ira/wip.git
13 years agoSplit up tests a bit, output subunit.
Jelmer Vernooij [Sat, 9 Feb 2008 18:04:14 +0000 (19:04 +0100)]
Split up tests a bit, output subunit.

13 years agoAdd docstring.
Jelmer Vernooij [Sat, 9 Feb 2008 16:42:31 +0000 (17:42 +0100)]
Add docstring.

13 years agoIgnore Python apidocs, vim swap files.
Jelmer Vernooij [Sat, 9 Feb 2008 16:41:50 +0000 (17:41 +0100)]
Ignore Python apidocs, vim swap files.

13 years agoFix syntax of docstrings, set project name when generating Python API documentation.
Jelmer Vernooij [Sat, 9 Feb 2008 16:37:42 +0000 (17:37 +0100)]
Fix syntax of docstrings, set project name when generating Python API documentation.

13 years agoregistry: Fix warning.
Jelmer Vernooij [Sat, 9 Feb 2008 16:18:51 +0000 (17:18 +0100)]
registry: Fix warning.

13 years agoMove nmblookup blackbox test closer to the code it tests.
Jelmer Vernooij [Sat, 9 Feb 2008 16:17:21 +0000 (17:17 +0100)]
Move nmblookup blackbox test closer to the code it tests.

13 years agoMove first_prereq test to m4 code.
Jelmer Vernooij [Sat, 9 Feb 2008 16:07:02 +0000 (17:07 +0100)]
Move first_prereq test to m4 code.

13 years agoMove some static makefile variables to rules.mk.
Jelmer Vernooij [Sat, 9 Feb 2008 15:41:44 +0000 (16:41 +0100)]
Move some static makefile variables to rules.mk.

13 years agopidl: Allow fixed size arrays inside unions.
Julien Kerihuel [Sat, 9 Feb 2008 15:26:16 +0000 (16:26 +0100)]
pidl: Allow fixed size arrays inside unions.

13 years agoFix test count after pointer_default_top() removal.
Jelmer Vernooij [Sat, 9 Feb 2008 15:25:34 +0000 (16:25 +0100)]
Fix test count after pointer_default_top() removal.

13 years agoMove more make code to rules.mk
Jelmer Vernooij [Sat, 9 Feb 2008 15:23:02 +0000 (16:23 +0100)]
Move more make code to rules.mk

13 years agoUse mkinclude (makes the Makefile a bit shorter), use double-colon for phony makefile...
Jelmer Vernooij [Sat, 9 Feb 2008 15:05:55 +0000 (16:05 +0100)]
Use mkinclude (makes the Makefile a bit shorter), use double-colon for phony makefile targets.

13 years agoSupport 'mkinclude' command in .mk files.
Jelmer Vernooij [Sat, 9 Feb 2008 15:01:53 +0000 (16:01 +0100)]
Support 'mkinclude' command in .mk files.

13 years agoRemove unused path update.
Jelmer Vernooij [Sat, 9 Feb 2008 14:43:28 +0000 (15:43 +0100)]
Remove unused path update.

13 years agoRe-enable python support in BECOME-DC test.
Jelmer Vernooij [Sat, 9 Feb 2008 14:42:19 +0000 (15:42 +0100)]
Re-enable python support in BECOME-DC test.

13 years agoMerge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial
Jelmer Vernooij [Sat, 9 Feb 2008 14:04:03 +0000 (15:04 +0100)]
Merge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial

13 years agoIgnore kinit with pkinit test failures for now.
Jelmer Vernooij [Sat, 9 Feb 2008 13:42:55 +0000 (14:42 +0100)]
Ignore kinit with pkinit test failures for now.

13 years agotorture: Merge util.c into util_smb.c
Jelmer Vernooij [Sat, 9 Feb 2008 13:26:46 +0000 (14:26 +0100)]
torture: Merge util.c into util_smb.c

13 years agoMove nt_time_equal() to libutil.
Jelmer Vernooij [Sat, 9 Feb 2008 13:24:24 +0000 (14:24 +0100)]
Move nt_time_equal() to libutil.

