Tim Potter [Sat, 4 Sep 2004 01:31:02 +0000 (01:31 +0000)]
r2221: Experiment with extra properties for telling the ethereal parser what
to do. There is a patch to the grammar and idl files for this but I
won't commit that just yet.
Andrew Bartlett [Sat, 4 Sep 2004 00:02:43 +0000 (00:02 +0000)]
r2220: Updates to the NETLOGON torture test. This copes with 'long'
passwords - where the LM hash is invalid.
Also, we now drive all the logon levels and validation levels from the
outer loop, so we can check the expected return values (rather than
overwriting them).
Andrew Bartlett
Jeremy Allison [Fri, 3 Sep 2004 20:06:27 +0000 (20:06 +0000)]
r2217: Ad-hoc tests to allow me to work out the correct error code
for the bad path algorithm.
Jeremy.
Jeremy Allison [Fri, 3 Sep 2004 17:44:07 +0000 (17:44 +0000)]
r2212: Optimisation. Passes masktest against W2K3.
Jeremy.
Andrew Tridgell [Fri, 3 Sep 2004 13:36:58 +0000 (13:36 +0000)]
r2209: patch from volker to add EnumPorts spoolss IDL and test code
the ndr->offset=0; stuff is ugly. We need a better way to handle this.
Volker Lendecke [Fri, 3 Sep 2004 13:21:52 +0000 (13:21 +0000)]
r2208: little attempt to (almost) fix spoolss relative
Andrew Tridgell [Fri, 3 Sep 2004 13:20:31 +0000 (13:20 +0000)]
r2207: this bug caued valgrind to consume infinite memory till the kernel killed it :(
Andrew Tridgell [Fri, 3 Sep 2004 12:58:19 +0000 (12:58 +0000)]
r2206: another (untested) attempt to make RELATIVE_CURRENT work for volker,
plus removed unused ndr_pull_relative() function
once volker commits a test for this I'll be able to ensure it really works as expected
Andrew Tridgell [Fri, 3 Sep 2004 12:52:04 +0000 (12:52 +0000)]
r2205: fixed an incorrect cast that broke relative strings in spoolss
Andrew Tridgell [Fri, 3 Sep 2004 12:10:34 +0000 (12:10 +0000)]
r2204: added [flag(RELATIVE_CURRENT)] to change [relative] pointer behaviour
for this struct and all sub-structures to be like spoolss relative
pointers (where offset is relative to current position).
volker will test this for me :)
Andrew Tridgell [Fri, 3 Sep 2004 09:17:32 +0000 (09:17 +0000)]
r2203: delete the key after testing, so as not to clutter the server with random keys
Andrew Tridgell [Fri, 3 Sep 2004 09:09:59 +0000 (09:09 +0000)]
r2202: don't close the smb pipe after the puzzle test
Andrew Tridgell [Fri, 3 Sep 2004 08:29:07 +0000 (08:29 +0000)]
r2201: removed an exit I accidentially left in
Andrew Tridgell [Fri, 3 Sep 2004 08:28:24 +0000 (08:28 +0000)]
r2200: solved another piece of the lsakey puzzle - the session key for lsa
encryption on ncacn_ip_tcp is a fixed buffer! I don't yet know what
the buffer is, but this code proves its the same buffer for different
w2k3 servers and different user passwords, plus it is independent of
the negotiated NTLMSSP session key.
Andrew Tridgell [Fri, 3 Sep 2004 05:16:25 +0000 (05:16 +0000)]
r2199: the unknown 16 bit number in lsa_LookupPrivDisplayName() is a language
ID, so the client can choose what language they get the privilege
description in.
this is the first time I've seen a language ID on the wire in CIFS.
