ira/wip.git
8 years agomidltests: add midltests-pipe-sync-ndr32-downgrade-01.idl example
Stefan Metzmacher [Wed, 29 Sep 2010 00:36:51 +0000 (02:36 +0200)]
midltests: add midltests-pipe-sync-ndr32-downgrade-01.idl example

metze

8 years agomidltests: add some usefull defines to midltests.idl
Stefan Metzmacher [Wed, 29 Sep 2010 00:50:19 +0000 (02:50 +0200)]
midltests: add some usefull defines to midltests.idl

metze

8 years agomidltests: make it possible to allow downgrades to NDR32
Stefan Metzmacher [Wed, 29 Sep 2010 00:35:54 +0000 (02:35 +0200)]
midltests: make it possible to allow downgrades to NDR32

metze

8 years agomidltests: add a midltests_tcp.exe tool
Stefan Metzmacher [Tue, 28 Sep 2010 09:04:59 +0000 (11:04 +0200)]
midltests: add a midltests_tcp.exe tool

This uses a man in the middle approach in order to dump the
request and response pdus.

It also tests NDR32 and NDR64.

metze

8 years agomidltests: move the current implementation to midltests_simple.exe
Stefan Metzmacher [Tue, 28 Sep 2010 08:50:05 +0000 (10:50 +0200)]
midltests: move the current implementation to midltests_simple.exe

metze

8 years agotestprogs/win32: add vs2010-metze.cmd
Stefan Metzmacher [Tue, 28 Sep 2010 07:47:55 +0000 (09:47 +0200)]
testprogs/win32: add vs2010-metze.cmd

metze

8 years agos3-printing: skip metadata entry when traversing printerlist.
Günther Deschner [Tue, 28 Sep 2010 23:18:07 +0000 (01:18 +0200)]
s3-printing: skip metadata entry when traversing printerlist.

We were creating a new printer (with a very broken name) out of the
lasttimestamp entry all the time.

Simo, please check.

Guenther

8 years agopidl: add support for pointers in typedefs
Stefan Metzmacher [Wed, 5 Aug 2009 11:43:49 +0000 (13:43 +0200)]
pidl: add support for pointers in typedefs

metze

8 years agopidl:NDR/Parser: remove unused code for array element index
Stefan Metzmacher [Tue, 21 Sep 2010 08:34:30 +0000 (10:34 +0200)]
pidl:NDR/Parser: remove unused code for array element index

metze

8 years agopidl:NDR/Parser: simplify logic in ParseMemCtxPullFlags()
Stefan Metzmacher [Tue, 21 Sep 2010 01:48:09 +0000 (03:48 +0200)]
pidl:NDR/Parser: simplify logic in ParseMemCtxPullFlags()

metze

8 years agopidl:NDR/Client: make the generated code look a bit nicer
Stefan Metzmacher [Tue, 21 Sep 2010 01:41:03 +0000 (03:41 +0200)]
pidl:NDR/Client: make the generated code look a bit nicer

metze

8 years agolibrpc/ndr: remove 'async' from ndr_interface_call
Stefan Metzmacher [Mon, 20 Sep 2010 22:44:30 +0000 (00:44 +0200)]
librpc/ndr: remove 'async' from ndr_interface_call

metze

8 years agopidl: remove unused async property handling
Stefan Metzmacher [Mon, 20 Sep 2010 22:41:29 +0000 (00:41 +0200)]
pidl: remove unused async property handling

metze

8 years agopidl/Python: use has_property($d, "noopnum") helper function
Stefan Metzmacher [Tue, 21 Sep 2010 01:10:10 +0000 (03:10 +0200)]
pidl/Python: use has_property($d, "noopnum") helper function

metze

8 years agopidl:NDR/Client.pm: remove unreached code
Stefan Metzmacher [Tue, 21 Sep 2010 01:05:41 +0000 (03:05 +0200)]
pidl:NDR/Client.pm: remove unreached code

metze

8 years agopidl/Python: remove todo handling from PythonFunction(), it's done by the caller
Stefan Metzmacher [Tue, 21 Sep 2010 00:17:21 +0000 (02:17 +0200)]
pidl/Python: remove todo handling from PythonFunction(), it's done by the caller

metze

8 years agopidl/Typelist: let typeIs() do TYPEDEF dereference in the HASH case
Stefan Metzmacher [Mon, 20 Sep 2010 23:40:56 +0000 (01:40 +0200)]
pidl/Typelist: let typeIs() do TYPEDEF dereference in the HASH case

metze

8 years agos3-waf: add in a little hack to deal with the ECHO rpc module for non-developer builds.
Günther Deschner [Tue, 28 Sep 2010 20:53:08 +0000 (22:53 +0200)]
s3-waf: add in a little hack to deal with the ECHO rpc module for non-developer builds.

