Commit Graph

226 Commits

Author SHA1 Message Date
Philip Withnall
8e963e0e31 Port internal uses to use g_source_set_static_name()
This should reduce allocations.

Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
2021-07-26 11:01:07 +01:00
Philip Withnall
a2e38fd28e gsocket: Use gsize to track native sockaddr’s size
Don’t use an `int`, that’s potentially too small. In practical terms,
this is not a problem, since no socket address is going to be that big.

By making these changes we can use `g_memdup2()` without warnings,
though. Fewer warnings is good.

Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #2319
2021-02-04 16:17:21 +00:00
Francesco Tamagni
f6ce5739f8 gsocket: Fix SO_NOSIGPIPE regression on Darwin
Where the early call to g_socket_set_option() fails because of
check_socket() failing due to `inited` still being FALSE.

This brings 634b692 back into working order, by fixing the regression
introduced in 39f047e.

Co-authored-by: Ole André Vadla Ravnås <oleavr@gmail.com>
2021-01-21 21:17:24 +01:00
s1341
9204c346d4 gsocket: Decrease msg_control buffer size for QNX 2021-01-19 23:17:30 +01:00
Emmanuel Fleury
b6f7e4678b Another fix on g_socket_send_message()
We forgot to take into account the case where num_vectors is '-1'.
2021-01-13 13:29:52 +01:00
Michael Catanzaro
dfdab13682 Merge branch 'more_on_g_socket_send_message' into 'master'
Fixing g_socket_send_message() documentation to make it clearer

See merge request GNOME/glib!1876
2021-01-12 18:10:05 +00:00
Philip Withnall
8d78b92794 Merge branch 'fix_more_warnings' into 'master'
Fix more warnings

See merge request GNOME/glib!1823
2021-01-12 18:03:06 +00:00
Emmanuel Fleury
7e00091b60 Fixing g_socket_send_message() documentation to make it clearer 2021-01-12 18:16:50 +01:00
Emmanuel Fleury
26ec52b9a2 Fix signedness warnings in gio/gsocket.c
gio/gsocket.c: In function ‘g_socket_send_message_with_timeout’:
gio/gsocket.c:4528:23: error: comparison of integer expressions of different signedness: ‘gint’ {aka ‘int’} and ‘guint’ {aka ‘const unsigned int’}
 4528 |         for (i = 0; i < _message->num_vectors; i++) \
      |                       ^

gio/gsocket.c: In function ‘g_socket_send_message_with_timeout’:
gio/gsocket.c:4543:19: error: comparison of integer expressions of different signedness: ‘gint’ {aka ‘int’} and ‘guint’ {aka ‘const unsigned int’}
 4543 |     for (i = 0; i < _message->num_control_messages; i++) \
      |                   ^

gio/gsocket.c: In function ‘g_socket_send_messages_with_timeout’:
gio/gsocket.c:5133:19: error: comparison of integer expressions of different signedness: ‘gint’ {aka ‘int’} and ‘guint’ {aka ‘unsigned int’}
 5133 |     for (i = 0; i < num_messages; ++i)
      |                   ^
gio/gsocket.c:5152:33: error: comparison of integer expressions of different signedness: ‘gint’ {aka ‘int’} and ‘guint’ {aka ‘unsigned int’}
 5152 |     for (num_sent = 0; num_sent < num_messages;)
      |                                 ^
2021-01-12 15:48:19 +01:00
Emmanuel Fleury
0705dbd17b Fix missing initializer warnings in gio/gsocket.c
gio/gsocket.c:3761:1: error: missing initializer for field ‘closure_callback’ of ‘GSourceFuncs’ {aka ‘struct _GSourceFuncs’}
 3761 | };
      | ^
gio/gsocket.c:4071:1: error: missing initializer for field ‘closure_marshal’ of ‘GSourceFuncs’ {aka ‘struct _GSourceFuncs’}
 4071 | };
      | ^
2021-01-12 15:48:19 +01:00
Emmanuel Fleury
7197ad3c40 Fix possible integer overflow of g_socket_send_message()
The explanation of this bug has been mentioned in !1823, basically
it fixes some possible integer overflow when message buffer size
is more than G_MAXSSIZE.
2021-01-12 15:17:02 +01:00
Sebastian Dröge
65271eb93d Merge branch 'fix/default-windows-socket-udp-behavior' into 'master'
gsocket: Improve default UDP behavior on Windows

