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.
For session bind, and the channel is only to be bound to the given
session just now, so it is not valid. The early request validation
code can hence not check it, and hence validation is defered to the
actual session setup code, which can look at the session binding flags.
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Michael Adam <obnox@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Michael Adam <obnox@samba.org>
Autobuild-Date(master): Wed Jul 29 21:31:09 CEST 2015 on sn-devel-104
This is in preparation of adding a variant that operates
on the client and does in particular not verify that the
connection belongs to a session as a channel.
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Michael Adam <obnox@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
this means we:
- accept unencrypted requests if encryption only desired
and not required,
- but we always send encrypted responses in the desired
case, not only when the request was encrypted.
For this purpose, the do_encryption in the request
structure is separated into was_encrypted and do_encryption.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11372
Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reuse of nonces with AES-CCM and AES-GCM leads to catastrophic failure,
so make sure the server drops the connection if that ever happens.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=11300
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Simo Sorce <simo@redhat.com>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Simo Sorce <idra@samba.org>
Autobuild-Date(master): Fri May 29 22:38:50 CEST 2015 on sn-devel-104
Following Windows behaviour, the last command in a compound request
should be padded to an 8 byte boundary and OS X clients crash badly if
we don't pad.
[MS-SMB2] 3.3.4.1.3, "Sending Compounded Responses", doesn't make it
clear whether the padding requirement governs the last command in a
compound response, a future MS-SMB2 update will document Windwows
product behaviour in a footnote.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=11277
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Used for:
- preauthentication validation
- negotiation of ciphers for sigingn and encryprtion
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Michael Adam <obnox@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Apr 22 01:04:02 CEST 2015 on sn-devel-104
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
What?
This patch gets rid of the central shared memory segment referenced by
"profile_p". Instead, every smbd gets a static profile_area where it collects
profiling data. Once a second, every smbd writes this profiling data into a
record of its own in a "smbprofile.tdb". smbstatus -P does a tdb_traverse on this
database and sums up what it finds.
Why?
At least in my perception sysv IPC has not the best reputation on earth. The
code before this patch uses shmat(). Samba ages ago has developed a good
abstraction of shared memory: It's called tdb.
The main reason why I started this is that I have a request to become
more flexible with profiling data. Samba should be able to collect data
per share or per user, something which is almost impossible to do with
a fixed structure. My idea is to for example install a profile area per
share and every second marshall this into one tdb record indexed by share
name. smbstatus -P would then also collect the data and either aggregate
them or put them into individual per-share statistics. This flexibility
in the data model is not really possible with one fixed structure.
But isn't it slow?
Well, I don't think so. I can't really prove it, but I do believe that on large
boxes atomically incrementing a shared memory value for every SMB does show up
due to NUMA effects. With this patch the hot code path is completely
process-local. Once a second every smbd writes into a central tdb, this of
course does atomic operations. But it's once a second, not on every SMB2 read.
There's two places where I would like to improve things: With the current code
all smbds wake up once a second. With 10,000 potentially idle smbds this will
become noticable. That's why the current only starts the timer when something has
changed.
The second place is the tdb traverse: Right now traverse is blocking in the
sense that when it has to switch hash chains it will block. With mutexes, this
means a syscall. I have a traverse light in mind that works as follows: It
assumes a locked hash chain and then walks the complete chain in one run
without unlocking in between. This way the caller can do nonblocking locks in
the first round and only do blocking locks in a second round. Also, a lot of
syscall overhead will vanish. This way smbstatus -P will have almost zero
impact on normal operations.
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Volker Lendecke <vl@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Bug: https://bugzilla.samba.org/show_bug.cgi?id=9702
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Thu Jan 29 14:58:40 CET 2015 on sn-devel-104
It turns out that the effective credits_requested is always at least 1,
even if the client sends credits_requested == 0.
This means the client is not able to reduce the amount of credits
itself.
Without this fix a client (e.g. Windows7) would reach the case
where it has been granted all credits it asked for.
When copying a large file with a lot of parallel requests,
all these requests have credits_requested == 0.
This means the amount of granted credits where reduced by each
request and only when the granted credits reached 0,
the server granted one credit to allow the client to go on.
The client might require more than one credit ([MS-SMB2] says
Windows clients require at least 4 credits) and freezes
with just 1 credit.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=9702
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If signing is not required we should not require it for reauthentication.
Windows clients would otherwise fail to reauthenticate.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=10958
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
In future we want to use this for lease breaks and they're not attached
to a session.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This improves profiling and corrently counts the total and idle time for async
requests.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
SMB 3.10 and later won't have SMB2_CAP_ENCRYPTION anymore.
xconn->smb2.server.cipher == 0 is the indication that we don't support encryption on the connection.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This will find the correct signing key for the connection.