Andrew Tridgell [Thu, 2 Sep 2004 21:55:24 +0000 (21:55 +0000)]
r2192: removed an erroneous free() call on an error path
Andrew Tridgell [Thu, 2 Sep 2004 12:54:48 +0000 (12:54 +0000)]
r2186: setting [ref] output pointers in dcerpc calls is pointless. Removed it
for test_LookupPrivDisplayName
Andrew Tridgell [Thu, 2 Sep 2004 12:03:16 +0000 (12:03 +0000)]
r2185: add a callback function to the dcerpc async API
also add a demonstration of its use in the netlogon async example
Andrew Tridgell [Thu, 2 Sep 2004 12:02:28 +0000 (12:02 +0000)]
r2184: use the smb.conf socket options for client code too
Tim Potter [Thu, 2 Sep 2004 11:54:13 +0000 (11:54 +0000)]
r2183: Tidy up naming of ethereal fields. The field names are taken from the
structure elements with underscores replaced with spaces and words
capitalised.
Fix small buglet where we confuse the name of the interface with the name
of the idl file.
Just discovered security descriptors don't display property anymore. )-:
Andrew Tridgell [Thu, 2 Sep 2004 11:26:58 +0000 (11:26 +0000)]
r2182: force the torture test domain join to happen on SMB to prevent the
LSAKEY problem holding up other tests
Andrew Tridgell [Thu, 2 Sep 2004 11:06:23 +0000 (11:06 +0000)]
r2181: an rpc async test on the netlogon pipe
Andrew Tridgell [Thu, 2 Sep 2004 10:45:58 +0000 (10:45 +0000)]
r2180: added RPC flags "padcheck" which enables checking of all received pad
bytes to make sure they are zero. Non-zero values usually indicate one
of two things:
- the server is leaking data through sending uninitialised memory
- we have mistaken a real field in the IDL for padding
to differentiate between the two you really need to run with
"print,padcheck" and look carefully at whether the non-zero pad bytes
are random or appear to be deliberate.
Andrew Tridgell [Wed, 1 Sep 2004 23:36:22 +0000 (23:36 +0000)]
r2179: two more lsa torture tests from Richard Renard. Thanks!
Volker Lendecke [Wed, 1 Sep 2004 14:49:43 +0000 (14:49 +0000)]
r2173: Fix asn1 BOOLEANs. Thanks to Love Hornquist-Astrand.
Volker
Andrew Tridgell [Wed, 1 Sep 2004 09:45:33 +0000 (09:45 +0000)]
r2170: if we don't have a native iconv library then we can't build this test
Andrew Tridgell [Wed, 1 Sep 2004 09:45:00 +0000 (09:45 +0000)]
r2169: switch core iconv code to use talloc
Tim Potter [Wed, 1 Sep 2004 09:40:09 +0000 (09:40 +0000)]
r2168: Move these files into a separate ethereal-pidl-plugin svn repository.
Not only will this allow the plugin to be built easily, but will also
get rid of the duplicate function names that muck up etags.
Tim Potter [Wed, 1 Sep 2004 09:16:39 +0000 (09:16 +0000)]
r2167: Implement some stubbed out functions.
Stefan Metzmacher [Wed, 1 Sep 2004 07:29:02 +0000 (07:29 +0000)]
r2166: sync the asn1 stuff with trunk
metze
Andrew Tridgell [Wed, 1 Sep 2004 06:59:48 +0000 (06:59 +0000)]
r2165: generalise the charset torture test to add testing of CP850
potentially we can test any charset
Andrew Tridgell [Wed, 1 Sep 2004 05:19:00 +0000 (05:19 +0000)]
r2164: put the latest "accept either form" utf-16 iconv code in samba4
Andrew Tridgell [Wed, 1 Sep 2004 04:39:06 +0000 (04:39 +0000)]
r2159: converted samba4 over to UTF-16.
I had previously thought this was unnecessary, as windows doesn't use
standards compliant UTF-16, and for filesystem operations treats bytes
as UCS-2, but Bjoern Jacke has pointed out to me that this means we
don't correctly store extended UTF-16 characters as UTF-8 on
disk. This can be seen with (for example) the gothic characters with
codepoints above 64k.