13 years agoA couple more ipsec classes for the schema.
Andrew Bartlett [Sat, 9 Feb 2008 03:48:56 +0000 (14:48 +1100)]
A couple more ipsec classes for the schema.

Andrew Bartlett

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local
Andrew Bartlett [Sat, 9 Feb 2008 03:13:58 +0000 (14:13 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local

13 years agoRemove the forced 'krb5' from the NET-API-BECOME-DC test.
Andrew Bartlett [Sat, 9 Feb 2008 03:13:23 +0000 (14:13 +1100)]
Remove the forced 'krb5' from the NET-API-BECOME-DC test.

If we want to try this authenication mechanism, we can always manually
specify the binding string.

Andrew Bartlett

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local
Andrew Bartlett [Sat, 9 Feb 2008 03:03:25 +0000 (14:03 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local

13 years agoGive a more useful error when the templates.ldb can't be found.
Andrew Bartlett [Sat, 9 Feb 2008 03:02:14 +0000 (14:02 +1100)]
Give a more useful error when the templates.ldb can't be found.

Andrew Bartlett

13 years agoMake the provision test more real, use a foo.example.com for the realm.
Andrew Bartlett [Sat, 9 Feb 2008 03:01:37 +0000 (14:01 +1100)]
Make the provision test more real, use a foo.example.com for the realm.

Andrew Bartlett

13 years agoFix upgrade after provision parameter rename.
Jelmer Vernooij [Sat, 9 Feb 2008 02:59:45 +0000 (03:59 +0100)]
Fix upgrade after provision parameter rename.

13 years agoMerge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial
Jelmer Vernooij [Sat, 9 Feb 2008 02:11:04 +0000 (03:11 +0100)]
Merge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial

13 years agoFix provision python test.
Jelmer Vernooij [Sat, 9 Feb 2008 02:09:56 +0000 (03:09 +0100)]
Fix provision python test.

13 years agoAdd tests for findnss(), add some docstrings.
Jelmer Vernooij [Sat, 9 Feb 2008 01:10:49 +0000 (02:10 +0100)]
Add tests for findnss(), add some docstrings.

13 years agoRemove unused class members.
Jelmer Vernooij [Sat, 9 Feb 2008 00:42:38 +0000 (01:42 +0100)]
Remove unused class members.

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local
Andrew Bartlett [Fri, 8 Feb 2008 23:00:14 +0000 (10:00 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local

13 years agoFrom a request from Peter Huang, include IPsec sechema.
Andrew Bartlett [Fri, 8 Feb 2008 22:59:45 +0000 (09:59 +1100)]
From a request from Peter Huang, include IPsec sechema.

(We may need to include more defaults in the template, but I want to
start small for now).

Andrew Bartlett

13 years agoReformat SMB_EXT_LIB_FROM_PKGCONFIG: fix spacing and indentation.
Michael Adam [Fri, 8 Feb 2008 15:28:04 +0000 (16:28 +0100)]
Reformat SMB_EXT_LIB_FROM_PKGCONFIG: fix spacing and indentation.

Sorry, could not resist: While trying to understand this bit of magic,
I had to bring indentation into shape to see more clearly what is going on.
Also, mixed tabs/spaces are changed to tabs.

Michael

13 years agobuild: Remove unused functions Import() and GetInfo().
Jelmer Vernooij [Fri, 8 Feb 2008 13:12:46 +0000 (14:12 +0100)]
build: Remove unused functions Import() and GetInfo().

