ira/wip.git
18 years agoUpdate to the modules system. Fixed:
Jelmer Vernooij [Tue, 25 Nov 2003 02:09:40 +0000 (02:09 +0000)]
Update to the modules system. Fixed:
 - get rid of smb_probe_module
 - merge older updates from 3.0
 - introduced register_subsystem() and register_backend() functions
 - adapt ntvfs and charset to use new register functions
 - made smb_load_modules() work recursively (e.g. 'preload modules = /usr/lib/samba')
 - got rid of some old remains

Things that still need work:
 - Did I break tankFS? I don't think so, but I can't test it here :-(
 - Add 'postload modules = ' (for modules that need to be loaded after fork() in smbd, if applicable)
 - Convert RPC, auth, passdb, etc to use new register_{subsystem,backend}() functions
 - Accept wildcards in 'preload modules' option, instead of loading recursively

18 years ago * fixed byte order in epmapper parsing
Andrew Tridgell [Mon, 24 Nov 2003 13:19:00 +0000 (13:19 +0000)]
 * fixed byte order in epmapper parsing

 * allow rpc transport to be specified on command line in smbtorture

18 years agoinitial implementation of dcerpc over tcp. RPC-EPMAPPER works, now to
Andrew Tridgell [Mon, 24 Nov 2003 12:40:47 +0000 (12:40 +0000)]
initial implementation of dcerpc over tcp. RPC-EPMAPPER works, now to
add epm_Map calls and support the rest of the pipes

18 years ago * prepared the dcerpc subsystem for adding the RPC over TCP transport
Andrew Tridgell [Mon, 24 Nov 2003 11:45:33 +0000 (11:45 +0000)]
 * prepared the dcerpc subsystem for adding the RPC over TCP transport

 * fixed a uninitialised variable bug in pidl (found by valgrind)

18 years agoremoved the STFS specific flags in the Makefile.
Andrew Tridgell [Mon, 24 Nov 2003 10:37:24 +0000 (10:37 +0000)]
removed the STFS specific flags in the Makefile.

Jim, we are going to need to redo the STFS backend as a module, as
otherwise maintaining it outside of the samba.org tree will become too
awkward.

18 years agoadded tests for epm_Map endpointer map calls
Andrew Tridgell [Mon, 24 Nov 2003 10:15:17 +0000 (10:15 +0000)]
added tests for epm_Map endpointer map calls

18 years agogive far more detail in the EPMAPPER results
Andrew Tridgell [Mon, 24 Nov 2003 05:28:10 +0000 (05:28 +0000)]
give far more detail in the EPMAPPER results

18 years agomake sure we don't try to update a constant
Andrew Tridgell [Mon, 24 Nov 2003 03:31:03 +0000 (03:31 +0000)]
make sure we don't try to update a constant

18 years agoadded tests for the remaining calls on the rpc management interface
Andrew Tridgell [Mon, 24 Nov 2003 03:21:49 +0000 (03:21 +0000)]
added tests for the remaining calls on the rpc management interface

18 years agoadded the dcerpc remote management interfaces as mgmt.idl, and wrote a
Andrew Tridgell [Mon, 24 Nov 2003 01:24:29 +0000 (01:24 +0000)]
added the dcerpc remote management interfaces as mgmt.idl, and wrote a
test suite. The test suite dumps all of the interfaces available on
all pipes. There sure are a lot more interfaces on w2k3 than w2k !

18 years ago * better diagnostics in ndrdump
Andrew Tridgell [Sun, 23 Nov 2003 21:51:24 +0000 (21:51 +0000)]
 * better diagnostics in ndrdump

 * added samr_Connect2()

18 years agoadded a tool called 'ndrdump' that allows you to dump NDR data
Andrew Tridgell [Sun, 23 Nov 2003 13:44:19 +0000 (13:44 +0000)]
added a tool called 'ndrdump' that allows you to dump NDR data
according to the current IDL taking the data from a file. In
combination with a little hack to ethereal to extract data this is a
quite powerful IDL development tool.