See merge request GNOME/glib!1844
2021-01-07 14:09:19 +00:00
Ole André Vadla Ravnås
17c53b5f16 gsocket: Don't call WSAEnumNetworkEvents if socket is closed
Makes Application Verifier happy.
2021-01-05 21:20:38 +01:00
Ole André Vadla Ravnås
d01588f37e gsocket: Don't bother updating select mask if socket is closed 2021-01-05 21:20:34 +01:00
Pascal Buhler
9cd134d9fb gsocket: Improve default UDP behavior on Windows
An ICMP port unreachable will result in a socket error, which is a
really bad default for UDP sockets.
2021-01-05 21:08:15 +01:00
Ole André Vadla Ravnås
9ac3a27f03 gsocket: Fix credentials error-handling on Apple OSes
- When querying a TCP socket, getsockopt() may succeed but the resulting
  `optlen` will be zero. This means we'd previously be reading
  uninitialized stack memory in such cases.
- After a file-descriptor has gone through FD-passing, getsockopt() may
  fail with EINVAL. At least this is the case with TCP sockets.
- While at it also use SOL_LOCAL instead of hard-coding its value.
2021-01-04 12:29:47 +00:00
Marco Mastropaolo
43b13d7a1a Windows: fix FD_READ condition flag still set on recoverable UDP socket errors.
Contrary to what the WSARecvFrom seem to imply, a UDP socket is perfectly recoverable and usable after a WSAECONNRESET error (and, I assume, WSAENETRESET).
However GSocket condition has the FD_READ bit set after a UDP socket fails with WSAECONNRESET, even if no data is available on the socket anymore; this causes select calls to report the socket as readable when, in fact, it's not.

The change resets FD_READ flag on a socket upon the above error conditions; there's no 'if' to filter between datagram and stream sockets as the change should be harmless in the case of stream sockets which are, however, very unlikely to be usable after a WSAECONNRESET.
2020-12-24 09:08:40 +00:00
Phil Clayton
4a848a0aca gio: add missing (out) annotations to g_socket_receive*
This makes the introspection interface to g_socket_receive* consistent
with the interface to g_input_stream_read*.

Closes #2176
2020-08-11 18:34:16 +01:00
Nirbheek Chauhan
4e485d76ac gio: Remove broken support for XP
We now require Windows 7 or newer, and the networking code hasn't
worked in a long time since we directly use symbols from iphlapi.dll
now.
2020-07-26 21:30:05 +05:30
Philip Withnall
39f047e821 gsocket: Add missing check_socket() checks
This makes no great difference, just increases robustness a bit.

Signed-off-by: Philip Withnall <withnall@endlessm.com>

Closes: #1339
2020-06-18 12:52:39 +01:00
Philip Withnall
00bfb3ab44 tree: Fix various typos and outdated terminology
This was mostly machine generated with the following command:
```
codespell \
    --builtin clear,rare,usage \
    --skip './po/*' --skip './.git/*' --skip './NEWS*' \
    --write-changes .
```
using the latest git version of `codespell` as per [these
instructions](https://github.com/codespell-project/codespell#user-content-updating).

Then I manually checked each change using `git add -p`, made a few
manual fixups and dropped a load of incorrect changes.

There are still some outdated or loaded terms used in GLib, mostly to do
with git branch terminology. They will need to be changed later as part
of a wider migration of git terminology.

If I’ve missed anything, please file an issue!

Signed-off-by: Philip Withnall <withnall@endlessm.com>
2020-06-12 15:01:08 +01:00
Simon McVittie
44c004c84e Normalize C source files to end with exactly one newline
Some editors automatically remove trailing blank lines, or
automatically add a trailing newline to avoid having a trailing
non-blank line that is not terminated by a newline. To avoid unrelated
whitespace changes when users of such editors contribute to GLib,
let's pre-emptively normalize all files.

