Andrew Bartlett [Tue, 12 Sep 2023 00:28:49 +0000 (12:28 +1200)]
CVE-2023-42670 s3-rpc_server: Strictly refuse to start RPC servers in conflict with AD DC
Just as we refuse to start NETLOGON except on the DC, we must refuse
to start all of the RPC services that are provided by the AD DC.
Most critically of course this applies to netlogon, lsa and samr.
This avoids the supression of these services being the result of a
runtime epmapper lookup, as if that fails these services can disrupt
service to end users by listening on the same socket as the AD DC
servers.
Andrew Bartlett [Tue, 12 Sep 2023 07:01:03 +0000 (19:01 +1200)]
CVE-2023-42669 s3-rpc_server: Disable rpcecho for consistency with the AD DC
The rpcecho server in source3 does have samba the sleep() feature that
the s4 version has, but the task architecture is different, so there
is not the same impact. Hoever equally this is not something that
should be enabled on production builds of Samba, so restrict to
selftest builds.
Andrew Bartlett [Tue, 12 Sep 2023 06:59:44 +0000 (18:59 +1200)]
CVE-2023-42669 s4-rpc_server: Disable rpcecho server by default
The rpcecho server is useful in development and testing, but should never
have been allowed into production, as it includes the facility to
do a blocking sleep() in the single-threaded rpc worker.
Andrew Bartlett [Tue, 8 Aug 2023 05:58:27 +0000 (17:58 +1200)]
CVE-2023-4154: Unimplement the original DirSync behaviour without LDAP_DIRSYNC_OBJECT_SECURITY
This makes LDAP_DIRSYNC_OBJECT_SECURITY the only behaviour provided by
Samba.
Having a second access control system withing the LDAP stack is unsafe
and this layer is incomplete.
The current system gives all accounts that have been given the
GUID_DRS_GET_CHANGES extended right SYSTEM access. Currently in Samba
this equates to full access to passwords as well as "RODC Filtered
attributes" (often used with confidential attributes).
Rather than attempting to correctly filter for secrets (passwords) and
these filtered attributes, as well as preventing search expressions for
both, we leave this complexity to the acl_read module which has this
facility already well tested.
The implication is that callers will only see and filter by attribute
in DirSync that they could without DirSync.
The aim here is to document the expected (even if not implemented)
SEARCH_FLAG_RODC_ATTRIBUTE vs SEARCH_FLAG_CONFIDENTIAL, behaviour, so
that any change once CVE-2023-4154 is fixed can be noted.
Andrew Bartlett [Tue, 8 Aug 2023 02:30:19 +0000 (14:30 +1200)]
CVE-2023-4154 dsdb/tests: Add test for SEARCH_FLAG_RODC_ATTRIBUTE behaviour
SEARCH_FLAG_RODC_ATTRIBUTE should be like SEARCH_FLAG_CONFIDENTIAL,
but for DirSync and DRS replication. Accounts with
GUID_DRS_GET_CHANGES rights should not be able to read this
attribute.
Andrew Bartlett [Mon, 7 Aug 2023 23:18:46 +0000 (11:18 +1200)]
CVE-2023-4154 dsdb/tests: Speed up DirSync test by only checking positive matches once
When we (expect to) get back a result, do not waste time against a potentially
slow server confirming we also get back results for all the other attribute
combinations.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 2c02378029fff6636b8f19e45af78b265f2210ed)
CVE-2023-4154 libcli/security: prepare security_descriptor_acl_add() to place the ace at a position
Often it is important to insert an ace at a specific position in the
ACL. As a default we still append by default by using -1, which is the
generic version of passing the number of existing aces.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit c3cb915a67aff6739b72b86d7d139609df309ada)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 9d8ff0d1e0b2ba7c84af36e1931f5bc99902a44b)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 4627997ddae44265ad35b3234232eb74458c6c34)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 9ea06aaf9f57e3c7094553d9ac40fb73057a9b74)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit a1109a9bf12e020636b8d66fc54984aac58bfe6b)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 8411e6d302e25d10f1035ebbdcbde7308566e930)
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 8b4e6f7b3fb8018cb64deef9b8e1cbc2e5ba12cf)
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
(cherry picked from commit b29793ffdee5d9b9c1c05830622e80f7faec7670)
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 2e5d08c908b3fa48b9b374279a331061cb77bce3)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
(cherry picked from commit 82d2ec786f7e75ff6f34eb3357964345b10de091)
Ralph Boehme [Tue, 1 Aug 2023 11:04:36 +0000 (13:04 +0200)]
CVE-2023-4091: smbd: use open_access_mask for access check in open_file()
If the client requested FILE_OVERWRITE[_IF], we're implicitly adding
FILE_WRITE_DATA to the open_access_mask in open_file_ntcreate(), but for the
access check we're using access_mask which doesn't contain the additional
right, which means we can end up truncating a file for which the user has
only read-only access via an SD.
