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.
'return req' should do the same as 'return true' for a bool function,
it's implicitly expanded as 'return (req!=NULL)?true:false.
There's no point in that as 'req' is always a valid pointer.
This was most likely just a copy and paste bug.
So we make this explicit now and avoid that Coverity reports this:
CID 1438158: Null pointer dereferences (REVERSE_INULL)
Null-checking "req" suggests that it may be null, but it has already
been dereferenced on all paths leading to the check.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Jul 31 14:20:49 CEST 2018 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>
This paves the way for adding async functions into the enumeration loop.
Best viewed with: git show -w
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
In future this will an impersonation wrapper tevent_context based on the
user session.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
HPUX has this problem.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=13270
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): Fri Feb 23 22:56:35 CET 2018 on sn-devel-144
The offset calculation for the offset that got passed to
fetch_write_time_send() in the enumeration loop was wrong as it passed
the offset before smbd_dirptr_lanman2_entry() added required padding.
This resulted in broken timestamps in the find response.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=13024
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Tue Sep 12 02:45:46 CEST 2017 on sn-devel-144
On 32-bit freebsd11, size_t is 32 bit. %zu does not cover
64 bits.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
This is just a hack for selftest that will be used in subsequent commits
for torturing compound find requests.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Finally use the new async dbwrap_parse_record_send/recv() functions
respectively the fetch_share_mode_send/recv wrappers for fetching the
write time from locking.tdb.
Previously for a directory with n files we would sit idle in the
directory enumeration loop fo n * m seconds waiting for responses from
ctdb, where m is the response time in seconds for a dbwrap request via
ctbd.
This is known to kill performance and we even have a parameter
"smbd:search ask sharemode" that can be used to disable fetching the
write time from locking.tdb.
Using fetch_write_time_send() works this way: in the directory
enumeration loop that calls smbd_dirptr_lanman2_entry() to marshall the
directory entries we
1. call fetch_write_time_send() after calling smbd_dirptr_lanman2_entry
passing a pointer to the current position in the marshall buffer.
2. If fetch_write_time_send() has set the out parameter "stop", we exit
the enumeration loop. This is necessary because we only send dbwrap
requests but don't consume the results. This has the potential to
deadlock so we must stop sending requests as soon as our ctdb send
queue is full.
3. In the fetch_write_time_done() callback, if the recv function got a
locking.tdb record, we push the write time into the marshall buffer
at the offet saved in the request state.
This new feature is still off by default as it doesn't
give any improvement in the non-clustered usecase.
"smbd:async search ask sharemode" can be used to activate it,
which makes only sense with "clustering = yes" (execept for testing).
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Not used for now, needed for async write_time updates in
smbd_smb2_query_directory_send().
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
dptr_CloseDir() will close and invalidate the fsp's file descriptor, we
have to reopen it.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=12496
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Uri Simchoni <uri@samba.org>
Also internally change path storage inside struct dptr_struct
to a struct smb_filename *.
This allows me to remove several of the synthetic_smb_fname()
calls I had to add in the previous patches, as we're now
dealing with struct smb_filename * throughout the dptr and
OpenDir code.
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): Tue Mar 1 18:34:24 CET 2016 on sn-devel-144
Currently SMB2/3 doesn't do posix pathname processing, leave this
as a placeholder for when SMB2 unix extensions are added.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <Volker.Lendecke@SerNet.DE>
This will allow us to move lp_posix_pathnames() out of unix_convert()
later.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <rb@sernet.de>
Prepare to remove lp_posix_pathnames() out of ms_has_wild().
Check before calls to ms_has_wild().
Fix smbd_smb2_query_directory_send().
No SMB2 client uses unix extensions yet, but this is a placeholder
for when we move the POSIX pathnames bit into the SMB2 request
when moving to handle based code.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Richard Sharpe <rsharpe@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Fri Mar 27 01:24:47 CET 2015 on sn-devel-104