18 years agoModify WINREG test program to recursively enumerate keys in all known
Tim Potter [Sun, 23 Nov 2003 11:57:15 +0000 (11:57 +0000)]
Modify WINREG test program to recursively enumerate keys in all known
hives.  Limit the recursion depth to ensure the test doesn't run too long.

18 years agoRemoved unused variable.
Tim Potter [Sun, 23 Nov 2003 11:05:31 +0000 (11:05 +0000)]
Removed unused variable.

18 years agoCheck NT_STATUS code from dcerpc call function before checking the
Tim Potter [Sun, 23 Nov 2003 10:51:39 +0000 (10:51 +0000)]
Check NT_STATUS code from dcerpc call function before checking the
WERROR code.

18 years agoAdd a strlen_m_term() function for returning the length of a string
Tim Potter [Sun, 23 Nov 2003 10:50:52 +0000 (10:50 +0000)]
Add a strlen_m_term() function for returning the length of a string
including the termination.  Using value(strlen_m((r->name)+1)*2) gives
the wrong answer for the NULL string.

18 years agofixed the handling of value() attributes on scalars in IDL that
Andrew Tridgell [Sun, 23 Nov 2003 07:14:17 +0000 (07:14 +0000)]
fixed the handling of value() attributes on scalars in IDL that
reference other elements of the same structure

18 years agoActually call right function name.
Tim Potter [Sun, 23 Nov 2003 06:54:18 +0000 (06:54 +0000)]
Actually call right function name.

18 years agoCall enumkey for tridge.
Tim Potter [Sun, 23 Nov 2003 06:38:37 +0000 (06:38 +0000)]
Call enumkey for tridge.

18 years agoString termination fix.
Tim Potter [Sun, 23 Nov 2003 06:31:10 +0000 (06:31 +0000)]
String termination fix.

Re-arrange winreg tests - currently it's failing with an input validation
error on tridge's test pull/push test code.

18 years agoStrings in the winreg pipe are terminated.
Tim Potter [Sun, 23 Nov 2003 06:29:11 +0000 (06:29 +0000)]
Strings in the winreg pipe are terminated.

18 years agoooh, this is fun!
Andrew Tridgell [Sun, 23 Nov 2003 06:28:12 +0000 (06:28 +0000)]
ooh, this is fun!

I have recoded the core dcerpc packet structures (all the PDUs etc) in
terms of IDL, which means we now use pidl to generate all the code for
handling the most basic dcerpc packets. This is not normally possible
as it isn't completely valid NDR, but pidl has a number of extensions
that make it quite easy.

This also means we get the server side dcerpc
marshalling/unmarshalling code for free.

18 years agosave about 35% of the time for "make idl" by processing multiple IDL
Andrew Tridgell [Sun, 23 Nov 2003 03:42:20 +0000 (03:42 +0000)]
save about 35% of the time for "make idl" by processing multiple IDL
files at once, which means less perl startup time.

18 years agoget rid of some more unused headers
Andrew Tridgell [Sun, 23 Nov 2003 03:10:50 +0000 (03:10 +0000)]
get rid of some more unused headers

18 years agoby using a single proto.h we gain another factor of 4 in the speed of
Andrew Tridgell [Sun, 23 Nov 2003 03:03:27 +0000 (03:03 +0000)]
by using a single proto.h we gain another factor of 4 in the speed of
"make proto"

18 years agofinal bit of tidyup and speedup
Andrew Tridgell [Sun, 23 Nov 2003 02:44:23 +0000 (02:44 +0000)]
final bit of tidyup and speedup

18 years agomuch faster inner loop and neater code
Andrew Tridgell [Sun, 23 Nov 2003 02:33:46 +0000 (02:33 +0000)]
much faster inner loop and neater code

18 years agofixed loadparm handling properly
Andrew Tridgell [Sun, 23 Nov 2003 02:11:55 +0000 (02:11 +0000)]
fixed loadparm handling properly

18 years agonicer base type handling
Andrew Tridgell [Sun, 23 Nov 2003 02:00:40 +0000 (02:00 +0000)]
nicer base type handling

18 years agoreduced the number of magic types we need in mkproto.pl
Andrew Tridgell [Sun, 23 Nov 2003 01:53:54 +0000 (01:53 +0000)]
reduced the number of magic types we need in mkproto.pl

In general I prefer "struct foo" to just "foo" for most
structures. There are exceptions.