Joseph Sutton [Mon, 4 Sep 2023 01:20:34 +0000 (13:20 +1200)]
s4:kdc: Add correct Asserted Identity SID in response to an S4U2Self request
I’m not sure exactly how this check was supposed to work. But in any
case, within fast_unwrap_request() the Heimdal KDC replaces the outer
padata with the padata from the inner FAST request. Hence, this check
does not accomplish anything useful: at no point should the KDC plugin
see the outer padata.
A couple of unwanted consequences resulted from this check. One was that
a client who sent empty FX‐FAST padata within the inner FAST request
would receive the *Authentication Authority* Asserted Identity SID
instead of the *Service* Asserted Identity SID. Another consequence was
that a client could in the same manner bypass the restriction on
performing S4U2Self with an RODC‐issued TGT.
Overall, samba_wdc_is_s4u2self_req() is somewhat of a hack. But the
Heimdal plugin API gives us nothing better to work with.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 5c580dbdb3e6a70c8d2f5059e2b7293a7e780414)
Joseph Sutton [Thu, 16 Mar 2023 20:16:17 +0000 (09:16 +1300)]
s4:kdc: Avoid copying data if not needed
krb5_pac_add_buffer() makes its own copy of the data we pass in. We
don't need to make yet another copy.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit fa901e7346d36ae64a7ceab5dcf76bc210a67c93)
Joseph Sutton [Thu, 16 Mar 2023 20:25:52 +0000 (09:25 +1300)]
s4:kdc: Don't pass a NULL pointer into krb5_pac_add_buffer()
Heimdal contains an assertion that the data pointer is not NULL. We need
to pass in a pointer to some dummy data instead.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 47ef49fd91f050ce4a79a8471b3e66c808f48752)
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Sep 20 02:43:18 UTC 2023 on atb-devel-224
nsswitch/wb_common.c: fix socket fd and memory leaks of global state
When we are called in wb_atfork_child() or winbind_destructor(),
wb_thread_ctx_destructor() is not called for the global state
of the current nor any other thread, which means we would
leak the related memory and socket fds.
Now we maintain a global list protected by a global mutex.
We traverse the list and close all socket fds, which are no
longer used (winbind_destructor) or no longer valid in the
current process (wb_atfork_child), in addition we 'autofree'
the ones, which are only visible internally as global (per thread)
context.
Tested-by: Krzysztof Piotr Oledzki <ole@ans.pl> 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): Thu Sep 14 18:53:07 UTC 2023 on atb-devel-224
nsswitch/wb_common.c: don't operate on a stale wb_global_ctx.key
If nss_winbind is loaded into a process that uses fork multiple times
without any further calls into nss_winbind, wb_atfork_child handler
was using a wb_global_ctx.key that was no longer registered in the
pthread library, so we operated on a slot that was potentially
reused by other libraries or the main application. Which is likely
to cause memory corruption.
So we better don't call pthread_key_delete() in wb_atfork_child().
Reported-by: Krzysztof Piotr Oledzki <ole@ans.pl> Tested-by: Krzysztof Piotr Oledzki <ole@ans.pl> Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 91b30a7261e6455d3a4f31728c23e4849e3945b9)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 836823e5047d0eb18e66707386ba03b812adfaf8)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 4faf806412c4408db25448b1f67c09359ec2f81f)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 62af25d44e542548d8cdecb061a6001e0071ee76)
Ralph Boehme [Tue, 29 Aug 2023 14:14:38 +0000 (16:14 +0200)]
mdssvc: better support for search with mdfind from Macs
When searching Samba via Spotlight from a Mac with mdfind, only 50 results are
returned.
