ira/wip.git
12 years agoAdd release script for tevent tevent-0.9.8
Simo Sorce [Fri, 4 Sep 2009 22:20:29 +0000 (18:20 -0400)]
Add release script for tevent

12 years agos4:configure: require tevent >= 0.9.8
Stefan Metzmacher [Fri, 4 Sep 2009 10:57:13 +0000 (12:57 +0200)]
s4:configure: require tevent >= 0.9.8

metze

12 years agotevent: change version to 0.9.8 after some critical bugs have been fixed
Stefan Metzmacher [Fri, 4 Sep 2009 10:56:39 +0000 (12:56 +0200)]
tevent: change version to 0.9.8 after some critical bugs have been fixed

metze

12 years agocifs.upcall: do a brute-force search for KRB5 credcache
Jeff Layton [Fri, 4 Sep 2009 10:29:44 +0000 (06:29 -0400)]
cifs.upcall: do a brute-force search for KRB5 credcache

A few weeks ago, I added some code to cifs.upcall to take the pid sent
by the kernel and use that to get the value of the $KRB5CCNAME
environment var for the process. That works fine on the initial mount,
but could be problematic on reconnect.

There's no guarantee on a reconnect that the process that initiates the
upcall will have $KRB5CCNAME pointed at the correct credcache. Because
of this, the current scheme isn't going to be reliable enough and we
need to use something different.

This patch replaces that scheme with one very similar to the one used by
rpc.gssd in nfs-utils. It searches the credcache dir (currently
hardcoded to /tmp) for a valid credcache for the given uid. If it finds
one then it uses that as the credentials cache. If it finds more than
one, it uses the one with the latest TGT expiration.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
12 years agos4:python fixed subunit tests of dcerpc
Andrew Tridgell [Fri, 4 Sep 2009 08:07:04 +0000 (18:07 +1000)]
s4:python  fixed subunit tests of dcerpc

The version of the unitest python module in Ubuntu Jaunty doesn't seem
to support this many level of subdirectories. Moving the tests up one
level solves the problem.

12 years agoldb: make ldb module programming less error prone
Andrew Tridgell [Fri, 4 Sep 2009 07:22:20 +0000 (17:22 +1000)]
ldb: make ldb module programming less error prone

When a top level method in a module returns an error, it is supposed
to call ldb_module_done(). We ran across a case where this wasn't
done, and then found that in fact that are hundreds of similar cases
in our modules. It took Andrew and I a full day to work out that this
was the cause of a subtle segv in another part of the code.

To try to prevent this happening again, this patch changes
ldb_next_request() to catch the error by checking if a module
returning an error has called ldb_module_done(). If it hasn't then the
call is made on behalf of the module.

12 years agos3:winbind: Fix Coverity ID 933: FORWARD_NULL
Volker Lendecke [Fri, 4 Sep 2009 05:59:51 +0000 (07:59 +0200)]
s3:winbind: Fix Coverity ID 933: FORWARD_NULL

12 years agos3:smbd: Fix Coverity ID 937, REVERSE_INULL
Volker Lendecke [Fri, 4 Sep 2009 05:57:19 +0000 (07:57 +0200)]
s3:smbd: Fix Coverity ID 937, REVERSE_INULL

12 years agos3:winbind: Fix Coverity ID 940: Resource Leak
Volker Lendecke [Fri, 4 Sep 2009 05:52:28 +0000 (07:52 +0200)]
s3:winbind: Fix Coverity ID 940: Resource Leak

12 years agoldb: ensure we cancel a ldb transaction
Andrew Tridgell [Fri, 4 Sep 2009 03:59:44 +0000 (13:59 +1000)]
ldb: ensure we cancel a ldb transaction

When we fail a ldbadd or ldbedit we should cancel the transaction to
prevent ldb giving a warning about having a open transaction in the
ldb destructor

12 years agos4: fixed a missing NULL termination in a attribute list passed to ldb_search
Andrew Tridgell [Fri, 4 Sep 2009 03:58:17 +0000 (13:58 +1000)]
s4: fixed a missing NULL termination in a attribute list passed to ldb_search

12 years agoreport the location of the original talloc_free on double free talloc-2.0.0
Andrew Tridgell [Fri, 4 Sep 2009 03:56:46 +0000 (13:56 +1000)]
report the location of the original talloc_free on double free

When we get a double free abort from talloc it is often hard to work
out where the first free came from. This patch takes advantage of the
fact that _talloc_free() now takes a location the free was called from
to allow the double free abort code to print the location of the first
free that conflicts.

