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.
If the fsp is forced closed by a SHUTDOWN_CLOSE whilst the
request is in flight (share forced closed by smbcontrol),
then we set state->req = NULL in the state destructor.
The existing state destructor prevents the state memory
from being freed, so when the thread completes and calls
vfs_gluster_pread_done(), just throw away the result if
state->req == NULL.
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>
Find the req we're finishing off by looking inside vfs_gluster_pread_state.
In a shutdown close the caller calls talloc_free(req), so we can't
access it directly as callback data.
The next commit will NULL out the vfs_gluster_pread_state->req pointer
when a caller calls talloc_free(req), and the request is still in
flight.
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>
We will need this to detect when this request is outstanding but
has been destroyed in a SHUTDOWN_CLOSE on this file.
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>
This reverts commit c9adf47ac5.
The fake fd is no longer necessary, as vfs_glusterfs now provides a
fcntl_fn hook.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14241
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
This hook is currently called via vfs_set_blocking(), so can safely be
ignored.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14241
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
GlusterFS currently doesn't have an API implementation to set flags on
open file descriptor. Thus we use pipe() to provide valid file descriptor
from the system.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14241
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Autobuild-User(master): Günther Deschner <gd@samba.org>
Autobuild-Date(master): Fri Jan 17 17:14:43 UTC 2020 on sn-devel-184
MS-SMB2 and the smbd code refer to this field as share_access. Use the
same name in the function argument.
Signed-off-by: Christof Schmitt <cs@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
On receiving an already initialized stat_ex buffer for readdir() call we
invoke readdirplus() GlusterFS API, an optimized variant of readdir(),
which then returns stat information along with dir entry result. But for
symlink entries we don't know if link or target info is needed. In that
case it is better to leave this decision back to caller by resetting
nlinks value inside stat information to make it invalid.
This was also preventing us from displaying msdfs link as directories
inside the share.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14182
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Günther Deschner <gd@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Dec 17 21:53:07 UTC 2019 on sn-devel-184
This implements two core changes:
* use NTTIME instead of struct timespec at the database layer
* use struct timespec { .tv_nsec = SAMBA_UTIME_OMIT } as special sentinel
value in smbd when processing timestamps
Using NTTIME at the database layer is only done to avoid storing the special
struct timespec sentinel values on disk. Instead, with NTTIME the sentinel value
for an "unset" timestamp is just 0 on-disk.
The NTTIME value of 0 gets translated by nt_time_to_full_timespec() to the
struct timespec sentinel value { .tv_nsec = SAMBA_UTIME_OMIT }.
The function is_omit_timespec() can be used to check this.
Beside nt_time_to_full_timespec(), there are various other new time conversion
functions with *full* in their name that can be used to safely convert between
different types with the changed sentinel value.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=7771
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Tue Nov 19 17:37:42 UTC 2019 on sn-devel-184
Note this isn't identical to unlink() as
this must cope with (flags & AT_REMOVEDIR),
which is identical to rmdir(). It calls
either unlink or rmdir depending on the
flags parameter.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14098
Signed-off-by: Poornima G <pgurusid@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Fri Aug 23 18:40:08 UTC 2019 on sn-devel-184
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14093
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Aug 20 19:25:28 UTC 2019 on sn-devel-184
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14090
RN: vfs_glusterfs: initialize st_ex_file_id, st_ex_itime and st_ex_iflags
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Fri Aug 16 01:07:23 UTC 2019 on sn-devel-184
The original implementation of the virtual xattr get_real_filename
in gluster was misusing the ENOENT errno as the authoritative anwer
that the file/dir that we were asking the real filename for does not
exist. But since the getxattr call is done on the parent directory,
this is a violation of the getxattr API which uses ENOENT for the
case that the file/dir that the getxattr call is done against does
not exist.
Now after a recent regression for fuse-mount re-exports due to
gluster mapping ENOENT to ESTALE in the fuse-bridge, the gluster
implementation is changed to more correctly return ENOATTR if the
requested file does not exist.
This patch changes the glusterfs vfs module to treat ENOATTR as ENOENT
to be fully functional again with latest gluster.
- Without this patch, samba against a new gluster will work correctly,
but the get_real_filename optimization for a non-existing entry
is lost.
- With this patch, Samba will not work correctly any more against
very old gluster servers: Those (correctly) returned ENOATTR
always, which Samba originally interpreted as EOPNOTSUPP, triggering
the expensive directory scan. With this patch, ENOATTR is
interpreted as ENOENT, the authoritative answer that the requested
entry does not exist, which is wrong unless it really does not exist.
Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
This is not called anymore, bump the VFS version number in a separate
commit
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): Thu Jun 20 18:34:20 UTC 2019 on sn-devel-184
error: assuming signed overflow does not occur when simplifying
conditional to constant [-Werror=strict-overflow]
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
VFS module for GlusterFS fails to compile due to recent changes done to
some API signatures. Therefore adding missing arguments to those APIs
adapting to new signatures.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=13330
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Sun Feb 3 17:00:33 CET 2019 on sn-devel-144
* ENODATA is not defined in FreeBSD
* ENOATTR is defined to be a synonym for ENODATA in Linux
* In its absence Samba already defines ENOATTR to either
ENODATA or ENOENT
Thus it is safe and correct to compare with ENOATTR rather
than ENODATA.
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Uri Simchoni <uri@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Jan 23 21:59:10 CET 2019 on sn-devel-144
Finally: use the new dos_mode_at_send() in the directory enumeration
loop. This means that fetching the DOS attributes for directory entries
is done asynchronously with regard to the enumeration loop.
As the DOS attribute is typically read from an extended attribute in the
filesytem, this avoids sequentially blocking on IO. If the IO subsystem
is slow servicing these request, enabling async processing can result in
performance improvements.
A parametric option
smbd:async dosmode = true | false (default: false)
can be used to enable the new async processing.
Simulating slow IO with usleep(5000) in the synchronous and asynchronous
versions of SMB_VFS_GET_DOS_ATTRIBUTES(), the results of enumerating a
directory with 10,000 files are:
smbd:async dosmode = no:
$ time bin/smbclient -U slow%x //localhost/test -c "ls dir\*" > /dev/null
real 0m59.597s
user 0m0.024s
sys 0m0.012s
smbd:async dosmode = yes:
$ time bin/smbclient -U slow%x //localhost/test -c "ls dir\*" > /dev/null
real 0m0.698s
user 0m0.038s
sys 0m0.025s
Performance gains in real world workloads depends on whether the actual
IO requests can be merged and parallelized by the kernel. Without such
wins at the IO layer, the async processing may even be slower then the
sync processing due to the additional overhead.
The following parameters can be used to adapt async processing behaviour
for specific workloads and systems:
aio max threads = X (default: 100)
smbd:max async dosmode = Y (default: "aio max threads" * 2)
By default we have at most twice the number of async requests in flight
as threads provided by the underlying threadpool. This ensures a worker
thread that finishes a job can directly pick up a new one without going
to sleep.
It may be advisable to reduce the number of threads to avoid scheduling
overhead while also increasing "smbd:max async dosmode".
Note that we disable async processing for certain VFS modules in the VFS
connect function to avoid the overhead of triggering the sync fallback
in dos_mode_at_send(). This is done for VFS modules that implement the
sync SMB_VFS_GET_DOS_ATTRIBUTES(), but not the async version (gpfs), and
for VFS modules that don't share a real filesystem where fchdir() can be
used (ceph, gluster). It is disabled for catia, because we realized that
the catia name translation macros used on
fsps (CATIA_FETCH_FSP_[PRE|POST]_NEXT) have a bug (#13547).
We use threadpool = smb_vfs_ev_glue_tp_chdir_safe() and then
pthreadpool_tevent_max_threads(threadpool) to get the number of maximum
worker threads which matches the pool used by the low level
SMB_VFS_GETXATTRAT_[SEND|RECV] implementation in vfs_default.
This is a terrible abstraction leak that should be removed in the future
by maybe making it possible to ask a VFS function which threadpool it
uses, internally suporting chaining so VFS function FOO that internally
uses BAR can forward the question to BAR.
On a hyphotetical system that had a getxattrat(dirfd, path, ...)
syscall and at the same time doesn't support per-thread current working
directories (eg FreeBSD doesn't have the latter) but has support for
per-thread-credentials, pthreadpool_tevent_max_threads() on the
tp_chdir_safe threadpool returns 1.
So when hooking the hyphotetical getxattrat() into the async
SMB_VFS_GETXATTRAT_[SEND|RECV] implementation in an VFS module, the
implementation could use the tp_path_safe threadpool, but the SMB2
layer would use the wrong threadpool in the call to
pthreadpool_tevent_max_threads(), resulting in no parallelism.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
The callback and _recv() functions should be called from the same
event context that was passed to the _send() function.
In future the completion pipe should be replaced by
tevent_threaded_context_create()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