It seems the changes for bug #15342 where one step in the right
direction. There, a status indicator meaning "search is still being processed"
was implemented, returning a special status indicator (0x23) in a response, when
the query was still running in the backend, eg Elasticsearch, and we haven't got
any result when when the Mac already comes along asking for results of a query.
Turns out, we should also return 0x23, ie "search is still being processed" when
we have some initial search results from the backend. Otherwise mdfind will stop
querying for more results. It works in Finder, as the Finder by default employs
a "live" search where it just keeps polling for more results even after the
server returned an empty result set. The Finder just keeps on querying in some
interval, typically 4 seconds, and a Mac server Spotlight server might return
new results if new files where created that match the query, hence "live"
search.
Signed-off-by: Ralph Boehme <slow@samba.org> Reviewed-by: Noel Power <noel.power@suse.com>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Wed Aug 30 10:51:40 UTC 2023 on atb-devel-224
MikeLiu [Tue, 22 Aug 2023 02:01:14 +0000 (10:01 +0800)]
vfs_aio_pthread: use SMB_VFS_NEXT_OPENAT() in aio_pthread_openat_fn()
1. Set 'aio_allow_open' to false if fsp->fsp_flags.is_pathref
2. Move !(how->flags & O_CREAT) and !(how->flags & O_EXCL) up and set 'aio_allow_open' to false
3. Use SMB_VFS_NEXT_OPENAT() instead of openat() for disable async opens case.
Signed-off-by: MikeLiu <mikeliu@qnap.com> Reviewed-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Aug 22 17:44:00 UTC 2023 on atb-devel-224
Autobuild-User(v4-18-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-18-test): Mon Sep 4 09:49:30 UTC 2023 on atb-devel-224
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): Thu Jul 27 10:52:50 UTC 2023 on atb-devel-224
Martin Schwenke [Tue, 15 Aug 2023 02:34:20 +0000 (12:34 +1000)]
ctdb-common: Set immediate mode for pcap capture
Fix a problem where ctdb_killtcp (almost always) fails to capture
packets with --enable-pcap and libpcap ≥ 1.9.1. The problem is due to
a gradual change in libpcap semantics when using
pcap_get_selectable_fd(3PCAP) to get a file descriptor and then using
that file descriptor in non-blocking mode.
pcap_set_immediate_mode(3PCAP) says:
pcap_set_immediate_mode() sets whether immediate mode should be set
on a capture handle when the handle is activated. In immediate
mode, packets are always delivered as soon as they arrive, with no
buffering.
and
On Linux, with previous releases of libpcap, capture devices are
always in immediate mode; however, in 1.5.0 and later, they are, by
default, not in immediate mode, so if pcap_set_immediate_mode() is
available, it should be used.
However, it wasn't until libpcap commit 2ade7676101366983bd4f86bc039ffd25da8c126 (before libpcap 1.9.1) that
it became a requirement to use pcap_set_immediate_mode(), even with a
timeout of 0.
Martin Schwenke [Tue, 15 Aug 2023 00:57:59 +0000 (10:57 +1000)]
ctdb-common: Replace pcap_open_live() by lower level calls
A subsequent commit will insert an additional call before
pcap_activate().
This sequence of calls is taken from the source for pcap_open_live(),
so there should be no change in behaviour.
Given the defaults set by pcap_create_common(), it would be possible
to omit the calls to pcap_set_promisc() and pcap_set_timeout().
However, those defaults don't seem to be well documented, so continue
to explicitly set everything that was set before.
Andrew Bartlett [Wed, 26 Jul 2023 02:27:16 +0000 (14:27 +1200)]
s4-rpc_server/drsupai: Avoid looping with Azure AD Connect by not incrementing temp_highest_usn for the NC root
We send the NC root first, as a special case for every chunk
that we send until the natural point where it belongs.
We do not bump the tmp_highest_usn in the highwatermark that
the client and server use (it is meant to be an opauqe cookie)
until the 'natural' point where the object appears, similar
to the cache for GET_ANC.
The issue is that without this, because the NC root was sorted
first in whatever chunk it appeared in but could have a 'high'
highwatermark, Azure AD Connect will send back the same
new_highwatermark->tmp_highest_usn, and due to a bug,
a zero reserved_usn, which makes Samba discard it.
