samba.git
16 years agor2743: fixed some errors in the description of talloc_reference(). Volker
Andrew Tridgell [Tue, 28 Sep 2004 23:31:52 +0000 (23:31 +0000)]
r2743: fixed some errors in the description of talloc_reference(). Volker
told me he found the description confusing, and given that some of it
was not correct I am not surprised!

added some more docs on the reporting calls

16 years agor2742: - fixed a bug in talloc_unreference()
Andrew Tridgell [Tue, 28 Sep 2004 23:30:14 +0000 (23:30 +0000)]
r2742: - fixed a bug in talloc_unreference()

- made the LOCAL-TALLOC smbtorture test much stricter, checking that
  block counts for every pointer are correct after every operation

16 years agor2741: Definition for ISystemActivator...
Jelmer Vernooij [Tue, 28 Sep 2004 22:58:56 +0000 (22:58 +0000)]
r2741: Definition for ISystemActivator...

16 years agor2740: Finish the RemoteActivation interface. Torture tests will follow
Jelmer Vernooij [Tue, 28 Sep 2004 22:04:35 +0000 (22:04 +0000)]
r2740: Finish the RemoteActivation interface. Torture tests will follow
after pidl has been fixed (to be able to use input variables
inside size_is() for output variables)

16 years agor2739: Start of torture test for winreg wrappers.
Tim Potter [Tue, 28 Sep 2004 21:44:05 +0000 (21:44 +0000)]
r2739: Start of torture test for winreg wrappers.

16 years agor2738: free up the session information as soon as it is invalidated in the
Andrew Tridgell [Tue, 28 Sep 2004 21:42:49 +0000 (21:42 +0000)]
r2738: free up the session information as soon as it is invalidated in the
RAW-CONTEXT test case

16 years agor2737: fixed up a corner case where talloc_unreference() and talloc_free()
Andrew Tridgell [Tue, 28 Sep 2004 21:41:33 +0000 (21:41 +0000)]
r2737: fixed up a corner case where talloc_unreference() and talloc_free()
might not place the pointer in the context specified in the docs. The
code was assuming that pointer was at the head of the child list,
which it may not be, depending on what other operations have happened
in between.

16 years agor2735: More DCOM updates:
Jelmer Vernooij [Tue, 28 Sep 2004 19:20:00 +0000 (19:20 +0000)]
r2735: More DCOM updates:
 - Several updates to the interface definitions after reading some more of the
 specs
  - Add Remote Activation interface
  - Add body extension uuids
 - Add oxidresolve torture test to list
 - Make pidl complain about object interfaces that don't inherit from IUnknown

16 years agor2734: the samdb_destructor can be static
Andrew Tridgell [Tue, 28 Sep 2004 14:38:13 +0000 (14:38 +0000)]
r2734: the samdb_destructor can be static

16 years agor2733: added a note on performance
Andrew Tridgell [Tue, 28 Sep 2004 14:37:28 +0000 (14:37 +0000)]
r2733: added a note on performance

16 years agor2731: use debug level 10 everywhere
Stefan Metzmacher [Tue, 28 Sep 2004 13:50:15 +0000 (13:50 +0000)]
r2731: use debug level 10 everywhere

metze

16 years agor2728: Break arg parsing stuff out of samr.py into a standalone program.
Tim Potter [Tue, 28 Sep 2004 12:49:05 +0000 (12:49 +0000)]
r2728: Break arg parsing stuff out of samr.py into a standalone program.

16 years agor2727: mark the password fields as hidden
Stefan Metzmacher [Tue, 28 Sep 2004 12:42:03 +0000 (12:42 +0000)]
r2727: mark the password fields as hidden

metze

16 years agor2726: added a -r option to ldbdel to allow easy delete of a whole
Andrew Tridgell [Tue, 28 Sep 2004 12:39:41 +0000 (12:39 +0000)]
r2726: added a -r option to ldbdel to allow easy delete of a whole
subtree. Useful when cleaning up a mess after testing.

