Stefan Metzmacher [Mon, 16 Jan 2023 18:46:13 +0000 (19:46 +0100)]
src/socket_wrapper.c: handle raw SYS_close, SYS_recvmmsg and SYS_sendmmsg syscall() invocations
This fixes a problem hit by 'nsupdate -g' and bind9 (dnsutils). If bind is built
against libuv <= 1.44.2 it will not use sendmmsg/recvmmsg functions from libc
but use the corresponding syscalls directly. Newer version of libuv removed
the syscall wrappers and use sendmmsg/recvmmsg from libc.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 16 Jan 2023 18:27:12 +0000 (19:27 +0100)]
tests: add test_echo_tcp_sendmmsg_recvmmsg
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 16 Jan 2023 16:00:48 +0000 (17:00 +0100)]
src/socket_wrapper.c: implement recvmmsg and sendmmsg
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 16 Jan 2023 15:58:04 +0000 (16:58 +0100)]
src/socket_wrapper.c: add some tracing for connected dgram sockets
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Sun, 6 Nov 2022 15:07:21 +0000 (16:07 +0100)]
src/socket_wrapper.c: don't leak unlink() errno (most likely ENOENT) in swrap_bind()
This fixes a problem hit by 'nsupdate -g' from the
bind9-dnsutils 1:9.18.1-1ubuntu1.2 package.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 16 Jan 2023 15:50:32 +0000 (16:50 +0100)]
src/socket_wrapper.c: pretty print ip addresses in tracing output.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 23 Nov 2022 11:12:12 +0000 (12:12 +0100)]
Bump version to 1.3.5
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 23 Nov 2022 10:46:45 +0000 (11:46 +0100)]
socket_wrapper: inject O_LARGEFILE to open[64|at]() if needed
On 32bit systems this is normally done by glibc if _FILE_OFFSET_BITS is 64,
but with socket wrapper we don't want to define _FILE_OFFSET_BITS=64,
as we need to overload open64 explicitly. But we need to inject
O_LARGEFILE for being transparent to the application.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15251
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Samuel Thibault [Thu, 10 Nov 2022 18:36:29 +0000 (18:36 +0000)]
Avoid dclose(RTLD_NEXT)
In case the libc was not found and RTLD_NEXT is used instead, we should not
dlclose it, otherwise mayhem happens.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15228
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 12 Sep 2022 09:25:17 +0000 (11:25 +0200)]
pkgconfig: Fix path to libsocket_wrapper.so
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Alex Richardson [Fri, 1 Oct 2021 09:00:32 +0000 (10:00 +0100)]
Fix -Wcast-qual warnings
Without this change I get the following -Werror build failure when building
samba on macOS:
```
../../third_party/socket_wrapper/socket_wrapper.c:5420:15: error: cast from 'const struct cmsghdr *' to 'unsigned char *' drops const qualifier [-Werror,-Wcast-qual]
__fds_in.p = CMSG_DATA(cmsg);
^
/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/usr/include/sys/socket.h:631:51: note: expanded from macro 'CMSG_DATA'
#define CMSG_DATA(cmsg) ((unsigned char *)(cmsg) + \
^
```
Signed-off-by: Alex Richardson <Alexander.Richardson@cl.cam.ac.uk>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Thu, 21 Jul 2022 12:48:02 +0000 (14:48 +0200)]
Bump version to 1.3.4
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 21 Jul 2022 05:22:08 +0000 (07:22 +0200)]
swrap: Fix a TOCTOU issue in swrap_auto_bind()
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 3 Mar 2022 08:24:09 +0000 (09:24 +0100)]
gitlab-ci: Introduce stages
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 3 Mar 2022 08:18:11 +0000 (09:18 +0100)]
gitlab-ci: Fix formatting
Complaints from yamllint.