The reserved_usn is now much less likely to ever be set because
the tmp_higest_usn is now always advancing.
RN: Avoid infinite loop in initial user sync with Azure AD Connect
when synchronising a large Samba AD domain.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 79ca6ef28a6f94965cb030c4a7da8c1b9db7150b)
Autobuild-User(v4-18-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-18-test): Mon Aug 21 09:12:14 UTC 2023 on atb-devel-224
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 17359afa627a3086ec8d6862f007a3479574a8b4)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 2aba9e230ea62efcbd829f6f073894dfa3180c91)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 2ed9815eeacfcf3a58871bafe0212398cc34c39e)
Andrew Bartlett [Tue, 27 Jun 2023 02:39:18 +0000 (14:39 +1200)]
s4-rpc_server/drsuapi: Avoid modification to ncRoot input variable in GetNCChanges
This tries to avoid it appearing that ncRoot is a value that can
be trusted and used internally by not updating it and instead leaving
it just as an input/echo-back value.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 548f141f11e89d335d8f9d74ab6925fa6b90fb84)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit fe7418e1765b79f60945b787536b4d84a548fe02)
Andrew Bartlett [Mon, 26 Jun 2023 04:53:10 +0000 (16:53 +1200)]
s4-rpc_server/drsuapi: Only keep and invalidate replication cycle state for normal replication
This changes the GetNCChanges server to use a per-call state for
extended operations like RID_ALLOC or REPL_OBJ and only maintain
and (more importantly) invalidate the state during normal replication.
This allows REPL_OBJ to be called during a normal replication cycle
that continues using after that call, continuing with the same
highwatermark cookie.
Azure AD will do a sequence of (roughly)
* Normal replication (objects 1..100)
* REPL_OBJ (of 1 object)
* Normal replication (objects 101..200)
However, if there are more than 100 (in this example) objects in the
domain, and the second replication is required, the objects 1..100
are sent, as the replication state was invalidated by the REPL_OBJ call.
RN: Improve GetNChanges to address some (but not all "Azure AD Connect")
syncronisation tool looping during the initial user sync phase.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 99579e706312192f46df33d55949db7f1475d0d0)
Andrew Bartlett [Mon, 24 Jul 2023 00:05:18 +0000 (12:05 +1200)]
s4-torture/drs: Add test showing that if present in the set the NC root leads and tmp_highest_usn moves
The NC root, on any replication when it appears, is the first object to be
replicated, including for all subsequent chunks in the replication.
However the tmp_highest_usn is not updated by that USN, it must
only be updated for the non-NC changes (to match Windows exactly),
or at least only updated with the non-NC changes until it would
naturally appear.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 87414955212143b8502b4c02aca150bc72cb8de5)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit b323169d6ff8357f7c999ae346137166c98218ac)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit db16366b0bbefcdb91a0b36c903ed63456a081b8)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 40f831e67e1f312b1db52c74c119899245d03e32)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 628eab11b3c2e82875bf602e363b781d3e5eb96d)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit c30bb8769ff2c4eba2d8f8a2bd3a56946b7d9d5e)
Andrew Bartlett [Tue, 27 Jun 2023 02:22:52 +0000 (14:22 +1200)]
s4-rpc_server/drsuapi: Remove rudundant check for valid and non-NULL ncRoot_dn
This check was valuable before aee2039e63ceeb5e69a0461fb77e0f18278e4dc4
but now only checks things we know to be true, as the value has come
from Samba via drs_ObjectIdentifier_to_dn_and_nc_root() either on this
or a previous call.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 0550e469eda4022659718ae9a56f5deaa9f9a307)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 63843a22c8db73d459bee61e73bb1f0d31e3d427)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit a12bcce89d26ae05bbaeed560cf8fcc7b5bcfdab)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit d0c1ce53add2fd3b3a4186581f4e214029cbcf1a)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 0d9ea6c559317e19642662220c089e2d59ef3ecd)
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Tue Aug 15 12:06:36 UTC 2023 on atb-devel-224
Jeremy Allison [Sat, 12 Aug 2023 00:14:38 +0000 (17:14 -0700)]
s3: smbd: init_smb1_request() isn't being passed zero'ed memory from any codepath.
