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.
Add the checks in reopen_from_procfd() into an if-condition, remove
the MORE_PROCESSING_REQUIRED logic that confused me.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Pass down a dummy "file_created" variable in open_directory(), remove
a unneeded if-condition.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We know the buffer size up-front, create a struct for that. Also, I
think if we ever hit another /proc pattern this is very likely on a
different OS that could be handled by an #ifdef.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Bug: https://bugzilla.samba.org/show_bug.cgi?id=15505
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): Fri Oct 27 21:19:35 UTC 2023 on atb-devel-224
By asking for TEVENT_FD_ERROR we're able to fail early
when a connection to a client is broken.
In that case it does not make any sense to process
pending requests in the recv queue as it's not
possible to deliver the response to the client anyway.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Tue Oct 24 10:32:56 UTC 2023 on atb-devel-224
This is accidentally removed by 6874ed6a9d.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Wed Oct 18 16:46:00 UTC 2023 on atb-devel-224
We cannot free ‘enc’ — it may be a copy of ‘buffer’, in which parameter
both construct_reply_chain() and smb_request_done() pass arrays of
automatic storage duration!
Fixes CID 1505354.
This reverts commit a395f752f0.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Together with the "dosmode | FILE_ATTRIBUTE_ARCHIVE" a line below this
is more obvious to me.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
To me these macros hide more than they clarify. In a lot of places we
already directly check for these flags without those macros. Unify
that.
Also, check for the dosmode bits first, lp_map_* is a bit more effort
to evaluate.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If the client requested FILE_OVERWRITE[_IF], we're implicitly adding
FILE_WRITE_DATA to the open_access_mask in open_file_ntcreate(), but for the
access check we're using access_mask which doesn't contain the additional
right, which means we can end up truncating a file for which the user has
only read-only access via an SD.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15439
Signed-off-by: Ralph Boehme <slow@samba.org>
In the fd_close() fsp->fsp_flags.fstat_before_close code path.
If this is a stream and delete-on-close was set, the
backing object (an xattr from streams_xattr) might
already be deleted so fstat() fails with
NT_STATUS_NOT_FOUND. So if fsp refers to a stream we
ignore the error and only bail for normal files where
an fstat() should still work. NB. We cannot use
fsp_is_alternate_stream(fsp) for this as the base_fsp
has already been closed at this point and so the value
fsp_is_alternate_stream() checks for is already NULL.
Remove knownfail.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=15487
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Oct 10 09:39:27 UTC 2023 on atb-devel-224
Indicate BACKUP_INTENT to vfs_openat(). Why? I have a customer request
who wants to add O_NOATIME in this case to avoid metadata updates when
a backup or virus-checking application comes along.
This does not fully handle BACKUP_INTENT correctly, this would require
become_root() appropriately. We might want to do that later after a
lot of careful security audit, but this patch independently might
already provide some infrastructure for it.
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): Thu Oct 5 14:00:33 UTC 2023 on atb-devel-224
We want to pass BACKUP_INTENT down into reopen_from_fsp, and the
elegant way is to do this via vfs_open_how.resolve.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
This needs close review. I could not see where we were actually
referencing the original flags in a way that would not be available in
local_flags. The reason for this patch is that I want to pass in
vfs_open_how into open_file(), and the distinction between flags and
local_flags made this significantly harder to understand for me.
The only place where we really used both versions is the DBG_NOTICE in
the last hunk, and this will come back in the next patch.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Move adding O_RDWR before the check for read only shares. I haven't
been able to pass this condition through SMB, but in any case we
should not accidentially open with O_RDWR in the !CAN_WRITE(conn)
case.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Simplify an if-condition:
We have to return NT_STATUS_OBJECT_NAME_INVALID even if we're not
creating. In fact, we probably should not end up in open_file() if
we're open a Windows file with a wildcard.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
We handle O_TRUNC further down anyway by passing local_flags&~O_TRUNC to
reopen_from_fsp(). No need for this FIFO special case.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
We use the plain (flags&O_TRUNC) a few lines above, make the
if-condition a bit more readable.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
We can unconditionally just and-out O_CREAT from local_flags, so
remove an if-condition.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
This will be part of a const struct vfs_open_how soon. Further down in
this function we don't look at O_CREAT or O_EXCL of "flags" anymore
anyway.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
We directly look at the flags in many other places in this function,
so do this also for O_ACCMODE for clarity.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Replace "truncating" variable reference with what it was defined
as. We use "(flags & O_TRUNC)" a few lines above, so it can't be that
bad.