13 years agoMerge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial
Jelmer Vernooij [Fri, 8 Feb 2008 13:02:22 +0000 (14:02 +0100)]
Merge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-trivial

Conflicts:

source/scripting/python/samba/provision.py

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local
Andrew Bartlett [Fri, 8 Feb 2008 06:20:03 +0000 (17:20 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-local

13 years agoRemove unused parameter from provision()
Andrew Bartlett [Fri, 8 Feb 2008 06:12:54 +0000 (17:12 +1100)]
Remove unused parameter from provision()

Andrew Bartlett

13 years agoLDAP now works with the python provision!
Andrew Bartlett [Fri, 8 Feb 2008 06:12:25 +0000 (17:12 +1100)]
LDAP now works with the python provision!

Andrew Bartlett

13 years agoFix LDAP backend with python
Andrew Bartlett [Fri, 8 Feb 2008 06:11:26 +0000 (17:11 +1100)]
Fix LDAP backend with python

We were still setting the hard-coded users.ldb etc, rather than
switching to the ldapi:// URI

Andrew Bartlett

13 years agoReset error strings
Andrew Bartlett [Fri, 8 Feb 2008 06:09:49 +0000 (17:09 +1100)]
Reset error strings

Avoid leaking error strings up to the application, when we are ignoring them.

13 years agoMerge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-test
Andrew Tridgell [Fri, 8 Feb 2008 03:13:48 +0000 (14:13 +1100)]
Merge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-test

13 years agomerge tdb changes from ctdb
Andrew Tridgell [Fri, 8 Feb 2008 03:13:19 +0000 (14:13 +1100)]
merge tdb changes from ctdb

13 years agoFix syntax in LDAP test
Andrew Bartlett [Fri, 8 Feb 2008 02:23:45 +0000 (13:23 +1100)]
Fix syntax in LDAP test

We don't get this far in the test at the moment however.

Andrew Bartlett

13 years agoFix formatting.
Jelmer Vernooij [Fri, 8 Feb 2008 02:03:44 +0000 (03:03 +0100)]
Fix formatting.

13 years agoRevert to ejs for 'provision'
Andrew Bartlett [Fri, 8 Feb 2008 01:54:09 +0000 (12:54 +1100)]
Revert to ejs for 'provision'

Andrew Bartlett

13 years agoRemove unused argument to provision().
Andrew Bartlett [Fri, 8 Feb 2008 01:52:54 +0000 (12:52 +1100)]
Remove unused argument to provision().

Andrew Bartlett

13 years agoLDAP backend provision now works with python.
Andrew Bartlett [Fri, 8 Feb 2008 01:49:30 +0000 (12:49 +1100)]
LDAP backend provision now works with python.

Andrew Bartlett

13 years agoRemove unused ldap_backend argument from provision.py
Andrew Bartlett [Fri, 8 Feb 2008 01:48:58 +0000 (12:48 +1100)]
Remove unused ldap_backend argument from provision.py

Andrew Bartlett

13 years agoFix LDAP backend with python provision
Andrew Bartlett [Fri, 8 Feb 2008 01:48:11 +0000 (12:48 +1100)]
Fix LDAP backend with python provision

The backend databases were still hard-coded as 'users.ldb' etc.

Andrew Bartlett

13 years agoEnsure expected errors do not leak up to the application.
Andrew Bartlett [Fri, 8 Feb 2008 01:45:51 +0000 (12:45 +1100)]
Ensure expected errors do not leak up to the application.

(reset the error string to be sure).

Andrew Bartlett

13 years agoSkip offline tests.
Jelmer Vernooij [Fri, 8 Feb 2008 00:19:14 +0000 (01:19 +0100)]
Skip offline tests.

13 years agoPut temporary test files in test directory.
Jelmer Vernooij [Thu, 7 Feb 2008 23:34:29 +0000 (00:34 +0100)]
Put temporary test files  in test directory.

13 years agoPut ldb test output in prefix.
Jelmer Vernooij [Thu, 7 Feb 2008 23:19:50 +0000 (00:19 +0100)]
Put ldb test output in prefix.