16 years agor2725: fixed ldbtest to give the basedn to ldb_search()
Andrew Tridgell [Tue, 28 Sep 2004 12:38:59 +0000 (12:38 +0000)]
r2725: fixed ldbtest to give the basedn to ldb_search()

16 years agor2724: - use ldapsrv_service and set it up with the rootDSE and default partition
Stefan Metzmacher [Tue, 28 Sep 2004 12:36:20 +0000 (12:36 +0000)]
r2724: - use ldapsrv_service and set it up with the rootDSE and default partition
  (this is not complete yet)

- call asn1_free() after each call

metze

16 years agor2723: fix some debug messages
Stefan Metzmacher [Tue, 28 Sep 2004 12:30:42 +0000 (12:30 +0000)]
r2723: fix some debug messages

metze

16 years agor2722: remove tmp debug messages
Stefan Metzmacher [Tue, 28 Sep 2004 12:17:03 +0000 (12:17 +0000)]
r2722: remove tmp debug messages

metze

16 years agor2721: added a -b option to ldbtest so it can be used with the new smbd ldap server
Andrew Tridgell [Tue, 28 Sep 2004 12:06:22 +0000 (12:06 +0000)]
r2721: added a -b option to ldbtest so it can be used with the new smbd ldap server
without changing realms

16 years agor2720: -implement sldb_Modify() call
Stefan Metzmacher [Tue, 28 Sep 2004 11:57:56 +0000 (11:57 +0000)]
r2720: -implement sldb_Modify() call

metze

16 years agor2719: an additional note on talloc_unreference()
Andrew Tridgell [Tue, 28 Sep 2004 11:57:40 +0000 (11:57 +0000)]
r2719: an additional note on talloc_unreference()

16 years agor2718: - added a talloc_unreference() function as requested by metze.
Andrew Tridgell [Tue, 28 Sep 2004 11:54:17 +0000 (11:54 +0000)]
r2718: - added a talloc_unreference() function as requested by metze.

- added documentation for talloc_unreference()

- made the abandoned child logic in talloc_free() clearer and more consistent

16 years agor2717: added talloc_p() docs
Andrew Tridgell [Tue, 28 Sep 2004 11:08:43 +0000 (11:08 +0000)]
r2717: added talloc_p() docs

16 years agor2716: created a separate detailed talloc_guide.txt document, after volker
Andrew Tridgell [Tue, 28 Sep 2004 11:04:55 +0000 (11:04 +0000)]
r2716: created a separate detailed talloc_guide.txt document, after volker
complained it was all too confusing :-)

I recommend that everyone who wants to work on Samba4 have a read of this.

16 years agor2714: - add sldb_Add() implementation
Stefan Metzmacher [Tue, 28 Sep 2004 10:44:44 +0000 (10:44 +0000)]
r2714: - add sldb_Add() implementation
- fix some errstr settings

metze

16 years agor2713: better handling of binary values in index key creation
Andrew Tridgell [Tue, 28 Sep 2004 09:34:49 +0000 (09:34 +0000)]
r2713: better handling of binary values in index key creation

16 years agor2712: fixed a bug in ldbtest to make it cope with an existing index
Andrew Tridgell [Tue, 28 Sep 2004 08:17:20 +0000 (08:17 +0000)]
r2712: fixed a bug in ldbtest to make it cope with an existing index

16 years agor2711: added a simple talloc speed tester. I get the following on my laptop:
Andrew Tridgell [Tue, 28 Sep 2004 06:12:07 +0000 (06:12 +0000)]
r2711: added a simple talloc speed tester. I get the following on my laptop:

  MEASURING TALLOC VS MALLOC SPEED
  talloc: 279154 ops/sec
  malloc: 318758 ops/sec

which I think is an acceptable overhead for the increased functionality

16 years agor2710: continue with the new style of providing a parent context whenever
Andrew Tridgell [Tue, 28 Sep 2004 05:44:59 +0000 (05:44 +0000)]
r2710: continue with the new style of providing a parent context whenever
possible to a structure creation routine. This makes for much easier
global cleanup.

16 years agor2709: finally solved the talloc reference problem.
Andrew Tridgell [Tue, 28 Sep 2004 05:42:02 +0000 (05:42 +0000)]
r2709: finally solved the talloc reference problem.

The problem was that the simple "uint_t ref_count;" in a talloc chunk
did not give enough information. It told us that a pointer was
referenced more than once, but it didn't say who it was referenced
by. This means that when the pointer was freed we had no sane way to
clean up the reference.