This commit also adds a LOCAL-ICONV torture test that tests the first
1 million codepoints against the system iconv library, and tests 5
million random UTF-16LE buffers for identical error handling to the
system iconv library.
the lib/iconv.c changes need backporting to samba3
Andrew Tridgell [Wed, 1 Sep 2004 04:31:15 +0000 (04:31 +0000)]
r2158: removed a misleading comment (the extra uint16 is just padding)
Tim Potter [Tue, 31 Aug 2004 23:57:14 +0000 (23:57 +0000)]
r2154: Merge of Jim's format string fixup in r2130:2131
Jeremy Allison [Tue, 31 Aug 2004 21:34:14 +0000 (21:34 +0000)]
r2151: Added some more ad-hoc tests. Found bugs in Samba3 with these :-).
Jeremy.
Andrew Tridgell [Tue, 31 Aug 2004 09:29:51 +0000 (09:29 +0000)]
r2128: netlogon DELTA_POLICY fix from rrenard
Andrew Tridgell [Tue, 31 Aug 2004 08:43:28 +0000 (08:43 +0000)]
r2127: more lsa IDL updates from Richard Renard
Andrew Tridgell [Tue, 31 Aug 2004 08:22:51 +0000 (08:22 +0000)]
r2126: two more lsa functions worked out by richard renard
Andrew Tridgell [Tue, 31 Aug 2004 08:21:29 +0000 (08:21 +0000)]
r2125: the lp_use_mmap() in map_file() is inappropriate for 2 reasons, so I have removed it.
- lp_use_mmap() is really meant to cope with systems that have broken
mmap coherence, but map_file() doesn't need coherence, as its maps
read only
- map_file() is used to map the charset files before loadparm has
loaded, so lp_use_mmap() is always returning false for the major
use of map_file()
Stefan Metzmacher [Tue, 31 Aug 2004 07:14:34 +0000 (07:14 +0000)]
r2124: merge from trunk (-r 2123):
Argl. I could never get the naming right. Having the most significant byte at
the lowest memory address is big endian, at least according to the google
search for 'big endian'....
Volker
Stefan Metzmacher [Tue, 31 Aug 2004 06:21:14 +0000 (06:21 +0000)]
r2122: merge from trunk (-r 2120):
Fix bug found by Love H?\195?\182rnquist ?\195?\133strand: asn1_write_Integer needs to push
stuff little endian.
Stefan Metzmacher [Tue, 31 Aug 2004 06:18:48 +0000 (06:18 +0000)]
r2121: fix compiler warning
metze
Andrew Bartlett [Tue, 31 Aug 2004 02:57:49 +0000 (02:57 +0000)]
r2119: Noticed by jra:
Clean up use of unitilaised variable.
Andrew Bartlett
Andrew Tridgell [Tue, 31 Aug 2004 01:16:50 +0000 (01:16 +0000)]
r2118: fixed the receipt of bigendian rpc packets with the new async code.
The fix uses the new talloc_increase_ref_count() function in an
interesting way. I suspect this sort of technique will become quite
common.
Andrew Tridgell [Mon, 30 Aug 2004 13:05:03 +0000 (13:05 +0000)]
r2107: added a SAMR async test - this one seems to work
Andrew Tridgell [Mon, 30 Aug 2004 12:03:01 +0000 (12:03 +0000)]
r2106: try to cope with a wider range of UTF-16 characters when we are using
an external libiconv library.
Andrew Tridgell [Mon, 30 Aug 2004 07:36:16 +0000 (07:36 +0000)]
r2105: added a TestSleep() operation to the echo pipe and extended the
RPC-ECHO test to use it to test asynchronous rpc operations.
Andrew Tridgell [Mon, 30 Aug 2004 06:14:18 +0000 (06:14 +0000)]
r2104: fixed typo that causes a segv
Andrew Tridgell [Mon, 30 Aug 2004 05:35:30 +0000 (05:35 +0000)]
r2103: in the conversion to async rpc I simplified the smb backend to only
use readx/writex instead of the more efficient SMBtrans calls. This
patch restores the efficiency by using SMBtrans when possible.