18 years agowe don't need awk now
Andrew Tridgell [Sun, 23 Nov 2003 01:31:40 +0000 (01:31 +0000)]
we don't need awk now

18 years agoget rid of our awk scripts
Andrew Tridgell [Sun, 23 Nov 2003 01:30:58 +0000 (01:30 +0000)]
get rid of our awk scripts

18 years agochanged to perl for mkproto, patch from vance
Andrew Tridgell [Sun, 23 Nov 2003 01:29:42 +0000 (01:29 +0000)]
changed to perl for mkproto, patch from vance

18 years agoadded support for enumerated types in IDL files. This makes unions
Andrew Tridgell [Sat, 22 Nov 2003 22:53:08 +0000 (22:53 +0000)]
added support for enumerated types in IDL files. This makes unions
easier to work with.

18 years agoadded support for 'const' in IDL files. This makes it easy to define
Andrew Tridgell [Sat, 22 Nov 2003 22:09:10 +0000 (22:09 +0000)]
added support for 'const' in IDL files. This makes it easy to define
symbolic names for bitfields etc.

18 years agoadded the beginnings of an IDL validator, to give clearer errors when
Andrew Tridgell [Sat, 22 Nov 2003 12:25:20 +0000 (12:25 +0000)]
added the beginnings of an IDL validator, to give clearer errors when
IDL is not valid

18 years ago * fixed null terminated string handling
Andrew Tridgell [Sat, 22 Nov 2003 11:49:22 +0000 (11:49 +0000)]
 * fixed null terminated string handling

 * fixed nested relative offsets in push functions

the spoolss torture test now passes!

18 years ago * fixed handling of relative subcontext unions
Andrew Tridgell [Sat, 22 Nov 2003 10:37:00 +0000 (10:37 +0000)]
 * fixed handling of relative subcontext unions

 * fixed GetPrinter in spoolss after OpenPrinter

18 years agodfs torture test now passes
Andrew Tridgell [Sat, 22 Nov 2003 09:45:40 +0000 (09:45 +0000)]
dfs torture test now passes

18 years ago * fixed NDR flag inheritance across push subcontexts
Andrew Tridgell [Sat, 22 Nov 2003 09:32:35 +0000 (09:32 +0000)]
 * fixed NDR flag inheritance across push subcontexts

 * don't consider not doing lsa_QueryInfoPolicy level 11 a failure
   (w2k3 doesn't have this level, w2k does)

 * on a NDR validation failure dump the failed data at level 3

18 years ago * enable RPC/NDR validation in all smbtorture code
Andrew Tridgell [Sat, 22 Nov 2003 08:41:40 +0000 (08:41 +0000)]
 * enable RPC/NDR validation in all smbtorture code

 * allow us to distinguish between null arrays and zero length arrays,
   by making the allocation size a minimum of 1 on array pull

18 years agoa fairly major upgrade to the dcerpc system
Andrew Tridgell [Sat, 22 Nov 2003 08:11:32 +0000 (08:11 +0000)]
a fairly major upgrade to the dcerpc system

 * added a NDR validator. The way it works is that when the
   DCERPC_DEBUG_VALIDATE_* flags are set the dcerpc system will
   perform NDR buffer validation. On sending a request the packet is
   first marshalled, then unmarahslled, then marshalled again, and it is
   confirmed that the two marshalling results are idential. This
   ensures that our pull and push routines are absolutely in sync, so
   that we can be very confident that if a routine works in the client
   then the corresponding routine must work on the server side. A
   similar validation is performed on all replies.

