18 years ago(merge from 3.0)
Andrew Bartlett [Thu, 22 Jan 2004 10:22:03 +0000 (10:22 +0000)]
(merge from 3.0)

Fix for debian Bug#225328 by LaMont Jones <>, where
the failure of our LFS test on ia64 caused the _GNU_SOURCE define not to be
added, causeing strndup() not to be defined...

This was due to strdup() bein implicitly declared, and casting

I'll into when we really should define _GNU_SOURCE shortly, but the
fix is correct anyway.

Andrew Bartlett

18 years ago- move the vars used by the backtrace stuff into the #ifdef
Stefan Metzmacher [Thu, 22 Jan 2004 02:28:17 +0000 (02:28 +0000)]
- move the vars used by the backtrace stuff into the #ifdef

- handle SIBABRT with a backtrace


18 years agofix last commit, which went in by mistake
Stefan Metzmacher [Thu, 22 Jan 2004 02:13:13 +0000 (02:13 +0000)]
fix last commit, which went in by mistake
('cvs ci filename' should be my friend:-)

get rid of some more global unconst vars.
I didn't find a referenz of this vars:
SIG_ATOMIC_T reload_after_sighup = 0;
SIG_ATOMIC_T got_sig_term = 0;

tridge: is it ok to remove them?
it compiles for me


18 years agosorry, the last commit went in by mistake:-(
Stefan Metzmacher [Thu, 22 Jan 2004 02:02:27 +0000 (02:02 +0000)]
sorry, the last commit went in by mistake:-(

- implement the interface_by_name() and interface_by_uuid() hooks
  for the autogenerated dcerpc server boilerplates.


18 years agoinitilize ev->maxfd = EVENT_INVALID_MAXFD; before the while() loop.
Stefan Metzmacher [Thu, 22 Jan 2004 01:56:31 +0000 (01:56 +0000)]
initilize ev->maxfd = EVENT_INVALID_MAXFD; before the while() loop.


18 years ago- make the recalculation of ev->maxfd a bit more efficient.
Stefan Metzmacher [Thu, 22 Jan 2004 01:35:51 +0000 (01:35 +0000)]
- make the recalculation of ev->maxfd a bit more efficient.
- fix the case when we have no fd_events left on the event_context
  we now exit after doing the timed events and not block in select()
  waiting for no fd's to become ready...


18 years agomake dcerpc_pipes[] completely const.
Stefan Metzmacher [Thu, 22 Jan 2004 01:29:44 +0000 (01:29 +0000)]
make dcerpc_pipes[] completely const.
so it now apears in the text section of
'size librpc/gen_ndr/tables.o'


18 years agoadded a little bit of const magic to get rid of the data in librpc/gen_ndr/*.o
Andrew Tridgell [Thu, 22 Jan 2004 01:12:43 +0000 (01:12 +0000)]
added a little bit of const magic to get rid of the data in librpc/gen_ndr/*.o

18 years agolet the event_add_XXX() function return a pointer of the allocated event structure
Stefan Metzmacher [Wed, 21 Jan 2004 23:19:59 +0000 (23:19 +0000)]
let the event_add_XXX() function return a pointer of the allocated event structure
and NULL on allacation error.


18 years agocope with a wider range of broken servers in the RAW-QFILEINFO test
Andrew Tridgell [Wed, 21 Jan 2004 03:30:03 +0000 (03:30 +0000)]
cope with a wider range of broken servers in the RAW-QFILEINFO test

18 years agocheck for the correct fault code when the server fails a use of a
Andrew Tridgell [Tue, 20 Jan 2004 06:10:15 +0000 (06:10 +0000)]
check for the correct fault code when the server fails a use of a
incorrect policy handle

18 years agoadded code to the RPC-SPOOLSS test that demonstrates that policy
Andrew Tridgell [Tue, 20 Jan 2004 06:07:09 +0000 (06:07 +0000)]
added code to the RPC-SPOOLSS test that demonstrates that policy
handles are not shared between open dcerpc connections, even when
those connections are on the same SMB socket. I have tested this with
w2k3, w2k and NT4. It seems that policy handles have a strict scope of
the dcerpc connection on which they were opened.