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 3 Mar 2022 08:22:01 +0000 (09:22 +0100)]
Add editorconfig
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 21 Jul 2022 06:59:32 +0000 (08:59 +0200)]
swrap: Remove _close symbol alias as this breaks 'make test' on FreeBSD
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 21 Jul 2022 05:33:11 +0000 (07:33 +0200)]
tests: Fix compiler warning that arrays might not be initialized
tests/test_echo_tcp_sendmsg_recvmsg_fd.c: In function 'test_tcp_sendmsg_recvmsg_fd_mixed':
tests/test_echo_tcp_sendmsg_recvmsg_fd.c:477:9: error: '<unknown>' may be used uninitialized [-Werror=maybe-uninitialized]
477 | test_tcp_sendmsg_recvmsg_fd_array(fd_array, num_fds);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tests/test_echo_tcp_sendmsg_recvmsg_fd.c:103:13: note: by argument 1 of type 'const int *' to 'test_tcp_sendmsg_recvmsg_fd_array' declared here
103 | static void test_tcp_sendmsg_recvmsg_fd_array(const int *fds, size_t num_fds)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Thu, 21 Jul 2022 05:50:27 +0000 (07:50 +0200)]
cmake: Disable -Werror=strict-overflow till glibc is fixed
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Joseph Sutton [Thu, 29 Jul 2021 23:12:12 +0000 (11:12 +1200)]
swrap: Don't pass NULL pointers to memcpy()
Doing so is undefined behaviour.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Mon, 15 Mar 2021 07:24:57 +0000 (08:24 +0100)]
Bump version to 1.3.3
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 15 Mar 2021 07:22:33 +0000 (08:22 +0100)]
cmake: Define a library version for the noop lib
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 15 Mar 2021 07:18:15 +0000 (08:18 +0100)]
cmake: Export cmake config for socket_wrapper_noop
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 15 Mar 2021 07:17:34 +0000 (08:17 +0100)]
cmake: Set public include dir for socket_wrapper_noop target
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 15 Mar 2021 07:09:29 +0000 (08:09 +0100)]
cmake: Set PRIVATE for linking libraries
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 09:58:29 +0000 (10:58 +0100)]
swrap: wrap __close_nocancel() if available
While it's no possible to inject swrap__close_nocancel() into
libc.so.6 directly, because it's no weak symbol, it seems to
be possible to inject it to other glibc libraries like
libpthread.so.0, which is better than nothing.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 11:29:27 +0000 (12:29 +0100)]
swrap: export a public socket_wrapper_indicate_no_inet_fd() helper function
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 24 Feb 2021 11:45:26 +0000 (12:45 +0100)]
swrap: introduce a socket_wrapper_noop.so and socket_wrapper.h to provide noop stubs
Applications with the need to call socket_wrapper_enabled() should link
against -lsocket_wrapper_noop in order to resolve the symbol at
link time.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 11:14:06 +0000 (12:14 +0100)]
swrap: split out swrap_remove_wrapper() to handle swrap_close() and swrap_remove_stale()
Except of closing the fd, both should do the same, even indicating a TCP
close in the pcap file.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 11:14:06 +0000 (12:14 +0100)]
swrap: remember the libc_close() errno in swrap_close()
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 10:00:53 +0000 (11:00 +0100)]
swrap: call libc_write() directly for internal fds
Otherwise we may deadlock with a backtrace like this:
swrap_accept():
...
SWRAP_LOCK_SI(si);
swrap_pcap_dump_packet() ->
write() ->
swrap_write() ->
SWRAP_LOCK_SI(si) -> abort()
This can happen if libc_open() called from swrap_pcap_get_fd()
return a stale fd. This may happen if glibc calls socket()
and closes it with __close_nocancel() instead of close().