13 years agoMark RAW-OFFLINE as known failing for Samba 4 (because it doesn't have offline suppor...
Jelmer Vernooij [Thu, 7 Feb 2008 23:03:02 +0000 (00:03 +0100)]
Mark RAW-OFFLINE as known failing for Samba 4 (because it doesn't have offline support yet).

13 years agoOverride KRB5CCNAME so existing Kerberos tickets in the users environment don't get...
Jelmer Vernooij [Thu, 7 Feb 2008 22:47:18 +0000 (23:47 +0100)]
Override KRB5CCNAME so existing Kerberos tickets in the users environment don't get overwritten.

13 years agomerge growing tdb for tdb_wipe_all() fix from ctdb
Andrew Tridgell [Thu, 7 Feb 2008 12:06:44 +0000 (23:06 +1100)]
merge growing tdb for tdb_wipe_all() fix from ctdb

13 years agomake the multiplier in the RAW-OFFLINE test settable
Andrew Tridgell [Thu, 7 Feb 2008 12:05:44 +0000 (23:05 +1100)]
make the multiplier in the RAW-OFFLINE test settable

13 years agowkssvc.idl: fix idl for wkssvc_NetrGetJoinableOus[2]()
Stefan Metzmacher [Sun, 3 Feb 2008 01:54:23 +0000 (02:54 +0100)]
wkssvc.idl: fix idl for wkssvc_NetrGetJoinableOus[2]()

metze

13 years agopidl/Samba4::NDR::Parser: fix support for embedded "ref" pointers
Stefan Metzmacher [Thu, 31 Jan 2008 14:04:22 +0000 (15:04 +0100)]
pidl/Samba4::NDR::Parser: fix support for embedded "ref" pointers

The memory allocation of embedded "ref" pointers needs to be the
same as for all other embedded pointers.

metze

13 years agopidl/NDR: fix handling of multilevel pointers in function elements
Stefan Metzmacher [Fri, 1 Feb 2008 22:09:37 +0000 (23:09 +0100)]
pidl/NDR: fix handling of multilevel pointers in function elements

The 2nd or higher level of wire pointers needs to be marked as deferred.

metze

13 years agopidl: get the pointer types correct when an element has multiple pointers
Stefan Metzmacher [Fri, 1 Feb 2008 09:30:47 +0000 (10:30 +0100)]
pidl: get the pointer types correct when an element has multiple pointers

Only the first level gets the pointer type from the
pointer property, the others get them from
the pointer_default() interface property

see http://msdn2.microsoft.com/en-us/library/aa378984(VS.85).aspx
(Here they talk about the rightmost pointer, but testing shows
they mean the leftmost pointer.)

metze

13 years agowin32-tests: add a very usefull w32 programm to explore midl features
Stefan Metzmacher [Wed, 6 Feb 2008 15:34:36 +0000 (16:34 +0100)]
win32-tests: add a very usefull w32 programm to explore midl features

You have to define idl, call the client function
and add a server implementation.

Then you need to compile midltests.exe
and run it, it will dump the generated NDR on the
screen.

metze

13 years agoprefork: also reload stuff in the client processes
Stefan Metzmacher [Tue, 5 Feb 2008 08:16:57 +0000 (09:16 +0100)]
prefork: also reload stuff in the client processes

metze

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet
Andrew Bartlett [Wed, 6 Feb 2008 05:14:09 +0000 (16:14 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet

13 years agosidmap: Some source code cleanups.
Kai Blin [Tue, 5 Feb 2008 10:40:23 +0000 (11:40 +0100)]
sidmap: Some source code cleanups.

13 years agoClarify nomaclature of socket names in process_single and process_prefork
Andrew Bartlett [Tue, 5 Feb 2008 03:51:01 +0000 (14:51 +1100)]
Clarify nomaclature of socket names in process_single and process_prefork

Fix talloc_steal in both cases to steal connected socket (from accept)
onto it's private structure, rather than stealing the bound socket.

Remove termination code from the prefork modal, we want the process to
still stay around, to serve future clients.