 * a result of this change is that pidl is fussier about the [ref]
   tag. You can only use it on pointers (which is the only place it
   makes sense)

 * fixed a basic alignment bug in the push side of the NDR code

 * added server side pull/push support. Our dcerpc system is now fully
   ready to be used on the server side.

 * fixed the relative offset pointer list. It must be traversed in
   reverse order on push

 * added automatic value setting for the size parameter in outgoing
   SdBuf structures.

 * expanded the ndr debugging code to always give a message on any
   failure

 * fixed the subcontext push code

 * fixed some memory leaks in smbtorture RPC tests

18 years ago * you can't have two parameters with the same name (pidl doesn't enforce
Andrew Tridgell [Sat, 22 Nov 2003 06:10:32 +0000 (06:10 +0000)]
 * you can't have two parameters with the same name (pidl doesn't enforce
   that currently, but will in a couple of hours when I've finished my
   NDR validator code)

 * only a pointer can be [ref]

18 years agoConvert all NTSTATUS return codes to WERROR.
Tim Potter [Sat, 22 Nov 2003 05:34:25 +0000 (05:34 +0000)]
Convert all NTSTATUS return codes to WERROR.

Implement all OpenXXX calls, EnumKey, OpenKey, QueryInfoKey.

Started EnumValue but it doesn't work yet.

18 years agoadded Parse::RecDescent module into pidl to ensure we all use the same
Andrew Tridgell [Sat, 22 Nov 2003 01:40:24 +0000 (01:40 +0000)]
added Parse::RecDescent module into pidl to ensure we all use the same
version. This should fix the problems Richard was seeing.

18 years agoadded some explanations for epmapper IDL and dom_sid2
Andrew Tridgell [Sat, 22 Nov 2003 01:27:40 +0000 (01:27 +0000)]
added some explanations for epmapper IDL and dom_sid2

18 years agofixed a bug with pushing non-pointer unions
Andrew Tridgell [Fri, 21 Nov 2003 22:34:45 +0000 (22:34 +0000)]
fixed a bug with pushing non-pointer unions

18 years agofix a smbtorture memory leak
Andrew Tridgell [Fri, 21 Nov 2003 22:00:38 +0000 (22:00 +0000)]
fix a smbtorture memory leak

18 years agocleaner handling of relative pointers to strings
Andrew Tridgell [Fri, 21 Nov 2003 22:00:00 +0000 (22:00 +0000)]
cleaner handling of relative pointers to strings

18 years agodon't ignore .h files here
Andrew Tridgell [Fri, 21 Nov 2003 21:25:48 +0000 (21:25 +0000)]
don't ignore .h files here

18 years agoforgot to commit this, sorry
Andrew Tridgell [Fri, 21 Nov 2003 21:25:03 +0000 (21:25 +0000)]
forgot to commit this, sorry

18 years agoprint out the UUIDs from the end point mapper
Andrew Tridgell [Fri, 21 Nov 2003 13:27:58 +0000 (13:27 +0000)]
print out the UUIDs from the end point mapper

18 years ago* changed the way strings are handled in pidl to a much more general
Andrew Tridgell [Fri, 21 Nov 2003 13:14:17 +0000 (13:14 +0000)]
* changed the way strings are handled in pidl to a much more general
  interface. We now support an arbitrary set of flags to each parser,
  and these can be used to control the string types. I have provided
  some common IDL string types in librpc/idl/idl_types.h which needs to
  be included in every IDL file.

* added IDL for the endpoint mapper. Added a test suite that enumerates
  all endpoints on the server.

18 years agoImplement DeleteKey, DeleteValue, FlushKey.
Tim Potter [Fri, 21 Nov 2003 06:14:14 +0000 (06:14 +0000)]
Implement DeleteKey, DeleteValue, FlushKey.

