s3: VFS: vfs_fruit: Convert fruit_rmdir_internal() to UNLINKAT().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit: Add dirfsp files_struct pointer parameter to fruit_rmdir_internal().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit: Wrap fruit_rmdir() so we can add parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit. Change fruit_unlink_internal() to call UNLNKAT.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit. Add files_struct *dirfsp to fruit_unlink_internal() so we can call UNLINKAT.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit. Wrap Convert fruit_unlink() -> fruit_unlink_internal() so we can add parameters.

Make fruit_unlink() a wrapper calling the internal version.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change rmdir_internals() to call SMB_VFS_UNLINKAT() in 2 places.

Use conn->cwd_fsp as current fsp. Rework to modern standards.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change recursive_rmdir() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp. Rework to modern standards.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change close_remove_share_mode() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp. Rework to modern standards.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change delete_all_streams() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change smb_set_file_unix_basic() to call SMB_VFS_UNLINKAT() in 3 places.

Use conn->cwd_fsp as current fsp. This is horrible code and must
be removed.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: pysmbd: Change py_smbd_unlink() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change remove_msdfs_link() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp. Rework to modern standards.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: smbd: Change create_msdfs_link() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp. Rework to modern standards.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: cmd_vfs: Change cmd_pathfunc() to call SMB_VFS_UNLINKAT().

Use conn->cwd_fsp as current fsp.

No logic change for now.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_acl_common: Convert unlink_acl_common(), acl_common_remove_object() to UNLINKAT.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_acl_common: Add dirfsp, flags parameters to unlink_acl_common().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_acl_xattr: Don't call unlink_acl_common() directly.

Use a wrapper function so we can change the parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_virusfilter: Change UNLINK -> UNLINKAT in infected_file_action_delete()

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_posix_eadb: Wrap posix_eadb_unlink() into a call to posix_eadb_unlink_internal().

Allows UNLINKAT parameters to be passed. Change posix_eadb_unlink_internal()
to call UNLINKAT instead of UNLINK.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_audit. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_virusfilter. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_unityed_media. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_time_audit. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_syncops. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_xattr: Convert streams_xattr_unlink_internal() to use UNLINKAT.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_xattr. Wrap Convert streams_xattr_unlink() -> streams_xattr_unlink_internal() so we can add parameters.

Make streams_xattr_unlink() a wrapper calling the internal version.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_xattr. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_depot: Change streams_depot_unlink_internal() to call UNLINKAT().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_depot. Wrap Convert streams_depot_unlink() -> streams_depot_unlink_internal() so we can add parameters.

Make streams_depot_unlink() a wrapper calling the internal version.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_streams_depot. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_snapper. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_shadow_copy2. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_recycle: Change recycle_unlink_internal() to call UNLINKAT().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_recycle: Wrap recycle_unlink() into a call to recycle_unlink_internal().

Allows UNLINKAT parameters to be passed.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_recycle. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_media_harmony. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_glusterfs. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_full_audit. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_fruit. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_extd_audit. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_ceph_snapshots. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_ceph. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_catia. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_cap. Implement unlinkat().

This is identical to unlink(), as there
are no special cases needed for rmdir().

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_xattr_tdb. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_posix_eadb. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_acl_xattr. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: vfs_acl_tdb. Implement unlinkat().

Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

s3: VFS: Add SMB_VFS_UNLINKAT().

Currently identical to SMB_VFS_UNLINK().

Next, add to all VFS modules that implement
unlink and eventually remove unlink.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>

ctdb-tests: Update README

Bring this up to date.  Drop descriptions of command-line options
because these tend to bit-rot - refer to "run_tests.sh -h" instead.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Thu Sep 26 06:01:33 UTC 2019 on sn-devel-184

ctdb-tests: Move simple tests to INTEGRATION/ subdirectory

Split some tests out into database/ and failover/ subdirectories.
Rename the remaining tests in simple/.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move complex tests to CLUSTER/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Convert local daemons include file into top-level include

Do the same with the alternative code for real clusters.

Both of these can now be used by other test suites.

Fix some basic shellcheck warnings (e.g. avoid word-splitting by
quoting) while moving code and add the new files to the shellcheck
test.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Drop use of array in run_tests()

This doesn't accomplish anything.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Drop custom handling for unit tests

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Skip README files in test directories/collections

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Recurse into directories instead of explicitly looping

run_tests() already has a loop, so use it.

This means collections of test suites can be handled - but explicitly
check valid collection names to avoid running junk.