Unlike more intrusive whitespace normalization like removing trailing
whitespace from each line, this seems unlikely to cause significant
issues with cherry-picking changes to stable branches.

Implemented by:

    find . -name '*.[ch]' -print0 | \
    xargs -0 perl -0777 -p -i -e 's/\n+\z//g; s/\z/\n/g'

Signed-off-by: Simon McVittie <smcv@collabora.com>
2020-06-10 09:48:02 +01:00
Dr. Michael Lauer
ec2f60a008 gio: add gcredential support for macOS
[smcv: Apply my review feedback from
<https://bugzilla.gnome.org/show_bug.cgi?id=668866>]

Co-authored-by: Simon McVittie <smcv@collabora.com>
Resolves: https://gitlab.gnome.org/GNOME/glib/issues/507
2020-05-07 14:19:16 +01:00
Simon McVittie
e08dffb71b gsocket: Improve diagnostics on bind() failure
This is in an attempt to diagnose GNOME/glib#1912.

Signed-off-by: Simon McVittie <smcv@collabora.com>
2019-11-04 15:09:15 +00:00
Philip Withnall
8f385b8cf5 Merge branch 'master' into 'master'
windows: fix multicast socket binding to specific network interfaces

Closes #1635

See merge request GNOME/glib!887
2019-07-29 11:10:40 +00:00
Riccardo Bortolato
2b1a9219f1 windows: fix multicast socket binding to specific network interfaces
v7, based on a patch by mrgard (GNOME/glib#1635)
make w32_adapter_ipv4_addr() C90-compliant
check for ERROR_BUFFER_OVERFLOW when calling GetAdaptersAddresses()
code-style fixes
indentation fixes
use g_try_(re)alloc and g_free
style suggestions by pwithnall
drop uni_count variable
cap maximum allowed interface name string length according to windows documentation

Fixes: #1635
2019-07-12 15:21:53 +02:00
Simon McVittie
5e24d7cd14 Document where we expect credentials-passing to be supported
This is useful information for implementors of portable software to know
whether they can rely on credentials-passing.

Signed-off-by: Simon McVittie <smcv@collabora.com>
2019-06-11 11:06:35 +01:00
Sebastian Dröge
20a2c379c4 Clamp number of vectors to IOV_MAX / UIO_MAXIOV for GOutputStream writev() calls in the implementations
More vectors will give an error and we can simply clamp here and
consider it like a short write instead.

In case of GSocketOutputStream this is done here instead of inside
GSocket before calling sendmsg() because we we can't generically handle
short writes when sending messages on a socket, e.g. for datagram
sockets this causes only part of the datagram to be sent and an error
would be more useful in this case than sending corrupted data.

Also reduce the fallback limit to 16 in gsocket.c as that's the minimum
value required by POSIX and add a static assertion that the limit is
never bigger than G_MAXINT as that's the type recvmmsg/sendmmsg take.
2019-05-31 12:51:59 +03:00
Philip Withnall
e9389efe5b gsocket: Clarify in docs that flags arguments can be platform specific
As suggested by Philip Chimento.

Signed-off-by: Philip Withnall <withnall@endlessm.com>
2019-04-29 13:21:08 +01:00
Philip Withnall
49b807c022 gsocket: Remove (type) annotation from flags arguments
This is essentially a revert of commit
cc7f2f6b28. While those `flags` arguments
do accept values of type GSocketMsgFlags, they also accept OS-specific
flags which are not defined in GSocketMsgFlags.

The use of (type GSocketMsgFlags) makes language bindings like GJS
rightfully assert that values passed in to the argument only contain
flags from GSocketMsgFlags, which precludes the use of OS-specific
flags, and hence breaks various bits of code.

See https://gitlab.gnome.org/GNOME/gjs/issues/227#note_460136 and
https://gitlab.gnome.org/GNOME/glib/merge_requests/710#note_460249.