I realise that this goes against existing folk-law in the team, but it
seems that the previous testing (I'm not sure who did this?) was
wrong. Perhaps clients do send us policy handles from other
connections, but if they do then the correct thing to do is to fail
the operation with a dcerpc fault. I suspect that failing it with
exactly the right dcerpc fault code is important.

18 years agodcerpc server output now copes with the client blocking part way
Andrew Tridgell [Tue, 20 Jan 2004 05:54:17 +0000 (05:54 +0000)]
dcerpc server output now copes with the client blocking part way
through a read. This happens to also avoid a memcpy on output for
dcerpc over tcp.

18 years agoavoid a copy of the data being input to the dcerpc server in the most
Andrew Tridgell [Tue, 20 Jan 2004 04:32:17 +0000 (04:32 +0000)]
avoid a copy of the data being input to the dcerpc server in the most
common case of there being no pending partial data and a full dcerpc
packet being received.

We should use this same model for the smb server. It gives us
efficient memory usage while allowing for completely async socket

18 years agoadd make [un]installtorture
Stefan Metzmacher [Mon, 19 Jan 2004 14:57:56 +0000 (14:57 +0000)]
add make [un]installtorture
and let this be called by make [un]install

we can later change this and the make all
to not do the torture stuff, but for know only developers will
use samba4...


18 years agoremove all unused stuff from
Stefan Metzmacher [Mon, 19 Jan 2004 14:22:32 +0000 (14:22 +0000)]
remove all unused stuff from
and some stuff from and dynconfig.[ch]

make all

bin/gentest   bin/masktest   bin/smbd
bin/locktest  bin/smbclient  bin/smbtorture



18 years agoAdd test to set single fields in fields_present with info21.
Simo Sorce [Mon, 19 Jan 2004 08:59:14 +0000 (08:59 +0000)]
Add test to set single fields in fields_present with info21.

18 years agomake sure we initialise r.out.handle in openeventlog
Andrew Tridgell [Sat, 17 Jan 2004 00:16:18 +0000 (00:16 +0000)]
make sure we initialise r.out.handle in openeventlog

18 years ago* fixed a segv when -U is not used in smbtorture.
Andrew Tridgell [Thu, 15 Jan 2004 23:19:07 +0000 (23:19 +0000)]
* fixed a segv when -U is not used in smbtorture.

* fixed the handling of anonymous logins

18 years agoa script which find unused or undefined vars in Makefile or
Stefan Metzmacher [Tue, 13 Jan 2004 22:24:56 +0000 (22:24 +0000)]
a script which find unused or undefined vars in Makefile or

call it like:

script/ Makefile




18 years agoAdd script for finding unused function checks in
Jelmer Vernooij [Tue, 13 Jan 2004 22:22:43 +0000 (22:22 +0000)]
Add script for finding unused function checks in

18 years agoFix --with-fhs for swatdir
Jelmer Vernooij [Sun, 11 Jan 2004 13:40:06 +0000 (13:40 +0000)]
Fix --with-fhs for swatdir

18 years agoadded dom_sid_string() function
Andrew Tridgell [Sun, 11 Jan 2004 05:16:07 +0000 (05:16 +0000)]
added dom_sid_string() function

18 years agoUpdate Copyright to 1992-2004
Stefan Metzmacher [Fri, 9 Jan 2004 06:59:48 +0000 (06:59 +0000)]
Update Copyright to 1992-2004

Jerry should I change this in 3_0 and HEAD as well?


18 years agofixed the rpc epmapper server. This fixes rpc over tcp.
Andrew Tridgell [Fri, 9 Jan 2004 02:43:23 +0000 (02:43 +0000)]
fixed the rpc epmapper server. This fixes rpc over tcp.

18 years agoThis patch adds a better dcerpc server infastructure.
Stefan Metzmacher [Thu, 8 Jan 2004 22:55:27 +0000 (22:55 +0000)]
This patch adds a better dcerpc server infastructure.

