autobuild: remove unused variables

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/tests/source: remove useless local variable

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/tests/credentials: remove unused variables

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/xattr: remove unused variable

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/upgradehelpers: remove unused variable

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

samba-tool dbcheck: remove unused variable

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

dbchecker: remove unused variables

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

samba-tool spn: remove unused variables

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

samba-tool rodc: remove unused variable

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/samba3: import passdb in the manner it is used

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/kcc/graph: import KCCError, which is used somewhere

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

tests/rodc_rwdc: fix message format

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

traffic_replay: Exception has no .message

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

py/tests/dcerpc_rpc: Py3 compat integer types

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

waflib: fix syntax error in string format

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

waflib: add necessary imports

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/remove_dc: fix Exception construction

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/remove_dc: fix formating syntax (missing %s)

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

provision/backend: LDAPBackendResult takes no creds

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

provision/backend: fix formating syntax

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

provision: fix string formatting (number of args)

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

provision: fix string format syntax

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

Python provision: remove "ExistingBackend"

This was unused and broken. e.g. here:

-    def init(self):
-        # Check to see that this 'existing' LDAP backend in fact exists
-        ldapi_db = Ldb(self.ldapi_uri)

there is no attribute self.ldapi_uri, so this would always raise an
exception.

It was being left around in case it became useful, but that doesn't
seem to be happening.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python dbcheck: use real exception name

and conventional indent

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python: avoid useless work in dsdb_Dn.__cmp__

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python join: use the sd_utils we imported

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/dsdb/pytest/sites: do not use variables out of scope

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

ldb/tests/py/index: remove dup value in dictionary

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

ldb/tests/py/api: use proper name for ldb.LdbError

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/tests/dnscmd: don't use undefined name

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/tests/dsdb: use correct variable names

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python/subnets: use the correct variable name

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

samba-tool user: fix message format

There were 2 % formats and 3 arguments.
Also reformat for line length

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

samba-tool: let self.usage() find argv[0] by itself

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

dbcheck: fix function call (right arguments)

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

dbcheck: fix message formatting

previously these would have raised an exception

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

python: PY3 Exceptions don't have .message

but str(e) is the same as str(e.message), so we can use that
on 2 and 3.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

script/show_test_time: attempt py3 compat

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/scripting/pfm_verify: PY3: use compat.text_type

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/scripting/samba_upgradedns: avoid .has_key()

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/scripting/minschema: whitespace reformat and py3 compatible print

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/scripting/autoidl: py3 compatible except

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/script/depfilter.py: use py3 compatible regex import

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

tests/python/sec_descriptor: safer use of super()

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

tests/python/notification: safer use of super()

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

s4/dsdb/pytest/ldap_schema: fix typo in docs

Commit 311727947799e896e05d644103c9db80a665de88 removed a duplicate
test but it removed the wrong one, leaving this dreadful typo.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

selftest/format_subunit_json: py3 print()

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

tests/samba_tool/provision_password_check: follow super inheritance

We were skipping a level in the inheritance chain, which had no effect
in this case (no .setUps or .tearDowns were missed) but it would be
confusing if the parents ever changed.

Note: in python 3, you just call super() with no args, and it works
out the right thing.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Noel Power <noel.power@suse.com>

nsswitch: Run nsswitch thread test

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Nov  1 05:06:23 CET 2018 on sn-devel-144

nsswitch: add test for parallel NSS & libwbclient calls

Signed-off-by: Ralph Wuerthner <ralph.wuerthner@de.ibm.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nsswitch: protect access to wb_global_ctx by a mutex

This change will make libwbclient thread safe for all API calls not using a
context. Especially there are no more conflicts with threads using nsswitch
and libwbclient in parallel.

Signed-off-by: Ralph Wuerthner <ralph.wuerthner@de.ibm.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nsswitch: make wb_global_ctx private add add get/put functions to access global context

Signed-off-by: Ralph Wuerthner <ralph.wuerthner@de.ibm.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nsswitch: use goto to have only one function return

Signed-off-by: Ralph Wuerthner <ralph.wuerthner@de.ibm.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s3: winbind: Remove fstring from wb_acct_info struct

The group enumeration backend functions try to allocate an array of
wb_acct_info structs with a number of elements equal to the number of
groups. In domains with a large number of groups this allocation may
fail due to the size of the chunk.

Found while trying to enumerate the groups in a domain with more than
700k groups.

Signed-off-by: Samuel Cabrero <scabrero@suse.de>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

vfs_fruit: let fruit_open_meta() with O_CREAT return a fake-fd

This is the final step in implementing the needed macOS semantics on the
FinderInfo stream: as long as the client hasn't written a non-zero
FinderInfo blob to the stream, there mustn't be a visible filesystem
entry for other openers.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Nov  1 01:14:23 CET 2018 on sn-devel-144

vfs_fruit: don't check for delete-on-close on the FinderInfo stream

macOS SMB server doesn't filter out the FinderInfo stream if it has
delete-on-close set.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: let fruit_pwrite_meta_stream also ftruncate empty FinderInfo

fruit_streaminfo currently filters out the FinderInfo stream is
delete-on-close is set. We set it here internally, but the client may
also set it over SMB. Turns out that the macOS SMB server does NOT
filter out FinderInfo stream with delete-on-close set, so we must change
the way filtering is done in fruit_streaminfo.