18 years agoStart of winreg idl. Implement OpenHKLM, GetVersion and CloseKey.
Tim Potter [Fri, 21 Nov 2003 05:28:36 +0000 (05:28 +0000)]
Start of winreg idl.  Implement OpenHKLM, GetVersion and CloseKey.

18 years agoMore work on eventlog - still doesn't work. (-:
Tim Potter [Fri, 21 Nov 2003 03:07:02 +0000 (03:07 +0000)]
More work on eventlog - still doesn't work.  (-:

18 years agoavoid calling the print routines completely if debug level < 2
Andrew Tridgell [Fri, 21 Nov 2003 02:50:40 +0000 (02:50 +0000)]
avoid calling the print routines completely if debug level < 2

18 years agoonly display really verbose packet dumps when smbtorture is run at
Andrew Tridgell [Fri, 21 Nov 2003 02:44:31 +0000 (02:44 +0000)]
only display really verbose packet dumps when smbtorture is run at
debug level 2 or above. This is more useful for checking that
something hasn't broken without swamping yourself with output.

18 years agoadded 4 more levels to samr_QueryDisplayInfo()
Andrew Tridgell [Fri, 21 Nov 2003 02:19:47 +0000 (02:19 +0000)]
added 4 more levels to samr_QueryDisplayInfo()

18 years agoadded support for 'ascstr', a ascii string in MSRPC !
Andrew Tridgell [Fri, 21 Nov 2003 02:19:09 +0000 (02:19 +0000)]
added support for 'ascstr', a ascii string in MSRPC !

18 years agoignore generated idl.pm
Andrew Tridgell [Thu, 20 Nov 2003 22:51:02 +0000 (22:51 +0000)]
ignore generated idl.pm

18 years agoTodd Sabin pointed out that a couple of the values I marked as hyper_t
Andrew Tridgell [Thu, 20 Nov 2003 22:47:20 +0000 (22:47 +0000)]
Todd Sabin pointed out that a couple of the values I marked as hyper_t
are in fact 4 byte aligned. Create a ULONG8 type in samr for
those. This type should probably be made global later.

18 years agowe now use a copy of pidl inside the samba4 source tree at
Andrew Tridgell [Thu, 20 Nov 2003 22:46:01 +0000 (22:46 +0000)]
we now use a copy of pidl inside the samba4 source tree at
source/build/pidl/

I imported the code using direct respository manipulation to preserve
the commit history

18 years agoAdd initial work on eventlog - doesn't quite work yet.
Tim Potter [Thu, 20 Nov 2003 21:52:40 +0000 (21:52 +0000)]
Add initial work on eventlog - doesn't quite work yet.

18 years agoadded samr_QueryDisplayInfo() (only level 1 so far)
Andrew Tridgell [Thu, 20 Nov 2003 12:10:41 +0000 (12:10 +0000)]
added samr_QueryDisplayInfo() (only level 1 so far)

18 years ago* fixed libndr.h header
Andrew Tridgell [Thu, 20 Nov 2003 11:17:06 +0000 (11:17 +0000)]
* fixed libndr.h header

* "make clean" now removed the generated files

18 years agomoved the pidl auto-generated files out of CVS
Andrew Tridgell [Thu, 20 Nov 2003 11:09:53 +0000 (11:09 +0000)]
moved the pidl auto-generated files out of CVS

18 years agoextensive samr_SetUserInfo/samr_QueryUserInfo testing, with
Andrew Tridgell [Thu, 20 Nov 2003 10:53:08 +0000 (10:53 +0000)]
extensive samr_SetUserInfo/samr_QueryUserInfo testing, with
cross-checking of all settable fields

18 years ago * fixed level2 of QueryUserInfo
Andrew Tridgell [Thu, 20 Nov 2003 10:29:54 +0000 (10:29 +0000)]
 * fixed level2 of QueryUserInfo

 * added per-field testing of SetUserInfo

 * fixed strlen_m()