I have now replaced ref_count with a "refs" list, which means that
references point to the pointer, and the pointer has a linked list of
references. So now we can cleanup from either direction without losing track of anything.

I've also added a LOCAL-TALLOC smbtorture test that tests talloc
behaviour for some common uses.

16 years agor2705: Don't try RemQueryInterface2 for now (doesn't validate)
Jelmer Vernooij [Tue, 28 Sep 2004 01:12:55 +0000 (01:12 +0000)]
r2705: Don't try RemQueryInterface2 for now (doesn't validate)

16 years agor2704: Complain about 'object interfaces' that don't have version 0.0 (the standard
Jelmer Vernooij [Tue, 28 Sep 2004 01:11:40 +0000 (01:11 +0000)]
r2704: Complain about 'object interfaces' that don't have version 0.0 (the standard
doesn't allow them to! I think the idea is that you just create a new
interface that inherits your old interface, thus ensuring backwards-compatibility)
Re-enable to validator

16 years agor2699: Correct handle ServerAlive() and ServerAlive2() + add torture tests
Jelmer Vernooij [Mon, 27 Sep 2004 22:06:27 +0000 (22:06 +0000)]
r2699: Correct handle ServerAlive() and ServerAlive2() + add torture tests

16 years agor2696: DCOM updates:
Jelmer Vernooij [Mon, 27 Sep 2004 16:37:41 +0000 (16:37 +0000)]
r2696: DCOM updates:
 - Start working on OXIDResolver interface
 - Add torture test for SimplePing()

16 years agor2695: revert "Del" renaming
Simo Sorce [Mon, 27 Sep 2004 15:40:12 +0000 (15:40 +0000)]
r2695: revert "Del" renaming

16 years agor2694: DN -> DC these are domain components ...
Simo Sorce [Mon, 27 Sep 2004 15:37:43 +0000 (15:37 +0000)]
r2694: DN -> DC these are domain components ...

16 years agor2693: - send a reply when no attributes there
Stefan Metzmacher [Mon, 27 Sep 2004 15:11:42 +0000 (15:11 +0000)]
r2693: - send a reply when no attributes there
- add some debug messages

metze

16 years agor2690: deleted by mistake
Simo Sorce [Mon, 27 Sep 2004 14:18:25 +0000 (14:18 +0000)]
r2690: deleted by mistake

16 years agor2689: Use consistent naming Del -> Delete
Simo Sorce [Mon, 27 Sep 2004 14:11:11 +0000 (14:11 +0000)]
r2689: Use consistent naming Del -> Delete
Add delete functionality to ldb simple lda server backend
add some const in ldap.h

16 years agor2688: - fix case where listed attributes are asked
Stefan Metzmacher [Mon, 27 Sep 2004 13:20:59 +0000 (13:20 +0000)]
r2688: - fix case where listed attributes are asked
- use the return code of the functions
  and only call ldapsrv_terminate_connection from ldapsrv_recv() or ldapsrv_send()
- the rootdse is now a normal partition

metze

16 years agor2687: Add ORPC flag (used by DCOM)
Jelmer Vernooij [Mon, 27 Sep 2004 12:57:59 +0000 (12:57 +0000)]
r2687: Add ORPC flag (used by DCOM)

16 years agor2686: remove unused gtk+ check
Stefan Metzmacher [Mon, 27 Sep 2004 11:38:36 +0000 (11:38 +0000)]
r2686: remove unused gtk+ check

metze

16 years agor2685: ALLOC_CHECK() after talloc_steal() isn't needed
Stefan Metzmacher [Mon, 27 Sep 2004 10:14:24 +0000 (10:14 +0000)]
r2685: ALLOC_CHECK() after talloc_steal() isn't needed
(thanks simo:-)

metze

16 years agor2684: Free the right talloc context (don't panic when encountering illegal multibyte
Jelmer Vernooij [Mon, 27 Sep 2004 10:13:41 +0000 (10:13 +0000)]
r2684: Free the right talloc context (don't panic when encountering illegal multibyte
sequences)

16 years agor2683: Fix a couple of compile warnings, depend on gtk+-2.4
Jelmer Vernooij [Mon, 27 Sep 2004 09:26:37 +0000 (09:26 +0000)]
r2683: Fix a couple of compile warnings, depend on gtk+-2.4

16 years agor2682: as sambdb holds all search data, don't double free the data
Stefan Metzmacher [Mon, 27 Sep 2004 09:21:52 +0000 (09:21 +0000)]
r2682: as sambdb holds all search data, don't double free the data

metze

16 years agor2681: commit the first semi working search implementation
Stefan Metzmacher [Mon, 27 Sep 2004 08:52:31 +0000 (08:52 +0000)]
r2681: commit the first semi working search implementation
which exports data from a ldb.

