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.
We can use renameat2() without ifdef as libreplace always provides it.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15693
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
This will support renameat2-like operations in future.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15693
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Currently we do this in mkdir_internal():
mkdirat(client_name);
if (EEXIST) {
return EEXIST;
}
prepare_acls(client_name);
Note 'prepare_acls()' is a placeholder for the complex steps
it is doing to prepare the directory. During these steps
we have the problem that other clients already see
the directory and are able to create files or subdirectories
in it and these may not inherit the correct acls as
the their parent directory is not created completely.
I think I found a good strategie even without relying on
renameat2(RENAME_NOREPLACE).
We would do this instead:
tmp_name = ".::TMPNAME:D:$PID:client_name"
mkdirat(tmp_name, mode=client_mode);
prepare_acls(tmp_name);
mkdirat(client_name, mode=0);
if (EEXIST) {
unlinkat(tmp_name);
return EEXIST;
}
renameat(tmp_name, client_name);
So instead of having a long windows during prepare_acls,
we just have a short window between mkdirat(client_name, mode=0)
and renameat(tmp_name, client_name);
And in that short window the directory with the client_name
has a mode of 0, so it's not possible for other clients
to create files or subdirs in it.
As the mkdirat(client_name, mode=0) still catches
EEXIST the race where two clients try to create
the same client_name is closed as before,
so we don't need any other protection.
Following patches will make use of renameat2(RENAME_NOREPLACE),
but this already a very good improvement.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15693
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
We should have preparation steps as close as possible together.
We could move it for files as well, but that's a task for another
day...
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15693
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
All callers which looked at this are gone
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Aug 20 10:57:12 UTC 2024 on atb-devel-224
We are already in the !NT_STATUS_IS_OK if-branch, so this call was
redundant
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
No caller looks at it, and if someone should ever want to return a
more informative error, we now have NTSTATUS
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
If I see it correctly, this case was already covered by checking the
return code of cli_trans() a few lines above.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Not that it matters much, but I want to make cli_api() static in
clirap2.c.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
cli_oem_change_password() returns a bool, net_rap_password must return
0 for success
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
"(rdrcnt < 2)" was wrong, we don't look at rdata but at rparam. Use
"min_rparam" to cli_trans and the fact that if that succeeds "rparam"
is not NULL.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
In order to support certain Windows applications that make use of copy reflink,
we need some way to allow configuring these values. According to testing, the
application somehow uses the value of phys_bytes_per_sector_atomic for some check
when requesting server-side reflink copies, eg for ZFS the following is needed
block size = 131072
fs:aligned bytes per sector = 131072
For some reason "block size" must also be set to the value of fs:aligned bytes
per sector, but fs:logical bytes per sector, which according to the spec should
match "block size", must stay at the default of 512, otherwise the application
does not work.
As the whole client behaviour could not be fully understood, I'm proposing to
introduce these options as undocumented parametric options, so we can at least
start testing with them.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Tue Aug 20 07:01:19 UTC 2024 on atb-devel-224
According to MS-FSA 2.1.5.9.4 FSCTL_DUPLICATE_EXTENTS_TO_FILE ReFS on Windows
does not check for byte range lock conflicts:
* The object store SHOULD<81> check for byte range lock conflicts ...
* <81>: The ReFS file system in Windows Server 2016 and later does not check
for byte range lock conflicts
To match Windows behaviour we also don't check for conflicts.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
This ensures the values we return via SMB_FS_ATTRIBUTE_INFORMATION is the same
we use internally via conn->fs_capabilities.
This deliberately preserves existing behaviour as much as possible and leaves
possible improvements as a future excercize. Particularily FILE_VOLUME_QUOTAS is
already set insided SMB_VFS_STATVFS() depending on backend filesystem flags
which is probably the correct way to do it instead of just setting the
capability when Samba was built with quota support.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15689
Signed-off-by: Pavel Filipenský <pfilipensky@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Pavel Filipensky <pfilipensky@samba.org>
Autobuild-Date(master): Mon Aug 19 13:21:08 UTC 2024 on atb-devel-224
winbind_ctdb_updatekeytab.sh assumes the presence `onnode` utility to
execute `net ads` command on all nodes in the cluster. But `onnode`
is only built when configured with clustering support. Therefore perform
the script installation only with ctdb configuration. Also fix the
installation path to /usr/share/ctdb/scripts.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15689
Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Pavel Filipenský <pfilipensky@samba.org>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15696
Signed-off-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 13 22:29:28 UTC 2024 on atb-devel-224
Use explicit 'ceph_new' prefix to each of the ceph specific config
parameters to avoid confusion with legacy 'vfs_ceph' module. Hence,
users will have in their smb.conf a format similar to:
...
[smbshare]
vfs objects = ceph_new
ceph_new: config_file = /etc/ceph/ceph.conf
ceph_new: user_id = user1
ceph_new: filesystem = fs1
...