This will be removed once we have the rpc modules subsystem in place.

Guenther

8 years agoautobuild: use git notes for autobuild messages
Andrew Tridgell [Tue, 28 Sep 2010 18:24:37 +0000 (11:24 -0700)]
autobuild: use git notes for autobuild messages

This avoids changing the commit ID when we add a note that the
autobuild has passed

thanks to Jelmer for this suggestion!

8 years agoselftest: enable FAIL_IMMEDIATELY in autobuild make test
Andrew Tridgell [Tue, 28 Sep 2010 18:23:35 +0000 (11:23 -0700)]
selftest: enable FAIL_IMMEDIATELY in autobuild make test

this should reduce the time we wait for previous failing builds.

Right now this will only work for s4, as we need a makefile change for
s3 support

8 years agos4-drs: added support for DRSUAPI_EXOP_REPL_OBJ
Andrew Tridgell [Tue, 28 Sep 2010 17:48:38 +0000 (10:48 -0700)]
s4-drs: added support for DRSUAPI_EXOP_REPL_OBJ

this extended getncchanges operation replicates a single object

8 years agoldb-tdb: ignore failure to register control on rootdse
Andrew Tridgell [Tue, 28 Sep 2010 17:46:03 +0000 (10:46 -0700)]
ldb-tdb: ignore failure to register control on rootdse

this is expected for non-sam LDBs

8 years agos4-drs: use drs_ObjectIdentifier_*() calls in getncchanges
Andrew Tridgell [Tue, 28 Sep 2010 17:40:18 +0000 (10:40 -0700)]
s4-drs: use drs_ObjectIdentifier_*() calls in getncchanges

this allows for replication by GUID or SID

8 years agos4-drs: moved the drs_ObjectIdentifier handling to dsdb_dn.c
Andrew Tridgell [Tue, 28 Sep 2010 17:39:52 +0000 (10:39 -0700)]
s4-drs: moved the drs_ObjectIdentifier handling to dsdb_dn.c

this will be used outside of the drs server.

This also fixes the handling of the ndr_size elements of the
drs_ObjectIdentifier

8 years agowaf: we don't need the preprocessor recursion limit any more
Andrew Tridgell [Tue, 28 Sep 2010 17:38:40 +0000 (10:38 -0700)]
waf: we don't need the preprocessor recursion limit any more

thanks to ita for this

8 years agos4-drs: Added check for drs-manage-topology to updateRefs.
Nadezhda Ivanova [Mon, 27 Sep 2010 04:16:47 +0000 (21:16 -0700)]
s4-drs: Added check for drs-manage-topology to updateRefs.

8 years agos4-drs: Added drs_security_access_check function
Nadezhda Ivanova [Mon, 27 Sep 2010 04:14:45 +0000 (21:14 -0700)]
s4-drs: Added drs_security_access_check function

It takes a security token, an ldb_context, and the desired CAR and checks
if the principal has this CAR granted

8 years agos4-dsdb: adapted check_access_on_dn for use in drs.
Nadezhda Ivanova [Mon, 27 Sep 2010 04:12:48 +0000 (21:12 -0700)]
s4-dsdb: adapted check_access_on_dn for use in drs.

8 years agoheimdal Fix DNS name qualification to not mangle IP addresses
Andrew Bartlett [Tue, 28 Sep 2010 17:59:15 +0000 (03:59 +1000)]
heimdal Fix DNS name qualification to not mangle IP addresses

If the host running this code used IPv6 forms for IPv4 addreses
then the check for '.' would not be sufficient to determine that this
isn't a name we should mangle.  Instead, check if it can be parsed
as a numeric address first, and only then mangle.