1.) We now register endpoint servers add startup via register_backend()
    and later use the smb.conf 'dcerpc endpoint servers' parameter to setup the dcesrv_context

2.) each endpoint server can register at context creation time as much interfaces as it wants
    (multiple interfaces on one endpoint are supported!)
    (NOTE:  there's a difference between 'endpoint server' and 'endpoint'!
    for details look at rpc_server/dcesrv_server.h)

3.) one endpoint can have a security descriptor registered to it self
    this will be checked in the future when a client wants to connect
    to an smb pipe endpoint.

4.) we now have a 'remote' endpoint server, which works like the ntvfs_cifs module
    it takes this options in the [globals] section:

    dcerpc remote:interfaces = srvsvc, winreg, w32time, epmapper
    dcerpc remote:binding = ...
    dcerpc remote:user = ...
    dcerpc remote:password = ...

5.) we currently have tree endpoint servers: epmapper, rpcecho and remote

    the default for the 'dcerpc endpiont servers = epmapper, rpcecho'

    for testing you can also do
    dcerpc endpoint servers = rpcecho, remote, epmapper
    dcerpc remote:interfaces = srvsvc, samr, netlogon

6,) please notice the the epmapper now only returns NO_ENTRIES
    (but I think we'll find a solution for this too:-)

7.) also there're some other stuff left, but step by step :-)

This patch also includes updates for the
register_subsystem() , ntvfs_init(), and some other funtions
to check for duplicate subsystem registration


(hmmm, my first large commit...I hope it works as supposed :-)

18 years agoremove a useless assignment
Stefan Metzmacher [Thu, 8 Jan 2004 06:48:54 +0000 (06:48 +0000)]
remove a useless assignment


18 years agoAdd note about new configure structure.
Jelmer Vernooij [Fri, 2 Jan 2004 13:47:13 +0000 (13:47 +0000)]
Add note about new configure structure.

18 years agoAutodetect service_type in cli_tree_full_connection() if the caller
Tim Potter [Fri, 2 Jan 2004 01:04:59 +0000 (01:04 +0000)]
Autodetect service_type in cli_tree_full_connection() if the caller
passes in NULL.

18 years agothe endpoint mapper now works in bigendian mode
Andrew Tridgell [Wed, 31 Dec 2003 01:32:33 +0000 (01:32 +0000)]
the endpoint mapper now works in bigendian mode

18 years ago"subsystems" should be static
Andrew Tridgell [Tue, 30 Dec 2003 22:55:12 +0000 (22:55 +0000)]
"subsystems" should be static

18 years agoremove some unused files
Andrew Tridgell [Mon, 22 Dec 2003 02:52:33 +0000 (02:52 +0000)]
remove some unused files

18 years agominor updates to make the srvsvc test compile
Andrew Tridgell [Sun, 21 Dec 2003 07:57:49 +0000 (07:57 +0000)]
minor updates to make the srvsvc test compile

18 years agowkssvc test updates from metze
Andrew Tridgell [Sun, 21 Dec 2003 07:53:57 +0000 (07:53 +0000)]
wkssvc test updates from metze

18 years agolatest srvsvc and wkssvc IDL from metze
Andrew Tridgell [Sat, 20 Dec 2003 21:50:51 +0000 (21:50 +0000)]
latest srvsvc and wkssvc IDL from metze

18 years agotim, I'm guessing you really didn't want to add this :)
Andrew Tridgell [Sat, 20 Dec 2003 21:09:04 +0000 (21:09 +0000)]
tim, I'm guessing you really didn't want to add this :)

18 years agoRemove Makefile targets for old python extensions.
Tim Potter [Sat, 20 Dec 2003 09:23:48 +0000 (09:23 +0000)]
Remove Makefile targets for old python extensions.

18 years agoaddition of samr_SetSecurity() from kai
Andrew Tridgell [Fri, 19 Dec 2003 23:44:26 +0000 (23:44 +0000)]
addition of samr_SetSecurity() from kai

we needed to adjust the alignment of [relative] buffers for this to
work. I wonder if they are always 4 byte aligned?