Signed-off-by: Philip Withnall <withnall@endlessm.com>

Fixes: https://gitlab.gnome.org/GNOME/gjs/issues/227
2019-03-15 10:28:14 +00:00
Andrea Azzarone
d1cb974932 socket: Fix annotation for flags in g_socket_receive_message
Closes: https://gitlab.gnome.org/GNOME/gjs/issues/227
2019-03-06 18:16:52 +00:00
Chun-wei Fan
bb73a22448 Win32: gio/gsocket.c: Set WSAEWOULDBLOCK on G_POLLABLE_RETURN_WOULD_BLOCK
To make things consistent across the board as that is the WinSock2 error
code that is received by g_socket_send_message_with_timeout() when it
returns G_POLLABLE_RETURN_WOULD_BLOCK.
2019-02-05 11:15:50 +08:00
Sebastian Dröge
9ae40d982a Rename timeout variables in gsocket.[ch] to include the unit as suffix
Makes it clearer which unit we work with. We have timeouts in seconds,
milliseconds and microseconds in here.
2019-01-24 16:25:50 +02:00
Sebastian Dröge
cc7f2f6b28 Add (type GSocketMsgFlags) to int-typed flags parameters in GSocket 2019-01-24 16:25:48 +02:00
Sebastian Dröge
f0a11b2727 Add g_socket_send_message_with_timeout() 2019-01-24 16:25:46 +02:00
Silvio Lazzeretti
9e89749e52 Partially revert "GSocket: Fix race conditions on Win32 if multiple threads are waiting on conditions for the same socket"
This partially reverts commit 799f8dcd46.
This patch seems to break the writability status of the server socket: once
somebody writes to it with success, then it reports it is not writable
anymore. Also, when the client socket has the flag FD_CONNECT set once,
it is never cleared and then it reports it is always writable, also when
it is not.
2018-12-14 16:26:59 +01:00
Ting-Wei Lan
293c103a7d socket: Fix get_available_bytes on systems other than Linux and Windows
FIONREAD ioctl on Linux reports the size of payload on UDP sockets.
However, other systems usually add internal header size to the reported
size, which vary between different operating systems and socket types.
To make it work on more systems, we should follow what we do on Windows
instead of using this unreliable FIONREAD ioctl.

This fixes socket test on FreeBSD.
2018-06-09 10:02:50 +08:00
Christoph Reiter
97c28f7fe1 ci: fix warnings and enable --werror for the mingw build
Fix various warnings regarding unused variables, duplicated
branches etc by adjusting the ifdeffery and some missing casts.

gnulib triggers -Wduplicated-branches in one of the copied files,
disable as that just makes updating the code harder.

The warning indicating missing features are made none fatal through
pragmas. They still show but don't abort the build.

https://bugzilla.gnome.org/show_bug.cgi?id=793729
2018-04-25 17:23:50 +02:00
Xavier Claessens
7efd76dd67 struct ip_mreq_source definition is broken on Android NDK <= r16
This fix the build on Android r16 and older, see:
https://issuetracker.google.com/issues/36987220

https://bugzilla.gnome.org/show_bug.cgi?id=740791
2018-04-24 15:00:17 -04:00
Xavier Claessens
51e5324493 Revert "gsocket: Fix build error on Android"
This was the wrong fix, the real cause is a bug in Android NDK r16:
https://issuetracker.google.com/issues/36987220

This reverts commit 994dd17ee5.
2018-04-20 16:37:35 -04:00
Xavier Claessens
994dd17ee5 gsocket: Fix build error on Android
imr_interface.s_addr is not defined in the HAVE_IP_MREQN case.

https://bugzilla.gnome.org/show_bug.cgi?id=740791
2018-04-20 16:09:43 -04:00
Philip Withnall
d8fe926ba4 Fix various strict aliasing problems with sockaddr
Fix various strict aliasing problems caused by casting between (struct
sockaddr *) and (struct sockaddr_storage *): the correct code here is to
keep the two in a union.