I commit this code, so that someone can help me to find a strange
bug

metze

16 years agor2680: switched the libcli/raw/ code over to use talloc_reference(), which simplifies...
Andrew Tridgell [Mon, 27 Sep 2004 08:41:39 +0000 (08:41 +0000)]
r2680: switched the libcli/raw/ code over to use talloc_reference(), which simplifies things quite a bit

16 years agor2679: fixed an uninitialised variable found with valgrind
Andrew Tridgell [Mon, 27 Sep 2004 08:39:51 +0000 (08:39 +0000)]
r2679: fixed an uninitialised variable found with valgrind

16 years agor2678: from_name and to_name aren't needed in smb_iconv_t
Andrew Tridgell [Mon, 27 Sep 2004 08:39:06 +0000 (08:39 +0000)]
r2678: from_name and to_name aren't needed in smb_iconv_t

16 years agor2677: - fixed a bug in the recursive logic talloc_free() when there are
Andrew Tridgell [Mon, 27 Sep 2004 08:38:13 +0000 (08:38 +0000)]
r2677: - fixed a bug in the recursive logic talloc_free() when there are
  circular references (circular references are allowed, they just need
  to be handled carefully inside talloc)

- mark talloc_reference() pointers nicely in the --leak-report-full
  code, so you see what has a reference to what in a useful manner

16 years agor2676: add a test of the reference counting logic in the SAMR server into the
Andrew Tridgell [Mon, 27 Sep 2004 05:15:14 +0000 (05:15 +0000)]
r2676: add a test of the reference counting logic in the SAMR server into the
RPC-SAMR torture test. This closes the samr connection before working
on a open domain handle. The server is supposed to know that the open
domain handle still holds a reference to the connection, so the
connection remains valid even though it has been closed.

16 years agor2675: added a convenience function
Andrew Tridgell [Mon, 27 Sep 2004 05:13:00 +0000 (05:13 +0000)]
r2675: added a convenience function

  void *talloc_reference(const void *context, const void *ptr);

this function makes a secondary reference to ptr, and hangs it off the
given context. This greatly simplifies some of the current reference
counting code in the samr server and I suspect it will be widely used
in other places too.

the way you use it is like this:

domain_state->connect_state = talloc_reference(domain_state, connect_state);

that makes the element connect_state of domain_state a secondary
reference to connect_state. The connect_state structure will then only
be freed when both domain_state and the original connect_state go
away, allowing you to free them independently and in any order.

you could do this alrady using a talloc destructor, and that is what
the samr server did previously, but that meant this construct was
being reinvented in several places. So this convenience function sets
up the destructor for you, giving a much more convenient and less
error prone API.

16 years agor2674: I have realised that talloc() should have its context marked const, as
Andrew Tridgell [Mon, 27 Sep 2004 04:20:18 +0000 (04:20 +0000)]
r2674: I have realised that talloc() should have its context marked const, as
a const pointer really means that "the data pointed to by this pointer
won't change", and that is certainly true of talloc(). The fact that
some behind-the-scenes meta-data can change doesn't matter from the
point of view of const.

this fixes a number of const warnings caused by const data structures
being passed as talloc contexts. That will no longer generate a
warning.

also changed the talloc leak reporting option from --leak-check to
--leak-report, as all it does is generate a report on exit. A new
--leak-report-full option has been added that shows the complete tree
of memory allocations, which is is quite useful in tracking things down.

NOTE: I find it quite useful to insert talloc_report_full(ptr, stderr)
calls at strategic points in the code while debugging memory
allocation problems, particularly before freeing a major context (such
as the connection context). This allows you to see if that context has
been accumulating too much data, such as per-request data, which
should have been freed when the request finished.