Andrew Bartlett

8 years agos4-kdc Handle the case where we may be given a ticket from an RODC in db layer
Andrew Bartlett [Tue, 28 Sep 2010 03:13:28 +0000 (13:13 +1000)]
s4-kdc Handle the case where we may be given a ticket from an RODC in db layer

This includes rewriting the PAC if the original krbtgt isn't to be
trusted, and reading different entries from the DB for the krbtgt
depending on the krbtgt number.

Andrew Bartlett

8 years agoheimdal Add an error code for use in the RODC
Andrew Bartlett [Tue, 28 Sep 2010 03:10:24 +0000 (13:10 +1000)]
heimdal Add an error code for use in the RODC

In this case, the whole request packet should be forwarded to
a real KDC, with full secrets, as we don't have the password.

This could also be used to implement 'play dead when the LDAP
server is down'.

Andrew Bartlett

8 years agoheimdal Add support for extracting a particular KVNO from the database
Andrew Bartlett [Tue, 28 Sep 2010 03:07:53 +0000 (13:07 +1000)]
heimdal Add support for extracting a particular KVNO from the database

This should allow master key rollover.

(but the real reason is to allow multiple krbtgt accounts, as used by
Active Directory to implement RODC support)

Andrew Bartlett

8 years agos4-kdc Add common setup, handle RODC setup case
Andrew Bartlett [Tue, 28 Sep 2010 03:05:37 +0000 (13:05 +1000)]
s4-kdc Add common setup, handle RODC setup case

This means we just set up the system_session etc in one place
and don't diverge between the MIT and Heimdal plugins.

We also now determine if we are an RODC and store some details
that we will need later.

Andrew Bartlett

8 years agos4-dsdb Add ldb_reset_err_string() when we set error codes.
Andrew Bartlett [Tue, 28 Sep 2010 02:57:15 +0000 (12:57 +1000)]
s4-dsdb Add ldb_reset_err_string() when we set error codes.

If we don't we could show an old, incrorrect error

8 years agos4-dsdb Make samdb_reference_dn() use dsdb_search() and DSDB_SEARCH_ONE_ONLY
Andrew Bartlett [Tue, 28 Sep 2010 02:55:48 +0000 (12:55 +1000)]
s4-dsdb Make samdb_reference_dn() use dsdb_search() and DSDB_SEARCH_ONE_ONLY

This simplifies the function.  While doing so, also change the error
string setting to set a really clear error string for the failure to find
and failure to parse cases.

Andrew Bartlett

8 years agos4-kdc Add function to determine if a hdb entry is a RODC
Andrew Bartlett [Tue, 28 Sep 2010 02:53:06 +0000 (12:53 +1000)]
s4-kdc Add function to determine if a hdb entry is a RODC

This is important, as we must ignore the PAC from an RODC.

Andrew Bartlett

8 years agos4-kdc Use msDS-SecondaryKrbTgtNumber to fill in the full KVNO
Andrew Bartlett [Tue, 28 Sep 2010 02:49:44 +0000 (12:49 +1000)]
s4-kdc Use msDS-SecondaryKrbTgtNumber to fill in the full KVNO

Andrew Bartlett

8 years agos4-dsdb Fix segfault in error case in rootdse module
Andrew Bartlett [Mon, 27 Sep 2010 04:43:33 +0000 (14:43 +1000)]
s4-dsdb Fix segfault in error case in rootdse module

8 years agoMake upgrade procedure more explicit.
Michael Wood [Sun, 26 Sep 2010 22:05:05 +0000 (00:05 +0200)]
Make upgrade procedure more explicit.

Add in a compile step.
Change the tar command to include the recommended dirs.

8 years agos3-waf: add AUTH_SCRIPT module to AUTH subsystem (which is build as shared
Günther Deschner [Tue, 28 Sep 2010 18:00:47 +0000 (20:00 +0200)]
s3-waf: add AUTH_SCRIPT module to AUTH subsystem (which is build as shared
module by default).

Guenther

8 years agos3-waf: add vfs_linux_xfs_sgid to the list of default shared modules.
Günther Deschner [Tue, 28 Sep 2010 18:00:12 +0000 (20:00 +0200)]
s3-waf: add vfs_linux_xfs_sgid to the list of default shared modules.

