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.
This is helpful if you are in a listening loop with the same receiver
for many sockets doing the same thing.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
I just came across this failure with a new test:
==16654== Invalid read of size 4
==16654== at 0x4950947: tevent_req_is_in_progress (tevent_req.c:270)
==16654== by 0x5AEEE8F: writev_trigger (async_sock.c:375)
==16654== by 0x494F9E7: tevent_queue_immediate_trigger (tevent_queue.c:149)
==16654== by 0x494F53C: tevent_common_invoke_immediate_handler (tevent_immediate.c:166)
==16654== by 0x494F642: tevent_common_loop_immediate (tevent_immediate.c:203)
==16654== by 0x4959E5E: epoll_event_loop_once (tevent_epoll.c:918)
==16654== by 0x495665A: std_event_loop_once (tevent_standard.c:110)
==16654== by 0x494DFCE: _tevent_loop_once (tevent.c:772)
==16654== by 0x4950A6A: tevent_req_poll (tevent_req.c:300)
==16654== by 0x4D166C9: tevent_req_poll_ntstatus (tevent_ntstatus.c:109)
==16654== by 0x18C98B: run_readdir_timestamp (test_readdir_timestamp.c:489)
==16654== by 0x161BC5: run_test (torture.c:14896)
==16654== by 0x162726: main (torture.c:15136)
==16654== Address 0x91bb878 is 216 bytes inside a block of size 853 free'd
==16654== at 0x48369AB: free (vg_replace_malloc.c:530)
==16654== by 0x49B405E: _tc_free_internal (talloc.c:1221)
==16654== by 0x49B4116: _talloc_free_internal (talloc.c:1247)
==16654== by 0x49B547C: _talloc_free (talloc.c:1789)
==16654== by 0x50ECE3B: smb2cli_req_writev_done (smbXcli_base.c:3468)
==16654== by 0x4950648: _tevent_req_notify_callback (tevent_req.c:141)
==16654== by 0x49507A9: tevent_req_finish (tevent_req.c:193)
==16654== by 0x49507D6: _tevent_req_done (tevent_req.c:199)
==16654== by 0x5AEEE28: writev_do (async_sock.c:363)
==16654== by 0x5AEEE83: writev_trigger (async_sock.c:374)
==16654== by 0x494F9E7: tevent_queue_immediate_trigger (tevent_queue.c:149)
==16654== by 0x494F53C: tevent_common_invoke_immediate_handler (tevent_immediate.c:166)
==16654== by 0x494F642: tevent_common_loop_immediate (tevent_immediate.c:203)
==16654== by 0x4959E5E: epoll_event_loop_once (tevent_epoll.c:918)
==16654== by 0x495665A: std_event_loop_once (tevent_standard.c:110)
==16654== by 0x494DFCE: _tevent_loop_once (tevent.c:772)
==16654== by 0x4950A6A: tevent_req_poll (tevent_req.c:300)
==16654== by 0x4D166C9: tevent_req_poll_ntstatus (tevent_ntstatus.c:109)
==16654== by 0x18C98B: run_readdir_timestamp (test_readdir_timestamp.c:489)
==16654== by 0x161BC5: run_test (torture.c:14896)
==16654== by 0x162726: main (torture.c:15136)
==16654== Block was alloc'd at
==16654== at 0x483577F: malloc (vg_replace_malloc.c:299)
==16654== by 0x49B300F: __talloc_with_prefix (talloc.c:782)
==16654== by 0x49B31E6: _talloc_pool (talloc.c:837)
==16654== by 0x49B3394: _talloc_pooled_object (talloc.c:905)
==16654== by 0x49501A6: _tevent_req_create (tevent_req.c:79)
==16654== by 0x5AEE956: writev_send (async_sock.c:266)
==16654== by 0x50ECBCA: smb2cli_req_compound_submit (smbXcli_base.c:3396)
==16654== by 0x50ECD49: smb2cli_req_send (smbXcli_base.c:3447)
==16654== by 0x50FE34F: smb2cli_create_send (smb2cli_create.c:153)
==16654== by 0x490325E: cli_smb2_create_fnum_send (cli_smb2_fnum.c:273)
==16654== by 0x48D0146: cli_ntcreate_send (clifile.c:2504)
==16654== by 0x18B737: create_ts_send (test_readdir_timestamp.c:59)
==16654== by 0x18BF77: create_ts_files_send (test_readdir_timestamp.c:253)
==16654== by 0x18C35C: create_files_send (test_readdir_timestamp.c:336)
==16654== by 0x18C953: run_readdir_timestamp (test_readdir_timestamp.c:482)
==16654== by 0x161BC5: run_test (torture.c:14896)
==16654== by 0x162726: main (torture.c:15136)
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 Apr 23 21:53:38 UTC 2020 on sn-devel-184
All callers now use nonblocking sockets, so that we can optimize by
doing early writes
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Mar 25 10:41:29 UTC 2020 on sn-devel-184
We can only reasonably cancel a writev request that is still
queued. Once writing has started, cancel is pointless. Simplify the
if-conditions.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Now that open_socket_out*() doesn't do the strange reentrant calls,
we can remove support for this in async_connect_send(), so that we'll
never get any new users of this.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
While it is true that [e]poll() only needs POLLOUT
and POLLERR/POLLHUP are added implicitly.
