rafal Szczesniak [Fri, 11 May 2007 21:51:53 +0000 (21:51 +0000)]
r22811: two more memory allocation checks.
rafal
Rafal Szczesniak [Fri, 11 May 2007 21:48:29 +0000 (21:48 +0000)]
r22810: when a test runs against domain controller use domain name
instead of target host name specified in binding string.
metze: this fixes the problem with passing name resolution
down the socket wrapper calls. NET tests now run against
$SERVER flawlessly.
rafal
Rafal Szczesniak [Fri, 11 May 2007 21:45:03 +0000 (21:45 +0000)]
r22809: use buffer size specified in libnet context.
rafal
Rafal Szczesniak [Fri, 11 May 2007 21:44:18 +0000 (21:44 +0000)]
r22808: store default buffer size for samr operations in libnet context.
This allows not requiring it as an argument in some function calls
and still enables specifying any size if it's necessary via libnet
context.
rafal
Rafal Szczesniak [Fri, 11 May 2007 19:10:34 +0000 (19:10 +0000)]
r22804: convert libnet_rpc_userinfo function to use "continue" functions
instead of a single handler.
rafal
Stefan Metzmacher [Fri, 11 May 2007 10:15:20 +0000 (10:15 +0000)]
r22793: use a state struct for r_do_release_demand to avoid a memleak on errors
metze
Stefan Metzmacher [Fri, 11 May 2007 10:13:10 +0000 (10:13 +0000)]
r22792: only don't free on success
metze
Stefan Metzmacher [Fri, 11 May 2007 10:05:13 +0000 (10:05 +0000)]
r22791: make it possible to use smb2_create_blob_add() in the server code too
metze
Stefan Metzmacher [Fri, 11 May 2007 10:03:04 +0000 (10:03 +0000)]
r22790: fix initialization for chained requests
metze
Stefan Metzmacher [Fri, 11 May 2007 09:24:41 +0000 (09:24 +0000)]
r22789: fix loading of internal samba4 modules
metze
Stefan Metzmacher [Fri, 11 May 2007 08:59:14 +0000 (08:59 +0000)]
r22788: fix typo
metze
Stefan Metzmacher [Thu, 10 May 2007 08:39:44 +0000 (08:39 +0000)]
r22774: correctly initialize the smbsrv_connection when switching to SMB2
metze
Jelmer Vernooij [Wed, 9 May 2007 11:51:39 +0000 (11:51 +0000)]
r22769: Fix include location.
Rafal Szczesniak [Tue, 8 May 2007 22:08:26 +0000 (22:08 +0000)]
r22764: - replace talloc_zero with composite_create
- use event context provided with libnet context instead of creating
a new one
rafal
Rafal Szczesniak [Tue, 8 May 2007 22:04:28 +0000 (22:04 +0000)]
r22763: replace talloc_zero calls with composite_create and add more
allocation checks.
rafal
Jelmer Vernooij [Tue, 8 May 2007 21:17:58 +0000 (21:17 +0000)]
r22762: Some ldb_map changes:
* Change license to LGPL, so it can be used by non-Samba users of
LDB (cleared with Martin as well).
* Include ldb_map in standalone build.
* Move ldb_map to its own directory
Stefan Metzmacher [Tue, 8 May 2007 09:54:01 +0000 (09:54 +0000)]
r22758: - add missing _talloc_set_name_const(ret, ret); to talloc_append_string()
- update manpage and add comments about _talloc_set_name_const(ret, ret);
in all places which use them
metze
Stefan Metzmacher [Tue, 8 May 2007 08:49:52 +0000 (08:49 +0000)]
r22757: test if calling talloc_free() works on parent elements from within a destrutor
metze
Andrew Bartlett [Tue, 8 May 2007 04:38:16 +0000 (04:38 +0000)]
r22756: Make it easier to setup an LDAP replica. Provision with
--partitions-only (suggestions for a better name welcome) will setup
the partitions records, but no any data in those partitions. This can
then point at the already configured remote LDAP server.
Andrew Bartlett
Stefan Metzmacher [Mon, 7 May 2007 15:43:40 +0000 (15:43 +0000)]
r22752: bail out on talloc failure
metze
Jelmer Vernooij [Mon, 7 May 2007 15:29:25 +0000 (15:29 +0000)]
r22750: dlopen() is always available now (and returns a correct error if not supported by the system), thanks to libreplace.