Signed-off-by: Philip Withnall <withnall@endlessm.com>

https://bugzilla.gnome.org/show_bug.cgi?id=791622
2018-01-08 11:50:26 +00:00
Michael Catanzaro
4b07869462 socket: actually remove fd from poll when socket is closed
In my previous patch, I failed to call g_source_remove_unix_fd() in
order to actually stop polling the fd of the closed socket.

The test did not catch this, because the test only checks that the right
source callback is dispatched properly. I don't know how to test this.

https://bugzilla.gnome.org/show_bug.cgi?id=723655
2017-10-25 09:36:07 -05:00
Michael Catanzaro
f99045fd03 socket: Don't poll the socket fd after close
This prevents polling on file descriptors that are no longer in use
or have been reused for something else.

Based on a patch by Mikhail Zabaluev

https://bugzilla.gnome.org/show_bug.cgi?id=723655
2017-10-24 09:02:02 -05:00
Julien Isorce
ea725a6414 gio: add g_socket_join_multicast_group_ssm (IGMPv3 SSM)
It adds support for source-specific multicast IGMPv3.

Allow receiving data only from a specified source when joining
a multicast group.

g_socket_join_multicast_group_ssm can be called multiple times
to allow receiving data from more than one source.

Support IPv4 and IPv6.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=740791
2017-10-16 11:14:57 +01:00
Руслан Ижбулатов
6abdc06da6 W32: Bump target NT version to 0x601 (7 or newer)
Also remove now-unnecessary if_nametoindex() implementation
(the HAVE_IF_NAMETOINDEX configure check didn't work correctly, it turned out),
which prevents glib from building. if_nametoindex() is available in lphlpapi since
Vista[1], so we don't need a compatibility function for it anymore, as Windows 7
is the new minimally-required version.

[1] https://msdn.microsoft.com/en-us/library/windows/desktop/bb408409(v=vs.85).aspx

https://bugzilla.gnome.org/show_bug.cgi?id=788180
2017-10-11 12:29:55 +01:00
grindhold
5b64522fcb socket: fix typo in get_remote_address docs
fixed the following typo:
"remove address" → "remote address"

https://bugzilla.gnome.org/show_bug.cgi?id=788766
2017-10-10 11:31:49 +01:00
Philip Withnall
5cddde1fb2 Consistently save errno immediately after the operation setting it
Prevent the situation where errno is set by function A, then function B
is called (which is typically _(), but could be anything else) and it
overwrites errno, then errno is checked by the caller.

errno is a horrific API, and we need to be careful to save its value as
soon as a function call (which might set it) returns. i.e. Follow the
pattern:
  int errsv, ret;
  ret = some_call_which_might_set_errno ();
  errsv = errno;

  if (ret < 0)
    puts (strerror (errsv));

This patch implements that pattern throughout GLib. There might be a few
places in the test code which still use errno directly. They should be
ported as necessary. It doesn’t modify all the call sites like this:
  if (some_call_which_might_set_errno () && errno == ESOMETHING)
since the refactoring involved is probably more harmful than beneficial
there. It does, however, refactor other call sites regardless of whether
they were originally buggy.

https://bugzilla.gnome.org/show_bug.cgi?id=785577
2017-08-03 10:21:13 +01:00
Colin Walters
017f78d77f gtype: Add private DEFINE_TYPE with prelude to workaround gtype deadlocks
And use it in GSocket, as it had a real-world case reported.

https://bugzilla.gnome.org/show_bug.cgi?id=674885
2017-06-14 14:45:45 -04:00
Sébastien Wilmet
3bf4a720c3 gio/: LGPLv2+ -> LGPLv2.1+
Sub-directories inside gio/ already processed in a previous commit:
- fam/
- gdbus-2.0/ (which contains only codegen/)
- gvdb/
- inotify/
- tests/
- win32/
- xdgmime/

Other sub-directories inside gio/:
- completion/: no license headers
- kqueue/: not LGPL, BSD-style license

https://bugzilla.gnome.org/show_bug.cgi?id=776504
2017-05-29 19:53:34 +02:00