For tevent we need TEVENT_FD_READ in order to see POLLERR/POLLHUP.
The socket becomes only readable when we hit an error.
Waiting for TEVENT_FD_WRITE is needed for getting success,
while TEVENT_FD_READ is required to get failures.
This matches what we have in tstream_bsd_connect_send().
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Set SOCKET_CLOEXEC on the sockets returned by accept. This ensures that
the socket is unavailable to any child process created by system().
Making it harder for malicious code to set up a command channel,
as seen in the exploit for CVE-2015-0240
Signed-off-by: Gary Lockyer <gary@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Allow callers to pass in socket fds that where already passed to an
earlier call of async_connect_send(). Callers expect this behaviour and
it was working until 05d4dbda83 broke it.
The proper fix would be to change callers to close the fd and start from
scratch with a fresh socket.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=12105
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): Thu Aug 4 05:03:21 CEST 2016 on sn-devel-144
According to Stevens UNIX Network Programming and various other sources,
the correct handling for non-blocking connect() is:
- when the initial connect() return -1/EINPROGRESS polling the socket
for *writeability*
- in the poll handler call getsocktopt() with SO_ERROR to get the
finished connect() return value
Simply calling connect() a second time without error checking is
probably wrong and not portable. For a successfull connect() Linux
returns 0, but Solaris will return EISCONN:
24254: 0.0336 0.0002 connect(4, 0xFEFFECAC, 16, SOV_DEFAULT) Err#150 EINPROGRESS
24254: AF_INET name = 10.10.10.143 port = 1024
24254: 0.0349 0.0001 port_associate(3, 4, 0x00000004, 0x0000001D,0x080648A8) = 0
24254: 0.0495 0.0146 port_getn(3, 0xFEFFEB50, 1, 1, 0xFEFFEB60) = 1 [0]
24254: 0.0497 0.0002 connect(4, 0x080646E4, 16, SOV_DEFAULT) Err#133 EISCONN
24254: AF_INET name = 10.10.10.143 port = 1024
Bug: https://bugzilla.samba.org/show_bug.cgi?id=11564
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Add an option to wait_for_read_send(), so that the request, upon
calling back, report whether the socket actually contains data
or is in EOF/error state. EOF is signalled via the EPIPE error.
This is useful for clients which do not expect data to arrive but
wait for readability to detect a closed socket (i.e. they do not
intend to actually read the socket when it's readable). Actual data
arrival would indicate a bug in this case, so the check can
be used to print an error message.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11397
Signed-off-by: Uri Simchoni <urisimchoni@gmail.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
This one might be a bit controversial. I don't see from man fcntl how this
could fail. But if it does, we definitely do want to know about it. And here we
don't have any good way to tell our caller, so abort.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: "Stefan (metze) Metzmacher" <metze@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Jun 16 19:22:52 CEST 2015 on sn-devel-104
This makes sure we remove the tevent_fd as soon as possible
and always reset the old_sockflags.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11316
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
This will facilitiate [un]become_root for smbd to connect safely to ctdbd.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Christian Ambach <ambi@samba.org>
Poll and select behave differently regarding error handling. When doing the
connect(2), we can not rely on poll telling us both readability and writability
upon error. Just always try a second connect(2). At least on Linux it returns 0
when it succeeded.
Signed-off-by: Jeremy Allison <jra@samba.org>
Autobuild-User: Jeremy Allison <jra@samba.org>
Autobuild-Date: Thu Aug 25 19:39:12 CEST 2011 on sn-devel-104
Only waiting for writability doesn't get fd errors back with poll.
So always begin by selecting for readability, and if we get it then
see if bytes were available to read or it really is an error condition.
If bytes were available, remove the select on read as we know we
will retrieve the error when we've finished writing and start
reading the reply (or the write will timeout or fail).
Metze and Volker please check.
Autobuild-User: Jeremy Allison <jra@samba.org>
Autobuild-Date: Mon Jun 6 21:53:16 CEST 2011 on sn-devel-104
In order to suppress a build warning.
Acked-by: Volker and Metze
Autobuild-User: Matthias Dieter Wallnöfer <mdw@samba.org>
Autobuild-Date: Wed Mar 23 10:26:23 CET 2011 on sn-devel-104
On my Linux box, this is definitely the more reliable strategy with unix domain
sockets, and according to my tests it also works correctly with TCP sockets.
A socket where the other side has closed only becomes readable. To catch
errors early when sitting in a pure writev, we need to also test for
readability.