Filtering is now done based on the FinderInfo stream being 0-bytes large which
is why I'm adding the ftruncate here.

No idea why the tests that check the filtering passed the commits
leading up to this one, but if you revert this commit after applying the
whole patchset, the "delete AFP_AfpInfo by writing all 0" test will fail.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: pass stream size to delete_invalid_meta_stream()

delete_invalid_meta_stream() is meant to guard against random data being
present in the FinderInfo stream. If the stream size is 0, it's likely a
freshly created stream where no data has been written to yet, so don't
delete it.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: let fruit handle all aio on the FinderInfo metadata stream

This will be required to support using fake fds for the FinderInfo
metadata stream.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: do ino calculation

As we'll start returning fake fds in open shortly, we can't rely on the
next module to calculat correct inode numbers for streams and must take
over that responsibility.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: prepare fruit_pread_meta() for reading on fake-fd

If the read on the stream fails we may have hit a handle on a just
created stream (fio->created=true) with no data written yet.

If that's the case return an empty initialized FinderInfo blob.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: prepare fruit_pwrite_meta() for on-demand opening and writing

This avoid creating files or blobs in our streams backend when a client
creates a stream but hasn't written anything yet. This is the only sane
way to implement the following semantics:

* client 1: create stream "file:foo"

* client 2: open stream "file:foo"

The second operation of client 2 must fail with NT_STATUS_NOT_FOUND.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: prepare struct fio for fake-fd and on-demand opening

Not used for now, that comes in the subsequent commits.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: add fio->created

fio->created tracks whether a create created a stream.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: remove resource fork special casing

Directly unlinking a file with open handles is not good, don't do it.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: add some debugging of dev/ino

Aids in debugging dev/ino mismatch failures in open_file_ntcreate.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: add test "empty_stream"

One to rule them all: consistently test critical operations on all
streams relevant to macOS clients: the FinderInfo stream, the Resource
Fork stream and an arbitrary stream that macOS maps to xattrs when
written to on a macOS SMB server.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: add check_stream_list_handle()

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/util: add torture_smb2_open()

This seems to be missing: a simple wrapper to just open a file without
fancy options.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: filter empty streams

First step in achieving macOS compliant behaviour wrt to empty streams:
- hide empty streams in streaminfo
- prevent opens of empty streams

This means that we may carry 0-byte sized streams in our streams
backend, but this shouldn't really hurt.

The previous attempt of deleting the streams when an SMB setinfo eof to
0 request came in, turned out be a road into desaster.

We could set delete-on-close on the stream, but that means we'd have to
check for it for every write on a stream and checking the
delete-on-close bits requires fetching the locking.tdb record, so this
is expensive and I'd like to avoid that overhead.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: use check on global_fruit_config.nego_aapl for macOS specific behaviour

Ensure any non MS compliant protocol behaviour targetted at supporting
macOS clients are only effective if the client negotiated AAPL.

Currently this only guards the resource fork which only macOS client are
going to use, but subsequent commits add more this at this place.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: enable AAPL extensions in a bunch of tests

These tests check for macOS SMB server specific behaviour. They work
currently against Samba without enabling AAPL because in vfs_fruit we're
currently don't check whether AAPL has been negotiated in one place. A
subsequent commit will change that and this commit prepares for that
change.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: don't unlink 0-byte size truncated streams

This caused all sort of havoc with subsequent SMB request that acted on
the handle of the then deleted backend storage (file or blob, depending
on the used streams module).

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: write some data to a just created teststream

Doesn't currently make a difference, but this prepares for a later
change in vfs_fruit that will filter out empty streams (which is the
macOS behaviour).

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: expand test "setinfo eof stream"

o Adds checks verifying that after setting eof to 0 on a stream, a
  subsequent open gets ENOENT, before and after closing the handle that
  had been used to set eof to 0.

o Verify that a write to a handle succeeds after that handle has been
  used to set eof to 0 on a stream.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: update handling of read-only creation of resource fork

macOS SMB server versions supports this since 10.12, so we adapt our
behaviour.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: update test "creating rsrc with read-only access" for newer macOS versions

While this operation failed against older macOS versions, it passes
against versions 10.12 and newer. Update the test accordingly, a
subsequent commit will then update our implementation.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: expand existing vfs_test "null afpinfo"

This adds a check that a read on a seperate handle also sees the
previously created AFP_AfpInfo stream.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: expand existing test "setinfo delete-on-close AFP_AfpInfo" a little bit

Add a check that verifies a create on a stream gets
NT_STATUS_DELETE_PENDING after delete-on-close has been set.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: update test "read open rsrc after rename" to work with macOS

macOS SMB server seems to return NT_STATUS_SHARING_VIOLATION in this
case while Windows 2016 returns NT_STATUS_ACCESS_DENIED.