Guenther

8 years agos3: Attempt to fix bug 7518
Volker Lendecke [Tue, 28 Sep 2010 16:40:49 +0000 (18:40 +0200)]
s3: Attempt to fix bug 7518

If select returns -1, we can't rely on the fd sets. The current code might loop
endlessly because when putting an invalid fd (the closed socket?) on the read
set, a select implementation might choose not to touch it but directly return
with EINVAL. Thus run_events will see the socket readable, which leads to a
"return true", and thus a NT_STATUS_RETRY -> same game again.

We should never get into this situation, but to me the logfiles given in bug
7518 do not reveal enough information to understand how this can happen.

8 years agos3: Increase the debuglevel for connection termination msgs
Volker Lendecke [Tue, 28 Sep 2010 08:38:20 +0000 (10:38 +0200)]
s3: Increase the debuglevel for connection termination msgs

8 years agos3-waf: fix dependencies in most of our module subsystems.
Günther Deschner [Tue, 28 Sep 2010 07:27:54 +0000 (09:27 +0200)]
s3-waf: fix dependencies in most of our module subsystems.

Guenther

8 years agos3-waf: add pam_smbpass.
Günther Deschner [Tue, 28 Sep 2010 06:54:39 +0000 (08:54 +0200)]
s3-waf: add pam_smbpass.

Guenther

8 years agos3-auth_util: make sure the system server info actually contains S-1-5-18.
Günther Deschner [Tue, 28 Sep 2010 05:45:47 +0000 (07:45 +0200)]
s3-auth_util: make sure the system server info actually contains S-1-5-18.

Without this, all security descriptor checks for the winreg spoolss backend fail
and make our spoolss system in its current shape basically unusable.

Andreas, please check.

Guenther

8 years agos3-printing: remove unused old structs.
Günther Deschner [Tue, 28 Sep 2010 03:39:42 +0000 (05:39 +0200)]
s3-printing: remove unused old structs.

Guenther

8 years agoldb: Fix path to alternative buildtools.
Jelmer Vernooij [Tue, 28 Sep 2010 07:16:03 +0000 (09:16 +0200)]
ldb: Fix path to alternative buildtools.

8 years agosamba4: Don't update Makefile/configure from autogen.sh.
Jelmer Vernooij [Tue, 28 Sep 2010 07:08:10 +0000 (09:08 +0200)]
samba4: Don't update Makefile/configure from autogen.sh.

8 years agoldb/tevent: Fix detection of waf paths.
Jelmer Vernooij [Tue, 28 Sep 2010 07:00:26 +0000 (09:00 +0200)]
ldb/tevent: Fix detection of waf paths.

8 years agotevent: Remove make targets that are not relevant for tevent.
Jelmer Vernooij [Tue, 28 Sep 2010 06:42:26 +0000 (08:42 +0200)]
tevent: Remove make targets that are not relevant for tevent.

8 years agotevent: Don't update Makefile/configure files.
Jelmer Vernooij [Tue, 28 Sep 2010 06:41:11 +0000 (08:41 +0200)]
tevent: Don't update Makefile/configure files.

8 years agoldb: Update autogen-waf.sh to no longer overwrite existing files.
Jelmer Vernooij [Tue, 28 Sep 2010 06:38:33 +0000 (08:38 +0200)]
ldb: Update autogen-waf.sh to no longer overwrite existing files.

8 years agoldb: Remove samba-specific targets from Makefile.
Jelmer Vernooij [Tue, 28 Sep 2010 06:34:10 +0000 (08:34 +0200)]
ldb: Remove samba-specific targets from Makefile.

8 years agoldb: Bump version because of addition of ldb_req_location.
Jelmer Vernooij [Tue, 28 Sep 2010 06:15:55 +0000 (08:15 +0200)]
ldb: Bump version because of addition of ldb_req_location.

8 years agoformat-subunit: Display number of failed tests even if there are no
Jelmer Vernooij [Tue, 28 Sep 2010 05:53:02 +0000 (07:53 +0200)]
format-subunit: Display number of failed tests even if there are no
failed testsuites.