Stefan Metzmacher [Mon, 7 May 2007 15:27:50 +0000 (15:27 +0000)]
r22749: fix memory leak of nbt_name_request structure which are used to send replies
and never have an async callback that could free it.
we only had the memory leak in the error path the
standard path was ok.
metze
Stefan Metzmacher [Mon, 7 May 2007 15:19:53 +0000 (15:19 +0000)]
r22748: fix memleaks by passing an mem_ctx to
irpc_servers_byname()
metze
Stefan Metzmacher [Mon, 7 May 2007 15:05:04 +0000 (15:05 +0000)]
r22746: fix the build of the sqlite module
metze
Stefan Metzmacher [Mon, 7 May 2007 13:32:34 +0000 (13:32 +0000)]
r22743: set the talloc name to the string...
metze
Jelmer Vernooij [Mon, 7 May 2007 12:41:23 +0000 (12:41 +0000)]
r22742: Fix tests.
Jelmer Vernooij [Mon, 7 May 2007 12:31:42 +0000 (12:31 +0000)]
r22741: Add MSVC-specific pragma's for Wireshark. Patch from Ronnie.
Rafal Szczesniak [Mon, 7 May 2007 05:55:40 +0000 (05:55 +0000)]
r22735: correct some comments
rafal
Rafal Szczesniak [Mon, 7 May 2007 05:42:26 +0000 (05:42 +0000)]
r22734: - use samr pipe if it is already opened
- close connection handle after domains enumeration
- collect domain names in subsequent rounds of enumeration
(if there are more than one)
rafal
Stefan Metzmacher [Sun, 6 May 2007 21:28:03 +0000 (21:28 +0000)]
r22723: fix compiler warnings
metze
Stefan Metzmacher [Sun, 6 May 2007 21:27:35 +0000 (21:27 +0000)]
r22722: fix dependencies
metze
Stefan Metzmacher [Sun, 6 May 2007 21:26:42 +0000 (21:26 +0000)]
r22721: use only included talloc and tdb untill jelmer
fixes the code to support both...
the problem is that we always add EXT::TALLOC in config.mk
before 'include lib/talloc/config.mk'
metze
Stefan Metzmacher [Sun, 6 May 2007 21:19:22 +0000 (21:19 +0000)]
r22718: remove autoconf warnings, we should only include
.m4 files only once
metze
Jelmer Vernooij [Sun, 6 May 2007 15:58:12 +0000 (15:58 +0000)]
r22699: Reduce amount of magic required for popt detection.
Jelmer Vernooij [Sun, 6 May 2007 15:46:24 +0000 (15:46 +0000)]
r22698: Support using system tdb, when available.
Jelmer Vernooij [Sun, 6 May 2007 15:31:03 +0000 (15:31 +0000)]
r22697: Support building with external talloc.
Simo Sorce [Sun, 6 May 2007 15:17:14 +0000 (15:17 +0000)]
r22696: Make sure this is an ascii char
spotted by Volker
Volker Lendecke [Sun, 6 May 2007 11:03:33 +0000 (11:03 +0000)]
r22694: It seems that AIX 5.3 with XLC has difficulties with <ctype.h>. This is
an attempt to work around this: Maybe it helps if we include other stuff
first.
This raises a question however: Do we want the DN handling to be locale
dependent? isalpha() can return different things depending on the
current locale.
Jelmer Vernooij [Sat, 5 May 2007 20:41:21 +0000 (20:41 +0000)]
r22687: use DESTDIR.
Jelmer Vernooij [Sat, 5 May 2007 19:54:21 +0000 (19:54 +0000)]
r22684: Fix native Samba build.
Stefan Metzmacher [Sat, 5 May 2007 19:05:42 +0000 (19:05 +0000)]
r22683: fix the logic for skipping the pipe_dead()
code on the 2nd run.
thanks volker for finding this!
metze
Jelmer Vernooij [Sat, 5 May 2007 18:59:52 +0000 (18:59 +0000)]
r22682: Fix standalone ldb build when parent directory name != ldb.
Jelmer Vernooij [Sat, 5 May 2007 18:50:56 +0000 (18:50 +0000)]
r22681: Fix standalone ldb build when parent directory name != ldb.
Volker Lendecke [Sat, 5 May 2007 17:21:10 +0000 (17:21 +0000)]
r22680: Fix a talloc ctx name
Rafal Szczesniak [Fri, 4 May 2007 18:59:51 +0000 (18:59 +0000)]
r22672: use composite_create calls instead of talloc_zero.
rafal
Andrew Tridgell [Fri, 4 May 2007 12:41:28 +0000 (12:41 +0000)]
r22670: changed the RAW-NOTIFY test to support clustered testing (two nodes)