18 years agofixed removal of moe than one ncacn_* option from option list
Andrew Tridgell [Fri, 19 Dec 2003 04:26:26 +0000 (04:26 +0000)]
fixed removal of moe than one ncacn_* option from option list

18 years agofixed the AddAliasMem test code
Andrew Tridgell [Fri, 19 Dec 2003 04:13:39 +0000 (04:13 +0000)]
fixed the AddAliasMem test code

18 years agoadded a bunch of alias functions in samr.idl based on work by Kai.
Andrew Tridgell [Fri, 19 Dec 2003 03:59:27 +0000 (03:59 +0000)]
added a bunch of alias functions in samr.idl based on work by Kai.

18 years agoadded code to the IDL validator to check for common errors with
Andrew Tridgell [Fri, 19 Dec 2003 00:40:40 +0000 (00:40 +0000)]
added code to the IDL validator to check for common errors with
pointers are arrays

18 years agofixed a segv in RPC-* when debug level > 2
Andrew Tridgell [Wed, 17 Dec 2003 21:37:34 +0000 (21:37 +0000)]
fixed a segv in RPC-* when debug level > 2
thanks to Kai for spotting this!

18 years agoFix typo.
Tim Potter [Wed, 17 Dec 2003 03:38:06 +0000 (03:38 +0000)]
Fix typo.

18 years agofixed a bug in the acct_flags checking on CreateUser2 in the RPC-SAMR test
Andrew Tridgell [Wed, 17 Dec 2003 02:35:04 +0000 (02:35 +0000)]
fixed a bug in the acct_flags checking on CreateUser2 in the RPC-SAMR test

18 years agoDelete old python bindings.
Tim Potter [Wed, 17 Dec 2003 02:18:36 +0000 (02:18 +0000)]
Delete old python bindings.

18 years agoadded a smb.conf flag "rpc big endian" that tells our rpc server to
Andrew Tridgell [Wed, 17 Dec 2003 02:06:44 +0000 (02:06 +0000)]
added a smb.conf flag "rpc big endian" that tells our rpc server to
send packets in bigendian format.

18 years agoremoved some unused functions
Andrew Tridgell [Tue, 16 Dec 2003 12:59:20 +0000 (12:59 +0000)]
removed some unused functions

18 years agofixed formatting of uuids in debug output
Andrew Tridgell [Tue, 16 Dec 2003 12:11:00 +0000 (12:11 +0000)]
fixed formatting of uuids in debug output

18 years agoadded a define for the DCERPC little-endian data representation flag
Andrew Tridgell [Tue, 16 Dec 2003 11:24:28 +0000 (11:24 +0000)]
added a define for the DCERPC little-endian data representation flag

18 years agofixed the RPC-MGMT and RPC-SCANNER tests to work with the new
Andrew Tridgell [Tue, 16 Dec 2003 10:57:17 +0000 (10:57 +0000)]
fixed the RPC-MGMT and RPC-SCANNER tests to work with the new
dcerpc_binding_string code

18 years agono longer require the pipe name or tcp port number to be the first
Andrew Tridgell [Tue, 16 Dec 2003 10:15:21 +0000 (10:15 +0000)]
no longer require the pipe name or tcp port number to be the first
option in the ncacn_* syntax

18 years agouse bigendian mode to fix the idl for samr_LogonHours
Andrew Tridgell [Tue, 16 Dec 2003 10:09:02 +0000 (10:09 +0000)]
use bigendian mode to fix the idl for samr_LogonHours

18 years agomuch better docs on the smbtorture ncacn syntax, including how to
Andrew Tridgell [Tue, 16 Dec 2003 09:55:09 +0000 (09:55 +0000)]
much better docs on the smbtorture ncacn syntax, including how to
choose bigendian smbtorture on the command line

18 years agoit turns out that a wire policy handle isn't a blob either, its a
Andrew Tridgell [Tue, 16 Dec 2003 09:50:49 +0000 (09:50 +0000)]
it turns out that a wire policy handle isn't a blob either, its a
uint32 followed by a GUID. I needed to fix this to support running in
mixed-mode rpc (where smbtorture is bigendian and w2k3 is
little-endian). Otherwise when you send back a policy handle the
server doesn't recognise it.

18 years agoadded support for big-endian ucs2 strings (as used by big-endian
Andrew Tridgell [Tue, 16 Dec 2003 09:20:34 +0000 (09:20 +0000)]
added support for big-endian ucs2 strings (as used by big-endian

this was easier than I expected!

