Volker Lendecke [Thu, 19 Mar 2020 16:00:59 +0000 (17:00 +0100)]
lib: Try nonblocking writes in writev_send()
All callers now use nonblocking sockets, so that we can optimize by
doing early writes
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Mar 25 10:41:29 UTC 2020 on sn-devel-184
Volker Lendecke [Fri, 20 Mar 2020 10:46:43 +0000 (11:46 +0100)]
libsmb: Make sure that the TCP socket is non-blocking
All traffic goes through smbXcli_base.c, and that is prepared to deal
with short writes via the conn->outgoing queue. Instead of making sure
that all callers properly set the socket nonblocking, do it here, so
that we can later optimize sending out data to the server.
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
third_party: Update resolv_wrapper to version 1.1.6
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Mar 25 08:33:17 UTC 2020 on sn-devel-184
Ralph Boehme [Mon, 23 Mar 2020 09:44:08 +0000 (10:44 +0100)]
vfs: remove SMB_VFS_OPENDIR()
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): Tue Mar 24 21:23:43 UTC 2020 on sn-devel-184
Ralph Boehme [Wed, 18 Mar 2020 15:01:17 +0000 (16:01 +0100)]
smbd: use open_internal_dirfsp_at() and OpenDir_fsp() in OpenDir()
This allows consolidating symlink safe processing of pathnames to the low-level
function non_widelink_open() used in the file open codepath via fd_open() and
the new utility function open_internal_dirfsp_at().
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Ralph Boehme [Wed, 18 Mar 2020 14:59:11 +0000 (15:59 +0100)]
smbd: remove fdopendir() fallback
In order to get rid of SMB_VFS_OPENDIR, we have to require
fdopendir(). Everybody but macOS seems to have it, so RIP opendir() fallback.
This also prepares for a subsequent commit where we're going to start calling
OpenDir_fsp() from inside OpenDir(). Without removing the fallback this would
result in a recursion.
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Ralph Boehme [Tue, 17 Mar 2020 16:35:44 +0000 (17:35 +0100)]
smbd: simplify non_widelink_open()
As fsp->is_directory is already correctly populated by our callers, we can drop
the complicated and possibly broken logic that relies on O_DIRECTORY being
defined.
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Ralph Boehme [Wed, 18 Mar 2020 16:40:29 +0000 (17:40 +0100)]
vfstest: use OpenDir()
Note that as ReadDirName() returns translated names (in Windows "encoding"), in
cmd_translate_name() test we have to translate back to UNIX "encoding" to check
if the filename matches the user requested path which is also in UNIX
"encoding".
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Ralph Boehme [Thu, 19 Mar 2020 10:19:48 +0000 (11:19 +0100)]
smbd: use canonicalize_connect_path() in create_conn_struct_as_root()
This fix ensures create_conn_struct_as_root() works the same way as
make_connection_snum() used in smbd for SMB sessions. Without this any caller of
create_conn_struct_as_root() will fail to use work on shares if the sharepath
contains a symlink.
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Currently fruit does a cleanup run on a directory when it is being removed: it
lists the directory and does an explicit unlink on any found file that is a
genuine AppleDouble file.
This is not realy needed: the ._ AppleDouble sidecar files are either vetoed by
default, so cleanup can be turned on by the built-in "delete veto files"
options. Or, if the user sets "fruit:veto_appledouble", the client will see the
file in a directory list as ordinary files, so the client is responsible for
deleting them first before removing the parent directory.
tl;dr
One caller of SMB_VFS_OPENDIR() less. :)
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
selftest: Fix string compare in DnsHandler() of dns_hub.py
dns_hub.py:115: SyntaxWarning: "is" with a literal. Did you mean "=="?
if forwarder is 'ignore':
dns_hub.py:117: SyntaxWarning: "is" with a literal. Did you mean "=="?
elif forwarder is 'fail':
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Tue Mar 24 14:31:25 UTC 2020 on sn-devel-184
Instead of attempting to trace all of the cases, just avoid leaving
rec->nodemap set to NULL. Attempting to use an old value is generally
harmless, especially since it will be the same as the new value in
most cases.
Reported-by: Volker Lendecke <vl@samba.org> Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Mar 24 01:22:45 UTC 2020 on sn-devel-184
Martin Schwenke [Fri, 21 Feb 2020 01:24:39 +0000 (12:24 +1100)]
ctdb-recovery: Replace use of ctdb_dbid_map with local db_list
This will be used to build a merged list of databases from all nodes,
allowing the recovery helper to create missing databases.
It would be possible to also include the db_name field in this
structure but that would cause a lot of churn. This field is used
locally in the recovery of each database so can continue to live in
the relevant state structure(s).
Martin Schwenke [Wed, 26 Feb 2020 00:50:09 +0000 (11:50 +1100)]
ctdb-daemon: Respect CTDB_CTRL_FLAG_ATTACH_RECOVERY when attaching databases
This is currently only set by the recovery daemon when it attaches
missing databases, so there is no obvious behaviour change. However,
attaching missing databases can now be moved to the recovery helper as
long as it sets this flag.
Martin Schwenke [Fri, 21 Feb 2020 00:13:05 +0000 (11:13 +1100)]
ctdb-recovery: Use CTDB_CTRL_FLAG_ATTACH_RECOVERY to attach during recovery
ctdb_ctrl_createdb() is only called by the recovery daemon, so this is
a safe, temporary change. This is temporary because
ctdb_ctrl_createdb(), create_missing_remote_databases() and
create_missing_local_databases() will all go away soon.