If a client does a SMB1 NEGPROT followed by SMB1 TCON
then req->session is left uninitialized.
Show this causes a crash by deliberately initializing
req->session to an invalid pointer. This will be removed
once the test shows the crash, and the fix is added to
cause init_smb1_request() to zero the memory passed in.
Signed-off-by: Jeremy Allison <jra@samba.org> Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Mon Aug 14 19:52:49 UTC 2023 on atb-devel-224
Jeremy Allison [Fri, 11 Aug 2023 17:42:41 +0000 (10:42 -0700)]
s3: smbd: Ensure srvstr_pull_req_talloc() always NULLs out *dest.
Robert Morris <rtm@lcs.mit.edu> noticed that in the case
where srvstr_pull_req_talloc() is being called with
buffer remaining == 0, we don't NULL out the destination
pointed which is *always* done in the codepaths inside
pull_string_talloc(). This prevents a crash in the caller.
Jones Syue [Mon, 7 Aug 2023 09:08:29 +0000 (17:08 +0800)]
mdssvc: Do an early talloc_free() in _mdssvc_open()
Environment setup:
When macOS Finder connect to a samba server with 'spotlight = yes',
macOS would issue mdssvc open (mdssvc.opnum == 0) to samba and it goes
through api _mdssvc_open().
After applied 578e434a94147dc2d7dbfc006d2ab84807859c1d,
(this is reported by jaywei@qnap.com)
this line 'talloc_free(path);' is deleted if _mdssvc_open() normal exit,
so memory is lazy de-allocate: delayed to
smbd_tevent_trace_callback() @ smb2_process.c. [1]
Supposed to explicitly free 'path' in _mdssvc_open() @ srv_mdssvc_nt.c[2]
just like abnormal exit, do not wait for main loop to free 'path' which is
no longer used, this is more consistent while reading source code.
[1] gdb tracing 'path' address 0x56204ccc67e0 to know how it is freed.
Breakpoint 2, _tc_free_children_internal (tc=0x56204ccc6780, ptr=0x56204ccc67e0, location=0x7ff430d96410 "../../lib/talloc/talloc.c:1714") at ../../lib/talloc/talloc.c:1656
1656 while (tc->child) {
(gdb) bt
0 _tc_free_children_internal (tc=0x56204ccc6780, ptr=0x56204ccc67e0, location=0x7ff430d96410 "../../lib/talloc/talloc.c:1714") at ../../lib/talloc/talloc.c:1656
1 0x00007ff430d92b14 in _tc_free_internal (tc=0x56204ccc6780, location=0x7ff430d96410 "../../lib/talloc/talloc.c:1714") at ../../lib/talloc/talloc.c:1183
2 0x00007ff430d93b71 in _tc_free_children_internal (tc=0x56204ccc6720, ptr=0x56204ccc6780, location=0x7ff430d96410 "../../lib/talloc/talloc.c:1714") at ../../lib/talloc/talloc.c:1668
3 0x00007ff430d93d66 in talloc_free_children (ptr=0x56204ccc6780) at ../../lib/talloc/talloc.c:1714
4 0x00007ff432235aca in talloc_pop (frame=0x56204ccc6780) at ../../lib/util/talloc_stack.c:125
5 0x00007ff430d92959 in _tc_free_internal (tc=0x56204ccc6720, location=0x7ff431f358d0 "../../source3/smbd/process.c:3726") at ../../lib/talloc/talloc.c:1157
6 0x00007ff430d92cd5 in _talloc_free_internal (ptr=0x56204ccc6780, location=0x7ff431f358d0 "../../source3/smbd/process.c:3726") at ../../lib/talloc/talloc.c:1247
7 0x00007ff430d93f96 in _talloc_free (ptr=0x56204ccc6780, location=0x7ff431f358d0 "../../source3/smbd/process.c:3726") at ../../lib/talloc/talloc.c:1791
8 0x00007ff431d81292 in smbd_tevent_trace_callback (point=TEVENT_TRACE_AFTER_LOOP_ONCE, private_data=0x7ffe46591e30) at ../../source3/smbd/process.c:3726
<...cut...>
[2] gdb tracing 'path' address 0x55a6d66deed0 to know how it is freed.