8 years agoselftest: Abort early on SIGPIPE.
Jelmer Vernooij [Tue, 28 Sep 2010 05:40:27 +0000 (07:40 +0200)]
selftest: Abort early on SIGPIPE.

8 years agoAdd dedicated exception for immediate failure in filter-subunit, don't raise it on...
Jelmer Vernooij [Tue, 28 Sep 2010 05:10:43 +0000 (07:10 +0200)]
Add dedicated exception for immediate failure in filter-subunit, don't raise it on known exceptions.

8 years agos4-selftest: added a --fail-immediately option to s4 test
Andrew Tridgell [Mon, 27 Sep 2010 01:58:05 +0000 (18:58 -0700)]
s4-selftest: added a --fail-immediately option to s4 test

this can be used to force an immediate test failure on the first
failed test case. You can also use:

  make test FAIL_IMMEDIATELY=1

8 years agofilter-subunit: added a --fail-immediately option
Andrew Tridgell [Mon, 27 Sep 2010 01:56:50 +0000 (18:56 -0700)]
filter-subunit: added a --fail-immediately option

8 years agos4-provision: fixed the authority response for our SOA record
Andrew Tridgell [Tue, 28 Sep 2010 05:53:06 +0000 (22:53 -0700)]
s4-provision: fixed the authority response for our SOA record

some clients rely on this being the hostname, not the domain

Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Tue Sep 28 06:39:19 UTC 2010 on sn-devel-104

8 years agos4-dns: implemented RODC DNS update in dns update task
Andrew Tridgell [Tue, 28 Sep 2010 04:08:43 +0000 (21:08 -0700)]
s4-dns: implemented RODC DNS update in dns update task

Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>

8 years agos4-netlogon: added RODC DNS update call fwded to dnsupdate task
Andrew Tridgell [Tue, 28 Sep 2010 04:08:23 +0000 (21:08 -0700)]
s4-netlogon: added RODC DNS update call fwded to dnsupdate task

when we get a netlogon RODC DNS update, we send it to the dnsupdate
task

8 years agos4-dns: added --update-list option to samba_dnsupdate
Andrew Tridgell [Tue, 28 Sep 2010 04:07:17 +0000 (21:07 -0700)]
s4-dns: added --update-list option to samba_dnsupdate

this allows us to use it for RODC netlogon updates

8 years agopidl: added ifdef guards around ndr headers
Andrew Tridgell [Tue, 28 Sep 2010 04:03:45 +0000 (21:03 -0700)]
pidl: added ifdef guards around ndr headers

this prevents us parsing the leading headers needlessly

8 years agos4-kdc: added ifdef guards in kdc.h
Andrew Tridgell [Tue, 28 Sep 2010 04:03:14 +0000 (21:03 -0700)]
s4-kdc: added ifdef guards in kdc.h

this prevents too much recursion in the compiler preprocessor

8 years agos4-ldb: removed an unused variable
Andrew Tridgell [Tue, 28 Sep 2010 02:47:14 +0000 (19:47 -0700)]
s4-ldb: removed an unused variable

8 years agos4-kcc: fixed a incorrect context to kcctpl_get_all_bridgehead_dcs
Andrew Tridgell [Tue, 28 Sep 2010 02:47:01 +0000 (19:47 -0700)]
s4-kcc: fixed a incorrect context to kcctpl_get_all_bridgehead_dcs

8 years agos4-dsdb: added samdb_find_site_for_computer() and samdb_find_ntdsguid_for_computer()
Andrew Tridgell [Tue, 28 Sep 2010 02:46:37 +0000 (19:46 -0700)]
s4-dsdb: added samdb_find_site_for_computer() and samdb_find_ntdsguid_for_computer()

these will be used by the new RODC dns update code

Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>

8 years agos4-auth: removed unused variable dom_sid
Andrew Tridgell [Tue, 28 Sep 2010 02:45:56 +0000 (19:45 -0700)]
s4-auth: removed unused variable dom_sid

8 years agos3: Lift smbd_messaging_context() from open_sockets_smbd()
Volker Lendecke [Tue, 31 Aug 2010 21:17:11 +0000 (23:17 +0200)]
s3: Lift smbd_messaging_context() from open_sockets_smbd()