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Wed, 17 Feb 2021 10:41:38 +0000 (11:41 +0100)]
swrap: warn about unreachable addresses
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14640
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Wed, 10 Feb 2021 09:49:30 +0000 (10:49 +0100)]
Bump version to 1.3.2
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Wed, 10 Feb 2021 09:45:11 +0000 (10:45 +0100)]
swrap: don't read the callers msg_control buffer in swrap_recvmsg_before_unix()
For recvmsg() msg_control is a write only buffer, that is filled by the
kernel, but the kernel won't read from that buffer. So we shouldn't
read from (copy) it either.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Tue, 9 Feb 2021 15:09:59 +0000 (16:09 +0100)]
swrap: don't touch msg_tmp in swrap_recvmsg_after_unix() on error
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Tue, 9 Feb 2021 14:57:22 +0000 (15:57 +0100)]
swrap: fix possible memory leak between swrap_recvmsg_{before,after}_unix()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Fri, 5 Feb 2021 14:55:49 +0000 (15:55 +0100)]
Bump version to 1.3.1
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 15:25:43 +0000 (16:25 +0100)]
test_echo_tcp_sendmsg_recvmsg_fd: add test_tcp_sendmsg_recvmsg_fd_mixed() tests
Here we mix sockets and other valid file descriptors.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 13:40:45 +0000 (14:40 +0100)]
test_echo_tcp_sendmsg_recvmsg_fd: add test_tcp_sendmsg_recvmsg_fd_different() tests
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 13:20:16 +0000 (14:20 +0100)]
test_echo_tcp_sendmsg_recvmsg_fd: also test passing the same socket up to 6 times
Note SWRAP_MAX_PASSED_FDS is currently 6.
This test demonstrates that even 64-bit systems required commit:
"swrap: fix fd-passing without 4 padding bytes"
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 13:10:03 +0000 (14:10 +0100)]
test_echo_tcp_sendmsg_recvmsg_fd: split out test_tcp_sendmsg_recvmsg_fd_same()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 4 Feb 2021 16:04:30 +0000 (17:04 +0100)]
test_echo_tcp_sendmsg_recvmsg_fd: split out test_tcp_sendmsg_recvmsg_fd_array()
This will allow us to test more combinations in order to
get better coverage. For now we just test a single fd.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 14:41:27 +0000 (15:41 +0100)]
tests/echo_srv: allow more than once tcp connection at a time
We should not wait for the last connection to disconnect,
there would not be any reason to use fork at all.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 4 Feb 2021 15:20:13 +0000 (16:20 +0100)]
swrap: fix fd-passing without 4 padding bytes
We noticed the problem on 32 bit platforms and sending a single
application fd, the hidden pipe-fd doesn't fit into the padding
bytes. This can also happen on 64 bit platforms and an even number
of application fds.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 18:36:26 +0000 (19:36 +0100)]
swrap: fix invalid read in swrap_sendmsg_unix_scm_rights()
Here the fds_out array is larger than the fds_in array, so we can
only copy the fds_in array using size_fds_in, leaving the last slot
of fds_out untouched, which is filled by fds_out[num_fds_in] = pipefd[0]
later.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 4 Feb 2021 04:02:32 +0000 (05:02 +0100)]
swrap: fix copy on write leak of ~38M for every fork.
commit
0f8e90dd7e59c473be615dee08d445dca98fdab9
(src/socket_wrapper.c: fix mutex fork handling)
let us touch the whole sockets array on every fork,
because each element in the array has it's own mutex.
max_sockets=65535 * sizeof(struct socket_info_container)=592 =
38796720
This was designed for the use of robust shared mutexes
when moving the sockets array into a shared memory file.
Until we really move to shared memory, we can use a single
global mutex in order to avoid the copy on write leaking.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 10:50:17 +0000 (11:50 +0100)]
swrap: abort on mutex errors
There's no way to continue in a reliable way...
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 12:13:44 +0000 (13:13 +0100)]
swrap: fallback to libc_getpeername() when we get an empty sun_path from accept()
This hopefully hides the strange behaviour of FreeBSD (at least 12.1)
for already disconnected AF_UNIX sockets.