Note that this doesn't cause a change in behaviour. The main daemon
will still only defer attaches from non-recoverd processes during
recovery.
Dont allow clients to connect to databases untile we are well past
and through the initial recovery phase
It is unclear what this commit was attempting to do. The commit
message implies that more attaches should be deferred but the code
change adds a conjunction that causes less attaches to be deferred.
In particular, no attaches will be deferred after startup is complete.
This seems wrong.
To implement what seems to be stated in the commit message an "or"
needs to be used so that non-recovery daemon attaches are deferred
either when in recovery or before startup is complete. Making this
change highlights that attaches need to be allowed during the
"startup" event because this is when smbd is started.
Amitay Isaacs [Tue, 18 Feb 2020 05:17:00 +0000 (16:17 +1100)]
ctdb-recovery: Don't trust nodemap obtained from local node
It's possible to have a node stopped, but recovery master not yet
updated flags on the local ctdb daemon when recovery is started. So do
not trust the list of active nodes obtained from the local node. Query
the connected nodes to calculate the list of active nodes.
Aurelien Aptel [Fri, 13 Mar 2020 12:42:53 +0000 (13:42 +0100)]
s3: libsmbclient.h: add missing time.h include
A recent change added a struct with timespec members in the public
libsmbclient header. This type is defined in time.h which was not
included thus making users of libsmbclient not build properly.
/.../libsmbclient.h:158:18: error: field 'btime_ts' has incomplete type
struct timespec btime_ts;
Fixes: bf13fe0f222 ("s3: libsmbclient: Add internal/external structures needed for readdirplus.") Signed-off-by: Aurelien Aptel <aaptel@suse.com> Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Mon Mar 23 20:47:44 UTC 2020 on sn-devel-184
Andrew Bartlett [Sat, 21 Mar 2020 19:15:30 +0000 (08:15 +1300)]
autobuild: Merge samba-simpleserver into samba-nt4
This saves CI resources (20mins build time) at the expense of making
the samba-nt4 job take 45min (5 min longer).
The new maximum job time is 1:15 by samba-o3 on the different hosts.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Mon Mar 23 14:45:59 UTC 2020 on sn-devel-184
Andrew Bartlett [Sat, 21 Mar 2020 19:29:33 +0000 (08:29 +1300)]
selftest: Run smb2.compound_find against filesrever only, not nt4_dc and ad_dc
This is a slow test of SMB2 protocol features, it needs not to be repeated. It only
got run against both by accident at the original merge and then extended to
run on the [compount_find] share (with smbd:find async delay usec = 10000)
but really does not need to be run three times.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Andrew Bartlett [Thu, 12 Mar 2020 03:07:01 +0000 (16:07 +1300)]
autobuild: Merge the samba-ktest-heimdal and samba-fileserver jobs
This avoids a full compile of Samba just to test Kerberos with a system Heimdal
while still providing an environment to test other fileserver features.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Andrew Bartlett [Wed, 24 Jul 2019 04:22:12 +0000 (16:22 +1200)]
build: Require --without-ad-dc for --without-ads
Building an AD DC while setting --without-ads makes no sense and just wastes compile time on our build hosts.
To allow samba-nt4 to build --without-ad-dc we set rpc.spoolss.notify
(which is built on the NTVFS fileserver for the callbacks) to run in
the ad_member environment rather than nt4_dc and ad_dc.
This is also just more realistic in any case.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Jeremy Allison [Wed, 18 Mar 2020 23:09:50 +0000 (16:09 -0700)]
s3: Remove tevent_wait code. The last user was fsp->deferred_close.
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): Mon Mar 23 12:06:45 UTC 2020 on sn-devel-184
Jeremy Allison [Wed, 18 Mar 2020 22:09:51 +0000 (15:09 -0700)]
s3: smbd: Add async internals of reply_close().
Waits until all aio requests on the closing fsps
before returning to the client.
Slightly modified version of the existing async
reply_close code, updated to use the wait_queue
pattern standardized in reply_tdis, reply_ulogoffX
and reply_exit.
Done this way (commented out) so it is a clean
diff and it's clear what is being added.
The next commit will remove the old version.
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org>
Douglas Bagnall [Mon, 9 Mar 2020 22:31:12 +0000 (11:31 +1300)]
idl/drsblobs: do not overwrite number of schedules == 1
If the struct has zero or two schedules, that is what it has, and we
should let that be.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Sun Mar 22 06:19:51 UTC 2020 on sn-devel-184
Gary Lockyer [Mon, 16 Mar 2020 02:18:12 +0000 (15:18 +1300)]
ldb tests: Confirm lmdb free list handling
Add cmocka tests to confirm lmdb's handling of the free list.
As a result of lmdb's MVCC (Multiversion Concurrency Control) long
running read transactions or stale readers (read transactions where the
process exited without ending the transaction) can cause the database to
run out of space.
Items in the free list are only reused when they would not be visible in
a read transaction. So long running read transactions prevent entries
in the free list being reused, and the database can run out of space.
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Samuel Cabrero [Tue, 28 May 2019 15:15:19 +0000 (17:15 +0200)]
s3:rpc_server: Remove dead code
Signed-off-by: Samuel Cabrero <scabrero@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Samuel Cabrero <scabrero@samba.org>
Autobuild-Date(master): Fri Mar 20 17:11:28 UTC 2020 on sn-devel-184
projects / thirdparty / samba.git / log