Breakpoint 2, _tc_free_children_internal (tc=0x55a6d66deed0, ptr=0x55a6d66def30, location=0x7fc4cca84040 "../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189") at ../../lib/talloc/talloc.c:1656
1656 while (tc->child) {
(gdb) bt
0 _tc_free_children_internal (tc=0x55a6d66deed0, ptr=0x55a6d66def30, location=0x7fc4cca84040 "../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189") at ../../lib/talloc/talloc.c:1656
1 0x00007fc4cb892b14 in _tc_free_internal (tc=0x55a6d66deed0, location=0x7fc4cca84040 "../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189") at ../../lib/talloc/talloc.c:1183
2 0x00007fc4cb892cd5 in _talloc_free_internal (ptr=0x55a6d66def30, location=0x7fc4cca84040 "../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189") at ../../lib/talloc/talloc.c:1247
3 0x00007fc4cb893f96 in _talloc_free (ptr=0x55a6d66def30, location=0x7fc4cca84040 "../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189") at ../../lib/talloc/talloc.c:1791
4 0x00007fc4cc9396e4 in _mdssvc_open (p=0x55a6d66d5600, r=0x55a6d66edc60) at ../../source3/rpc_server/mdssvc/srv_mdssvc_nt.c:189
<...cut...>
Signed-off-by: Jones Syue <jonessyue@qnap.com> Reviewed-by: Noel Power <npower@samba.org> Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Mon Aug 14 18:11:37 UTC 2023 on atb-devel-224
If a client opens multiple connection with the same
client guid in parallel, our connection passing is likely
to hit a race.
Assume we have 3 processes:
smbdA: This process already handles all connections for
a given client guid
smbdB: This just received a new connection with an
SMB2 neprot for the same client guid
smbdC: This also received a new connection with an
SMB2 neprot for the same client guid
Now both smbdB and smbdC send a MSG_SMBXSRV_CONNECTION_PASS
message to smbdA. These messages contain the socket fd
for each connection.
While waiting for a MSG_SMBXSRV_CONNECTION_PASSED message
from smbdA, both smbdB and smbdC watch the smbXcli_client.tdb
record for changes (that also verifies smbdA stays alive).
Once one of them say smbdB received the MSG_SMBXSRV_CONNECTION_PASSED
message, the dbwrap_watch logic will wakeup smbdC in order to
let it recheck the smbXcli_client.tdb record in order to
handle the case where smbdA died or deleted its record.
Now smbdC rechecks the smbXcli_client.tdb record, but it
was not woken because of a problem with smbdA. It meant
that smbdC sends a MSG_SMBXSRV_CONNECTION_PASS message
including the socket fd again.
As a result smbdA got the socket fd from smbdC twice (or even more),
and creates two (or more) smbXsrv_connection structures for the
same low level tcp connection. And it also sends more than one
SMB2 negprot response. Depending on the tevent logic, it will
use different smbXsrv_connection structures to process incoming
requests. And this will almost immediately result in errors.
The typicall error is:
smb2_validate_sequence_number: smb2_validate_sequence_number: bad message_id 2 (sequence id 2) (granted = 1, low = 1, range = 1)
But other errors would also be possible.
The detail that leads to the long delays on the client side is
that our smbd_server_connection_terminate_ex() code will close
only the fd of a single smbXsrv_connection, but the refcount
on the socket fd in the kernel is still not 0, so the tcp
connection is still alive...
Now we remember the server_id of the process that we send
the MSG_SMBXSRV_CONNECTION_PASS message to. And just keep
watching the smbXcli_client.tdb record if the server_id
don't change. As we just need more patience to wait for
the MSG_SMBXSRV_CONNECTION_PASSED message.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Tue Aug 8 13:59:58 UTC 2023 on atb-devel-224
Moving it just above a possible messaging_filtered_read_send()
will allow us to only clear it if we actually create a new
request. That will help us in the next commits.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 50d61e5300250922bf36bb699306f82dff6a00b9)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 4028d6582907cf582730ceec56872d8584ad02e6)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 2b93058be3f6e5eaee239ad3b0e707c62089d18e)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit dc5a500f0a76720b2a5cb5b1142cf4c35cb6bdea)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit ade663ee6ca1a2813b203ea667d933f4dab9e7b7)
librpc/rpc: let dcerpc_read_ncacn_packet_next_vector() handle fragments without any payload
DCERPC_PKT_CO_CANCEL and DCERPC_PKT_ORPHANED don't have any payload by
default. In order to receive them via dcerpc_read_ncacn_packet_send/recv
we need to allow fragments with frag_len == DCERPC_NCACN_PAYLOAD_OFFSET.