The race is triggered when the following detects the usage of 'getpeername':
truss -o ./truss.out -f -H -a -e -D -s 160 ctest -V -R test_thread_echo_tcp_connect;
grep getpeername truss.out
In a simplified log the following is happening:
ECHO_SRV(parent): socket(PF_LOCAL,SOCK_STREAM,0) = 4 (0x4)
ECHO_SRV(parent): unlink("/tmp/w_E37bkf/T0A0007") ERR#2 'No such file or directory'
ECHO_SRV(parent): bind(4,{ AF_UNIX "/tmp/w_E37bkf/T0A0007" },106) = 0 (0x0)
ECHO_SRV(parent): listen(4,16) = 0 (0x0)
...
ECHO_SRV(parent): write(2,"SWRAP_ERROR[echo_srv (9792)] - swrap_accept: before accept(sa_socklen=106)\n",75) = 75 (0x4b)
ECHO_SRV(parent): accept4(0x4,0x7ffffffde158,0x7ffffffde150,0x0) = 5 (0x5)
ECHO_SRV(parent): write(2,"SWRAP_ERROR[echo_srv (9792)] - swrap_accept: after accept(sa_socklen=106, family=1)\n",84) = 84 (0x54)
ECHO_SRV(parent): getsockname(5,{ AF_UNIX "/tmp/w_E37bkf/T0A0007" },0x7ffffffde0c0) = 0 (0x0)
ECHO_SRV(parent): swrap_accept() returned a valid connection and a per connection child (pid=9793) handles it
TEST_THREAD: socket(PF_LOCAL,SOCK_STREAM,0) = 7 (0x7)
TEST_THREAD: bind(7,{ AF_UNIX "/tmp/w_E37bkf/T014D4F" },106) = 0 (0x0)
TEST_THREAD: connect(7,{ AF_UNIX "/tmp/w_E37bkf/T0A0007" },106) = 0 (0x0)
TEST_THREAD: close(7) = 0 (0x0)
ECHO_SRV(parent): wait4(-1,0x0,0x0,0x0) = 9793 (0x2641)
ECHO_SRV(parent): close(5) = 0 (0x0)
ECHO_SRV(parent): write(2,"SWRAP_ERROR[echo_srv (9792)] - swrap_accept: before accept(sa_socklen=106)\n",75) = 75 (0x4b)
ECHO_SRV(parent): accept4(0x4,0x7ffffffde158,0x7ffffffde150,0x0) = 5 (0x5)
TEST_THREAD: unlink("/tmp/w_E37bkf/T014D4F") = 0 (0x0)
ECHO_SRV(parent): write(2,"SWRAP_ERROR[echo_srv (9792)] - swrap_accept: after accept(sa_socklen=16, family=1)\n",83) = 83 (0x53)
ECHO_SRV(parent): getpeername(5,0x7ffffffde158,0x7ffffffde150) ERR#57 'Socket is not connected'
ECHO_SRV(parent): getsockname(5,{ AF_UNIX "/tmp/w_E37bkf/T0A0007" },0x7ffffffde0c0) = 0 (0x0)
ECHO_SRV(parent): getpeername(5,0x7ffffffde158,0x7ffffffde150) ERR#57 'Socket is not connected'
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 11:13:12 +0000 (12:13 +0100)]
swrap: make swrap_accept() more resilient against races related to already disconnected sockets
Callers of accept() expect to get ECONNABORTED instead of a disconnected
socket.
Even on Linux we have a potential race calling libc_getsockname()
after accept(), so we map ENOTCONN to ECONNABORTED.
We should do all syscalls in order to have peer and sockname, before
doing in memory things like calling sockaddr_convert_from_un().
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 12:13:07 +0000 (13:13 +0100)]
swrap: add better logging to convert_un_in()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Fri, 5 Feb 2021 11:22:47 +0000 (12:22 +0100)]
tests/echo_srv: make the main server logic resilient to ECONNABORTED from accept()
That should fix a race where the connect() directly followed by close()
in test_thread_echo_tcp_connect will cause the echo_srv to terminate
early, which results in connect() returning ECONNREFUSED in for other
threads.