16 years agor2673: in the rpc server, free up the old call when we decide to extend an
Andrew Tridgell [Mon, 27 Sep 2004 04:11:37 +0000 (04:11 +0000)]
r2673: in the rpc server, free up the old call when we decide to extend an
existing call rather than creating a new one. This prevents call
structures hanging around on the rpc connection context until it is
closed

16 years agor2672: don't call a variable "dup" as that conflicts with a standard system call...
Andrew Tridgell [Mon, 27 Sep 2004 04:00:23 +0000 (04:00 +0000)]
r2672: don't call a variable "dup" as that conflicts with a standard system call name

16 years agor2671: we're getting too many errors caused by the talloc_realloc() API not
Andrew Tridgell [Mon, 27 Sep 2004 01:36:19 +0000 (01:36 +0000)]
r2671: we're getting too many errors caused by the talloc_realloc() API not
taking a context (so when you pass a NULL pointer you end up with
memory in a top level context). Fixed it by changing the API to take a
context. The context is only used if the pointer you are reallocing is
NULL.

16 years agor2670: use a destructor to auto-close the samr ldb when the last user
Andrew Tridgell [Mon, 27 Sep 2004 00:48:01 +0000 (00:48 +0000)]
r2670: use a destructor to auto-close the samr ldb when the last user
disconnects. Previously the ldb was always kept open.

16 years agor2669: convert make_user_info() and associated functions from malloc to talloc
Andrew Tridgell [Sun, 26 Sep 2004 23:39:18 +0000 (23:39 +0000)]
r2669: convert make_user_info() and associated functions from malloc to talloc

16 years agor2668: steal the cli pointer into the pipe context so a single free destroys the...
Andrew Tridgell [Sun, 26 Sep 2004 23:38:14 +0000 (23:38 +0000)]
r2668: steal the cli pointer into the pipe context so a single free destroys the cli context too

16 years agor2667: Remove forward declaration of static function from function. GCC 3.5 and 4...
Jelmer Vernooij [Sun, 26 Sep 2004 22:19:48 +0000 (22:19 +0000)]
r2667: Remove forward declaration of static function from function. GCC 3.5 and 4.0 don't accept declarations of static functions inside other
functions, see http://gcc.gnu.org/ml/gcc-patches/2004-08/msg02514.html

16 years agor2664: fixed the final server leak for normal operation. We now get a clean report...
Andrew Tridgell [Sun, 26 Sep 2004 12:51:49 +0000 (12:51 +0000)]
r2664: fixed the final server leak for normal operation. We now get a clean report from --leak-check

16 years agor2663: fix an epmapper server leak - another talloc_realloc(NULL, ) leak
Andrew Tridgell [Sun, 26 Sep 2004 12:50:36 +0000 (12:50 +0000)]
r2663: fix an epmapper server leak - another talloc_realloc(NULL, ) leak

16 years agor2662: make --leak-check completely silent if not blocks are allocated
Andrew Tridgell [Sun, 26 Sep 2004 12:48:06 +0000 (12:48 +0000)]
r2662: make --leak-check completely silent if not blocks are allocated

16 years agor2661: fixed a client side memory leak in the clilist code.
Andrew Tridgell [Sun, 26 Sep 2004 11:45:14 +0000 (11:45 +0000)]
r2661: fixed a client side memory leak in the clilist code.

This sort of bug happens quite easily with the new talloc_realloc()
interface. talloc_realloc() now looks like this:

  void *talloc_realloc(void *ptr, size_t size);

and if ptr is not NULL then everything is fine. If ptr is NULL then
talloc_realloc() presumes you want to allocate in the NULL context,
which is probably not what is wanted.

For now the solution is to initialise ptr like this:
  ptr = talloc(mem_ctx, 0);
so when the realloc happens it has a context to get hold of.

