Volker Lendecke [Mon, 24 Feb 2020 17:42:46 +0000 (09:42 -0800)]
libsmb: Pass the correct path to cli_posix_stat()
This fixes doing strlen() on talloc_tos(), about which valgrind is pretty
unhappy. Without this patch we survive the tests because we have fallbacks to
the non-posix flavors of stat(). With this patch in place cli_posix_stat()
becomes functional in this code path. This creates conflicts with the readdir
libsmbclient tests, which need fixing separately.
Volker Lendecke [Sun, 23 Feb 2020 12:34:18 +0000 (04:34 -0800)]
selftest: Slightly simplify setting up libsmbclient tests
We'll add per-test arguments soon
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): Tue Feb 25 19:30:27 UTC 2020 on sn-devel-184
Volker Lendecke [Sun, 23 Feb 2020 11:22:47 +0000 (12:22 +0100)]
selftest: Consolidate planning libsmbclient tests to source4
We don't need to plan those tests both in source3 and source4's
tests.py. The source4 version seems more "advanced", it uses
smbtorture4_testsuites() to list all tests, making it less
error-prone, and it also does the tests for all SMB1 and SMB3
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Ralph Boehme [Mon, 24 Feb 2020 14:03:56 +0000 (15:03 +0100)]
VFS: default: add support for FILE_ATTRIBUTE_OFFLINE to async dosmode
This had been missing in the initial async dosmode implementation. It's the
responsibility of the sync and async dosmode functions to call
vfswrap_is_offline() since the offline functionality has been converted from a
first class VFS function to be a part of the DOS attributes VFS functions.
Ralph Boehme [Mon, 24 Feb 2020 13:29:01 +0000 (14:29 +0100)]
VFS: default: avoid a crash in vfswrap_getxattrat_do_sync()
Must use tevent_req_data() to get our tevent_req state, talloc_get_type_abort()
will just crash as struct tevent_req != struct vfswrap_getxattrat_state.
Ralph Boehme [Mon, 24 Feb 2020 13:24:12 +0000 (14:24 +0100)]
VFS: default: let vfswrap_is_offline() take conn, not handle
vfswrap_is_offline() has been converted to a "helper" function some time ago, it
had been a VFS interface function before. To make this change more obvious let
it take a struct connection_struct instead of a struct vfs_handle_struct which
is the canonical first parameter to VFS functions.
* Bug 14270:
Samba 4.11 and later give incorrect results for SCOPE_ONE searches
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Karolin Seeger <kseeger@samba.org>
Autobuild-Date(master): Tue Feb 25 12:59:02 UTC 2020 on sn-devel-184
The only difference is that the pull macros do the correct casting of
the integer in the end.
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Fri Feb 21 03:35:58 UTC 2020 on sn-devel-184
This is an implementation which doesn't have undefined behavior
problems. It casts correctly that calculations are don in the correct
integer space. Also the naming is less confusing than what we have in
byteorder.h.
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Volker Lendecke [Thu, 20 Feb 2020 13:13:35 +0000 (14:13 +0100)]
smbd: Separate aio_pthread indicator from normal EINTR
According to Posix and the Linux open(2) manpage, the open-syscall can
return EINTR. If that happens, core smbd saw this as an indication
that aio_pthread's open function was doing its job. With a real EINTR
without aio_pthread this meant we ended up in a server_exit after 20
seconds, because there was nobody to do the retry.
EINTR is mapped to NT_STATUS_RETRY. Handle this by just retrying after
a second.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=14285 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 Feb 20 22:14:25 UTC 2020 on sn-devel-184
Volker Lendecke [Fri, 14 Feb 2020 16:20:22 +0000 (17:20 +0100)]
autobuild: samba-ctdb does not need an AD DC
Speed up the build a bit
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Wed Feb 19 11:28:56 UTC 2020 on sn-devel-184
Martin Schwenke [Tue, 4 Feb 2020 08:44:11 +0000 (19:44 +1100)]
selftest:clusteredmember: Build a unclist for every share
This is fairly cheap and it is simple to do. This allows the Python
code to be able to specify a unclist quite simply. The level of
coupling doesn't seem worse than anything else in the
selftest/autobuild code.
