Refactor nm-bluez-device.c to use GDBus both to connect to
BlueZ 4 and BlueZ 4.
Also remove the unused property RSSI.
Also prefix every logline with the dbus path of the device.
Signed-off-by: Thomas Haller <thaller@redhat.com>
NMBluezManager is now a proxy and only delegates to either
NMBluez4Manager or NMBluez5Manager. It detects the running BlueZ
version at runtime, and once it decides for one version, it cannot be
changed anymore as long NetworkManager is running.
This means, when switching from BlueZ4 to BlueZ5 or vice versa you have
to restart NetworkManager. This should be acceptable, because it is
not a common use case (most systems won't have both versions installed
anyway) and it greatly simplifies implementation.
Also note that NMBluez4Manager and NMBluez5Manager do not implement a
common interface. NMBluezManager delegates to the correct manager.
Having them share an common interface or base class would not simplify
the code, because NMBluezManager not only delegates, but it also acts as
a proxy until it is decided which BlueZ version is running. So, this
proxy-like behaviour would still be needed. The alternative would be to
merge the functionality of all three NMBluez*Manager classes into one.
This also removes the --enable-bluez4 configure switch, because both
versions are now always enabled.
https://bugzilla.gnome.org/show_bug.cgi?id=709412
Signed-off-by: Thomas Haller <thaller@redhat.com>
If the WiMAX plugin isn't installed, or the WiMAX device isn't
recognized, NetworkManager shouldn't treat the interface as
regular ethernet since the device requires specific setup to
be ready for IP configuration, which of course NetworkManager
can't do because the WiMAX plugin isn't loaded. Ignore them
instead.
Plugin owns the object and callers must reference it if they wish to use it outside
of the function they called "add" from. Likewise, callers of the ConnectionProvider's
add_connection method must also reference the returned object if they wish to
continue using it.
Actually, this case should no longer happen, but just to be sure:
when a udev remove event without ifindex comes, get the ifindex from
the cache and announce the device removal.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Before NMPlatform landed, the old NMManager code looked at either
DEVTYPE=wlan or asked the internal wifi utilities whether the
device was WiFi or not. This got lost when moving to NMPlatform.
It turns out that only mac80211-based drivers set the DEVTYPE=wlan
flag in sysfs, while older WEXT, out-of-tree, and staging drivers
often do not (though they should).
To avoid breaking recognition of these crappy drivers that used
to work, re-add the wifi utils checks.
Although it's convenient in some places to have IP configs on all
connections, it makes more sense in other places to not have IP
configs on slaves. (eg, it's confusing for nmcli, etc, to report a
full NMSettingIP4Config on a slave device). So revert parts of the
earlier patch. However, it's still safe to assume that s_ip4 != NULL
if method != DISABLED, so some of the earlier simplifications can
stay.
Also, add nm_utils_get_ip_config_method(), which returns the correct
IP config method for a connection, whether the connection has IP4 and
IP6 settings objects or not, and use that to keep some more of the
simplifications from the earlier patch.
ifcfg-rh had the rule that if an ifcfg file had no BOOTPROTO and no
IPv4 addresses, then it should be treated as method=auto for
compatibility. But in fact, current ifup treats it as method=disabled,
so we should too.
https://bugzilla.gnome.org/show_bug.cgi?id=708875
Make sure that all connections returned from NMSettings or created via
AddAndActivateConnection have an NMSettingIP4Config and an
NMSettingIP6Config, with non-NULL methods, and get rid of
now-unnecessary checks for those.
Also move the slaves-can't-have-IP-config checks into the
platform-independent code as well. This also gets rid of spurious
"ignoring IP4/IP6 configuration" warnings in ifcfg-rh when reading a
slave ifcfg file.
Partly based on a patch from Pavel.
https://bugzilla.gnome.org/show_bug.cgi?id=708875
To present a consistent API to clients, the IP-related properties
are only valid when the device has finished IP configuration. But
they are set before that happens, and their change notifications
were emitted before the IP configuration was considered valid.
Re-emit the change notifications when the device enters the IP_CHECK
state (and thus has IP configuration) and also when the device
deactivates to enusre clients have up-to-date IP-related property
information.
For the changes to has_ip_config(), the priv->ipX_state checks are
not necessary since the device will have valid IP configuration
when it enters the IP_CHECK state. The other checks can be
consolidated into a single statement.