I might change the interface of talloc_realloc() later to prevent this
problem in a more robust manner

16 years agor2660: - converted the libcli/raw/ library to use talloc_increase_ref_count()
Andrew Tridgell [Sun, 26 Sep 2004 11:30:20 +0000 (11:30 +0000)]
r2660: - converted the libcli/raw/ library to use talloc_increase_ref_count()
  rather than manual reference counts

- properly support SMBexit in the cifs and posix backends

- added a logoff method to all backends

With these changes the RAW-CONTEXT test now passes against the posix backend

16 years agor2659: removed some extraneous debug msgs
Andrew Tridgell [Sun, 26 Sep 2004 11:27:01 +0000 (11:27 +0000)]
r2659: removed some extraneous debug msgs

16 years agor2658: fixed a couple of error codes found with RAW-CONTEXT
Andrew Tridgell [Sun, 26 Sep 2004 11:25:33 +0000 (11:25 +0000)]
r2658: fixed a couple of error codes found with RAW-CONTEXT

16 years agor2657: if we are already fully authenticated in session setup then the vuid is ignored
Andrew Tridgell [Sun, 26 Sep 2004 11:24:39 +0000 (11:24 +0000)]
r2657: if we are already fully authenticated in session setup then the vuid is ignored

16 years agor2656: moved the seteuid configure tests into the posix backend (these tests
Andrew Tridgell [Sun, 26 Sep 2004 07:44:15 +0000 (07:44 +0000)]
r2656: moved the seteuid configure tests into the posix backend (these tests
don't actually work yet, that will come later)

16 years agor2655: fixed an error in the shutdown of the sock->transport->session->tree
Andrew Tridgell [Sun, 26 Sep 2004 07:04:35 +0000 (07:04 +0000)]
r2655: fixed an error in the shutdown of the sock->transport->session->tree
smbcli raw context handling

16 years agor2654: fixed some more server memory leaks. We are now down to a single leak
Andrew Tridgell [Sun, 26 Sep 2004 06:44:08 +0000 (06:44 +0000)]
r2654: fixed some more server memory leaks. We are now down to a single leak
of 16 bytes, caused by the 16 byte data_blob in the smb_signing
code.

16 years agor2653: - data_blob() and data_blob_talloc() now get automatic names
Andrew Tridgell [Sun, 26 Sep 2004 06:41:59 +0000 (06:41 +0000)]
r2653: - data_blob() and data_blob_talloc() now get automatic names

- talloc_strdup() and related functions get automatic names

16 years agor2650: fixed a memory leak in make_server_info()
Andrew Tridgell [Sun, 26 Sep 2004 05:38:45 +0000 (05:38 +0000)]
r2650: fixed a memory leak in make_server_info()

16 years agor2649: - used some cpp tricks to make users of talloc() and talloc_realloc()
Andrew Tridgell [Sun, 26 Sep 2004 04:59:03 +0000 (04:59 +0000)]
r2649: - used some cpp tricks to make users of talloc() and talloc_realloc()
  to get auto-naming of pointers very cheaply.

- fixed a couple of memory leaks found with the new tricks

A typical exit report for smbd is now:

talloc report on 'null_context' (total 811 bytes in 54 blocks)
        auth/auth_sam.c:334            contains     20 bytes in   1 blocks
        struct auth_serversupplied_info contains    498 bytes in  33 blocks
        UNNAMED                        contains      8 bytes in   1 blocks
        lib/data_blob.c:40             contains     16 bytes in   1 blocks
        iconv(CP850,UTF8)              contains     61 bytes in   4 blocks
        iconv(UTF8,CP850)              contains     61 bytes in   4 blocks
        iconv(UTF8,UTF-16LE)           contains     67 bytes in   4 blocks
        iconv(UTF-16LE,UTF8)           contains     67 bytes in   4 blocks
        UNNAMED                        contains     13 bytes in   1 blocks

which is much better than before

16 years agor2648: - use a destructor on struct server_connection to simplify the
Andrew Tridgell [Sun, 26 Sep 2004 03:50:24 +0000 (03:50 +0000)]
r2648: - use a destructor on struct server_connection to simplify the
  connection termination cleanup, and to ensure that the event
  contexts are properly removed for every process model

- gave auth_context the new talloc treatment, which removes another
  source of memory leaks.