If there's already a channel session key, we need to use that
otherwise we need to use the global session key.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
These should be per client (based on the SMB >= 2.1 client_guid),
this is a preparation for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
If we have a channel session key, we also always have a global session key.
For multi-channel it's possible that the channel session key is not in place
yet, in that case the global session key needs to be used.
In both cases (reauth or session bind) we session setup requests need to be
signed.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
Calling exit_server() from within a destructor is a bit ugly...
This will result in smbd_server_connection_terminate() instead of
directly calling exit_server(), which will be useful for multi-channel in
future.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This will simplify future features like multi-channel and rdma.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This makes the code more consistent with other functions
using 'struct smbXsrv_connection *xconn' as local variables.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Each request needs to respond on the same connection,
where it arrived.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
As the list of pending requests moved from sconn->smb2.requests to xconn->smb2.requests,
it is more logical to use smbXsrv_connection as talloc parent.
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
This prepares the structures for multi-channel support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
Most of this routine can be re-used for sending lease breaks
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Jul 8 19:54:09 CEST 2014 on sn-devel-104
We don't need the assignment to state->vector[1+SMBD_SMB2_DYN_IOV_OFS],
this is zero-initialized by talloc_zero
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
This actually saves a few bytes in .text. Maybe due to the struct assignments?
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
MS-SMB2: 3.3.5.2.4 Verifying the Signature.
If the SMB2 header of the SMB2 NEGOTIATE
request has the SMB2_FLAGS_SIGNED bit set in the
Flags field, the server MUST fail the request
with STATUS_INVALID_PARAMETER.
Found and fix confirmed by Microsoft test tool.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ira Cooper <ira@samba.org>
This avoids recursion into smbd_smb2_io_handler(),
which avoids confusion when analysing out put of
performance analysing tools, e.g. callgrind.
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): Sat May 31 04:25:36 CEST 2014 on sn-devel-104
smb2req always comes from talloc_zero().
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): Fri Apr 11 23:55:17 CEST 2014 on sn-devel-104
This way the buffer will likely be allocated within the existing talloc_pool,
which avoids one malloc() per request.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
For recvfile we haven't read and may not allocated the dyn buffer.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
In the common case with just one request, we can use a preallocated
req->out.vector.
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): Thu Mar 6 00:59:29 CET 2014 on sn-devel-104
It is only important that the content of info->data stays alive
for the lifetime of the request, but the DATA_BLOB structure itself
can be on the stack, while passing it as 'dyn' to smbd_smb2_request_done_ex().
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Use a preallocated buffer for the first response in the compound chain.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We can use a preallocated buffer for the possible error
response of the first response in the compound chain.
This avoids a talloc_array_zero() call for the common case.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We can avoid a talloc_zero_array() call in the
common case (without compound requests) and use a
preallocated array instead.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If this isn't NT_STATUS_OK, we skip any io on the socket.
This avoids possible problems during shutdown.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Nov 27 16:31:44 CET 2013 on sn-devel-104
We need to pass the NBT header, SMB2 header and SMB2 Read header
as header blob to SMB_VFS_SENDFILE(). This allows the usage
of MSG_SEND or other tricks to avoid multiple TCP packets
on the wire.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
This matches the behavior for smb1 requests
and avoids an additional malloc() per request.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
If the request is already done we can avoid one iteration
of tevent_loop_once(), which means we avoids one
talloc_stackframe_pool/talloc_free pair.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Going via tevent_req_create/talloc_free at multiple layer costs
too much cpu cycles per request.
I tested downloading a 16GB (sparse) file with smbclient -b1 -mNT1,
and -mSMB2_02. Using smb2 max read = 64512, which means smb1 and smb2
will use the same read size.
I build with -O3 -g and compared the results with valgrind --tool=callgrind.
With -mNT1 the server uses about 2.000.000.000 cpu cycles.
This patch reduces the userspace cpu cycles for -mSMB2_02
from about ~ 8.000.000.000 down to ~ 4.000.000.000.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
smbd_smb2_request_error_ex() should return NTSTATUS and the caller
will terminate the connection.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
I.e. when the request is a session setup.
We replied with ACCESS_DENIED, but windows expects USER_SESSION_DELETED
Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Oct 2 22:07:44 CEST 2013 on sn-devel-104
Windows (at least the test suites) may skip the SMB2_HDR_FLAG_SIGNED
in a reauth session setup, but still provide a valid signature.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
Signed-off-by: Richard Sharpe <rsharpe@samba.org>
Autobuild-User(master): Richard Sharpe <sharpe@samba.org>
Autobuild-Date(master): Thu May 23 06:53:17 CEST 2013 on sn-devel-104