Acked-by: Dan Winship
If the user disabled IPv6 support in the kernel with "ipv6.disable=1" on the
kernel boot line, then any attempts to open IPv6 sockets (which libndp does)
will fail. This failed the entire connection, even if IPv6's "may-fail"
property was TRUE. Instead, just fail IPv6 and allow IPv4 to proceed. If
IPv4 fails or is disabled, then other logic will fail the entire connection.
connection_from_file() requires the 'error' parameter. Not passing a
valid 'error' parameter causes the function to fail and return NULL,
which mean that commit_changes() would always re-write the connection
instead of ignoring commits where nothing has actually changed.
connection_from_file() no longer requires the unmanaged, keyfile,
or routefile parameters, so remove them.
Some RA implementations (like radvd) dump whatever the user configures
onto the wire, accepting a prefix of "2001:db8:1:0::1/64" without
masking the host bits off.
This causes NetworkManager to send that route down to the kernel, which
*does* mask the host bits off. This causes a mismatch between the
route NetworkManager expects the kernel to create, and what the kernel
actually created, when searching for the kernel object in the platform's
refresh_object() function:
cache = choose_cache (platform, object);
cached_object = nl_cache_search (choose_cache (platform, object), object);
kernel_object = get_kernel_object (priv->nlh, object);
kernel_object is NULL since 'object' (a route which came from the RA
prefix) is not the same as the object the kernel actually did create.
Ensure we match kernel behavior by fixing up prefixes for dumb router
advertisement services.
g_child_watch_add only emits one signal and gets automatically
removed from the source. No need for double removal, so unset
teamd_process_watch before calling teamd_cleanup.
Signed-off-by: Thomas Haller <thaller@redhat.com>
g_bus_watch_name immediatly emits a vanished signal, if the name does not exist
at the begining. So the first signal, does not indicate an actual change of
state and must be ignored.
Signed-off-by: Thomas Haller <thaller@redhat.com>
teamd was being watched using g_bus_watch_name(). But when the NM team device
was destroyed NM tried to remove the watcher with g_source_remove() instead of
g_bus_unwatch_name(). Thus the watcher was not removed and
teamd_dbus_appeared() was called on disposed device.
https://bugzilla.redhat.com/show_bug.cgi?id=1013593#c28
Signed-off-by: Thomas Haller <thaller@redhat.com>
53e55aab36 mistakenly moved the address
prefix without moving the memset() to initialize the address. Make
sure the address is initialized before trying to do anything with it.
The variable teamd_on_dbus in nm-device-team was never set to TRUE.
Remove it. Also, before teamd_dbus_vanished did never cleanup anything.
Signed-off-by: Thomas Haller <thaller@redhat.com>
The assertion that priv->teamd_pid is set does not hold.
For example, if no teamd binary can be found, "Stage 1 of 5 (Device
Prepare)" fails with "Activation (nm-team) to start teamd: not found".
This causes the device state to become "failed" and later
"disconnected", which in turn calls teamd_stop without a valid teamd_pid
set.
Signed-off-by: Thomas Haller <thaller@redhat.com>
NMDeviceTeam dispose() did not chain up to parent dispose(). This prevented
parent dispose() to be called (only finalize()) and thus link_changed_cb()
and device_ip_changed() handlers were not disconnected. Later these handlers
were called on invalid device resulting in a crash.
https://bugzilla.redhat.com/show_bug.cgi?id=1013593
ndp_open() can fail, do not assume it cannot.
When 'ipv6.disable=1' kernel parameter is used, the whole IPv6 stack is
disabled and the attempt to create PF_INET6 socket fails.
https://bugzilla.redhat.com/show_bug.cgi?id=1012151
The errors were documented as org.freedesktop.NetworkManager.VPN.Error.*,
but the actual values were org.freedesktop.NetworkManager.VPN.Plugin.*
Also update the errors documentation.
The ConnectInteractive() -> Connect() fallback code doesn't work, because
_connect_internal() changes the state to NM_VPN_SERVICE_STATE_STARTING before
checking if it can implement ConnectInteractive(), and then when the Connect()
call comes in, the VPN is not in STOPPED or INIT, so it returns an error.
The commit moves setting state to STARTING after the ConnectInteractive() check
availability, in the plugin. We introduce new plugin error and set it when the
the plugin does not implement ConnectInteractive(). NetworkManager uses this
error for ConnectInteractive() -> Connect() fallback.
https://bugzilla.gnome.org/show_bug.cgi?id=708255
Don't unconditionally call disconnect during device deactivation, since
the device also gets deactivated in the UNAVAILABLE -> DISCONNECTED
state change, long before any Bluetooth connection has been made.