Andrew Bartlett and David Disseldorp

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet
Andrew Bartlett [Mon, 4 Feb 2008 22:16:37 +0000 (09:16 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet

13 years agopidl: remove 'pointer_default_top()' support
Stefan Metzmacher [Sat, 2 Feb 2008 09:18:33 +0000 (10:18 +0100)]
pidl: remove 'pointer_default_top()' support

metze

13 years agospoolss.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 21:25:32 +0000 (22:25 +0100)]
spoolss.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agosamr.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 21:11:43 +0000 (22:11 +0100)]
samr.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agonetlogon.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 20:53:41 +0000 (21:53 +0100)]
netlogon.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agowinreg.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 20:35:52 +0000 (21:35 +0100)]
winreg.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agosvcctl.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 20:24:54 +0000 (21:24 +0100)]
svcctl.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agowkssvc.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 20:22:39 +0000 (21:22 +0100)]
wkssvc.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agosrvsvc.idl: don't use pointer_default_top() and specify "unique" explizit
Stefan Metzmacher [Fri, 1 Feb 2008 19:41:26 +0000 (20:41 +0100)]
srvsvc.idl: don't use pointer_default_top() and specify "unique" explizit

Note: this doesn't change the generated output!

metze

13 years agoFix a few more breakages from our recent changes to the server_id
Andrew Bartlett [Mon, 4 Feb 2008 12:04:35 +0000 (23:04 +1100)]
Fix a few more breakages from our recent changes to the server_id
structure.  The BASE-TORTURE test found this problem - caused because
the messaging path was not unique.

If we didn't use a macro for cluster_id_equal(), we could make it
opaque, and avoid this...

Andrew Bartlett

13 years agoRemove useless layer of indirection, where every service called
Andrew Bartlett [Mon, 4 Feb 2008 10:58:29 +0000 (21:58 +1100)]
Remove useless layer of indirection, where every service called
task_service_init() manually.  Now this is called from service.c for
all services.

Andrew Bartlett

13 years agoRemaining changes to implement the prefork process model
Andrew Bartlett [Mon, 4 Feb 2008 06:59:16 +0000 (17:59 +1100)]
Remaining changes to implement the prefork process model

To use, run 'smbd -M prefork'

By default, only the smb service is preforked.  4 children are
created, and all listen for new connections.  The Linux Kernel 'wake
one' behaviour should ensure that only one is given the oportunity to
accept.  We need to look into the ideal number of worker children, as
well as load balancing behaviours.

To change, set:

prefork children : smb = 6

valid service names (smb in this example) match those in 'server services'.

Andrew Bartlett and David Disseldorp

13 years agoRework process_single.c to take advantage of cluster_id() now taking an additional...
Andrew Bartlett [Mon, 4 Feb 2008 06:53:01 +0000 (17:53 +1100)]
Rework process_single.c to take advantage of cluster_id() now taking an additional argument.

Andrew Bartlett and David Disseldorp

13 years agoRework cluster_id() to take an additional argument, as we need
Andrew Bartlett [Mon, 4 Feb 2008 06:51:38 +0000 (17:51 +1100)]
Rework cluster_id() to take an additional argument, as we need
<node>.<pid>.<fd> to be unique in a prefork process environment.

Andrew Bartlett and David Disseldorp

13 years agoRework service init functions to pass down service name. This is
Andrew Bartlett [Mon, 4 Feb 2008 06:48:51 +0000 (17:48 +1100)]
Rework service init functions to pass down service name.  This is
needed to change prefork behaviour based on what service is being
started.