16 years agor2646: - use a talloc destructor to ensure that sockets from the new socket
Andrew Tridgell [Sun, 26 Sep 2004 03:05:04 +0000 (03:05 +0000)]
r2646: - use a talloc destructor to ensure that sockets from the new socket
  library are closed on abnormal termination

- convert the service.h structures to the new talloc methods

16 years agor2645: converted the NTLMSSP code to the new style of talloc
Andrew Tridgell [Sun, 26 Sep 2004 02:16:25 +0000 (02:16 +0000)]
r2645: converted the NTLMSSP code to the new style of talloc

16 years agor2644: removed an unused function
Andrew Tridgell [Sun, 26 Sep 2004 02:14:25 +0000 (02:14 +0000)]
r2644: removed an unused function

16 years agor2643: convert more of the auth subsyystem to the new talloc methods. This
Andrew Tridgell [Sun, 26 Sep 2004 01:43:05 +0000 (01:43 +0000)]
r2643: convert more of the auth subsyystem to the new talloc methods. This
also fixes a memory leak found with --leak-check.

16 years agor2642: smb_iconv_t is a pointer, so checks against -1 errors should use a cast
Andrew Tridgell [Sun, 26 Sep 2004 01:41:55 +0000 (01:41 +0000)]
r2642: smb_iconv_t is a pointer, so checks against -1 errors should use a cast

16 years agor2641: talloc_p() now produces a named talloc pointer, with the name
Andrew Tridgell [Sun, 26 Sep 2004 01:41:04 +0000 (01:41 +0000)]
r2641: talloc_p() now produces a named talloc pointer, with the name
auto-derived from the type you are allocating. This is done with
basically zero overhead by relying on the stringify operator in cpp
producing string constants.

the result is that --leak-check nicely names all pointers that come
from talloc_p()

16 years agor2640: valgrind does a great job on some types of memory leaks, but is slow
Andrew Tridgell [Sun, 26 Sep 2004 01:14:26 +0000 (01:14 +0000)]
r2640: valgrind does a great job on some types of memory leaks, but is slow
and can't properly handle leaks of doubly linked lists which we use a
lot (as the memory is always reachable). Even with --show-reachable
its hard to track leaks down sometimes.

I realised that talloc does have the necessary information to track
these, and by using the cascading property of the new talloc it can
report on leaks in a much more succinct fashion than valgrind can.

I have added a new samba option --leak-check that applies to all Samba
tools. When enabled it prints a leak report summarising all top level
contexts that are present when the program exits. A typical report
looks like this:

talloc report on 'null_context' (total 1071 bytes in 52 blocks)
        iconv(CP850,UTF8)              contains     43 bytes in   3 blocks
        UNNAMED                        contains     24 bytes in   1 blocks
        UNNAMED                        contains     24 bytes in   1 blocks
        dcesrv_init                    contains    604 bytes in  26 blocks
        server_service                 contains    120 bytes in   6 blocks
        UNNAMED                        contains     24 bytes in   1 blocks
        UNNAMED                        contains     24 bytes in   1 blocks
        server_service                 contains    104 bytes in   4 blocks
        server_context                 contains     12 bytes in   2 blocks
        iconv(UTF8,UTF-16LE)           contains     46 bytes in   3 blocks
        iconv(UTF-16LE,UTF8)           contains     46 bytes in   3 blocks

the numbers are recursive summaries for all the memory hanging off each context.

this option is not thread safe when used, but the code is thread safe
if the option is not given, so I don't think thats a problem.

16 years agor2639: we doon't need the valid_table code, so get rid of it
Andrew Tridgell [Sun, 26 Sep 2004 01:09:04 +0000 (01:09 +0000)]
r2639: we doon't need the valid_table code, so get rid of it

16 years agor2638: do lazy initialisation of iconv handles, so we don't initialise a
Andrew Tridgell [Sun, 26 Sep 2004 01:08:29 +0000 (01:08 +0000)]
r2638: do lazy initialisation of iconv handles, so we don't initialise a
handle unless we use it. This saves quite a bit of memory (libc chews
a lot loading a handle). Typically smbd now loads 3 handles, instead
of 36.