12 years agowbinfo: fix various valgrind warnings and an invalid free.
Günther Deschner [Thu, 3 Sep 2009 22:02:13 +0000 (00:02 +0200)]
wbinfo: fix various valgrind warnings and an invalid free.

Kai, please check.

Guenther

12 years agos3-ntlm_auth: fix two segfaults in diagnostics mode.
Günther Deschner [Thu, 3 Sep 2009 21:49:50 +0000 (23:49 +0200)]
s3-ntlm_auth: fix two segfaults in diagnostics mode.

Guenther

12 years agoHopefully last part of the fix for bug 6651 - smbd SIGSEGV when breaking oplocks.
Jeremy Allison [Thu, 3 Sep 2009 14:40:48 +0000 (07:40 -0700)]
Hopefully last part of the fix for bug 6651 - smbd SIGSEGV when breaking oplocks.
This one is subtle. There is a race condition where a signal can be
queued for oplock break, and then the file can be closed by the client
before the signal can be processed. Currently if this occurs we panic
(we can't match an incoming signal fd with a fsp pointer). Simply log
the error (at debug level 10 right now, might be too much) and then
return without processing the break request. It looks like there is
another race condition with this fix, but here's why it won't happen.
If the signal was pending (caused by a kernel oplock break from a
local file open), and the client closed the file and then re-opened
another file which happened to use the same file descriptor as the
file just closed, then theoretically the oplock break requests could
be processed on the wrong fd. Here's why this should be very rare..
Processing a pending signal always take precedence over an incoming
network request, so as long as the client close request is non-chained
then the break signal should always be harmlessly processed *before*
the open can be called. If the open is chained onto the close, and
the fd on the new open is the same as the old closed fd, then it's
possible this race will occur. However, all that will happen is that
we'll lose the oplock on this file. A shame, but not a fatal event.
Jeremy.

12 years agoAnother part of the fix for bug 6651 - smbd SIGSEGV when breaking oplocks.
Jeremy Allison [Thu, 3 Sep 2009 14:38:21 +0000 (07:38 -0700)]
Another part of the fix for bug 6651 - smbd SIGSEGV when breaking oplocks.
SA_INFO_QUEUE_COUNT *MUST* be a power of 2, in order for the ring buffer
wrap to work correctly at the 32 bit boundary. Thanks to Petr
Vandrovec <petr@vandrovec.name> for this.

12 years agoFirst attempt to implement dcesrv_drsuapi_DsGetNCChanges
Anatoliy Atanasov [Wed, 2 Sep 2009 14:39:28 +0000 (17:39 +0300)]
First attempt to implement dcesrv_drsuapi_DsGetNCChanges

So far it returns the ctr6 responce without proper linked attributes
support and metadata. A couple of improvements are the filter in the search
uses '(uSNChanged>=N)', added extended dn search support, non-replicated attributes
are excluded from the result.

12 years agoFix the dsdb_syntax_OID_ldb_to_drsuapi function
Anatoliy Atanasov [Wed, 2 Sep 2009 14:20:16 +0000 (17:20 +0300)]
Fix the dsdb_syntax_OID_ldb_to_drsuapi function

This replace the dsdb_syntax_FOOBAR_ldb_to_drsuapi function,
which was left as a TODO code. Implementation in both added functions
is completely identical and probably should differ in the future.