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Take special care for readdir errno setting: in case of error, update
errno by libcephfs (and protect from possible over-write by debug
logging); in the case of successful result or end-of-stream restore
errno to its previous value before calling the readdir_fn VFS hook.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Autobuild-User(master): Günther Deschner <gd@samba.org>
Autobuild-Date(master): Wed Aug 7 14:20:02 UTC 2024 on atb-devel-224
There is code in Samba that expects errno from a previous system call
to be preserved through a subsequent system call. Thus, avoid setting
"errno = 0" in status_code() and lstatus_code() upon successful return
from libcephfs API call.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
With this in the future we can avoid some special cases in our callees
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Aug 6 17:37:39 UTC 2024 on atb-devel-224
We're in setpathinfo, so if there's without an fsp it's
OBJECT_NAME_NOT_FOUND, the last component is missing.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Tested manually, but OBJECT_NAME_NOT_FOUND makes much more sense given
the new semantics of filename_convert_dirfsp.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Skip the smb1-only SERVER_HAS_UNIX_CIFS(), chmod now also does SMB2
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Go via create/fchmod/close. Only fchmod has to be smb2-specific this way.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Do a posix-level fchmod on a fnum. This will be used for smb2 soon as
well which does not have setpathinfo.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Will be used in smb311 unix chmod soon: We should only do the special
setsd on real posix handles. Otherwise we would probably destroy a
valid acl.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Pass the persistent/volatile handle as uint64's. Why? I found the
talloc_memdup() slightly misleading, and smbXcli handles those 2 id's
separately. map_smb2_handle_to_fnum() is the function to create the
smb2_hnd.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Currently we don't have an explicit check for the presence of
ceph_select_filesystem() libcephfs API as it is always found to
be present with the minimum ceph version that is supported with
Samba right now. Therefore under this assumption directly call
ceph_select_filesystem() without any #ifdefs. Please note that
this change is already part of vfs_ceph via ef0068cd.
ref: https://gitlab.com/samba-team/samba/-/merge_requests/3715
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Autobuild-User(master): Anoop C S <anoopcs@samba.org>
Autobuild-Date(master): Mon Aug 5 16:06:47 UTC 2024 on atb-devel-224
If ‘num_addrs’ is equal to zero, name_queries_send() will pass an
invalid address to name_query_send().
Signed-off-by: Jo Sutton <josutton@catalyst.net.nz>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
For my taste this is a nicer configuration syntax than
/../username1/file1/../username2/file2/
Is this too expensive? I don't think so. The scanning only happens an
tcon time, and it only walks the parametric options. If this turns out
to be a performance problem, we should think about smarter data
structures for parametric options instead of just a linked list of
string triples for everything.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=15688
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
We'll have to add to an existing namearray soon.
This turns one talloc_array() into a set of reallocs. This is slower,
but set_namearray is only used for smb.conf entries where we don't
expect hundreds or more entries to add. I've done this to avoid array
length calculations, but if it turns out to be too slow we can get
smarter again.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=15688
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Add developer's debug-logging upon each call to libcephfs' low-level
APIs.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement extended-attributes operations using libcephfs' low-level
APIs. Whenever possible, use the open file-handle from fsp-extension to
resolve inode-reference and user-permissions. Otherwise, resolve both
on-the-fly.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement mknodat operations using libcephfs' low-level APIs. Requires
parent directory to have valid inode-ref associated with its fsp
extension.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement renameat operations using libcephfs' low-level APIs. Requires
both directories to have valid inode-ref associated with their fsp
extension.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement link operations using libcephfs' low-level APIs. Requires two
phase operation: resolve (by-lookup) reference to inode and then do the
actual (hard) link operation using parent dir-inode reference to the
locally-cached inode.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement read/write IO operations using libcephfs' low-level APIs.
Requires open ceph Fh* associated with fsp (extension) to complete both
pread/pwrite as well as async I/O operations.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement unlinkat using libcephfs low-level APIs. For readlink
operation need to resolve child inode by-lookup and then used the inode
reference for the actual low-level readlink.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement unlinkat using libcephfs low-level APIs. Operate using parent
directory's open file-handle. When flags has AT_REMOVEDIR bit set call
low-level rmdir; otherwise, do normal unlink.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement fntimes hook using libcephfs' low-level APIs. Convert
smb_file_time to ceph_statx plus proper field mask on-the-fly upon
issuing low-level call to libcephfs.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Use libcephfs' low-level APIs to implement 'fchown' and 'fchmod' using
open file-handle. If fsp does not have an open cephfs Fh reference,
set errno to EBADF and return -1 to VFS.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Error handling in the case of 'ceph_readdir' is done by setting 'errno'
deep within libcephfs code. In case of error, emit proper debug message
and re-update errno to avoid possible over-write by logging mechanism.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement readdir and rewinddir operations using libcephfs' low-level
APIs. Casts the opaque DIR pointer into struct vfs_ceph_dirp (the first
member of struct vfs_ceph_fh) to resolve the ceph_dir_result pointer
which libcephfs expects for readdir operations.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement 'mkdirat' hook using libcephfs' low-level APIs, via the open
file-handle reference to parent directory.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Implement fdopendir using libcephfs low-level API and cached (via fsp)
open file-handle. Embed the result within cached vfs_ceph_fh so it may
be used properly by closedir.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Use libcephfs' low-level APIs to do lookup-by-name via parent's open
reference followed by getattr on the inode itself.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Use libcephfs' low-level APIs and apply the same logic as stat, but
via explicit inode-reference.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15686
Signed-off-by: Shachar Sharon <ssharon@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