18 years agoa fairly large commit!
Andrew Tridgell [Tue, 16 Dec 2003 09:02:58 +0000 (09:02 +0000)]
a fairly large commit!

This adds support for bigendian rpc in the client. I have installed
SUN pcnetlink locally and am using it to test the samba4 rpc
code. This allows us to easily find places where we have stuffed up
the types (such as 2 uint16 versus a uint32), as testing both
big-endian and little-endian easily shows which is correct. I have now
used this to fix several bugs like that in the samba4 IDL.

In order to make this work I also had to redefine a GUID as a true
structure, not a blob. From the pcnetlink wire it is clear that it is
indeed defined as a structure (the byte order changes). This required
changing lots of Samba code to use a GUID as a structure.

I also had to fix the if_version code in dcerpc syntax IDs, as it
turns out they are a single uint32 not two uint16s.

The big-endian support is a bit ugly at the moment, and breaks the
layering in some places. More work is needed, especially on the server

18 years agothe out substructure is not empty if there is a return from the
Andrew Tridgell [Tue, 16 Dec 2003 04:15:51 +0000 (04:15 +0000)]
the out substructure is not empty if there is a return from the

18 years agoenable rep_inet_ntoa() for non-pthread builds
Andrew Tridgell [Tue, 16 Dec 2003 04:10:47 +0000 (04:10 +0000)]
enable rep_inet_ntoa() for non-pthread builds

18 years agomore mutex portability.
Andrew Tridgell [Tue, 16 Dec 2003 03:45:04 +0000 (03:45 +0000)]
more mutex portability.

18 years agosome compilers can't handle structures with no elements. Generate
Andrew Tridgell [Tue, 16 Dec 2003 03:44:32 +0000 (03:44 +0000)]
some compilers can't handle structures with no elements. Generate
dummy elements if need be.

18 years agomore portability fixes. We now almost compile on solaris
Andrew Tridgell [Tue, 16 Dec 2003 03:34:23 +0000 (03:34 +0000)]
more portability fixes. We now almost compile on solaris

18 years agouse smb_rwlock_t instead of rwlock_t to avoid conflicts with system
Andrew Tridgell [Tue, 16 Dec 2003 03:27:36 +0000 (03:27 +0000)]
use smb_rwlock_t instead of rwlock_t to avoid conflicts with system
types on some platforms (eg. solaris)

18 years agodon't rely on the ability of perl 5.6.x to remove elements from arrays
Andrew Tridgell [Tue, 16 Dec 2003 03:21:34 +0000 (03:21 +0000)]
don't rely on the ability of perl 5.6.x to remove elements from arrays
using delete(). This makes pidl portable back to perl 5.0.

18 years agomake sure we have a value for DEFAULT_PRINTING on all platforms
Andrew Tridgell [Tue, 16 Dec 2003 02:59:30 +0000 (02:59 +0000)]
make sure we have a value for DEFAULT_PRINTING on all platforms

18 years agoremove a redundent second check for PAM
Andrew Tridgell [Tue, 16 Dec 2003 01:10:03 +0000 (01:10 +0000)]
remove a redundent second check for PAM

18 years agocope with different cpp formatting on some systems, and make sure we
Andrew Tridgell [Tue, 16 Dec 2003 00:38:33 +0000 (00:38 +0000)]
cope with different cpp formatting on some systems, and make sure we
don't loop forever if we get a cpp format line we don't understand

18 years agouse the @CPP@ value from configure if possible when determining cpp in
Andrew Tridgell [Mon, 15 Dec 2003 22:06:18 +0000 (22:06 +0000)]
use the @CPP@ value from configure if possible when determining cpp in

18 years agolets see if "cc -E" keeps more of the build farm happy then "cpp"
Andrew Tridgell [Mon, 15 Dec 2003 21:55:42 +0000 (21:55 +0000)]
lets see if "cc -E" keeps more of the build farm happy then "cpp"

We'll probably need a configure entry for this

18 years agolots of shells don't have the -nt test, so use hackish find command
Andrew Tridgell [Mon, 15 Dec 2003 12:13:34 +0000 (12:13 +0000)]
lots of shells don't have the -nt test, so use hackish find command