After we set it to "false" further down, it was never used again.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Copy the logic from store_smb2_posix_info() to allow use of
ndr_push_smb3_file_posix_information().
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
‘tm’ must be initialized prior to calling strptime().
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Always offer it, it's a client thing to ask for it or not.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Sep 21 17:43:23 UTC 2023 on atb-devel-224
Matches file and directory closes.
Remove knownfail.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15423
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Sep 20 02:43:18 UTC 2023 on atb-devel-224
Make this DEVELOPER-only as it walks the entire open
file list on every file close (with associated aio).
This helps catch really subtle problems with orphaned
aio lnk structs.
Reproducer test case to follow.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15423
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
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 Sep 19 19:51:47 UTC 2023 on atb-devel-224
So it never looks at any passed in dst_dirfsp.
Prove this by passing NULL.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
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): Tue Sep 12 17:08:17 UTC 2023 on atb-devel-224
The main reason is the preparation of io_uring support,
as it can't be on the stack for async operations.
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): Wed Sep 6 20:31:04 UTC 2023 on atb-devel-224
The logic in smbd_smb2_advance_send_queue() will be reused for io_uring.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
In future we may use vectors with more elements, so we convert to
a single element array now...
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
The logic in smbd_smb2_advance_incoming() will be reused for io_uring.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This makes the code a little bit faster, but the main reason
is the preparation of io_uring support, as it can't be on the
stack for async operations.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We should avoid calling memset_s() in the core smbd processing,
we can use struct initializers instead.
This reduces the overhead...
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Remove the now unneeded req->xxx = NULL assignments (and the
deliberately bogus req->session = (void *)0xDEADBEEF one
used to demonstrate the bug).
Remove knownfail.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15432
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Tue Aug 15 12:06:36 UTC 2023 on atb-devel-224
If a client does a SMB1 NEGPROT followed by SMB1 TCON
then req->session is left uninitialized.
Show this causes a crash by deliberately initializing
req->session to an invalid pointer. This will be removed
once the test shows the crash, and the fix is added to
cause init_smb1_request() to zero the memory passed in.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15432
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Noel Power <npower@samba.org>
This augments 5c37615efa, the DBG macros already prepend the the function name
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Remove knownfail.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15430
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Noel Power <npower@samba.org>
Autobuild-User(master): Noel Power <npower@samba.org>
Autobuild-Date(master): Mon Aug 14 19:52:49 UTC 2023 on atb-devel-224
Now we've fixed srvstr_pull_req_talloc() this isn't
strictly needed, but ensuring pointers are initialized
is best practice to avoid future bugs.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15420
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Mon Aug 14 15:55:43 UTC 2023 on atb-devel-224
Rather than restore to uninitialized, set to NULL as per
modern coding practices.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15420
Reviewed-by: Volker Lendecke <vl@samba.org>
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Robert Morris <rtm@lcs.mit.edu> noticed that in the case
where srvstr_pull_req_talloc() is being called with
buffer remaining == 0, we don't NULL out the destination
pointed which is *always* done in the codepaths inside
pull_string_talloc(). This prevents a crash in the caller.
Remove knownfail.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15420
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
We will need this to show smbd crashing in the test code.
This will be removed once we're passing the test.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15420
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
The goal of this is to eventually remove reparse_symlink.c once we
have marshalling routines for symlinks in reparse.c
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If a client opens multiple connection with the same
client guid in parallel, our connection passing is likely
to hit a race.
Assume we have 3 processes:
smbdA: This process already handles all connections for
a given client guid
smbdB: This just received a new connection with an
SMB2 neprot for the same client guid
smbdC: This also received a new connection with an
SMB2 neprot for the same client guid
Now both smbdB and smbdC send a MSG_SMBXSRV_CONNECTION_PASS
message to smbdA. These messages contain the socket fd
for each connection.
While waiting for a MSG_SMBXSRV_CONNECTION_PASSED message
from smbdA, both smbdB and smbdC watch the smbXcli_client.tdb
record for changes (that also verifies smbdA stays alive).
Once one of them say smbdB received the MSG_SMBXSRV_CONNECTION_PASSED
message, the dbwrap_watch logic will wakeup smbdC in order to
let it recheck the smbXcli_client.tdb record in order to
handle the case where smbdA died or deleted its record.
Now smbdC rechecks the smbXcli_client.tdb record, but it
was not woken because of a problem with smbdA. It meant
that smbdC sends a MSG_SMBXSRV_CONNECTION_PASS message
including the socket fd again.