16 years agor2635: mem_ctx cleanups on the lsa and netlogon pipes in the rpc server
Andrew Tridgell [Sat, 25 Sep 2004 13:28:38 +0000 (13:28 +0000)]
r2635: mem_ctx cleanups on the lsa and netlogon pipes in the rpc server

16 years agor2634: use discard_const_p() in a few places
Andrew Tridgell [Sat, 25 Sep 2004 12:48:56 +0000 (12:48 +0000)]
r2634: use discard_const_p() in a few places

16 years agor2633: fixed some function types in the (unused) print backend
Andrew Tridgell [Sat, 25 Sep 2004 12:48:09 +0000 (12:48 +0000)]
r2633: fixed some function types in the (unused) print backend

16 years agor2632: a new approach to handling const errors. We have had huge numbers of
Andrew Tridgell [Sat, 25 Sep 2004 12:36:36 +0000 (12:36 +0000)]
r2632: a new approach to handling const errors. We have had huge numbers of
const warnings for a long time, and no real way to approach a
solution. Some of them are unavoidable due to the way the C standard
works (for example, any function that provides strchr() like
functionality _must_ produce a const warning)

I will be converting a bunch of places that currently produce const
warnings to use the discard_const_p(). Some of these will be
unavoidable const problems, some of them will be ones we will fix up
over time. At least this change means we will no longer be swamped
with const warnings, and we will easily be able to see when new
problems emerge.

16 years agor2631: the strchr family of functions should not return const strings.
Andrew Tridgell [Sat, 25 Sep 2004 12:31:49 +0000 (12:31 +0000)]
r2631: the strchr family of functions should not return const strings.

16 years agor2630: I missed a couple of places in the gensec talloc conversion
Andrew Tridgell [Sat, 25 Sep 2004 12:30:10 +0000 (12:30 +0000)]
r2630: I missed a couple of places in the gensec talloc conversion

16 years agor2629: convert gensec to the new talloc model
Andrew Tridgell [Sat, 25 Sep 2004 12:08:57 +0000 (12:08 +0000)]
r2629: convert gensec to the new talloc model

by making our gensec structures a talloc child of the open connection
we can be sure that it will be destroyed when the connection is
dropped.

16 years agor2628: got rid of some warnings and converted a few more places to use hierarchical...
Andrew Tridgell [Sat, 25 Sep 2004 11:48:30 +0000 (11:48 +0000)]
r2628: got rid of some warnings and converted a few more places to use hierarchical memory allocation

16 years agor2627: use the new talloc capabilities in a bunch more places in the rpc
Andrew Tridgell [Sat, 25 Sep 2004 11:24:10 +0000 (11:24 +0000)]
r2627: use the new talloc capabilities in a bunch more places in the rpc
server code. This fixes a number of memory leaks I found when testing
with valgrind and smbtorture, as the cascading effect of a
talloc_free() ensures that anything derived from the top level object
is destroyed on disconnect.

16 years agor2626: the symbol gai_error is defined in /usr/include, so don't use that name in...
Andrew Tridgell [Sat, 25 Sep 2004 11:19:29 +0000 (11:19 +0000)]
r2626: the symbol gai_error is defined in /usr/include, so don't use that name in our code

16 years agor2625: use talloc_p, not talloc when possible (when allocating a structure in particu...
Andrew Tridgell [Sat, 25 Sep 2004 11:18:50 +0000 (11:18 +0000)]
r2625: use talloc_p, not talloc when possible (when allocating a structure in particular), as it gives us type checking.

16 years agor2624: - save some system calls by only trying read/write operations that select...
Andrew Tridgell [Sat, 25 Sep 2004 11:18:04 +0000 (11:18 +0000)]
r2624: - save some system calls by only trying read/write operations that select has indicated are possible

- when a socket is dead, don't try to do anything more on it

16 years agor2623: don't do pointer arithmetic on void*, as it doesn't work with non-GNU compilers
Andrew Tridgell [Sat, 25 Sep 2004 11:17:09 +0000 (11:17 +0000)]
r2623: don't do pointer arithmetic on void*, as it doesn't work with non-GNU compilers

16 years agor2622: to implement the SOCKET_FLAG_BLOCK option in the socket library we
Andrew Tridgell [Sat, 25 Sep 2004 11:16:30 +0000 (11:16 +0000)]
r2622: to implement the SOCKET_FLAG_BLOCK option in the socket library we
need to add MSG_WAITALL to the recv() flags. This is needed by the
current server code or sometimes it will fail with a receive error.

16 years agor2621: - now that the client code is non-blocking, we no longer need
Andrew Tridgell [Sat, 25 Sep 2004 11:15:18 +0000 (11:15 +0000)]
r2621: - now that the client code is non-blocking, we no longer need
  write_data and read_data, which are inherently blocking operations

- got rid of some old NBT keepalive routines that are not needed