This mainly happens on FreeBSD, but it can also happen on Linux.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Tue, 2 Feb 2021 09:33:37 +0000 (10:33 +0100)]
Bump version to 1.3.0
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 2 Feb 2021 09:30:04 +0000 (10:30 +0100)]
doc: Add website to manpage
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 2 Feb 2021 09:28:59 +0000 (10:28 +0100)]
doc: Define doctype for manpage
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 2 Feb 2021 09:26:07 +0000 (10:26 +0100)]
doc: Rename socket_wrapper.1.txt to socket_wrapper.1.adoc
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Mon, 29 Jun 2020 12:10:54 +0000 (14:10 +0200)]
swrap: add basic support for fd-passing via SCM_RIGHTS
We only allow passing up to 6 fds in a single sendmsg call,
in order to keep the logic simple.
That's more than enough for Samba's use of fd-passing,
there we only pass a single fd and the fd will be
closed in the sender on success. It means it's ok
to keep the socket_info.io.pck_{snd,rcv} fields per process
and the PCAP generation will still work as expected.
If these constraints turn out to be a problem for other applications,
we need to change to a more complex design and move the socket_info array
into a shared memory file and use shared robust mutexes.
But for now we just want to support multi-channel testing in Samba.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11899
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Anoop C S [Thu, 23 Jul 2020 10:29:29 +0000 (15:59 +0530)]
tests: Add test for socket_wrapper fd-passing support
Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Tue, 30 Jun 2020 15:53:20 +0000 (17:53 +0200)]
swrap: always check new fd's against socket_fds_max and use libc_close() for cleanup
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Tue, 30 Jun 2020 15:51:29 +0000 (17:51 +0200)]
swrap: add const to swrap_add_socket_info()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 29 Jun 2020 07:39:22 +0000 (09:39 +0200)]
swrap: let swrap_sendmsg_before_unix() create a copy of msg_tmp.msg_control
With fd-passing we'll have to modify the content of it.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 29 Jun 2020 07:27:24 +0000 (09:27 +0200)]
swrap: add stubs for swrap_{sendmsg,recvmsg}_{before,after}_unix()
In order to implement fd-passing of socket_wrapper simulated sockets
we need to modify the msghdr structures from the callers.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 29 Jun 2020 11:07:26 +0000 (13:07 +0200)]
swrap: filter out SCM_{RIGHTS,CREDENTIALS} on inet socket
These are only valid on unix domain sockets and ignored otherwise (at
least on Linux).
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Mon, 29 Jun 2020 08:12:07 +0000 (10:12 +0200)]
swrap: add error checking/cleanup to swrap_sendmsg_filter_cmsghdr()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Anoop C S [Wed, 22 May 2019 05:15:37 +0000 (10:45 +0530)]
swrap: Fix MSGHDR check in sendmsg()
Check for msg_controllen and msg_control data members from msghdr
structure needs to be validated on the received omsg pointer rather
than on newly created msghdr struture inside the wrapper.
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 28 Jan 2021 11:40:14 +0000 (12:40 +0100)]
swrap: print out SOCKET_WRAPPER_PACKAGE and SOCKET_WRAPPER_VERSION on first use
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Thu, 28 Jan 2021 19:06:11 +0000 (20:06 +0100)]
src/socket_wrapper.c: Improve checks and debug output of socket_wrapper_dir()
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Tue, 26 Jan 2021 12:14:41 +0000 (13:14 +0100)]
src/socket_wrapper.c: fix mutex fork handling
We need to use pthread_mutex_init in the child handler...
See https://sourceware.org/bugzilla/show_bug.cgi?id=2745
Valgrind tools like helgrind and drd don't understand this
(at least in 3.15.0), they require a pthread_mutex_unlock()
in the child in order work.