Add special cases for simple and complex.  These will be removed when
those test suites are moved to collections.  This seems to be the
smallest amount of churn to support bisection.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Fold find_and_run_one_test() in to run_tests()

The additional function makes the logic harder to understand.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move tool tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move takeover_helper tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move takeover tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move shellcheck tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move onnode tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move eventscript tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move eventd tests to UNIT/ subdirectory

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb-tests: Move cunit tests to UNIT/ subdirectory

The UNIT subdirectory directory will contain all unit tests.  Use an
uppercase name to avoid conflicts with source code subdirectories.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

waf: Use waf function to add for -Wl,--as-needed

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Wed Sep 25 16:56:40 UTC 2019 on sn-devel-184

waf:replace: Do not link against libpthread if not necessary

On Linux we should avoid linking everything against libpthread. Symbols
used my most application are provided by glibc and code which deals with
threads has to explicitly link against libpthread.  This avoids setting
LDFLAGS=-pthread globally.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

third_party: Link uid_wrapper against pthread

uid_wrapper uses pthread_atfork() which is only provided by libpthread.                                                                                   │····················
So we need an explicit dependency.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

third_party: Link nss_wrapper against pthread

nss_wrapper uses pthread_atfork() which is only provided by libpthread.
So we need an explicit dependency.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

third_party: Only link cmocka against librt if really needed

cmocka also uses clock_gettime().

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

pthreadpool: Only link pthreadpool against librt if we have to

This calls clock_gettime() which is available in glibc on Linux. If the
wscript in libreplace detected that librt is needed for clock_gettime()
we have to link against it.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

replace: Only link against librt if really needed

fdatasync() and clock_gettime() are provided by glibc on Linux, so there
is no need to link against librt. Checks have been added so if there are
platforms which require it are still functional.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

s3:waf: Do not check for nanosleep() as we don't use it anywhere

We use usleep() in the meantime.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14140

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Pair-Programmed-With: Isaac Boukris <iboukris@gmail.com>
Reviewed-by: Matthias Dieter Wallnöfer <mdw@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>

s4:scripting: make samba-gpupdate heed --disable-python

Signed-off-by: Philipp Gesang <philipp.gesang@intra2net.com>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Noel Power<npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Wed Sep 25 11:06:04 UTC 2019 on sn-devel-184

s3/librpc/crypto: clang: Fixes Value stored to 'ret' is never read

Fixes:

source3/librpc/crypto/gse_krb5.c:63:3: warning: Value stored to 'ret' is never read <--[clang]
                ret = smb_krb5_kt_free_entry(krbctx, &kt_entry);
                ^     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 warning generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s3/libads: clang: Fix Array access results in a null pointer dereference

Fixes:

source3/libads/cldap.c:400:6: warning: Array access (from variable 'responses') results in a null pointer dereference <--[clang]
        if (responses[0] == NULL) {
            ^
1 warning generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s3/librpc/crypto: clang: warning: Value stored to 'gss_maj' is never read

Fixes: Value stored to 'gss_maj' is never read
source3/librpc/crypto/gse.c:562:3: warning: Value stored to 'gss_maj' is never read <--[clang]
                gss_maj = gss_release_buffer(&gss_min, &out_data);
                ^         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source3/librpc/crypto/gse.c:687:3: warning: Value stored to 'gss_maj' is never read <--[clang]
                gss_maj = gss_release_buffer(&gss_min, &out_data);
                ^         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source3/librpc/crypto/gse.c:739:3: warning: Value stored to 'gss_maj' is never read <--[clang]
                gss_maj = gss_release_buffer(&gss_min, &msg_min);
                ^         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source3/librpc/crypto/gse.c:742:3: warning: Value stored to 'gss_maj' is never read <--[clang]
                gss_maj = gss_release_buffer(&gss_min, &msg_maj);
                ^         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4 warnings generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/clideltree.c: clang: Value stored to 'status' is never read

Fixes:

source4/libcli/clideltree.c:113:3: warning: Value stored to 'status' is never read <--[clang]

status = smbcli_setatr(tree, dname, FILE_ATTRIBUTE_NORMAL, 0);
                ^        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source4/libcli/clideltree.c:131:3: warning: Value stored to 'status' is never read <--[clang]
                status = smbcli_setatr(dstate.tree, dname, FILE_ATTRIBUTE_NORMAL, 0);
                ^        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2 warnings generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/smb2: clang: Fix 'value stored to 'status' is never read'

Fixes:

source4/libcli/smb2/util.c:134:3: warning: Value stored to 'status' is never read <--[clang]
                status = smb2_util_setatr(tree, dname, FILE_ATTRIBUTE_NORMAL);
                ^        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source4/libcli/smb2/util.c:187:5: warning: Value stored to 'status' is never read <--[clang]
                                status = smb2_util_setatr(tree, name, FILE_ATTRIBUTE_NORMAL);
                                ^        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source4/libcli/smb2/util.c:209:3: warning: Value stored to 'status' is never read <--[clang]
                status = smb2_util_setatr(tree, dname, FILE_ATTRIBUTE_NORMAL);
                ^        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3 warnings generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/smb2: clang: Fix Passed-by-value arg contains uninitialized data