Lets stick with the Windows error code for now in the Samba fileserver,
but let the test pass against macOS.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: ensure a directory handle is closed in all code paths

Otherwise we get a sharing violation when running against Samba and
opening the directory a second time.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: update test "stream names" to work with macOS

o create the basefile before trying to create a stream on it, otherwise
  this fails on macOS

o write something to the stream, otherwise the stream is not listed as
  macOS hides 0-byte sized streams

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: update test "SMB2/CREATE context AAPL" to work against macOS

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: set share_access to NTCREATEX_SHARE_ACCESS_MASK in check_stream_list

Avoid sharing conflicts with other opens on the basefile.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: fix a few error checks in "delete AFP_AfpInfo by writing all 0"

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

s4:torture/vfs/fruit: skip a few tests when running against a macOS SMB server

These tests are designed to test specific vfs_fruit functionality.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_streams_xattr: fix open implementation

Since a long time the modules's open function happily returned success
when opening a non existent stream without O_CREAT.

This change fixes it to return -1 and errno=ENOATTR if

o get_ea_value() returns NT_STATUS_NOT_FOUND (eg mapped from
  getxattr() = -1, errno=ENOATTR) and

o flags doesn't contain O_CREAT

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13646

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

vfs_fruit: optionally delete AppleDouble files without Resourcefork data

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

vfs_fruit: add option "delete_empty_adfiles"

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

vfs_fruit: detect empty resource forks in ad_convert()

For some reason the macOS client often writes AppleDouble files with a
non-zero sized resource fork, but the resource fork data is just
boilerplate data with the following string close to the start

  This resource fork intentionally left blank

A dump with apple_dump looks like this:

Entry ID   : 00000002 : Resource Fork
Offset     : 00000052 : 82
Length     : 0000011E : 286

-RAW DUMP--:  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F : (ASCII)
00000000   : 00 00 01 00 00 00 01 00 00 00 00 00 00 00 00 1E : ................
00000010   : 54 68 69 73 20 72 65 73 6F 75 72 63 65 20 66 6F : This resource fo
00000020   : 72 6B 20 69 6E 74 65 6E 74 69 6F 6E 61 6C 6C 79 : rk intentionally
00000030   : 20 6C 65 66 74 20 62 6C 61 6E 6B 20 20 20 00 00 :  left blank   ..
00000040   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000050   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000060   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000070   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000080   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000090   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000A0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000B0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000C0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000D0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000E0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
000000F0   : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 : ................
00000100   : 00 00 01 00 00 00 01 00 00 00 00 00 00 00 00 1E : ................
00000110   : 00 00 00 00 00 00 00 00 00 1C 00 1E FF FF       : ..............

We can safely discard this Resource Fork data.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

vfs_fruit: add option "wipe_intentionally_left_blank_rfork"

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

s4:torture: add test for AppleDouble ResourceFork conversion

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

s3:selftest: list vfs testssuites one per line

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

docs:vfs_fruit: add "delete_empty_adfiles" option

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

docs:vfs_fruit: add "wipe_intentionally_left_blank_rfork" option

Bug: https://bugzilla.samba.org/show_bug.cgi?id=13642

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

s3:smbd: remove "0x" string prefix from dev/ino

We used %llu as conversion specifier which results in a decimal number
being printed, so remove the misleading "0x" prefix.

While at it, I'll change %llu to the terse %ju.

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

third_party: Update nss_wrapper to version 1.1.5

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

netcmd: Add 'samba-tool group stats' command

With large domains it's hard to get an idea of how many groups there
are, and how many users are in each group, on average. However, this
could have a big impact on whether a problem can be reproduced or not.

This patch dumps out some summary information so that you can get a
quick idea of how big the groups are.

Signed-off-by: Tim Beale <timbeale@catalyst.net.nz>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org>
Autobuild-Date(master): Wed Oct 31 03:40:41 CET 2018 on sn-devel-144

netcmd: Include num-members in 'samba-tool group list --verbose'

This adds an easy way for users to see (via samba-tool) how many members
are in various groups, without querying the members for each individual
group.

For example, you could pipe this output to grep to check for groups with
zero or one members (i.e. historic groups that may no longer make
sense).

Signed-off-by: Tim Beale <timbeale@catalyst.net.nz>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>

selftest: Specify different DB backends for restored testenvs

Vary the DB backend that we use for the renamed DCs. The labdc and
renamedc are fairly similar, so let's have each of them use a different
backend.

Signed-off-by: Tim Beale <timbeale@catalyst.net.nz>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>