Björn Jacke [Wed, 23 Jan 2019 13:01:26 +0000 (14:01 +0100)]
sambaundoguididx: use the right escaped oder unescaped sam ldb files
the correct filename is taken from the partition database before, we should not
unescape that because this can result in a new unescaped ldb file being created
and the script not to work at all.
Joe Guo [Thu, 7 Feb 2019 22:25:44 +0000 (11:25 +1300)]
.gitignore: add ignore rules for a few dev tools
Signed-off-by: Joe Guo <joeg@catalyst.net.nz> Reviewed-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): Fri Feb 15 06:07:07 CET 2019 on sn-devel-144
Tim Beale [Thu, 14 Feb 2019 23:17:49 +0000 (12:17 +1300)]
autobuild: Replace samba-buildpy2-only with samba-nopython-py2
For Samba 4.11, the minimum python2 functionality we will support (for
now, at least - we may change our minds) is for the --disable-python
target, i.e. if you're excluding all the python functionality from
samba, then WAF should still support being built with python2.
The use case here is old unix platforms that want to use smbd, but don't
have python3 support.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Tim Beale [Tue, 12 Feb 2019 23:41:34 +0000 (12:41 +1300)]
autobuild: Update variable name to make more sense
When we switched from python2 being the default to python3, we didn't
update this variable name. It's now handling the python2 case, but it's
a boolean flag named 'py3', which is rather confusing.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Tim Beale [Tue, 12 Feb 2019 23:12:18 +0000 (12:12 +1300)]
autobuild: Drop py2 autobuild jobs
Samba v4.11 will no longer support python2, so let's drop the autobuild
jobs. This will save some gitlab/sn-devel time and money, as it's less
work for CI to do.
Note that this highlights some previous inconsistencies:
- samba-none-env-py2 was being built for gitlab but not sn-devel.
- samba-nt4-py2 was being built for sn-devel but not gitlab
I've left samba-buildpy2-only for now, which will be addressed in a
subsequent patch.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
This is just a runc container, not tied to docker. See podman.io
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Björn Jacke <bj@sernet.de>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Thu Feb 14 17:32:38 CET 2019 on sn-devel-144
Gary Lockyer [Wed, 13 Feb 2019 21:53:37 +0000 (10:53 +1300)]
s4 dsdb util: samdb_client_site_name clean up
* Initialise pointers to NULL
* replace talloc_free with TALLOC_FREE
* add goto exit to ensure memory deallocated correctly
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Thu Feb 14 06:53:14 CET 2019 on sn-devel-144
Gary Lockyer [Wed, 13 Feb 2019 20:18:20 +0000 (09:18 +1300)]
s4 dsdb util: samdb_client_site_name use dsdb_domain_count
Replace the call to samdb_search_count with dsdb_domain_count. As this
is the only remaining caller of samdb_search_count, replacing it will
allow the removal of samdb_search_count.
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Gary Lockyer [Wed, 13 Feb 2019 20:33:57 +0000 (09:33 +1300)]
s4 rpc_server_samr: DomGeneralInformation use dsdb_domain_count
Use dsdb_domain_count instead of samdb_search_count to determine the
number of users, groups and aliases. This gives a performance gain of
around 10%, reduces the total memory allocated and fixes the incorrect
count returned for aliases.
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Tim Beale [Wed, 13 Feb 2019 23:26:26 +0000 (12:26 +1300)]
.gitlab-ci.yml: Make docker image name more explicit
The 'image' YAML tag implies a docker image, but for people who find
gitlab mysterious, let's make it blatantly obvious what we're doing
here.
+ added a comment
+ added 'DOCKER' to the variable names
+ removed 'BUILD', as we've now dropped this from all the job-names
+ tried to make the variable names consistent, both within the file and
WRT docker terminology
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Thu Feb 14 03:51:21 CET 2019 on sn-devel-144
Tim Beale [Wed, 13 Feb 2019 22:20:16 +0000 (11:20 +1300)]
.gitlab-ci.yml: Include the actual command used by gitlab
Someone who finds gitlab mysterious will have no idea what $CI_JOB_NAME
should be, if they wanted to reproduce the autobuild job manually. It
should be trivial to include the actual command being run in the logs.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
samba_dnsupdate: make rodc_dns_update() more robust against timing problems
Without this we had an interesting race!