8 years agos3: Remove two calls to procid_self()
Volker Lendecke [Tue, 31 Aug 2010 21:16:50 +0000 (23:16 +0200)]
s3: Remove two calls to procid_self()

8 years agos3: Remove talloc_autofree_context() from myhostname()
Volker Lendecke [Mon, 27 Sep 2010 05:05:43 +0000 (22:05 -0700)]
s3: Remove talloc_autofree_context() from myhostname()

No destructor needed, so we can as well use the NULL context

8 years agos3: Remove some remaining files.c globals to sconn
Volker Lendecke [Mon, 27 Sep 2010 04:06:02 +0000 (06:06 +0200)]
s3: Remove some remaining files.c globals to sconn

8 years agos3: Remove talloc_autofree_context() from files.c
Volker Lendecke [Mon, 27 Sep 2010 03:50:22 +0000 (05:50 +0200)]
s3: Remove talloc_autofree_context() from files.c

8 years agos3: Lift smbd_server_conn from file_find_fd
Volker Lendecke [Mon, 27 Sep 2010 02:54:29 +0000 (04:54 +0200)]
s3: Lift smbd_server_conn from file_find_fd

8 years agos3: Remove smbd_server_conn from file_fsp
Volker Lendecke [Mon, 27 Sep 2010 02:49:00 +0000 (04:49 +0200)]
s3: Remove smbd_server_conn from file_fsp

8 years agos3: Slightly simplify file_fnum
Volker Lendecke [Mon, 27 Sep 2010 02:46:18 +0000 (04:46 +0200)]
s3: Slightly simplify file_fnum

req==NULL should never happen, see the comment

8 years agos3: Remove smbd_server_conn from file_sync_all
Volker Lendecke [Mon, 27 Sep 2010 02:13:22 +0000 (04:13 +0200)]
s3: Remove smbd_server_conn from file_sync_all

8 years agos3: Remove smbd_server_conn from file_find_subpath
Volker Lendecke [Mon, 27 Sep 2010 02:12:15 +0000 (04:12 +0200)]
s3: Remove smbd_server_conn from file_find_subpath

8 years agos3: Lift smbd_server_conn from file_find_di_first
Volker Lendecke [Mon, 27 Sep 2010 02:05:25 +0000 (04:05 +0200)]
s3: Lift smbd_server_conn from file_find_di_first

8 years agos3: Lift smbd_server_conn from file_find_dif
Volker Lendecke [Mon, 27 Sep 2010 02:00:31 +0000 (04:00 +0200)]
s3: Lift smbd_server_conn from file_find_dif

8 years agos3: Remove smbd_server_conn from files_forall
Volker Lendecke [Mon, 27 Sep 2010 01:53:00 +0000 (03:53 +0200)]
s3: Remove smbd_server_conn from files_forall

8 years agos3: Remove smbd_server_conn from file_close_user
Volker Lendecke [Mon, 27 Sep 2010 01:46:12 +0000 (03:46 +0200)]
s3: Remove smbd_server_conn from file_close_user

8 years agos3: Remove smbd_server_conn from file_close_pid
Volker Lendecke [Mon, 27 Sep 2010 01:42:36 +0000 (03:42 +0200)]
s3: Remove smbd_server_conn from file_close_pid

8 years agos3: Remove smbd_server_conn from file_close_conn
Volker Lendecke [Mon, 27 Sep 2010 01:40:11 +0000 (03:40 +0200)]
s3: Remove smbd_server_conn from file_close_conn

8 years agos3: Remove smbd_server_conn from file_new and file_free
Volker Lendecke [Mon, 27 Sep 2010 01:38:14 +0000 (03:38 +0200)]
s3: Remove smbd_server_conn from file_new and file_free

8 years agos3: Lift smbd_server_conn from file_fnum
Volker Lendecke [Mon, 27 Sep 2010 00:37:59 +0000 (02:37 +0200)]
s3: Lift smbd_server_conn from file_fnum

8 years agos3: Move "Files" to smbd_server_connection
Volker Lendecke [Mon, 27 Sep 2010 00:29:36 +0000 (02:29 +0200)]
s3: Move "Files" to smbd_server_connection