Jones Syue [Wed, 2 Aug 2023 01:48:40 +0000 (09:48 +0800)]
vfs_aio_pthread: fix segfault if samba-tool ntacl get
If configured as AD DC and aio_pthread appended into 'vfs objects'[1],
run these commands would get segfault:
1. sudo samba-tool ntacl get .
2. sudo net vfs getntacl sysvol .
gdb said it goes through aio_pthread_openat_fn() @ vfs_aio_pthread.c[2],
and the fsp->conn->sconn->client is null (0x0).
'sconn->client' memory is allocated when a new connection is accpeted:
smbd_accept_connection > smbd_process > smbXsrv_client_create
While running local commands looks like it would not go through
smbXsrv_client_create so the 'client' is null, segfault might happen.
We should not dereference 'client->server_multi_channel_enabled',
if 'client' is null.
[1] smb.conf example, samba-4.18.5, ubuntu 22.04.2
[global]
dns forwarder = 127.0.0.53
netbios name = U22-JONES-88X1
realm = U22-JONES-88X1.X88X1.JONES
server role = active directory domain controller
workgroup = X88X1
idmap_ldb:use rfc2307 = yes
vfs objects = dfs_samba4 acl_xattr aio_pthread
[sysvol]
path = /var/lib/samba/sysvol
read only = No
[netlogon]
path = /var/lib/samba/sysvol/u22-jones-88x1.x88x1.jones/scripts
read only = No
[2] gdb
(gdb) run /usr/local/samba/bin/samba-tool ntacl get .
Starting program: /usr/local/Python3/bin/python3 /usr/local/samba/bin/samba-tool ntacl get .
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/libthread_db.so.1".
Program received signal SIGSEGV, Segmentation fault.
0x00007fffd0eb809e in aio_pthread_openat_fn (handle=0x8d5cc0, dirfsp=0x8c3070, smb_fname=0x18ab4f0, fsp=0x1af3550, flags=196608, mode=0)
at ../../source3/modules/vfs_aio_pthread.c:467
warning: Source file is more recent than executable.
467 if (fsp->conn->sconn->client->server_multi_channel_enabled) {
(gdb) bt
at ../../source3/modules/vfs_aio_pthread.c:467
at ../../source3/smbd/pysmbd.c:320
---Type <return> to continue, or q <return> to quit---
(gdb) f
at ../../source3/modules/vfs_aio_pthread.c:467
467 if (fsp->conn->sconn->client->server_multi_channel_enabled) {
(gdb) p fsp->conn->sconn->client
$1 = (struct smbXsrv_client *) 0x0
(gdb)
Signed-off-by: Jones Syue <jonessyue@qnap.com> Reviewed-by: Ralph Boehme <slow@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 8f4c1c67b4f118a9a47b09ac7908cd3d969b19c2)
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 3694f2ce6205a647eb5dab2115785fb45decaf0b)
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Fri Jul 21 06:16:30 UTC 2023 on atb-devel-224
RN: post-exec password redaction for samba-tool is more reliable for
fully random passwords as it no longer uses regular expressions
containing the password value itself.
Autobuild-User(v4-18-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-18-test): Fri Aug 4 08:05:00 UTC 2023 on atb-devel-224
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit 414b3803bb6a1b12c44b52ab1ff64a8b7f61fd03)
Andrew Bartlett [Fri, 21 Jul 2023 01:30:39 +0000 (13:30 +1200)]
samba-tool: Use samba.glue.get_burnt_cmdline rather than regex
This use avoids having two different methods to match on command-line
passwords. We already have a dependency on the setproctitle python
module, and this does not change as the (C) libbsd setproctitle()
can't be run from within a python module.
Signed-off-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
(cherry picked from commit a53ebc288f47329c997d52325eeeb5e91ce43b75)
projects / thirdparty / samba.git / log