The messaging_dgm code caches connected datagram sockets based on the
destination pid for 1 second.
The fact that samba_dnsupdate constantly recreates its messaging
context (and the underlying datagram socket) means that we the winbindd
messaging context may get a stale connection. As a result sending any
message from winbindd back to samba_dnsupdate will result in
ECONNREFUSED.
That means the IRPC response from winbindd never reaches
samba_dnsupdate, which will then hit a timeout.
In turn samba_dnsupdate on the RODC times out.
This was a workaround for the problem, by having just one global
IRPC handle and thus just one messaging_dgm context.
The actual problem is solved a few commits before
("messages_dgm: Properly handle receiver re-initialization").
But we keep this as an performance optimization, which hopefully
means that the overall samba_dnsupdate is less likely to
timeout after the hardcoded 20 seconds.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
This only properly covers the small-message nonblocking case. Covering
the large-message and the blocking case is a much larger effort assuming
we want to re-send the failed message if parts of the message has gone
through properly. Don't do that for now.
This was found by sanba_dnsupdate constantly recreating its irpc handle to
winbindd in the RODC case.
The messaging_dgm code cached connected datagram sockets based on the
destination pid for 1 second. Which means the IRPC responses from
winbindd are never delivered to samba_dnsupdate,
which will then hit a timeout.
Volker Lendecke [Thu, 7 Feb 2019 14:57:06 +0000 (15:57 +0100)]
messages_dgm: Use saved errno value
In this case this is just a cleanup, the value has just been set by
messaging_dgm_sendmsg. But as that already saves errno into a local
variable, use that.
Christof Schmitt [Tue, 12 Feb 2019 19:28:32 +0000 (12:28 -0700)]
waf: Check for libnscd
The check was in the old autoconf, but not in waf. As the code is still
in source3/lib/util_nscd.c, add the check for libnscd to allow building
and using the code.
Noel Power [Mon, 28 Jan 2019 16:57:17 +0000 (16:57 +0000)]
Make sure results from GetAttrString are decref'ed where needed
Signed-off-by: Noel Power <noel.power@suse.com> Reviewed-by: Andrew Bartlett abartlet@samba.org
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Wed Feb 13 14:51:12 CET 2019 on sn-devel-144
Noel Power [Mon, 28 Jan 2019 15:23:59 +0000 (15:23 +0000)]
Fix instances of PyDict_SetItem to decref the value
Although it would be better to use the BuildValue approach to
create the dictionares here, unfortunately the dictionaries created
here have key/values that are created dynamically (based on input params).
Signed-off-by: Noel Power <noel.power@suse.com> Reviewed-by: Andrew Bartlett abartlet@samba.org
As the Heimdal build does not need this, only specify for the MIT build.
Tested with an MIT AD DC build with:
make test TESTS="samba3.raw.session samba3.smb2.session"
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Gary Lockyer <gary@catalyst.net.nz>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Wed Feb 13 05:49:43 CET 2019 on sn-devel-144
Tim Beale [Mon, 4 Feb 2019 23:23:43 +0000 (12:23 +1300)]
selftest: Change backup/restore testenvs to use 1 prefork child
Recently the gitlab CI jobs were hitting memory resource limits and
using swap, which then caused test failures. The process model used in
the testenvs seemed to be contributing to this problem.
We can reduce the memory overhead of the restore/backup testenvs by
using 1 prefork child process instead of the default of 4 (kudos to
Garming for the idea). The tests run against these testenvs are basic
sanity-checks, rather than heavy-duty stress tests, so the number of
prefork workers shouldn't matter.