18 years agoadded " --template" to dump a rough template to save typing
Andrew Tridgell [Mon, 15 Dec 2003 10:55:10 +0000 (10:55 +0000)]
added " --template" to dump a rough template to save typing
when starting a pipe. Thanks to metze for a script that gave the idea.

do something like this to use it: --parse --template librpc/idl/XXX.idl > rpc_server/XXX/rpc_XXX.c

then fill in the functions in rpc_XXX.c

18 years agofixed srvsvc DiskEnum call
Andrew Tridgell [Mon, 15 Dec 2003 10:21:09 +0000 (10:21 +0000)]
fixed srvsvc DiskEnum call
strangely, the output doesn't seem to contain an info level. Some
programmer stuffed up the IDL :)

18 years agopatch from metze with updates to srvsvc and changes to pidl to allow
Andrew Tridgell [Mon, 15 Dec 2003 09:18:53 +0000 (09:18 +0000)]
patch from metze with updates to srvsvc and changes to pidl to allow
for the server rpc boilerplate to correctly generate for multiple
interfaces on an endpoint

18 years agomake sure we allow clients to negotiate ntlmssp seal if they want it
Andrew Tridgell [Mon, 15 Dec 2003 08:59:25 +0000 (08:59 +0000)]
make sure we allow clients to negotiate ntlmssp seal if they want it

18 years agomake pidl no longer dependent on Data::Dumper, which isn't installed
Andrew Tridgell [Mon, 15 Dec 2003 08:58:00 +0000 (08:58 +0000)]
make pidl no longer dependent on Data::Dumper, which isn't installed
on some systems

18 years agofixed a typo in the m4 file
Andrew Tridgell [Mon, 15 Dec 2003 08:37:42 +0000 (08:37 +0000)]
fixed a typo in the m4 file

18 years ago"make idl" is now incremental and "make idl_full" is a full idl build
Andrew Tridgell [Mon, 15 Dec 2003 05:41:05 +0000 (05:41 +0000)]
"make idl" is now incremental and "make idl_full" is a full idl build

18 years agoa script fix for the build farm
Andrew Tridgell [Mon, 15 Dec 2003 05:37:15 +0000 (05:37 +0000)]
a script fix for the build farm

18 years agodisable pam in samba4 until someone fixes it (hi andrew!)
Andrew Tridgell [Mon, 15 Dec 2003 04:35:56 +0000 (04:35 +0000)]
disable pam in samba4 until someone fixes it (hi andrew!)

18 years agodon't try and build the libclient stuff in samba4 yet
Andrew Tridgell [Mon, 15 Dec 2003 04:07:16 +0000 (04:07 +0000)]
don't try and build the libclient stuff in samba4 yet

18 years agomore flexible handling of [] in binding strings
Andrew Tridgell [Mon, 15 Dec 2003 03:41:08 +0000 (03:41 +0000)]
more flexible handling of [] in binding strings

18 years agoadded some ncacn_* binding string examples
Andrew Tridgell [Mon, 15 Dec 2003 03:40:45 +0000 (03:40 +0000)]
added some ncacn_* binding string examples

18 years agoallow the specification of full dcerpc endpoint binding strings on the
Andrew Tridgell [Mon, 15 Dec 2003 03:29:55 +0000 (03:29 +0000)]
allow the specification of full dcerpc endpoint binding strings on the
command line. This allows you to (for example) control signing/sealing
of smbtorture RPC sessions

18 years agoswitch off the default of ntlmssp on rpc on smb as some windows pipes
Andrew Tridgell [Mon, 15 Dec 2003 01:09:45 +0000 (01:09 +0000)]
switch off the default of ntlmssp on rpc on smb as some windows pipes
can't handle it (I'd switched it on while experimenting, and forgot to
switch it off before committing)

18 years agofixed the handling of zero-length top level arrays in pidl
Andrew Tridgell [Mon, 15 Dec 2003 01:07:08 +0000 (01:07 +0000)]
fixed the handling of zero-length top level arrays in pidl