There may be cleverer ways of doing this (e.g. a wrapper in
testprogs/blackbox/clusteredmember_smbtorture or similar) but cleverer
code isn't necessarily better code... and they'll probably involve
code duplication.
Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Volker Lendecke <vl@samba.org>
Jeremy Allison [Wed, 12 Feb 2020 21:54:08 +0000 (13:54 -0800)]
s3: DFS: Remove is_msdfs_link_internal() - no longer used.
All DFS links are now read through the VFS and not via
symlink calls.
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Tue Feb 18 22:34:16 UTC 2020 on sn-devel-184
s3:tests: Add test for a dropbox with dir mode 0733
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Feb 18 21:07:44 UTC 2020 on sn-devel-184
Volker Lendecke [Tue, 11 Feb 2020 21:10:32 +0000 (22:10 +0100)]
lib: Fix a shutdown crash with "clustering = yes"
This is a bit confusing now, sorry for that:
register_msg_pool_usage() in the ctdb case uses
messaging_ctdb_register_tevent_context(), which talloc_reference()s
the central struct messaging_ctdb_fde_ev of the
messaging_ctdb_context. In messaging_reinit(), we talloc_free only one
of those references and allocate a new messaging_ctdb_fde_ev. The
remaining messaging_ctdb_fde_ev should have been deleted as well, but
due to the second reference this does not happen. When doing the
shutdown messaging_ctdb_fde_ev_destructor() is called twice, once on
the properly reinitialized fde_ev, and once much later on the leftover
one which references invalid data structures.
By the way, this is not a problem with talloc_reference(), this would
have happened with explicit refcounting too.
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Björn Baumbach <bb@sernet.de>
Autobuild-Date(master): Tue Feb 18 13:05:53 UTC 2020 on sn-devel-184
Consolidate "msg_dgm_ref" and "msg_ctdb_ref": The only purpose of
those pointers was to TALLOC_FREE() them in messaging_reinit(). We'll
have a third entity to talloc_free() in the next commit, make that
simpler.
Martin Schwenke [Wed, 12 Feb 2020 23:09:08 +0000 (10:09 +1100)]
ctdb-docs: Provide example commands for "ctdb event ..."
The example output doesn't tell a user what command generated it.
Adding the command makes the examples much more useful.
Reported-by: Stefan Kania <stefan@kania-online.de> 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): Tue Feb 18 04:22:56 UTC 2020 on sn-devel-184
Martin Schwenke [Wed, 11 Dec 2019 10:44:28 +0000 (21:44 +1100)]
ctdb-tests: Rationalise node stop/start/restart
Separate functions are not needed for stopping/starting/restarting
individual nodes. The stop and start functions essentially just use
onnode, though for local daemons this is embedded in local_daemons.sh.
So, just provide one stop and one start function that takes an
optional nodespec, defaulting to all nodes.
Restarting becomes common.
Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Martin Schwenke [Mon, 10 Feb 2020 06:50:30 +0000 (17:50 +1100)]
ctdb-daemon: Fix signed/unsigned comparison
csbuild says:
ctdb/server/ctdb_lock.c: scope_hint: In function ‘ctdb_find_lock_context’
ctdb/server/ctdb_lock.c:671:33: warning: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Wsign-compare]
Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Amitay Isaacs <amitay@gmail.com>
The module makes use of the new io_uring infrastructure
(intruduced in linux 5.1), see https://lwn.net/Articles/778411/ and
http://git.kernel.dk/cgit/liburing/
Currently this only implements SMB_VFS_{PREAD,PWRITE,FSYNC}_SEND/RECV
and avoids the overhead of our userspace threadpool.
In future we'll hopefully make more use of more advanced io_uring
features.
For now we don't have automated tests as our test infrastructure
doesn't use a recent kernel. At least we're able to do compile tests
on fedora31.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Sat Feb 15 11:37:45 UTC 2020 on sn-devel-184
projects / thirdparty / samba.git / log