12 years agoIgnoring generated files:
Kamen Mazdrashki [Fri, 21 Aug 2009 11:39:09 +0000 (14:39 +0300)]
Ignoring generated files:

libcli/smb/smb_common_proto.h
source4/torture/ntp/proto.h

12 years agoanother large change to the linked_attribute module
Andrew Tridgell [Thu, 3 Sep 2009 08:36:02 +0000 (18:36 +1000)]
another large change to the linked_attribute module

This one copes with deleted objects where linked attributes have been
set on the module. We hit this when we do the ldb wipe at the start of
a provision, which trigers linked attribute updates, but for objects
that have disappeared. We need to ensure that the linked attribute
updates only happen on the right object, and if the object gets
re-created (as happens with a provision) then it is not the right
object.

To cope with this we record the GUID of the object when the operation
that triggered the linked attribute update comes in, and then find the
DN by suing that GUID when we apply the change in the prepare commit
hook.

12 years agohook on prepare_commit instead of transaction_end
Andrew Tridgell [Thu, 3 Sep 2009 08:33:17 +0000 (18:33 +1000)]
hook on prepare_commit instead of transaction_end

This allows for safe transaction end aborts

12 years agogreatly simplify the transaction processing in the partition module
Andrew Tridgell [Thu, 3 Sep 2009 08:32:42 +0000 (18:32 +1000)]
greatly simplify the transaction processing in the partition module

Now that ldb is calling prepare commit separately, the job of the
partition module on transaction end is much simpler (and more robust!)

12 years agoshow more reasonable object counts during a vampire
Andrew Tridgell [Thu, 3 Sep 2009 08:31:45 +0000 (18:31 +1000)]
show more reasonable object counts during a vampire

We now show the total number of objects we have processed, which gives
the user a better idea of how much has been done. A vampire on a large
domain can take an hour or more (which needs to be fixed btw, it is a
problem with the lack of scalability of the ltdb index code). Watching
the same msg for an hour makes you wonder if any progress is being
made!

12 years agoalways use prepare_commit in ldb transaction commits if possible
Andrew Tridgell [Thu, 3 Sep 2009 08:29:58 +0000 (18:29 +1000)]
always use prepare_commit in ldb transaction commits if possible

The reason we need this is to make multi-tdb transactions safe, with
the partition module. The linked_attributes and repl_meta_data modules
now do extra processing when the transaction ends, and that processing
can fail. When it fails we need to cancel the transaction, which we
can only do if the hook is on the prepare commit instead of the end
transaction call. Otherwise the partition module cannot ensure that no
commit has been done on another partition.

12 years agoadded dsdb_find_guid_by_dn()
Andrew Tridgell [Thu, 3 Sep 2009 08:27:29 +0000 (18:27 +1000)]
added dsdb_find_guid_by_dn()

This will be used by the linked_attribute module

12 years agoallow setting of the debug level in python from C
Andrew Tridgell [Thu, 3 Sep 2009 03:03:31 +0000 (13:03 +1000)]
allow setting of the debug level in python from C

12 years agochange repl_meta_data to process linked_attributes structures in end_transaction
Andrew Tridgell [Thu, 3 Sep 2009 02:55:57 +0000 (12:55 +1000)]
change repl_meta_data to process linked_attributes structures in end_transaction

When running at functional level 2 or above, the repl_meta_data module
can receive linked attribute structures from the repl replication
task. These attributes can come through DRS before the associated
objects have been created. To cope with this, we need to process
linked attributes in the end_transaction hook.

12 years agofixed transaction handling in linked_attributes module
Andrew Tridgell [Thu, 3 Sep 2009 02:53:36 +0000 (12:53 +1000)]
fixed transaction handling in linked_attributes module

We need to call down to the next transaction function when we finish
in linked_attributes.