18 years agozero length echo is not an error
Andrew Tridgell [Mon, 15 Dec 2003 01:06:01 +0000 (01:06 +0000)]
zero length echo is not an error

18 years agofixed the transfer syntax in the dcerpc bind reply
Andrew Tridgell [Mon, 15 Dec 2003 00:40:57 +0000 (00:40 +0000)]
fixed the transfer syntax in the dcerpc bind reply

18 years agoadded auto-generation of the server side boilerplate code for each
Andrew Tridgell [Sun, 14 Dec 2003 13:22:12 +0000 (13:22 +0000)]
added auto-generation of the server side boilerplate code for each

The server side code gets generated as librpc/gen_ndr/ndr_NAME_s.c and
gets included in the pipe module

18 years agotypo
Simo Sorce [Sun, 14 Dec 2003 12:27:51 +0000 (12:27 +0000)]

18 years agofixed fragmented signed connections to our rpc server over SMB
Andrew Tridgell [Sun, 14 Dec 2003 12:21:21 +0000 (12:21 +0000)]
fixed fragmented signed connections to our rpc server over SMB

18 years agoreduce double explanation of FIXED ARRAYS
Simo Sorce [Sun, 14 Dec 2003 12:04:55 +0000 (12:04 +0000)]
reduce double explanation of FIXED ARRAYS

18 years agontlmssp over rpc over tcp now fully works
Andrew Tridgell [Sun, 14 Dec 2003 10:45:50 +0000 (10:45 +0000)]
ntlmssp over rpc over tcp now fully works

I needed to hack the ntlmssp code a little, as the auth code in samba4
is out of date relative to the samba3 auth code. I need to do a merge :)

18 years agofix _ptr_ declaration
Andrew Tridgell [Sun, 14 Dec 2003 08:44:38 +0000 (08:44 +0000)]
fix _ptr_ declaration

18 years agofixed some memory leaks in the rpc server code
Andrew Tridgell [Sun, 14 Dec 2003 03:51:48 +0000 (03:51 +0000)]
fixed some memory leaks in the rpc server code

18 years agofixed a bug handling multiple PDUs being read from a socket at one
Andrew Tridgell [Sun, 14 Dec 2003 01:09:10 +0000 (01:09 +0000)]
fixed a bug handling multiple PDUs being read from a socket at one
time in the rpc server.

started on the framework for the dcerpc authentication server code

18 years agoadd the ntlmssp calls back into smbtorture on rpc over tcp. The samba4
Andrew Tridgell [Sat, 13 Dec 2003 23:30:43 +0000 (23:30 +0000)]
add the ntlmssp calls back into smbtorture on rpc over tcp. The samba4
rpc server doesn't do ntlmssp yet, so this breaks the tests, but that
is the next step.

18 years agocompleted the linkage between the endpoint mapper and the dcerpc
Andrew Tridgell [Sat, 13 Dec 2003 23:25:15 +0000 (23:25 +0000)]
completed the linkage between the endpoint mapper and the dcerpc
server endpoints. We can now successfully setup listening endpoints on
high ports, then use our endpoint mapper redirect incoming clients to
the right port.

also greatly cleanup the rpc over tcp session handling.

18 years agomake the IO in the dcerpc over TCP server completely async, handling
Andrew Tridgell [Sat, 13 Dec 2003 11:44:28 +0000 (11:44 +0000)]
make the IO in the dcerpc over TCP server completely async, handling
partial packets on both input and output

18 years agodcerpc over tcp in the samba4 server now works to some extent. It
Andrew Tridgell [Sat, 13 Dec 2003 10:58:48 +0000 (10:58 +0000)]
dcerpc over tcp in the samba4 server now works to some extent. It
needs quite a bit more work to get it finished. The biggest missing
feature is the lack of NTLMSSP which is needed for basic
authentication over tcp

18 years agorpcdump.exe now works fine against a Samba4 server
Andrew Tridgell [Sat, 13 Dec 2003 04:46:50 +0000 (04:46 +0000)]
rpcdump.exe now works fine against a Samba4 server

for some reason the epm_Lookup replies can't be parsed by ethereal,
although w2k parses then fine as does the Samba4 NDR code.