s3:utils: let smbstatus report anonymous signing/encryption explicitly
We should mark sessions/tcons with anonymous encryption or signing
in a special way, as the value of it is void, all based on a
session key with 16 zero bytes.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Günther Deschner <gd@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Thu May 23 13:37:09 UTC 2024 on atb-devel-224
s3:smbd: allow anonymous encryption after one authenticated session setup
I have captures where a client tries smb3 encryption on an anonymous session,
we used to allow that before commit da7dcc443f45d07d9963df9daae458fbdd991a47
was released with samba-4.15.0rc1.
Testing against Windows Server 2022 revealed that anonymous signing is always
allowed (with the session key derived from 16 zero bytes) and
anonymous encryption is allowed after one authenticated session setup on
the tcp connection.
https://bugzilla.samba.org/show_bug.cgi?id=15412
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Günther Deschner <gd@samba.org>
Douglas Bagnall [Wed, 22 May 2024 21:40:00 +0000 (09:40 +1200)]
ldb: move struct ldb_debug_ops to ldb_private.h
Only accessed through struct ldb_context -> debug_ops, which is already private.
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): Thu May 23 00:19:30 UTC 2024 on atb-devel-224
Douglas Bagnall [Sun, 12 May 2024 22:39:44 +0000 (10:39 +1200)]
lib/fuzzing: add fuzz_strncasecmp_ldb
As well as checking for the usual overflows, this asserts that
strncasecmp_ldb is always transitive, by splitting the input into 3
pieces and comparing all pairs.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Thu, 16 May 2024 02:09:46 +0000 (14:09 +1200)]
ldb-samba: use ldb_comparison_fold_utf8()
This means ldb-samba/dsdb comparisons will be case-insensitive for
non-ASCII UTF-8 characters (within the bounds of the 16-bit casefold
table). And they will remain transitive.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Sun, 12 May 2024 23:32:26 +0000 (11:32 +1200)]
util:charset: strncasecmp_ldb degrades to ASCII strncasecmp
If strncasecmp_ldb() encounters invalid utf-8 bytes, it compares those
as greater than any valid bytes (that is, it sorts them to the end of
the list).
If an invalid sequence is encountered in both strings at once, the
rest of the strings are now compared using the default ldb_comparison_fold
rules, as implemented in ldb_comparison_fold_ascii(). That is, each
byte is compared individually, [a-z] are translated to [A-Z], and runs of
spaces are collapsed into single spaces.
There is no perfect answer in this case, but this solution is stable,
fine-grained, and probably close to what is expected. This
byte-by-byte comparison is equivalent to a utf-8 comparison without
case-folding of multibyte codes.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Tue, 30 Apr 2024 00:41:43 +0000 (12:41 +1200)]
util:charset: add strncasecmp_ldb()
This is a function for comparing strings in a way that suits a
case-insenstive syntaxes in LDB.
We have it here, rahter than in LDB itself, because it needs the
upcase table. By default uses ASCII-only comparisons. SSSD and
OpenChange use it in that configuration, but Samba replaces the
comparison and casefold functions with Unicode aware versions.
Until now Samba has done that in a bad way; this will allow it to do
better.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Thu, 9 May 2024 05:21:29 +0000 (17:21 +1200)]
ldb: ldb_comparison_fold_ascii sorts unsigned
Typically in 8-bit character sets, those with the 0x80 bit set are
seen as 288-255, not negative numbers. This will sort them after 'Z',
not before 'A'.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Wed, 15 May 2024 08:51:08 +0000 (20:51 +1200)]
ldb: add ldb_comparison_fold_ascii() for default comparisons
This function is made from the ASCII-only bits of the old
ldb_comparison_fold() -- that is, what you get if you never follow a
`goto utf8str` jump. It comparse the bytes, but collapses spaces and
maps [a-z] to [A-Z].
This does exactly what ldb_comparison_fold_utf8_broken() would do in
situations where ldb_casfold() calls ldb_casefold_default(). That
means SSSD.
The comparison is probably using signed char, so high bytes are
actually low bytes.
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Thu, 16 May 2024 23:34:35 +0000 (11:34 +1200)]
ldb: add ldb_set_utf8_functions() for setting casefold functions
This replaces ldb_set_utf8_fns(), which will be deprecated really soon.