This also changes linked_attributes to use the common
dsdb_find_dn_by_guid() function

12 years agoadd the the linked attributes elements to the repl structure
Andrew Tridgell [Thu, 3 Sep 2009 02:52:31 +0000 (12:52 +1000)]
add the the linked attributes elements to the repl structure

This exposes the linked_attributes to the repl_meta_data module

12 years agotell the server that we support linked attribute replication
Andrew Tridgell [Thu, 3 Sep 2009 02:51:42 +0000 (12:51 +1000)]
tell the server that we support linked attribute replication

12 years agoadded dsdb_find_dn_by_guid()
Andrew Tridgell [Thu, 3 Sep 2009 02:51:22 +0000 (12:51 +1000)]
added dsdb_find_dn_by_guid()

This came from the linked_attributes module, but now the
repl_meta_data module needs the same functionality, so move it to a
common routine.

12 years agofix the ndr print routines for samba4
Andrew Tridgell [Thu, 3 Sep 2009 02:50:19 +0000 (12:50 +1000)]
fix the ndr print routines for samba4

This uses a hackish #if just for samba4. The proper fix is much more
complex.

12 years agosupport config files in the current directory
Andrew Tridgell [Wed, 2 Sep 2009 11:51:02 +0000 (21:51 +1000)]
support config files in the current directory

12 years agoFix for DSSYNC test against Windows 2003
Anatoliy Atanasov [Tue, 25 Aug 2009 15:39:06 +0000 (18:39 +0300)]
Fix for DSSYNC test against Windows 2003

12 years agos3:libsmb: Attempt to fix bug 6665
Volker Lendecke [Thu, 3 Sep 2009 07:18:10 +0000 (09:18 +0200)]
s3:libsmb: Attempt to fix bug 6665

Before the async libsmb rewrites, we sent tid==0 on negprot. With the rewrite,
we send 0xffff. This *should* not matter, but this is one difference in the
sniffs I see.

12 years agos3-net: allow to exit "net rpc sh" with "q" as well.
Günther Deschner [Wed, 2 Sep 2009 19:09:13 +0000 (21:09 +0200)]
s3-net: allow to exit "net rpc sh" with "q" as well.

Guenther

12 years agos3-netlogon: use WERRORs in NETLOGON_INFO structures.
Günther Deschner [Wed, 2 Sep 2009 13:29:44 +0000 (15:29 +0200)]
s3-netlogon: use WERRORs in NETLOGON_INFO structures.

Guenther

12 years agoDo not redefine strupr
Matt Kraai [Wed, 26 Aug 2009 21:50:28 +0000 (14:50 -0700)]
Do not redefine strupr

Signed-off-by: Günther Deschner <gd@samba.org>
12 years agos3-pdb_ldap: Fix bug #4296: Clean up group membership while deleting a user.
Günther Deschner [Tue, 23 Jun 2009 22:33:44 +0000 (00:33 +0200)]
s3-pdb_ldap: Fix bug #4296: Clean up group membership while deleting a user.

Note that this only is tried with editposix=yes.

Guenther

12 years agos4-smbtorture: test netr_ServerSetPassword2 against Samba3.
Günther Deschner [Fri, 28 Aug 2009 14:04:08 +0000 (16:04 +0200)]
s4-smbtorture: test netr_ServerSetPassword2 against Samba3.

Guenther

12 years agos3-netlogon: implement _netr_ServerPasswordSet2.
Günther Deschner [Thu, 27 Aug 2009 21:30:50 +0000 (23:30 +0200)]
s3-netlogon: implement _netr_ServerPasswordSet2.

Guenther

12 years agos3-netlogon: rework _netr_ServerPasswordSet.
Günther Deschner [Thu, 27 Aug 2009 21:30:14 +0000 (23:30 +0200)]
s3-netlogon: rework _netr_ServerPasswordSet.

Guenther

12 years agoshow the full set of command line options for ldb tools
Andrew Tridgell [Wed, 2 Sep 2009 08:31:22 +0000 (18:31 +1000)]
show the full set of command line options for ldb tools

I always found it hard to remember some of the options. We might as
well use popt to give us the full list

12 years agos3-netlogon: implement _netr_LogonSamLogonWithFlags().
Günther Deschner [Mon, 31 Aug 2009 15:13:05 +0000 (17:13 +0200)]
s3-netlogon: implement _netr_LogonSamLogonWithFlags().