This is a bit of a tradeoff between testing the defaults that will
actually be used in production vs using limited resources efficiently on
shared CI runner machines.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Wed, 9 Jan 2019 02:12:43 +0000 (15:12 +1300)]
dsdb: make get_parsed_dns_trusted() a common helper function
We are already using it in two places, and are about to add a third.
The version in repl_meta_data.c did more work in the case that the
parsed_dns can't really be trusted to conform to the expected format;
this is now a wrapper called get_parsed_dns_trusted_fallback().
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
s3:utils: Add missing NULL check in rpc_fetch_domain_aliases()
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): Wed Feb 13 00:52:25 CET 2019 on sn-devel-144
Ralph Boehme [Wed, 23 Jan 2019 08:43:33 +0000 (09:43 +0100)]
CI: split out "samba-ad-dc-ntvfs[-py2]" test targets
Many AD tests currently use the "samba" target. Split out a new target
"samba-ad-dc-ntvfs" and have all tests that use the "ad_dc_ntvfs" env
use the new target. This should greatly speed up the runtime for the "samba"
target and avoid swapping.
This reduces the total CI time by ~ 55%, I got an autobuild and a gitlab
pipeline finished in just ~ 100 mins!
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org> Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Mon Feb 11 14:10:12 CET 2019 on sn-devel-144
Tim Beale [Sun, 3 Feb 2019 20:28:07 +0000 (09:28 +1300)]
selftest: Make dns_hub socket timeout match DNS_REQUEST_TIMEOUT
I was hitting the recv_packet = s.recv(2048, 0) exception because
the socket timeout was reached. We've seen it before, but it seemed more
common after changing the default process-model to prefork. This patch
makes the socket timeout used by the python code consistent with the C
code.
Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
docs-xml/smbdotconf: document export of SAMBA_CPS_{ACCOUNT,USER_PRINCIPAL,FULL}_NAME for check password script
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Mon Feb 11 11:03:58 CET 2019 on sn-devel-144
s3:modules: Fix compilation of nfs41acl_xdr.c when building outside src
If the Samba build directory is outside its source directory, generation
of nfs41acl_xdr.c by rpcgen leads to improper include paths to nfs41acl.h.
This happens because rpcgen is designed to produce its generated file in the
same directory as the input template. If the build directory is not located
under the source directory, this relative path will be invalid and the header
will not be found.
Example:
src dir is ~/samba-src
bld dir is ~/samba-bld
rpcgen will use path ../../samba-src/source3/modules/nfs41acl.x
running from ~/samba-bld/default and nfs41acl_xdr.c will contain:
#include "../../samba-src/source3/modules/nfs41acl.h"
This behaviour is fixed through an intermediate copy of the input file to
the build directory so that rpcgen receives the path as if located in src.
Also now we avoid generation of nfs41acl_xdr.c when HAVE_RPC_XDR_H is
not defined because it will not be used as part of the vfs_nfs4acl_xattr
module.
Signed-off-by: Aliaksei Karaliou <akaraliou@panasas.com> Reviewed-by: Garming Sam <garming@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
s3:util: Move static file_pload() function to lib/util
file_pload() is static private function in Samba3 library, however it
does not have any special dependencies and might be widely used as
common function, so moving it into common samba-util library.
Signed-off-by: Aliaksei Karaliou <akaraliou@panasas.com> Reviewed-by: Garming Sam <garming@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Garming Sam [Fri, 1 Feb 2019 01:11:18 +0000 (14:11 +1300)]
dnsserver: Return access denied to the caller if the user was not a DNS admin
This is not a proper fix to match Windows, but at the very least, it
should be more obvious to users (using samba-tool for instance), that
the user needs to be given more access or that they should use the
administrator.
Windows seems to deny access altogether by returning a fault after they
have bound to the pipe and actually sent an operation.