Pair-Programmed-With: Andreas Schneider <asn@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Wed, 12 Feb 2020 09:56:13 +0000 (10:56 +0100)]
gitlab-ci: Add job for ThreadSanitizer
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Wed, 12 Feb 2020 09:56:13 +0000 (10:56 +0100)]
cmake: Add ThreadSanitizer compile flags
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Thu, 28 Jan 2021 13:31:31 +0000 (14:31 +0100)]
src/socket_wrapper.c: always go through swrap_bind_symbol_all() protected by pthread_once()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 28 Jan 2021 12:11:51 +0000 (13:11 +0100)]
src/socket_wrapper.c: split out _swrap_bind_symbol_generic()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Stefan Metzmacher [Thu, 28 Jan 2021 12:11:16 +0000 (13:11 +0100)]
src/socket_wrapper.c: remove unused swrap_bind_symbol_libnsl()
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Björn Jacke [Fri, 30 Oct 2020 14:44:53 +0000 (15:44 +0100)]
socket_wrapper.c/tests: fall back to pragma init/fini for constructor/destructor if possible
Signed-off-by: Bjoern Jacke <bjacke@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Björn Jacke [Fri, 30 Oct 2020 14:31:42 +0000 (15:31 +0100)]
configure: check for pragma init/fini for constructors/destructors
Signed-off-by: Bjoern Jacke <bjacke@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Andreas Schneider [Mon, 22 Jun 2020 14:53:36 +0000 (16:53 +0200)]
Bump version to 1.2.5
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Mon, 8 Jun 2020 12:21:25 +0000 (14:21 +0200)]
socket_wrapper.c: let swrap_vioctl() handle SIOCOUTQ/TIOCOUTQ/FIONWRITE explicitly
They are used to ask for the number of unacked bytes in the send queue,
with AF_UNIX sockets get strange result, on linux 5.3 I get more bytes
reported than I sent into the socket. All bytes reach the destination
directly, so we can just always report 0 unacked bytes.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit
f317ebcdcdd626ed9e06de2eb60031306994c803)
Stefan Metzmacher [Mon, 8 Jun 2020 12:18:44 +0000 (14:18 +0200)]
socket_wrapper.c: make FIONREAD handling more robust in swrap_vioctl()
We should only dereference the va args when the kernel already checked
they are valid.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit
c95b7cb1d7b9348472276edceff71889aa676d25)
Stefan Metzmacher [Fri, 19 Jun 2020 18:52:23 +0000 (20:52 +0200)]
test_echo_tcp_socket_options.c: add tests for TCP_INFO
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit
a37c0175492fb1b35257b785c71dea4e4f6d4750)
Stefan Metzmacher [Mon, 8 Jun 2020 08:32:28 +0000 (10:32 +0200)]
socket_wrapper.c: implement getsockopt(TCP_INFO) if the platform supports it
This just implements a few basics, which are required by Samba.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit
300de6e099ea82ee5361918de8c3abb389e0782d)
Martin Schwenke [Wed, 13 May 2020 23:57:24 +0000 (09:57 +1000)]
doc: Document SOCKET_WRAPPER_DIR_ALLOW_ORIG
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Martin Schwenke [Thu, 14 May 2020 00:01:54 +0000 (10:01 +1000)]
swrap: Add SOCKET_WRAPPER_DIR_ALLOW_ORIG to allow fall back
Instead of failing when the path returned by realpath(3) is too long,
if SOCKET_WRAPPER_DIR_ALLOW_ORIG is set then fall back to the
original value. If this original path is too long or something else
fails then abort.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Martin Schwenke [Wed, 13 May 2020 03:45:02 +0000 (13:45 +1000)]
swrap: Abort if socket wrapper directory is too long to be usable
If the socket wrapper directory path is too long to allow reliable
construction of the required Unix domain socket paths then
convert_in_un_alloc() can return ENFILE if paths are truncated in
unfortunate ways. This can be very hard to debug since, for example,
bind(2) should never return ENFILE.
Instead, abort if the path returned by realpath(3) is unusable.
The code structure is slightly weird but this accommodates an
additional change.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Martin Schwenke [Wed, 13 May 2020 03:23:21 +0000 (13:23 +1000)]
swrap: Add abstractions to construct Unix domain socket paths
These include overflow checks but the results of the checks are not
yet used.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Martin Schwenke [Sat, 13 Jun 2020 10:50:42 +0000 (20:50 +1000)]
swrap: Abort on failure to use SOCKET_WRAPPER_DIR
If SOCKET_WRAPPER_DIR is set the intention is to use socket wrapper.