Andrew Tridgell [Mon, 30 Aug 2004 05:33:49 +0000 (05:33 +0000)]
r2102: fixed a race condition when handling dos errors that are in our
table. Should get rid of the static buffer completely at some point.
Andrew Tridgell [Mon, 30 Aug 2004 05:32:00 +0000 (05:32 +0000)]
r2101: fixed a signed/unsigned char warning
Andrew Tridgell [Mon, 30 Aug 2004 03:10:43 +0000 (03:10 +0000)]
r2100: rework the dcerpc client side library so that it is async. We now
generate a separate *_send() async function for every RPC call, and
there is a single dcerpc_ndr_request_recv() call that processes the
receive side of any rpc call. The caller can use
dcerpc_event_context() to get a pointer to the event context for the
pipe so that events can be waited for asynchronously.
The only part that remains synchronous is the initial bind
calls. These could also be made async if necessary, although I suspect
most applications won't need them to be.
Andrew Bartlett [Mon, 30 Aug 2004 01:34:01 +0000 (01:34 +0000)]
r2099: Get rid of another private ARCFOUR implementation from the codebase.
Andrew Bartlett
Andrew Bartlett [Mon, 30 Aug 2004 01:26:28 +0000 (01:26 +0000)]
r2098: The first 8 bytes of this sig is not used in the 'is it correct' calculation.
Andrew Bartlett
Andrew Bartlett [Sun, 29 Aug 2004 11:28:31 +0000 (11:28 +0000)]
r2096: Enable use of NTLM2 for connections that do not got on to be NTLMSSP
signed or sealed.
This allows NTLM2 for SMB connections, and NTLMSSP over HTTP for example.
Andrew Bartlett
Stefan Metzmacher [Fri, 27 Aug 2004 18:00:55 +0000 (18:00 +0000)]
r2092: fix the build(don't catch main() by make proto)
metze
Stefan Metzmacher [Fri, 27 Aug 2004 15:56:04 +0000 (15:56 +0000)]
r2089: fix standalone build with LDAP, this needs -llber too
metze
Stefan Metzmacher [Fri, 27 Aug 2004 15:52:54 +0000 (15:52 +0000)]
r2088: add tdbtorture tdbdump and tdbtest to the build
NOTE: tdbbackup and tdbtool seems to be broken...
NOTE: I also added SMB_EXT_LIB(GDBM,...)
because that is needed by tdbtest
metze
Stefan Metzmacher [Fri, 27 Aug 2004 14:35:34 +0000 (14:35 +0000)]
r2087: the ldb tools really don't need LIBSMB!
that reduces the size of thet binaries with '-g' and gcc 3.4
from 5.3 MB to 745 KB:-)
metze
Andrew Tridgell [Wed, 25 Aug 2004 23:10:30 +0000 (23:10 +0000)]
r2074: fixed a typo
Stefan Metzmacher [Wed, 25 Aug 2004 14:31:59 +0000 (14:31 +0000)]
r2071: - change smbtorture to use the popt_common stuff
this means -U DOM\\user is know allowed
- torture:userdomain is a new smb.conf parameter
because lp_workgroup is not the domain of the user
- we use torture:userdomain now in the tests instad of lp_workgroup
- for backward compat the userdomain is lp_workgroup() by default and
not lp_netbios_name(), which my change later to match 'net' and 'smbclient'..
- we now have dublicate options e.g. -N -s ...
tridge: can we change this?
metze
Stefan Metzmacher [Wed, 25 Aug 2004 12:14:06 +0000 (12:14 +0000)]
r2066: add 'net time' command
use it like:
net time <server>
net time <server> -U "" -N
net time <server> -U DOM\\user
...
metze
Stefan Metzmacher [Wed, 25 Aug 2004 12:11:28 +0000 (12:11 +0000)]
r2065: add libnet_RemoteTOD() call with levels GENERIC and SRVSVC
metze
Stefan Metzmacher [Wed, 25 Aug 2004 11:04:47 +0000 (11:04 +0000)]
r2064: fix some error cases
metze
Andrew Bartlett [Wed, 25 Aug 2004 10:46:28 +0000 (10:46 +0000)]
r2063: Ensure the first argument to a printf() like function (talloc_init()
in this case) is constant.