Guenther

12 years agotraverse the ac list in reverse order
Andrew Tridgell [Wed, 2 Sep 2009 08:18:33 +0000 (18:18 +1000)]
traverse the ac list in reverse order

items are added to the linked attribute list using DLIST_ADD(), which
means to commit them to the database in the same order they came from
the server we need to walk the list backwards when we traverse it

12 years agorepl_meta_data should only be included when we are a DC
Andrew Tridgell [Wed, 2 Sep 2009 07:08:52 +0000 (17:08 +1000)]
repl_meta_data should only be included when we are a DC

12 years agos4:dsdb rewrite the linked_atrributes code to commit in the end_transaction hook
Andrew Tridgell [Wed, 2 Sep 2009 07:02:56 +0000 (17:02 +1000)]
s4:dsdb rewrite the linked_atrributes code to commit in the end_transaction hook

linked attribute changes can come in any order. This means it is
possible for a forward link to come over the wire in DRS before the
target even exists. To make this work this patch changed the linked
attributes module to gather up all the changes it needs to make in a
linked list, then execute the changes in the end_transaction hook for
the module.

During that commit phase we also fix up all the DNs that we got by
searching for their GUID, as the objects may have moved after the
linked attribute was sent, but before the end of the transaction

12 years agomove the repl_meta_data module up the ldb module stack
Andrew Tridgell [Wed, 2 Sep 2009 06:58:04 +0000 (16:58 +1000)]
move the repl_meta_data module up the ldb module stack

The repl_meta_data module needs to be above the linked_attributes
module, to allow linked_attributes to do its magic

12 years agowrap the entire vampire operation in a transaction
Andrew Tridgell [Wed, 2 Sep 2009 06:57:25 +0000 (16:57 +1000)]
wrap the entire vampire operation in a transaction

We want to grab the whole database, or none of it.
This is also needed to get linked attributes right

12 years agouse ldb_cmdline_help() in ldbsearch
Andrew Tridgell [Wed, 2 Sep 2009 06:56:39 +0000 (16:56 +1000)]
use ldb_cmdline_help() in ldbsearch

12 years agoadded ldb_cmdline_help()
Andrew Tridgell [Wed, 2 Sep 2009 06:56:20 +0000 (16:56 +1000)]
added ldb_cmdline_help()

This allows the ldb tools to show their full command line options

12 years agoadd the partition_control control to replication requests
Andrew Tridgell [Wed, 2 Sep 2009 06:55:57 +0000 (16:55 +1000)]
add the partition_control control to replication requests

We know the partition DN from the DRS objects, we need to pass this
down the modules below us to ensure they operate on the right
partition

12 years agochange the dsdb_control_current_partition to not include internal variables
Andrew Tridgell [Wed, 2 Sep 2009 03:36:54 +0000 (13:36 +1000)]
change the dsdb_control_current_partition to not include internal variables

This structures was used in two ways. In one way it held variables
that are logically internal to the partition module, and in the other
way it was used to pass the partition DN down to other modules. This
change makes the structure contain just the dn which is being passed
down.

This change is part of the support for linked attributes. We will be
passing this control down from above the partition module to force
which partition a request acts upon. The partition module now only
adds this control if it isn't already there.

12 years agodon't allow two controls to be added with the same OID
Andrew Tridgell [Wed, 2 Sep 2009 03:33:37 +0000 (13:33 +1000)]
don't allow two controls to be added with the same OID

Two controls with the same OID makes no sense, as they may
have different data attached

12 years agoDisplay ldif formatted versions of all DRS changes at log level 4
Andrew Tridgell [Wed, 2 Sep 2009 01:20:26 +0000 (11:20 +1000)]
Display ldif formatted versions of all DRS changes at log level 4

This helps a lot with debugging the DRS replication code

12 years agoWrap DRS changes in a transaction
Andrew Tridgell [Wed, 2 Sep 2009 01:17:43 +0000 (11:17 +1000)]
Wrap DRS changes in a transaction

We should always apply a whole set of DRS changes or none of them. See
[MS-DRSR] 3.3.2