Returning NULL means socket wrapper is disabled. The only sure way to
avoid running without socket wrapper is to abort.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 21:26:28 +0000 (23:26 +0200)]
Revert "socket_wrapper.c: implement getsockopt(TCP_INFO) if the platform supports it"
This reverts commit
300de6e099ea82ee5361918de8c3abb389e0782d.
This got merged automatically by gitlab without review.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 21:26:25 +0000 (23:26 +0200)]
Revert "test_echo_tcp_socket_options.c: add tests for TCP_INFO"
This reverts commit
a37c0175492fb1b35257b785c71dea4e4f6d4750.
This got merged automatically by gitlab without review.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 21:26:22 +0000 (23:26 +0200)]
Revert "socket_wrapper.c: make FIONREAD handling more robust in swrap_vioctl()"
This reverts commit
c95b7cb1d7b9348472276edceff71889aa676d25.
This got merged automatically by gitlab without review.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 21:25:40 +0000 (23:25 +0200)]
Revert "socket_wrapper.c: let swrap_vioctl() handle SIOCOUTQ/TIOCOUTQ/FIONWRITE explicitly"
This reverts commit
f317ebcdcdd626ed9e06de2eb60031306994c803.
This got merged automatically by gitlab without review.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 21:07:09 +0000 (21:07 +0000)]
Merge branch 'tcp_info' into 'master'
Add support for TCP_INFO and SIOCOUTQ/TIOCOUTQ/FIONWRITE
See merge request cwrap/socket_wrapper!11
Stefan Metzmacher [Mon, 8 Jun 2020 12:21:25 +0000 (14:21 +0200)]
socket_wrapper.c: let swrap_vioctl() handle SIOCOUTQ/TIOCOUTQ/FIONWRITE explicitly
They are used to ask for the number of unacked bytes in the send queue,
with AF_UNIX sockets get strange result, on linux 5.3 I get more bytes
reported than I sent into the socket. All bytes reach the destination
directly, so we can just always report 0 unacked bytes.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Mon, 8 Jun 2020 12:18:44 +0000 (14:18 +0200)]
socket_wrapper.c: make FIONREAD handling more robust in swrap_vioctl()
We should only dereference the va args when the kernel already checked
they are valid.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Fri, 19 Jun 2020 18:52:23 +0000 (20:52 +0200)]
test_echo_tcp_socket_options.c: add tests for TCP_INFO
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Stefan Metzmacher [Mon, 8 Jun 2020 08:32:28 +0000 (10:32 +0200)]
socket_wrapper.c: implement getsockopt(TCP_INFO) if the platform supports it
This just implements a few basics, which are required by Samba.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11897
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Mike Gilbert [Fri, 8 May 2020 03:28:50 +0000 (23:28 -0400)]
tests: do not truncate pid to 16 bits
On Linux, pid_t is a 32-bit type, and the kernel permits pids up to 22
bits in length.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14373
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Anoop C S [Tue, 7 Apr 2020 08:53:09 +0000 (14:23 +0530)]
swrap: Remove extra new line towards end of SWRAP_LOG()
Signed-off-by: Anoop C S <anoopcs@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Andreas Schneider [Mon, 23 Mar 2020 08:47:32 +0000 (09:47 +0100)]
Bump version to 1.2.4
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Mon, 23 Mar 2020 08:14:06 +0000 (09:14 +0100)]
Rename CHANGELOG
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 24 Mar 2020 07:44:20 +0000 (08:44 +0100)]
gitlab-ci: Add ubuntu runner
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 24 Mar 2020 07:54:59 +0000 (08:54 +0100)]
gitlab-ci: Remove debian cross mips build
This isn't supported by Debian anymore.
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Andreas Schneider [Tue, 24 Mar 2020 07:53:55 +0000 (08:53 +0100)]
tests: Fix a compile warning on Ubuntu
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>