Fixes:

source4/libcli/smb2/getinfo.c:222:11: warning: Passed-by-value struct argument contains uninitialized data (e.g., field: 'data') <--[clang]
        status = smb_raw_fsinfo_passthru_parse(b.out.blob, mem_ctx, io->generic.level, io);
                 ^
1 warning generated.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/raw: clang: Fix 'Value stored to 'p' is never read'

Fixes:

source4/libcli/raw/clisession.c:198:4: warning: Value stored to 'p' is never read <--[clang]
                        p += smbcli_req_pull_string(&req->in.bufinfo, mem_ctx, &parms->old.out.domain, p, -1, STR_TERMINATE);
                        ^    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source4/libcli/raw/clisession.c:212:5: warning: Value stored to 'p' is never read <--[clang]
                                p += smbcli_req_pull_string(&req->in.bufinfo, mem_ctx, &parms->nt1.out.domain, p, -1, STR_TERMINATE);
                                ^    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
source4/libcli/raw/clisession.c:232:3: warning: Value stored to 'p' is never read <--[clang]
                p += smbcli_req_pull_string(&req->in.bufinfo, mem_ctx, &parms->spnego.out.workgroup, p, -1, STR_TERMINATE);
                ^    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/raw: clang: Fix 'Dereference of null pointer'

Fixes:

source4/libcli/raw/rawfileinfo.c:597:2: warning: Dereference of null pointer <--[clang]
        SMBCLI_CHECK_WCT(req, 10);
        ^
source4/libcli/raw/rawfileinfo.c:639:2: warning: Dereference of null pointer <--[clang]
        SMBCLI_CHECK_WCT(req, 11);
        ^

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

s4/libcli/raw: clang: Fix 'Access to field results in a deref of a null ptr

Fixes:

source4/libcli/raw/rawfileinfo.c:301:56: warning: Access to field 'transport' results in a dereference of a null pointer (loaded from variable 'session') <--[clang]
    parms->standard.out.create_time = raw_pull_dos_date2(session->transport,
                                                                         ^
source4/libcli/raw/rawfileinfo.c:314:55: warning: Access to field 'transport' results in a dereference of a null pointer (loaded from variable 'session') <--[clang]
    parms->ea_size.out.create_time = raw_pull_dos_date2(session->transport,

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andreas Schneider <asn@samba.org>

vfs_fileid: add "fsname_norootdir_ext" option

This can be used to deliberately break lock coherency between all smbd processes
in the whole cluster for the root directory of a share.

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Wed Sep 25 00:48:45 UTC 2019 on sn-devel-184

vfs_fileid: add extid mapping hooks

For this always ends up calling fileid_extid_mapping_zero(), so no change in
behaviour. This will change in a subsequent commit.

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>

selftest/Samba3.pm: use "winbind use krb5 enterprise principals = yes" for ad_member

This demonstrates that can do krb5_auth in winbindd without knowning about trusted domains.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Autobuild-User(master): Günther Deschner <gd@samba.org>
Autobuild-Date(master): Tue Sep 24 19:51:29 UTC 2019 on sn-devel-184

selftest/Samba3.pm: use "winbind scan trusted domains = no" for ad_member

This demonstrates that we rely on knowning about trusted domains before
we can do krb5_auth in winbindd.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

selftest/tests.py: test pam_winbind for trusts domains

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

selftest: Export TRUST information in the ad_member target environment

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

selftest/tests.py: test pam_winbind with a lot of username variations

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

selftest/tests.py: test pam_winbind with krb5_auth

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

selftest/tests.py: prepare looping over pam_winbindd tests

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

test_pam_winbind.sh: allow different pam_winbindd config options to be specified

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

tests/pam_winbind.py: allow upn names to be used in USERNAME with an empty DOMAIN value

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

tests/pam_winbind.py: turn pypamtest.PamTestError into a failure

A failure generated by the AssertionError() checks can be added
to selftest/knownfail.d/*.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

s3:winbindd: implement the "winbind use krb5 enterprise principals" logic

We can use enterprise principals (e.g. upnfromB@B.EXAMPLE.COM@PRIMARY.A.EXAMPLE.COM)
and delegate the routing decisions to the KDCs.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

docs-xml: add "winbind use krb5 enterprise principals" option

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>

krb5_wrap: let smb_krb5_parse_name() accept enterprise principals

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14124

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>