12 years agoadd a ref to the WSPP docs
Andrew Tridgell [Wed, 2 Sep 2009 01:15:44 +0000 (11:15 +1000)]
add a ref to the WSPP docs

12 years agofixed spelling
Andrew Tridgell [Wed, 2 Sep 2009 01:15:29 +0000 (11:15 +1000)]
fixed spelling

12 years agoadded ldb_ldif_message_string()
Andrew Tridgell [Wed, 2 Sep 2009 01:14:33 +0000 (11:14 +1000)]
added ldb_ldif_message_string()

This function provides a easy function for displaying a ldb_message
structure in a human readable format. It is especially useful for
calling in gdb.

12 years agos3-smbpasswd: Fix Bug #6584: allow DOM\user when changing passwords remotely.
Simo Sorce [Sat, 29 Aug 2009 23:31:02 +0000 (19:31 -0400)]
s3-smbpasswd: Fix Bug #6584: allow DOM\user when changing passwords remotely.

Signed-off-by: Günther Deschner <gd@samba.org>
12 years agowbclient: Fix Bug #6680: always activate handling of large (> 256 byte) ntlmv2
Günther Deschner [Tue, 1 Sep 2009 09:58:05 +0000 (11:58 +0200)]
wbclient: Fix Bug #6680: always activate handling of large (> 256 byte) ntlmv2
blobs in wbcAuthenticateUserEx().

Guenther

12 years agos3-docs: Update build howto.
Karolin Seeger [Tue, 1 Sep 2009 10:55:23 +0000 (12:55 +0200)]
s3-docs: Update build howto.

This addresses bug #6661.

Karolin

12 years agonetlogon: fill in more details for the netr_LogonControlX family of calls.
Günther Deschner [Mon, 31 Aug 2009 15:12:07 +0000 (17:12 +0200)]
netlogon: fill in more details for the netr_LogonControlX family of calls.

Guenther

12 years agos3-netlogon: add DS_WEB_SERVICE_REQUIRED bit for dsgetdcname calls.
Günther Deschner [Mon, 31 Aug 2009 15:12:34 +0000 (17:12 +0200)]
s3-netlogon: add DS_WEB_SERVICE_REQUIRED bit for dsgetdcname calls.

Guenther

12 years agoldb: Depend on libtalloc 2.0.0
Simo Sorce [Sun, 30 Aug 2009 21:52:03 +0000 (17:52 -0400)]
ldb: Depend on libtalloc 2.0.0

12 years agoldb: cosmetic changes in ldb_dn
Simo Sorce [Sun, 30 Aug 2009 20:07:44 +0000 (16:07 -0400)]
ldb: cosmetic changes in ldb_dn

- remove trailing spaces and tabs
- shorten some variable names for readability
- try to break superlong lines for readability

12 years agos3: Fix libtalloc so version
Simo Sorce [Sun, 30 Aug 2009 17:53:14 +0000 (13:53 -0400)]
s3: Fix libtalloc so version

12 years agos3:libwbclient: Fix bug 6349, initialize domain info struct
Volker Lendecke [Sun, 30 Aug 2009 09:39:41 +0000 (11:39 +0200)]
s3:libwbclient: Fix bug 6349, initialize domain info struct

12 years agos3:winbind: Fix bug 5626
Volker Lendecke [Sun, 30 Aug 2009 09:12:17 +0000 (11:12 +0200)]
s3:winbind: Fix bug 5626

Apparently the AIX compiler can't deal with sizeless array declarations

12 years agow3:winbind: Convert WINBINDD_ALLOCATE_GID to the new API
Volker Lendecke [Sun, 30 Aug 2009 07:46:34 +0000 (09:46 +0200)]
w3:winbind: Convert WINBINDD_ALLOCATE_GID to the new API

12 years agow3:winbind: Convert WINBINDD_ALLOCATE_UID to the new API
Volker Lendecke [Sun, 30 Aug 2009 07:41:43 +0000 (09:41 +0200)]
w3:winbind: Convert WINBINDD_ALLOCATE_UID to the new API

12 years agoOne would expect I could spell my name...
Michael Adam [Sat, 29 Aug 2009 22:55:28 +0000 (00:55 +0200)]
One would expect I could spell my name...