For more info on Wireshark decryption support see
https://wiki.samba.org/index.php/Wireshark_Decryption
Signed-off-by: Aurelien Aptel <aaptel@suse.com> Reviewed-by: Noel Power <npower@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): David Disseldorp <ddiss@samba.org>
Autobuild-Date(master): Sat Feb 9 21:43:25 CET 2019 on sn-devel-144
s3:vfs: Correctly check if OFD locks should be enabled or not
Also the smb.conf options should only be checked once and a reload of
the config should not switch to a different locking mode.
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): Sat Feb 9 03:43:50 CET 2019 on sn-devel-144
Jeremy Allison [Thu, 7 Feb 2019 02:01:52 +0000 (18:01 -0800)]
s4: torture: vfs_fruit. Change test_fruit_locking_conflict() to match the vfs_fruit working server code.
Originally added for BUG: https://bugzilla.samba.org/show_bug.cgi?id=13584
to demonstrate a lock order violation, this test
exposed problems in the mapping of SMB1/2 share modes
and open modes to NetATalk modes once we moved to OFD locks.
Change the test slightly (and add comments)
so it demonstrates working NetATalk share modes
on an open file.
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Ralph Böhme <slow@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Fri Feb 8 23:26:46 CET 2019 on sn-devel-144
However, due to underlying bugs in the vfs_fruit
code the file locks were not being properly applied.
There are two problems in fruit_check_access().
Problem #1:
Inside fruit_check_access() we have:
flags = fcntl(fsp->fh->fd, F_GETFL);
..
if (flags & (O_RDONLY|O_RDWR)) {
We shouldn't be calling fcntl(fsp->fh->fd, ..) directly.
fsp->fh->fd may be a made up number from an underlying
VFS module that has no meaning to a system call.
Secondly, in all POSIX systems - O_RDONLY is defined as
*zero*. O_RDWR = 2.
Which means flags & (O_RDONLY|O_RDWR) becomes (flags & 2),
not what we actually thought.
Problem #2:
deny_mode is *not* a bitmask, it's a set of discrete values.
Inside fruit_check_access() we have:
if (deny_mode & DENY_READ) and also (deny_mode & DENY_WRITE)
so if deny_mode = DENY_WRITE, or if deny_mode = DENY_READ
then it's going to trigger both the if (deny_mode & DENY_READ)
*and* the (deny_mode & DENY_WRITE) conditions.
These problems allowed the original test test_netatalk_lock code to
pass (which was added for BUG: https://bugzilla.samba.org/show_bug.cgi?id=13584
to demonstrate the lock order violation).
This patch refactors the fruit_check_access()
code to be much simpler (IMHO) to understand.
Firstly, pass in the SMB1/2 share mode, not old
DOS deny modes.
Secondly, read all the possible NetAtalk locks
into local variables:
Then do the share mode/access mode checks
with the requested values against any stored
netatalk modes/access modes.
Finally add in NetATalk compatible locks
that represent our share modes/access modes
into the file, with an early return if we don't
have FILE_READ_DATA (in which case we can't
write locks anyway).
The patch is easier to understand by looking
at the completed patched fruit_check_access()
function, rather than trying to look at the
diff.
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Ralph Böhme <slow@samba.org>
Douglas Bagnall [Thu, 7 Feb 2019 04:36:02 +0000 (17:36 +1300)]
s4/registry/py: use unsigned ParseTuple format for unsigned value
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Fri Feb 8 17:09:51 CET 2019 on sn-devel-144
Douglas Bagnall [Thu, 7 Feb 2019 04:04:43 +0000 (17:04 +1300)]
s3/libsmb/py: match input argument types with C types
If PyArg_ParseTupleAndKeywords() is given, say, an "H" format (meaning
unsigned short int) but the referenced variable is a plain unsigned
int, the top 16 bits of the variable will be left undefined. In that
case we should use an "I" format (and/or initialize the variable).
In many cases the change is fairly innocuous, such as when "i" and "I"
are mixed (for signed and unsigned ints respectively), but the
resulting write is the same size and probably gives the same result in
practice.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Noel Power <npower@samba.org>
projects / thirdparty / samba.git / log