As a result smbdA got the socket fd from smbdC twice (or even more),
and creates two (or more) smbXsrv_connection structures for the
same low level tcp connection. And it also sends more than one
SMB2 negprot response. Depending on the tevent logic, it will
use different smbXsrv_connection structures to process incoming
requests. And this will almost immediately result in errors.
The typicall error is:
smb2_validate_sequence_number: smb2_validate_sequence_number: bad message_id 2 (sequence id 2) (granted = 1, low = 1, range = 1)
But other errors would also be possible.
The detail that leads to the long delays on the client side is
that our smbd_server_connection_terminate_ex() code will close
only the fd of a single smbXsrv_connection, but the refcount
on the socket fd in the kernel is still not 0, so the tcp
connection is still alive...
Now we remember the server_id of the process that we send
the MSG_SMBXSRV_CONNECTION_PASS message to. And just keep
watching the smbXcli_client.tdb record if the server_id
don't change. As we just need more patience to wait for
the MSG_SMBXSRV_CONNECTION_PASSED message.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15346
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Tue Aug 8 13:59:58 UTC 2023 on atb-devel-224
Commit 5d66d5b84f introduced a
'verify_again:' target, if we ever hit that, we would leak
the existing filter_subreq.
Moving it just above a possible messaging_filtered_read_send()
will allow us to only clear it if we actually create a new
request. That will help us in the next commits.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15346
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
time_to_asc() adds a trailing newline of its own.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Remove knownfail.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15419
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): Thu Jul 27 10:52:50 UTC 2023 on atb-devel-224
This was broken by commit 1f3f6e20dc because when
calling srv_init_signing() very early after accepting the connection in
smbd_add_connection(), conn->protocol is still PROTOCOL_NONE.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15397
Signed-off-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Jule Anger <janger@samba.org>
Autobuild-Date(master): Fri Jul 21 13:03:09 UTC 2023 on atb-devel-224
This is just going to bitrot. Anyone who's interested can just grep for
"signing_mandatory" and look up what it does.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15397
Signed-off-by: Ralph Boehme <slow@samba.org>
This can happen if DELETE-ON-CLOSE is set, but the deletion fails for some
reason.
The bug was introduced by 1808e5c133.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15417
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): Mon Jul 10 21:32:32 UTC 2023 on atb-devel-224
The parent smdb forwards SIGTERM to its process group in order to kill all
children like the scavenger. This happens from a function registered via
atexit() which means the signal forwarding is happening very briefly before the
main smbd process exits. When exiting the pipe between smbd and scavenger is
closed which triggers a file event in the scavenger.
However, due to kernel sheduling it is possible that the file descriptor event
is received before the signal, where we call exit_server() which call
smb_panic() at the end.
Change the exit to exit_server_cleanly() and just log this event at level 2
which we already do.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15275
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Jul 5 13:14:08 UTC 2023 on atb-devel-224
Probably not really measurable, but as this is a pretty frequently
accessed table, maybe this saves us some cache misses. And it's very
cheap to get :-)
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
3 times talloc_zero() and read_symlink_reparse() makes a nice separate
function.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Move filtering of entries, in particular symlinks, fully into
smbd_dirptr_get_entry(). Before, this was hidden in magic code inside
openat_pathref_fsp() and the mode_fn()s. Changing anything file open
code led to changes in very distant code paths because of unforeseen
consequences to directory listing. This change centralizes the
decision what directory entries to show into
smbd_dirptr_get_entry(). It uses openat_pathref_fsp_nosymlink()
without any symlink magic. It might need some tweaking when we also
want to show other special files, but this will hopefully be easier.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Turn it into openat_pathref_fsp_nosymlink() which opens not only
directories but normal files and symlinks too. If it finds a symlink,
return NT_STATUS_STOPPED_ON_SYMLINK and all the metadata we can find:
struct stat_ex plus the symlink target.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Prepare for allowing openat_pathref_fsp_nosymlink() to open real files
in the next patches.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
The goal of this patch is to make the next patches smaller. Next
patches will also make openat_pathref_dirfsp_nosymlink() open real
files, not only directories.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Fake up a symlink reparse point structure from an on-disk reparse
point.
Turn in-share absolute symlinks into relative ones for the client to
properly follow symlinks. Pass on everything else as-is. In particular
clients follow symlinks pointing at \\server\share\...