12 years agoOne would expect I could spell my name.
Michael Adam [Sat, 29 Aug 2009 22:54:35 +0000 (00:54 +0200)]
One would expect I could spell my name.

12 years agotdb: move the original abi_checks script to script/abi_checks_gcc.sh
Michael Adam [Fri, 28 Aug 2009 14:04:47 +0000 (16:04 +0200)]
tdb: move the original abi_checks script to script/abi_checks_gcc.sh

Michael

12 years agotdb: run abi_checks as part of "make test".
Michael Adam [Sat, 29 Aug 2009 22:40:18 +0000 (00:40 +0200)]
tdb: run abi_checks as part of "make test".

Strange: I had to place "test:: abi_checks" before the main
"test::" target here, otherwise the abi checks would not get run.

Michael

12 years agotdb: remove files generated by "make abi_checks" in "make clean".
Michael Adam [Sat, 29 Aug 2009 22:28:31 +0000 (00:28 +0200)]
tdb: remove files generated by "make abi_checks" in "make clean".

Michael

12 years agotdb: add a target "abi_checks" to tdb.
Michael Adam [Fri, 28 Aug 2009 14:03:06 +0000 (16:03 +0200)]
tdb: add a target "abi_checks" to tdb.

Simply run "make abi_checks" to call the abi check script appropriately.

Michael

12 years agotdb: add script/abi_checks.sh. check for abi changes without gcc magic.
Michael Adam [Fri, 28 Aug 2009 13:53:12 +0000 (15:53 +0200)]
tdb: add script/abi_checks.sh. check for abi changes without gcc magic.

USAGE: abi_checks.sh LIBRARY_NAME header1 [header2 ...]

This creates symbol signature lists using the mksyms and mksigs scripts
and compares them with the checked in lists.

Michael

12 years agotdb: add script to extract signatures from header files.
Michael Adam [Fri, 28 Aug 2009 13:08:19 +0000 (15:08 +0200)]
tdb: add script to extract signatures from header files.

This produces output like the output gcc produces when
invoked with the -aux-info switch.

Run like this: cat include/tdb.h | ./script/mksigs.pl

This simple parser is probably too coarse to handle all
possible header files, but it treats tdb.h correctly...

Michael

12 years agotdb: add scripts to extract library symbols (exports file) from headers
Michael Adam [Fri, 28 Aug 2009 13:01:17 +0000 (15:01 +0200)]
tdb: add scripts to extract library symbols (exports file) from headers

Michael

12 years agotalloc: move the original abi_checks script to script/abi_checks_gcc.sh
Michael Adam [Fri, 28 Aug 2009 14:04:47 +0000 (16:04 +0200)]
talloc: move the original abi_checks script to script/abi_checks_gcc.sh

Michael

12 years agotalloc: remove talloc_free from the exports
Michael Adam [Sat, 29 Aug 2009 22:06:58 +0000 (00:06 +0200)]
talloc: remove talloc_free from the exports

Michael

12 years agotalloc: run abi_checks as part of "make test"
Michael Adam [Sat, 29 Aug 2009 22:16:36 +0000 (00:16 +0200)]
talloc: run abi_checks as part of "make test"

Michael

12 years agotalloc: remove files generated by "make abi_check" in "make clean"
Michael Adam [Sat, 29 Aug 2009 22:25:24 +0000 (00:25 +0200)]
talloc: remove files generated by "make abi_check" in "make clean"

Michael

12 years agotalloc: add a target "abi_checks" to talloc.
Michael Adam [Fri, 28 Aug 2009 14:03:06 +0000 (16:03 +0200)]
talloc: add a target "abi_checks" to talloc.

Simply run "make abi_checks" to call the abi check script appropriately.

