IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Shows we must protect against a null fsp handle when doing POSIX chmod on a symlink,
whether the symlink points to a real object or is dangling.
Add to knownfail for now. Commit 9722732b18
removed the fsp == NULL protection for POSIX, and we need to put it back.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
This creates a directory, then a symlink to a directory,
and then checks we can POSIX create and delete file, directory,
symlink and hardlink filesystem objects under the symlink
parent directory.
Mark as knownfail until next commit.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
It shows this isn't done correctly for streams_xattr.
A common config is:
vfs_objects = streams_xattr acl_xattr
to store both streams and Windows ACLs in xattrs.
Unfortunately getting and setting ACLs using handles
opened on stream files isn't being done correctly
in Samba.
This test passes against Windows 10.
This adds tests that prove this doesn't work. Next
patch will add the fix and remove the knownfail.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Best reviewed with "git show -b". Right now lsarpc in the nt4_dc
environment is not available over ncalrpc, so instead of getusername
we need to use epmlookup for the rpcclient tests
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This allows us to test session binds with different users.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14512
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
$PREFIX is the the value from --with-selftest-prefix.
The result of the test should not depend on --with-selftest-prefix,
the 'long_path' test in particular.
If the path is to long smbclient (via libarchive) will only
put the full path into a PAX HEADER as 'path' keyword,
that's fine in general, modern tools handle it just fine.
But Perl's Archive::Tar don't handle it and only seems
truncated file names.
I have a fix for Archive::Tar, see:
https://git.samba.org/?p=metze/samba/wip.git;a=shortlog;h=c75037d0a06a96cdaca3f3b20a6d237e768b075b
But finishing that is a task for another day, for now I just want to remove
the dependency to --with-selftest-prefix.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Note that uses SMB2 for the "Windows client" (aka non-POSIX) connection as SMB1
directory listing code translates a directory listing with a search mask that
matches an existing file to a CREATE which won't cut it for our test as we're
targetting the directory listing code.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
The kernel of the gitlab shared runners container host has a bug in the
interaction between kernel oplocks and O_PATH opens which was fixed by
387e3746d01c34457d6a73688acd90428725070b in 5.3.1:
<https://kernel.googlesource.com/pub/scm/linux/kernel/git/jlayton/linux/+/refs/tags/locks-v5.3-1%5E%21/>
Don't actually start the OPLOCK5 test is kernel oplocks are not available,
instead of relying on the #ifdef HAVE_KERNEL_OPLOCKS_LINUX magic in torture.c.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
samba3.blackbox.smbclient_tar is marked as flapping so it
seems we have missed that it has stopped working. The local path
passed to script/tests/test_smbclient_tarmode.pl must point to a
valid share
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14581
Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
After this change both samba3.blackbox.smbclient_tar &
samba3.blackbox.smbclient_tarmode now use the same dedicated share
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14581
Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
The other tarmode torture test samba3.blackbox.smbclient_tar now uses a share
'tarmode' which uses the same source path as samba3.blackbox.smbclient_tarmode
Avoid conflicting paths and use a new subdir (of the test share) called
'smbclient_tarmode'
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14581
Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Use the notifyd "messaging" protocol to check if notifyd works at all
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Just don't run the tests instead of retrofitting them to the skiplist.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This fixes the test with fast disks where 20MB transfers are done in
less than a second.
This also cleans up the code to have less sleeping time!
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): Mon Oct 19 21:14:21 UTC 2020 on sn-devel-184
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Fri Oct 16 18:30:18 UTC 2020 on sn-devel-184
There is no reason to use the SMB1-based ntdeny2 test. It was just
an arbitrary test that depends on clustering to work.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Add a ZeroLogon test suite, to allow the ZeroLogon tests to be run against
the s3 and s4 netlogon servers.
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
samba3.smbtorture_s3.crypt_client.SMB1-WILD-MANGLE-RENAME(nt4_dc_smb1)
samba3.smbtorture_s3.plain.SMB1-WILD-MANGLE-RENAME(fileserver_smb1)
knownfail for now.
The recent wildcard changes broke something that used to work.