18 years agomake sure we don't try to use c++ style variable declaration
Andrew Tridgell [Thu, 20 Nov 2003 07:28:59 +0000 (07:28 +0000)]
make sure we don't try to use c++ style variable declaration

18 years agostarted on samr_SetUserInfo()
Andrew Tridgell [Thu, 20 Nov 2003 07:20:59 +0000 (07:20 +0000)]
started on samr_SetUserInfo()

cope with the 'samrtorturetest' user already existing in the samr test
(to cope with previously failed runs)

18 years agofixed pushing structures containing pointers inside unions
Andrew Tridgell [Thu, 20 Nov 2003 07:07:09 +0000 (07:07 +0000)]
fixed pushing structures containing pointers inside unions

18 years agoa bit neater way of emitting code
Andrew Tridgell [Thu, 20 Nov 2003 06:34:18 +0000 (06:34 +0000)]
a bit neater way of emitting code

18 years agofixed a comment
Andrew Tridgell [Thu, 20 Nov 2003 05:46:45 +0000 (05:46 +0000)]
fixed a comment

18 years ago * cope with pidl not being in $HOME/pidl
Andrew Tridgell [Thu, 20 Nov 2003 05:39:54 +0000 (05:39 +0000)]
 * cope with pidl not being in $HOME/pidl

 * cope with empty structures in unions

18 years agofixed srvsvc_NetShareEnumAll()
Andrew Tridgell [Thu, 20 Nov 2003 05:34:02 +0000 (05:34 +0000)]
fixed srvsvc_NetShareEnumAll()

18 years agoAdd a 1004 level for srvsvc.
Richard Sharpe [Thu, 20 Nov 2003 04:45:16 +0000 (04:45 +0000)]
Add a 1004 level for srvsvc.

18 years agouse a precompiled grammer in pidl. This speeds up pidl by about a
Andrew Tridgell [Thu, 20 Nov 2003 04:36:09 +0000 (04:36 +0000)]
use a precompiled grammer in pidl. This speeds up pidl by about a
factor of 2 on my system.

18 years agoAdded ATSVC pipe.
Tim Potter [Thu, 20 Nov 2003 03:27:56 +0000 (03:27 +0000)]
Added ATSVC pipe.

18 years agoadded samr_LookupRids() and test code
Andrew Tridgell [Thu, 20 Nov 2003 03:18:07 +0000 (03:18 +0000)]
added samr_LookupRids() and test code

18 years agoadded samr_LookupNames() and test code
Andrew Tridgell [Thu, 20 Nov 2003 03:09:19 +0000 (03:09 +0000)]
added samr_LookupNames() and test code

18 years agoavoid compiler warnings for unused variables with [relative] pointers
Andrew Tridgell [Thu, 20 Nov 2003 02:06:25 +0000 (02:06 +0000)]
avoid compiler warnings for unused variables with [relative] pointers

18 years agoits LoadStructure() in util.pm that doesn't like strict perl - any
Andrew Tridgell [Thu, 20 Nov 2003 02:03:36 +0000 (02:03 +0000)]
its LoadStructure() in util.pm that doesn't like strict perl - any
perl gurus out there know why?

18 years agodamn, "use strict;" in util.pm breaks pidl - but why?
Andrew Tridgell [Thu, 20 Nov 2003 01:55:32 +0000 (01:55 +0000)]
damn, "use strict;" in util.pm breaks pidl - but why?

