When a static function only has one caller, it is often simpler to not
have the code in a separate function. Drop need_private_key_password()
and move it to need_secrets_tls().
g_warning() for unexpected scheme is not right. Either, this should be an
assertion (and never be hit), or the library should be silent about conditions
that can happen regularly.
I think code is easier to understand, if the difference (between phase1
and phase2) is pushed to the bottom. Having one large "if(phase2){}else{}"
at the top makes it harder to compare the two branches and see where
they differ.
When no radio hardware is present in the system, "nmcli radio"
currently displays:
WIFI-HW WIFI WWAN-HW WWAN
enabled enabled enabled enabled
which is misleading. Use the new RadioFlags property to display
"missing" in the *-HW columns when there is no hardware for the
given radio technology.
https://bugzilla.redhat.com/show_bug.cgi?id=1996918
Introduce a RadioFlags property on the manager object. For now it
contains two bits WLAN_AVAILABLE, WWAN_AVAILABLE to indicate whether
any radio interface is present in the system. The presence of a radio
is detected by looking at devices and rfkill switches.
In future, any radio-related read-only boolean flag can be exposed via
this property, including the already existing WirelessHardwareEnabled
and WwanHardwareEnabled properties.
Previously, only the daemon was writing keyfiles, and it ensures
that they are always valid.
As we now have this function as public API of libnm, we should drop this
restriction and write the profile the best we can. Granted, an invalid
profile may not be expressed in keyfile format, and the result is
undefined. But make the best of it.
This is severe. We cache the list of names, and we must invalidate the
cache when the names change. Otherwise, out-of-bound access and crash.
Fixes: d0192b698e ('libnm: add nm_setting_option_set(), nm_setting_option_get_boolean(), nm_setting_option_set_boolean()')
Fixes: 150af44e10 ('libnm: add nm_setting_option_get_uint32(), nm_setting_option_set_uint32()')
If __VA_ARGS__ contains odd arguments, it's not clear that N_ARG() gives
the same as the array initialization. Add a static assert that the
numbers agree to catch wrong usage of the macro.
For example:
nm_gobject_notify_together(setting, a, b, );
If we can't find a connection for any reason other than that it doesn't
exist, we should error out immediately and consistently, regardless of
whether we already encountered a non-existent connection.
This will allow migrating a connection. If specified, the connection will
be confined to a particular settings plugin when written back. If the
plugin differs from the existing one, it will be removed from the old one.
We expect to read NUL terminated strings. Upon NUL, we should do
something. Treat it as a line break.
Fixes: 8ae9cf4698 ('Revert "libnm: buffer reads in nm_vpn_service_plugin_read_vpn_details()"')
This partially reverts commit 4a9fcb0fc3, which replaced one-byte
reads with buffered ones in the VPN service plugin.
Unfortunately the buffering means that commands coming after the magic
"DONE" string were being pulled into the buffer. Secrets agents expect
a "QUIT" to come after the "DONE", and since with buffering "QUIT" was
in the buffer, this led to a twenty-second delay on every VPN
connection using a secrets manager.
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/1164
Fixes: 4a9fcb0fc3 ('libnm: buffer reads in nm_vpn_service_plugin_read_vpn_details()')
A couple of argument documentation strings were long enough to be broken
up to span over multiple lines. gdbus-codegen, on the other hand, was
of differing opinion and promptly punished the offense by garbling the
resulting docbook.
Merge the string into single lines.
This one attempts to document an option dictionary in a manner that has
been deemed criminal by the high court of gdbus-codegen.
Beat it back into shape.
There has been a lot wrong with this one. Aside from the messy
capitalization, it broke the argument documentation into multiple lines,
baffling gdbus-codegen, which, in turn, generated garbage documentation.
Overhaul it.
There has been a lot wrong with this one. Aside from the messy
capitalization, it broke the argument documentation into multiple lines,
baffling gdbus-codegen, which, in turn, generated garbage documentation.
Overhaul it.
The GMainLoop instance (and the default GMainContext singleton) is not
required for trivial operations like --print-config, --version or
--help). If running as SysV daemon, the event file descriptor is
unnecessarily dup'ed from the parent to the child process.
Signed-off-by: Christian Eggers <ceggers@arri.de>
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/1160
`msgfmt -vc` warns:
po/be@latin.po:2: warning: header field 'Language' still has the initial default value
Check all files and update the Language manually.
The documentation ([1]) says that either "ll", "ll_CC" or "ll_CC@variant"
is valid. Update accordingly.
[1] https://www.gnu.org/software/gettext/manual/html_node/Header-Entry.html
`msgfmt -vc` warns:
po/as.po:9: warning: header field 'Project-Id-Version' still has the initial default value
Update them all:
$ git grep -l Project-Id-Version | xargs sed 's/^"Project-Id-Version: \(.*\)\\n"$/"Project-Id-Version: NetworkManager\\n"/' -i
The documentation ([1]) says:
Project-Id-Version
This is the name and version of the package. Fill it in if it has not already
been filled in by xgettext.
but don't add the "version". We have these files in git on the devel branch,
so whenever we branch a new release, the version would change. Just say
"NetworkManager".
[1] https://www.gnu.org/software/gettext/manual/html_node/Header-Entry.html