Andrew Bartlett
Andrew Bartlett [Wed, 25 Aug 2004 10:21:14 +0000 (10:21 +0000)]
r2062: Fix a couple more of the printf warnings (real bugs).
You should never pass a non-constant string as a format for a printf()
function - it could contain printf macros, and these need to be
checked.
Andrew Bartlett
Stefan Metzmacher [Wed, 25 Aug 2004 09:22:47 +0000 (09:22 +0000)]
r2061: - split libnet header files
- add LIB_RPC_CONNECT_STANDARD level which takes a server name
and the PIPE NAME,UUID and VERSION
metze
Stefan Metzmacher [Wed, 25 Aug 2004 09:09:17 +0000 (09:09 +0000)]
r2060: fix compiler warning
metze
Stefan Metzmacher [Wed, 25 Aug 2004 08:58:12 +0000 (08:58 +0000)]
r2059: abartlet: is there a better way to fix this compiler warning
(the same problem as in -r 2056)
metze
Günther Deschner [Wed, 25 Aug 2004 08:31:48 +0000 (08:31 +0000)]
r2058: Merge heimdal krb5_locate_kdc-fix over from trunk/3_0 although
krb5_locate_kdc is (yet) an unused function in Samba4.
Guenther
Andrew Bartlett [Wed, 25 Aug 2004 07:30:58 +0000 (07:30 +0000)]
r2056: Allow the compiler to check this format string.
Andrew Bartlett
Andrew Bartlett [Wed, 25 Aug 2004 07:15:21 +0000 (07:15 +0000)]
r2055: Add PRINTF_ATTRIBUTE to many more parts of the code, and a new
--enable-developer warning for when they are missing.
Andrew Bartlett
Andrew Bartlett [Wed, 25 Aug 2004 07:13:53 +0000 (07:13 +0000)]
r2054: Fix compile warnings/build failures on non-gcc.
Andrew Bartlett
Andrew Bartlett [Wed, 25 Aug 2004 07:13:01 +0000 (07:13 +0000)]
r2053: All RPC sessions 'want' a session key. Of course, the key they
currently get it bougs, but anyway...
Andrew Bartlett
Andrew Tridgell [Wed, 25 Aug 2004 07:10:40 +0000 (07:10 +0000)]
r2052: rewrote the talloc section of the programming guide
Andrew Tridgell [Wed, 25 Aug 2004 06:44:23 +0000 (06:44 +0000)]
r2051: switched the samdb over to using the new destructor and reference
count features of talloc, instead of re-implementing both those
features inside of samdb (which is what we did before).
This makes samdb considerably simpler, and also fixes some bugs, as I
found some error paths that didn't call samdb_close(). Those are now
handled by the fact that a talloc_free() will auto-close and destroy
the samdb context, using a destructor.
Andrew Tridgell [Wed, 25 Aug 2004 06:42:08 +0000 (06:42 +0000)]
r2050: fixed a case where code assumed you could Realloc the result of a data_blob()
Andrew Tridgell [Wed, 25 Aug 2004 06:40:58 +0000 (06:40 +0000)]
r2049: talloc now has destructors and reference counts
this means you can do:
talloc_set_destructor(ptr, my_destructor);
and your destructor will be called with the pointer as an argument
when the pointer is about to be freed. The destructor can refuse the
free by returning -1.
You can also increase the reference count on a pointer like this:
talloc_increase_ref_count(ptr);
and a talloc_free() will just reduce the reference count, only
actually freeing the memory when the count reaches zero.
Stefan Metzmacher [Wed, 25 Aug 2004 06:13:08 +0000 (06:13 +0000)]
r2048: fix memory leak found by abartlet
metze
Andrew Bartlett [Wed, 25 Aug 2004 05:58:33 +0000 (05:58 +0000)]
r2047: Warn on 'declaration after statement' (breaks non-gcc).