Andrew Bartlett and David Disseldorp

13 years agoMerge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet
Andrew Bartlett [Sun, 3 Feb 2008 22:56:24 +0000 (09:56 +1100)]
Merge branch 'v4-0-test' of git://git.samba.org/samba into 4-0-abartlet

13 years agopidl/Samba4::NDR::Parser: correctly get the name of an array element
Stefan Metzmacher [Thu, 31 Jan 2008 13:57:35 +0000 (14:57 +0100)]
pidl/Samba4::NDR::Parser: correctly get the name of an array element

When we have "*r->out.ous"
(char ***ous, a pointer to a pointer to an array of pointers).
we need to use "(*r->out.ous)[3]" to access the 3rd
element of the array "*r->out.ous[3]" was generated before,
but that's the same as "*(r->out.ous[3])" which would mean
the array would apply to a different level.

This patch prepares support for:

[out,ref,size_is(,num)] [string,charset(UTF16)] uint16 ***names;

It means a [ref] pointer to a [unique] pointer to an array
of [unique] pointers which point to an UTF16 string.

metze

13 years agopidl/Samba4::NDR::Parser: fix $var_name for arrays of scalar reference types
Stefan Metzmacher [Fri, 1 Feb 2008 08:54:25 +0000 (09:54 +0100)]
pidl/Samba4::NDR::Parser: fix $var_name for arrays of scalar reference types

uint32 num;
nstring strings[num];

this should use 'r->strings' instead of
'*r->strings' as the pointer to the array.

metze

13 years agopidl/Samba4::NDR::Parser: move logic for extra get_pointer_of() into a function
Stefan Metzmacher [Fri, 1 Feb 2008 08:49:54 +0000 (09:49 +0100)]
pidl/Samba4::NDR::Parser: move logic for extra get_pointer_of() into a function

metze

13 years agopidl: revert changes it didn't want to push...sorry!
Stefan Metzmacher [Sat, 2 Feb 2008 10:13:03 +0000 (11:13 +0100)]
pidl: revert changes it didn't want to push...sorry!

8ebf16c0741085fa769fcc2929f275ab49b1ea5d Works!!!...pidl/Samba4::NDR::Parser: fix support for embedded "ref" pointers
6fcf2456d0e81898b5779ef1650f38b4c5363a80 WORKS!!!...pidl/NDR: fix handling of multilevel pointers in function elements
0569139ca2960ec5478829c3e66f7ff69bdb55cd LOOKS OK... pidl: get the pointer types correct when an element has multiple pointe
rs
13afc89a87716063180723f0e9cb4f76daca837e CHECKED... pidl/Samba4::NDR::Parser: correctly get the name of an array element
29c104944bcad30c6a2a3fa70d527bf0ee8969de CHECKED... TODO:MSG pidl/Samba4::NDR::Parser: fix ...
3369015f5d8c425e1a9f9d861471028f03f163bb CHECKED... pidl/Samba4::NDR::Parser: move logic for extra get_pointer_of() into a f
unction

metze

13 years agoWorks!!!...pidl/Samba4::NDR::Parser: fix support for embedded "ref" pointers
Stefan Metzmacher [Thu, 31 Jan 2008 14:04:22 +0000 (15:04 +0100)]
Works!!!...pidl/Samba4::NDR::Parser: fix support for embedded "ref" pointers

The memory allocation of embedded "ref" pointers needs to be the
same as for all other embedded pointers.

metze

13 years agoWORKS!!!...pidl/NDR: fix handling of multilevel pointers in function elements
Stefan Metzmacher [Fri, 1 Feb 2008 22:09:37 +0000 (23:09 +0100)]
WORKS!!!...pidl/NDR: fix handling of multilevel pointers in function elements

The 2nd or higher level of wire pointers needs to be marked as deferred.

metze

13 years agoLOOKS OK... pidl: get the pointer types correct when an element has multiple pointers
Stefan Metzmacher [Fri, 1 Feb 2008 09:30:47 +0000 (10:30 +0100)]
LOOKS OK... pidl: get the pointer types correct when an element has multiple pointers

Only the first level gets the pointer type from the
pointer property, the others get them from
the pointer_default() interface property

see http://msdn2.microsoft.com/en-us/library/aa378984(VS.85).aspx
(Here they talk about the rightmost pointer, but testing shows
they mean the leftmost pointer.)