8 years agos3: Remove "server_fd" global variable
Volker Lendecke [Mon, 27 Sep 2010 00:15:18 +0000 (02:15 +0200)]
s3: Remove "server_fd" global variable

8 years agos3: Remove talloc_autofree_context() from receive_unexpected()
Volker Lendecke [Sun, 26 Sep 2010 23:44:06 +0000 (01:44 +0200)]
s3: Remove talloc_autofree_context() from receive_unexpected()

This is freed in this routine a few lines down

8 years agos4:gensec_tstream: remove plain socket handling
Stefan Metzmacher [Tue, 28 Sep 2010 02:50:32 +0000 (04:50 +0200)]
s4:gensec_tstream: remove plain socket handling

metze

Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue Sep 28 04:54:24 UTC 2010 on sn-devel-104

8 years agos4:lib/tls: add gnutls backend for tstream
Stefan Metzmacher [Wed, 3 Feb 2010 13:36:10 +0000 (14:36 +0100)]
s4:lib/tls: add gnutls backend for tstream

metze

Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue Sep 28 02:29:42 UTC 2010 on sn-devel-104

8 years agos4:gensec: add gensec_create_tstream()
Stefan Metzmacher [Wed, 22 Sep 2010 10:13:28 +0000 (12:13 +0200)]
s4:gensec: add gensec_create_tstream()

Based on the initial patch from Andreas Schneider <asn@redhat.com>.

metze

8 years agos4:wrepl_server: use SOCKET_FLAG_NOCLOSE instead of a dup()
Stefan Metzmacher [Tue, 28 Sep 2010 00:33:23 +0000 (02:33 +0200)]
s4:wrepl_server: use SOCKET_FLAG_NOCLOSE instead of a dup()

The key thing is that we might have to turn the incomming
connection into a outgoing connection.

This change makes sense anyway, because we donate the fd to
tstream.

metze

8 years agos4:rpc_server: use SOCKET_FLAG_NOCLOSE to avoid calling close() on the socket fd...
Stefan Metzmacher [Tue, 28 Sep 2010 00:13:12 +0000 (02:13 +0200)]
s4:rpc_server: use SOCKET_FLAG_NOCLOSE to avoid calling close() on the socket fd twice.

metze

8 years agoAdd torture test BAD-NBT-SESSION as regression fix for bug 7698 - Assert causes smbd...
Jeremy Allison [Tue, 28 Sep 2010 01:24:01 +0000 (18:24 -0700)]
Add torture test BAD-NBT-SESSION as regression fix for bug 7698 - Assert causes smbd to panic on invalid NetBIOS session request.

Jeremy.

8 years agoSmall tweak to bugfix for 7698 - Assert causes smbd to panic on invalid NetBIOS sessi...
Jeremy Allison [Tue, 28 Sep 2010 01:21:46 +0000 (18:21 -0700)]
Small tweak to bugfix for 7698 - Assert causes smbd to panic on invalid NetBIOS session request.

Don't just fail to reply on a bad NBT name, just don't do the
internal action.

Jeremy.

8 years agotsocket: make sure we delete the fd event before calling close()
Stefan Metzmacher [Mon, 27 Sep 2010 21:57:34 +0000 (23:57 +0200)]
tsocket: make sure we delete the fd event before calling close()

We got random double free errors, when getting events from
epoll_wait() and try to dereference the private talloc pointer
attached to it.

Before doing the close() in the tstream_disconnect_send() function
we need to delete the fd event.

metze

Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue Sep 28 01:02:55 UTC 2010 on sn-devel-104

8 years agoChange to using TDB_INCOMPATIBLE_HASH (the jenkins hash) on all
Jeremy Allison [Mon, 27 Sep 2010 12:46:07 +0000 (05:46 -0700)]
Change to using TDB_INCOMPATIBLE_HASH (the jenkins hash) on all
TDB_CLEAR_IF_FIRST tdb's. For tdb's like gencache where we open
without CLEAR_IF_FIRST and then with CLEAR_IF_FIRST if corrupt
this is still safe to use as if opening an existing tdb the new
hash will be ignored - it's only used on creating a new tdb not
opening an old one.

Jeremy.