Andrew Bartlett
Andrew Tridgell [Wed, 25 Aug 2004 04:16:48 +0000 (04:16 +0000)]
r2046: fixed two server packet format errors found with the RAW-* tests
Andrew Tridgell [Wed, 25 Aug 2004 04:16:14 +0000 (04:16 +0000)]
r2045: fixed a date format push in SMBsearch
Andrew Tridgell [Wed, 25 Aug 2004 04:03:28 +0000 (04:03 +0000)]
r2044: fixed two unin\11
Andrew Tridgell [Wed, 25 Aug 2004 03:23:39 +0000 (03:23 +0000)]
r2043: data_blob() now returns a talloc'd pointer. If everyone has been
following the data_blob() API properly then this will cause no
problems. I'm expecting chaos.
this is part of the general move towards using talloc for everything
in samba4
Andrew Tridgell [Wed, 25 Aug 2004 02:25:48 +0000 (02:25 +0000)]
r2042: missed a couple of places that should be talloc_free()
Andrew Bartlett [Wed, 25 Aug 2004 02:25:20 +0000 (02:25 +0000)]
r2041: Fix NTLMSSP RPC sealing, client -> win2k3 server.
The bug (found by tridge) is that Win2k3 is being tighter about the
NTLMSSP flags. If we don't negotiate sealing, we can't use it.
We now have a way to indicate to the GENSEC implementation mechanisms
what things we want for a connection.
Andrew Bartlett
Andrew Tridgell [Wed, 25 Aug 2004 02:07:53 +0000 (02:07 +0000)]
r2040: fixed a memory handling error in clisocket (caught with valgrind)
Andrew Tridgell [Wed, 25 Aug 2004 02:07:20 +0000 (02:07 +0000)]
r2039: got rid of the free() ptr in DATA_BLOB
I plan on replacing the concept by adding a generic destructor in all talloc ptrs, so you can do:
talloc_set_destructor(ptr, my_destructor);
to setup a function that will be called on free.
Andrew Tridgell [Wed, 25 Aug 2004 02:06:08 +0000 (02:06 +0000)]
r2038: get rid of the optimisation in the dcerpc server that tries to avoid a
data copy by playing internal games with DATA_BLOB and free().
Andrew Tridgell [Wed, 25 Aug 2004 02:05:02 +0000 (02:05 +0000)]
r2037: switched the asn.1 code to use talloc
Andrew Tridgell [Wed, 25 Aug 2004 02:04:35 +0000 (02:04 +0000)]
r2036: switched the spnego code to use talloc
now that talloc_free() doesn't need to take a context ptr, there is no
reason we can't use talloc everywhere that we currently use malloc().
Andrew Bartlett [Wed, 25 Aug 2004 01:58:08 +0000 (01:58 +0000)]
r2035: Fix spelling.
Andrew Bartlett
Jeremy Allison [Wed, 25 Aug 2004 01:12:59 +0000 (01:12 +0000)]
r2034: Allow user-selectable hash functions on open only.
Jeremy.
Andrew Tridgell [Wed, 25 Aug 2004 00:04:47 +0000 (00:04 +0000)]
r2031: add a check for a blank secret return in lsa secret tests
Andrew Tridgell [Tue, 24 Aug 2004 23:58:00 +0000 (23:58 +0000)]
r2030: quick hack to allow the simple NTVFS backend to handler base directories with mixed case names
Andrew Tridgell [Tue, 24 Aug 2004 23:56:59 +0000 (23:56 +0000)]
r2029: changed our client side dcerpc padding to match what w2k3 does - a 16
byte padding of the payload portion of the packet.
Stefan Metzmacher [Tue, 24 Aug 2004 16:28:08 +0000 (16:28 +0000)]
r2021: add password change fallback to SetUserInfo level 25,24,23
metze
Stefan Metzmacher [Tue, 24 Aug 2004 15:48:14 +0000 (15:48 +0000)]
r2020: fix compiler warnings
metze
Stefan Metzmacher [Tue, 24 Aug 2004 15:43:57 +0000 (15:43 +0000)]
r2019: fix compiler warnings
metze