The reason for this, as shown in surrounding commits, is that without
an explicit case-insensitive comparison we need to rely on the casefold,
and if the casefold can fail (because, e.g. bad utf-8) the comparison
ends up being a bit chaotic. The strings being compared are generally
user controlled, and a malicious user might find ways of hiding values
or perhaps fooling a binary search.
A case-insensitive comparisons that works gradually through the string
without an all-at-once casefold is better placed to deal with problems
where they happen, and we are able to separately specialise for the
ASCII case (used by SSSD) and the UTF-8 case (Samba).
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Douglas Bagnall [Sun, 19 May 2024 23:15:47 +0000 (11:15 +1200)]
ldb: add test_ldb_comparison_fold
Currently this fails like this:
test_ldb_comparison_fold_default_common: 118 errors out of 256
test_ldb_comparison_fold_default_ascii: 32 errors out of 100
test_ldb_comparison_fold_utf8_common: 40 errors out of 256
test_ldb_comparison_fold_utf8: 28 errors out of 100
Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Jo Sutton [Thu, 9 May 2024 01:16:50 +0000 (13:16 +1200)]
s4:kdc: Add comment about possible interaction between the krbtgt account and Group Managed Service Accounts
Signed-off-by: Jo Sutton <josutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Wed May 22 21:33:14 UTC 2024 on atb-devel-224
s3:utils: Use lp_dns_hostname() for 'net' dns updates
name_to_fqdn() requires /etc/hosts to be set up in a special way to find
out the fqdn for dns updates. They are not set up by default and the
DNS update fails. Normally the fqdn is just <lp_netbios_name>.<realm>
and we should just use that. However if it is different, you can set
it to the special value in the smb.conf now.
Signed-off-by: Andreas Schneider <asn@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
selftest/Samba4: make use of get_cmd_env_vars() to setup all relevant env variables
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): Wed May 22 05:26:48 UTC 2024 on atb-devel-224
s4:dsdb/repl: let drepl_out_helpers.c always go via dreplsrv_out_drsuapi_send()
I have customer backtraces showing that 'drsuapi' is NULL in
dreplsrv_op_pull_source_get_changes_trigger() called from the
WERR_DS_DRA_SCHEMA_MISMATCH retry case of
dreplsrv_op_pull_source_apply_changes_trigger(), while 'drsuapi' was
a valid pointer there.
From reading the code I don't understand how this can happen,
but it does very often on RODCs. And this fix prevents the problem.
Volker Lendecke [Thu, 16 May 2024 10:05:38 +0000 (12:05 +0200)]
smbd: Remove a no-op call to init_strict_lock_struct
"lock" is a variable on the stack, and initializing it is all that
init_strict_lock_struct does. We've done the corresponding checks
already in smbd_smb2_read_send
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Volker Lendecke [Wed, 15 May 2024 15:26:27 +0000 (17:26 +0200)]
smbd: Remove the ZERO_ZERO define
This code went in with e8e98c9ea06 in 2001 without any mention in the
commit message.
I don't remember all details, but there was some discussion around the
offset=0,length=0 lock being somehow special. These days [MS-FSA]
defines the required behaviour (don't conflict) and we implement that
behaviour.
If someone still remembers why this code is in and thinks it's
worthwhile, that's what version control is for.
Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
test_kinit_export_keytab: reset pw of the test account and test --only-current-keys
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): Wed May 22 04:07:02 UTC 2024 on atb-devel-224
s4:kdc: split out samba_kdc_fill_trust_keys() helper
This simplifies the logic in samba_kdc_trust_message2entry(),
is very similar to our samba_kdc_fill_user_keys() helper
and will make it trivial to provide the previous keys
in entry->old_keys in the next commit.
Review with: git show -p --patience
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
s4:kdc: let samba_kdc_trust_message2entry() ignore KRB5_PROG_ETYPE_NOSUPP
We already handle it in samba_kdc_fill_user_keys() mostly for DES keys,
but other encryption types might be from kerberos libraries in future.
And things like FIPS mode may also alter the runtime behaviour.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
projects / thirdparty / samba.git / log