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Implementing virtually empty directories for "dont descend" looks
easier to me this way. It should also be an optimization, because now
we don't walk the whole directory anymore after . and .., which always
come first anyway.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
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): Fri Jun 23 18:29:40 UTC 2023 on atb-devel-224
Nobody used that anymore, most callers had passed in NULL anyway.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Some log levels change because the macros don't cover all the previously used
log levels or because importance was slightly reconsidered.
Signed-off-by: Bjoern Jacke <bjacke@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Some log levels change because the macros don't cover all the previously used
log levels or because importance was slightly reconsidered.
Signed-off-by: Bjoern Jacke <bjacke@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
No caller uses this anymore. The only downside here now is that we
always go directly to mask_match instead of a trying strcasecmp_m
first. I very much doubt this makes a measurable difference because
this would have been called for non-wildcard
readdirs (a.k.a. qpathinfo), and there we do this only once per
complete directory read. Also I don't believe mask_match() is
measurably more expensive than strcasecmp_m() for the usually short
filenames we're looking at here.
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): Fri Jun 16 17:07:46 UTC 2023 on atb-devel-224
I don't think this is an effective optimization at all anymore. It was
intended to speed up non-wildcard readdirs after we found the correct
entry. Nowadays we do the non-wildcard readdirs by a direct fstatat,
and after we successfully found the entry dptr_ReadDirName()
immediately returns without any further action. So my very strong
guess is that this never really kicked in anymore. Not using this flag
can't be *that* bad, smb2 never used it.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We have that available in smbXsrv_connection that is available almost
everywhere.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
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): Wed Jun 14 00:26:55 UTC 2023 on atb-devel-224
Nobody does anything with this anymore, we just call ReadDirName() in
sequence or do a RewindDir(). So we don't have to look at offsets as
given by the file system anymore.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This saves the result of smbd_dirptr_get_entry() for later retrieval
in case we could not marshall it to the output buffer. Return this
entry when calling smbd_dirptr_get_entry() again.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
While trying to understand the ReadDirName() at the end of
dptr_ReadDirName() in a code path that was supposed to be just a
"stat"-style readdir with a non-wcard mask I came to the conclusion
that this was there to find dptr->wcard with a mangled
name. get_real_filename_at() already takes care of name mangling, so I
think I could eliminate a source of confusion by using it.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
We need to pass "conn" explicitly because dptr_struct is private to
dir.c.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
No caller does anything with the smb_fname upon overflow, so we might
as well do an early return.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If I'm not completely blind then there's no way how *pst can be a
valid stat. We did a SET_STAT_INVALID at the beginning of the
function, and there's no code path up to this if-statement that can
make *pst valid again.
Review with "git show -U40".
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
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 8 18:34:44 UTC 2023 on atb-devel-224
reply_search() is the only place in the code where we have to deal
with [MS-CIFS] 2.2.4.59.1 ResumeKey structures. This concentrates the
formatting of this to pure SMB1 code in reply_search(), moving away
knowledge about the format from smbd/dir.c's dptr_fill() and
dptr_fetch_fsp().
With this code we just count up the FileIndex from behaviour note
110. If the client is sane and sends us the last FileIndex we returned
to it in a subsequent search, we can completely avoid any
telldir/seekdir. If it skips back, with the new code we rewind and
re-readdir the directory. This will be slower for a very special
corner case, but it's a lot simpler to understand (at least to
me). Also, it avoids calling telldir/seekdir for every entry.
Tested both cases (sane and insane clients) manually with a modified
cli_list_old_done(). Not doing automated tests. If this breaks real
users, we'll fix it and write tests then.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Don't wait for the main loop to free file names no longer used.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
name_str can pile up when reading directories, we don't talloc_free()
our stackframe before we have filled the whole readdir response packet.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Return how many name entries have been returned from this dptr.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This never took off outside of a special OEM setup.
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): Mon Jun 5 18:13:05 UTC 2023 on atb-devel-224
Looks larger than it is, "git clang-format" added a few lines
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This allows us to support starting smbd under callgrind and only start
the overhead and instrumentation after the SMB2 negprot, this allows us
to profile only useful stuff and not all the smbd startup, forking and
multichannel handling.
This will do the trick:
valgrind --tool=callgrind --instr-atstart=no smbd
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
It means we can make the fd blocking, which will help
with io_uring support.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
This is not strictly needed, but it helps profiling
the core smb2_server.c code with the 'smb2.bench.echo'
test.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
With this ld.so does not have to relocate the string pointers
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