metze

13 years agoCHECKED... pidl/Samba4::NDR::Parser: correctly get the name of an array element
Stefan Metzmacher [Thu, 31 Jan 2008 13:57:35 +0000 (14:57 +0100)]
CHECKED... pidl/Samba4::NDR::Parser: correctly get the name of an array element

When we have "*r->out.ous"
(char ***ous, a pointer to a pointer to an array).
we need to use "(*r->out.ous)[3]" to access the 3rd
element of the array "*r->out.ous[3]" was generated before,
but that's the same as "*(r->out.ous[3])".

metze

13 years agoCHECKED... TODO:MSG pidl/Samba4::NDR::Parser: fix ...
Stefan Metzmacher [Fri, 1 Feb 2008 08:54:25 +0000 (09:54 +0100)]
CHECKED... TODO:MSG pidl/Samba4::NDR::Parser: fix ...

metze

13 years agoCHECKED... pidl/Samba4::NDR::Parser: move logic for extra get_pointer_of() into a...
Stefan Metzmacher [Fri, 1 Feb 2008 08:49:54 +0000 (09:49 +0100)]
CHECKED... pidl/Samba4::NDR::Parser: move logic for extra get_pointer_of() into a function

metze

13 years agopidl: cosmetic fix to use the same value everywhere
Stefan Metzmacher [Fri, 1 Feb 2008 17:20:32 +0000 (18:20 +0100)]
pidl: cosmetic fix to use the same value everywhere

metze

13 years agoSamba4 poor mans debug_ctx()
Amin Azez [Fri, 1 Feb 2008 16:19:36 +0000 (16:19 +0000)]
Samba4 poor mans debug_ctx()

DEBUG(), DEBUGADD() and friends can now use debug_ctx() in the
formatting expressions again,
e.g.

DEBUG(5,("Guid failed to match: %s\n", GUID_string(debug_ctx(), r->guid)));

Sadly it's done with macros (again) but when we need to save the 8 or 16
bytes of object code per DEBUG() expression we can do it the Samba 3 way
with added thread-safety for Samba 4. That could save up to 200K,
allowing 12 bytes for each occurrance of DEBUG...

Signed-off-by: Amin Azez <azez@ufomechanic.net>
13 years agopidl/IDL: don't strip ',' from the properties content
Stefan Metzmacher [Thu, 31 Jan 2008 14:30:50 +0000 (15:30 +0100)]
pidl/IDL: don't strip ',' from the properties content

metze

13 years agopidl/Samba4::NDR::Parse: move the calculation of NDR_PULL_SET_MEM_CTX() flags into...
Stefan Metzmacher [Fri, 1 Feb 2008 06:00:50 +0000 (07:00 +0100)]
pidl/Samba4::NDR::Parse: move the calculation of NDR_PULL_SET_MEM_CTX() flags into one function

metze

13 years agoFix sending of large nttrans responses.
Amin Azez [Thu, 31 Jan 2008 17:04:30 +0000 (17:04 +0000)]
Fix sending of large nttrans responses.

req_grow_data was growing the original req, not this_req which
was being used for the current fragment.

13 years agopidl/Samba4::NDR::Parser: pass down the correct $var_name to AllocateArrayLevel()
Stefan Metzmacher [Thu, 31 Jan 2008 13:48:15 +0000 (14:48 +0100)]
pidl/Samba4::NDR::Parser: pass down the correct $var_name to AllocateArrayLevel()

metze

13 years agoMerge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-test
Andrew Tridgell [Wed, 30 Jan 2008 22:49:15 +0000 (09:49 +1100)]
Merge branch 'v4-0-test' of ssh://git.samba.org/data/git/samba into v4-0-test

13 years agomerged tdb transaction fix
Andrew Tridgell [Wed, 30 Jan 2008 22:48:46 +0000 (09:48 +1100)]
merged tdb transaction fix