Consider a directory with 2 files:
dir/
foo
fo*
The 'fo*' file has a mangled name of FSHCRD~2.
SMB1rename("dir/FSHCRD~2", "dir/ba*") will rename *both* files
as the new 'rename has wildcard' check is done after
the name unmangle.
SMB2 doesn't allow wildcard renames so doesn't have this problem.
Fix to follow.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
samba3.smbtorture_s3.crypt_client.SMB1-WILD-MANGLE-UNLINK(nt4_dc_smb1)
samba3.smbtorture_s3.plain.SMB1-WILD-MANGLE-UNLINK(fileserver_smb1)
knownfail for now.
The recent wildcard changes broke something that used to work.
Consider a directory with 2 files:
dir/
a
*
The '*' file has a mangled name of _2X68P~X.
SMB1unlink("_2X68P~X") will delete *both* files
as the new 'unlink has wildcard' check is done after
the name unmangle.
SMB2 doesn't suffer from this problem, as it doesn't
allow wildcard unlinks.
Fix to follow.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
The mapping functions of the vfs_gpfs module can be easily unit tested.
Begin a cmocka test to cover those.
Signed-off-by: Christof Schmitt <cs@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We need to make sure that winbind's idmap_ad deals fine with an
expired krb ticket used to connect to AD via LDAP. In a customer
situation we have seen the RFC4511 section 4.4.1 unsolicited ldap exop
response coming through, but the TCP disconnect that Windows seems to
do after that did not make it. Winbind deals fine with a TCP
disconnect, but right now it does not handle just the section 4.4.1
response properly: It completely hangs.
This test requests a ticket valid for 5 seconds and makes the LDAP
server postpone the TCP disconnect after the ticket expiry for 10
seconds. The tests that winbind reacts to the ticket expiry exop
response by making sure in this situation the wbinfo call running into
the issue takes less than 8 seconds. If it did not look at the expiry
exop response, it would take more than 10 seconds.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=14465
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
When reading entries from gencache, wb_cache_rids_to_names() can
return STATUS_SOME_UNMAPPED, which _wbint_LookupRids() does not handle
correctly.
This test enforces this situation by filling gencache with one wbinfo
-R and then erasing the winbindd_cache.tdb. This forces winbind to
enter the domain helper process, which will then read from gencache
filled with the previous wbinfo -R.
Without having the entries cached this does not happen because
wb_cache_rids_to_names() via the do_query: path calls deep inside
calls dcerpc_lsa_lookup_sids_noalloc(), which hides the
STATUS_SOME_UNMAPPED that came in as lsa_LookupSids result value.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=14435
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Shows smbd panics if connection is terminated (torn down)
by killing the client with outstanding aio requests in the
queue. As we're closing smbd we should cope with this.
Followup-bugfix for:
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14301
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
The test is written for SAMR, but as the handle type is verified by the
server core library it also applies to other RPC services.
Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
raw.samba3rootdirfid tests with the share root directory as root_dir_fid handle,
that doesn't cover the case where the relative name has more then one path
component. It only works because in unix_convert() we run into the creating file
optimasation.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14380
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
SMB2 and NT1 fail this, CORE already returns NT_STATUS_INVALID_NETWORK_RESPONSE
on bad conversion.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14374
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14345
Signed-off-by: Amit Kumar <amitkuma@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Alexander Bokovoy <ab@samba.org>
Signed-off-by: David Mulder <dmulder@suse.com>
Reviewed-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 888abcaf8ffbec45fc47520bd3f544e3aa6f58f2)
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Apr 28 19:46:32 UTC 2020 on sn-devel-184
Signed-off-by: David Mulder <dmulder@suse.com>
Reviewed-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit eb167bc43dbe196ef5b3bfd24160c72c74113dea)
Signed-off-by: David Mulder <dmulder@suse.com>
Reviewed-by: Noel Power <noel.power@suse.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 9437b44668c9f7742d6d4fe0891ac4d9fda7c804)
Create -o files per -N client connections, set a specific timestamp, then write
a bit. This leads to the locking.tdb dmasters to be spread across all nodes.
Then list from one node. This makes sure that the async share mode fetch works
right.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