18 years agouse strict perl package everywhere for error checking
Andrew Tridgell [Thu, 20 Nov 2003 01:43:35 +0000 (01:43 +0000)]
use strict perl package everywhere for error checking

18 years agomake the echo idl match the win32 IDL again
Andrew Tridgell [Thu, 20 Nov 2003 01:32:43 +0000 (01:32 +0000)]
make the echo idl match the win32 IDL again

18 years agoEnumPrinterData in spoolss now works
Andrew Tridgell [Thu, 20 Nov 2003 01:28:33 +0000 (01:28 +0000)]
EnumPrinterData in spoolss now works

18 years ago * added a 'lstring' type for spoolss
Andrew Tridgell [Thu, 20 Nov 2003 01:27:17 +0000 (01:27 +0000)]
 * added a 'lstring' type for spoolss

 * added some notes on type equivalents in NOTES.txt

18 years agosome errors are expected in SAMR and LSA tests - don't fail the test
Andrew Tridgell [Thu, 20 Nov 2003 01:02:09 +0000 (01:02 +0000)]
some errors are expected in SAMR and LSA tests - don't fail the test
when they are given

18 years ago * changed to midl syntax using [case(x)] instead of case(x)
Andrew Tridgell [Thu, 20 Nov 2003 00:57:59 +0000 (00:57 +0000)]
 * changed to midl syntax using [case(x)] instead of case(x)

 * use empty defaults instead of a dummy empty structure

18 years ago * added support for empty case elements (including default cases)
Andrew Tridgell [Thu, 20 Nov 2003 00:57:00 +0000 (00:57 +0000)]
 * added support for empty case elements (including default cases)

 * changed to midl syntax using [case(x)] instead of case(x)

18 years agomake the socket send code a little clearer
Andrew Tridgell [Thu, 20 Nov 2003 00:36:10 +0000 (00:36 +0000)]
make the socket send code a little clearer

18 years agofixed wkssvc idl and test code for TransportEnum
Andrew Tridgell [Wed, 19 Nov 2003 23:18:35 +0000 (23:18 +0000)]
fixed wkssvc idl and test code for TransportEnum

18 years agoupdated copyright year
Andrew Tridgell [Wed, 19 Nov 2003 23:17:55 +0000 (23:17 +0000)]
updated copyright year

18 years agoswitched to a new way of handling unions, so that we can handle
Andrew Tridgell [Wed, 19 Nov 2003 22:11:17 +0000 (22:11 +0000)]
switched to a new way of handling unions, so that we can handle
alignment correctly for unions that have non-uint16 discriminants

added recursive structure alignment. This gets quite hairy, but I
think I've got it mostly right.

18 years agoswitched to a new way of handling unions, so that we can handle
Andrew Tridgell [Wed, 19 Nov 2003 22:10:20 +0000 (22:10 +0000)]
switched to a new way of handling unions, so that we can handle
alignment correctly for unions that have non-uint16 discriminants

fixed the union handling in srvsvc.idl. (metze, please take a look at
the changes, your IDL did match what was one the wire in most cases,
but isn't the way IDL is usually coded)

18 years agoFix up some things in the IDL, and fix up a mistake I made before.
Richard Sharpe [Wed, 19 Nov 2003 19:44:17 +0000 (19:44 +0000)]
Fix up some things in the IDL, and fix up a mistake I made before.

entriesreturned does not appear as a top-level IDL item. It appears in the
structure that is returned.

I have traces of TransportEnum that Ethereal properly dissects ...

18 years agoSome simple tweaks to the IDL to make it more correct ...
Richard Sharpe [Wed, 19 Nov 2003 19:32:28 +0000 (19:32 +0000)]
Some simple tweaks to the IDL to make it more correct ...

I am also concerned about the difference between NetShareEnum and
NetShareEnumAll. I would be interested to see a trace of NetShareEnumAll.

I also think we should stick to the names used in MSDN where possible, as
that makes it easier for people who are familiar with MSDN to help us get
this right.

18 years agostarted to expand the echo tests to include more interesting test
Andrew Tridgell [Wed, 19 Nov 2003 12:03:11 +0000 (12:03 +0000)]
started to expand the echo tests to include more interesting test
cases. We fail one of the alignment tests.

18 years agosrvsvc IDL and test code from metze. Thanks!
Andrew Tridgell [Wed, 19 Nov 2003 09:15:46 +0000 (09:15 +0000)]
srvsvc IDL and test code from metze. Thanks!