Michael

12 years agotalloc: add script/abi_checks.sh. check for abi changes without gcc magic.
Michael Adam [Fri, 28 Aug 2009 13:53:12 +0000 (15:53 +0200)]
talloc: add script/abi_checks.sh. check for abi changes without gcc magic.

USAGE: abi_checks.sh LIBRARY_NAME header1 [header2 ...]

This creates symbol signature lists using the mksyms and mksigs scripts
and compares them with the checked in lists.

Michael

12 years agotalloc: add script to extract signatures from header files.
Michael Adam [Fri, 28 Aug 2009 13:08:19 +0000 (15:08 +0200)]
talloc: add script to extract signatures from header files.

This produces output like the output gcc produces when
invoked with the -aux-info switch.

Run like this: cat talloc.h | ./script/mksigs.pl

This simple parser is probably too coarse to handle all possible
header files, but it does treat talloc.h correctly.

Michael

12 years agotalloc: add scripts to extract library symbols (exports file) from headers
Michael Adam [Fri, 28 Aug 2009 13:01:17 +0000 (15:01 +0200)]
talloc: add scripts to extract library symbols (exports file) from headers

Michael

12 years agos3:winbind: Convert WINBINDD_LIST_GROUPS to the new API
Volker Lendecke [Sat, 29 Aug 2009 15:17:47 +0000 (17:17 +0200)]
s3:winbind: Convert WINBINDD_LIST_GROUPS to the new API

12 years agos3:winbind: Convert WINBINDD_LIST_USERS to the new API
Volker Lendecke [Sat, 29 Aug 2009 14:05:02 +0000 (16:05 +0200)]
s3:winbind: Convert WINBINDD_LIST_USERS to the new API

12 years agos3:nsstest: Fix a very confusing behaviour in nsstest
Volker Lendecke [Sat, 29 Aug 2009 17:39:35 +0000 (19:39 +0200)]
s3:nsstest: Fix a very confusing behaviour in nsstest

Testing getgrent I thought I get the offset calculations wrong whereas
it was only nsstest printing stuff with spaces...

12 years agos3:winbind: add a fake NDR winbind cache validator
Volker Lendecke [Sat, 29 Aug 2009 17:28:42 +0000 (19:28 +0200)]
s3:winbind: add a fake NDR winbind cache validator

12 years agos3:winbind: Convert the getgrent functions to the new API
Volker Lendecke [Thu, 27 Aug 2009 21:33:45 +0000 (23:33 +0200)]
s3:winbind: Convert the getgrent functions to the new API

12 years agos3:winbind: Add a comment about fill_grent_mem_domusers...
Volker Lendecke [Sat, 29 Aug 2009 13:21:29 +0000 (15:21 +0200)]
s3:winbind: Add a comment about fill_grent_mem_domusers...

12 years agos3:winbind: Add some missing error checks
Volker Lendecke [Thu, 27 Aug 2009 21:27:37 +0000 (23:27 +0200)]
s3:winbind: Add some missing error checks

12 years agos3:winbind: Fix a typo
Volker Lendecke [Thu, 27 Aug 2009 19:54:58 +0000 (21:54 +0200)]
s3:winbind: Fix a typo

12 years agos3:winbind: Add async wb_next_grent
Volker Lendecke [Thu, 27 Aug 2009 19:53:15 +0000 (21:53 +0200)]
s3:winbind: Add async wb_next_grent

12 years agos3:winbind: Add async QueryGroupList
Volker Lendecke [Thu, 27 Aug 2009 17:54:18 +0000 (19:54 +0200)]
s3:winbind: Add async QueryGroupList

12 years agow3:winbind: Convert WINBINDD_LOOKUPRIDS to the new API
Volker Lendecke [Thu, 27 Aug 2009 15:11:24 +0000 (17:11 +0200)]
w3:winbind: Convert WINBINDD_LOOKUPRIDS to the new API

12 years agos3:winbind: Convert the GETDCNAME routines to the new API
Volker Lendecke [Wed, 26 Aug 2009 10:27:32 +0000 (12:27 +0200)]
s3:winbind: Convert the GETDCNAME routines to the new API