We have to copy the UUID (key) because otherwise the pointer can be invalidated
when the connection is destroyed and problems will start.
The issue showed up as an unability to delete a conenction via D-Bus.
Reproducer:
$ nmcli con add type eth con-name AAA ifname blah
$ nmcli con delete AAA
$ nmcli con add type eth con-name AAA ifname blah
$ nmcli con delete AAA
-- here the connection is not removed from NM
(even though ifcfg- file) was removed --
Add a "monitor-connection-files" config option, which can be set to
"false" to disable automatic reloading of connections on file change.
To go with this, add a new ReloadConnections method on
o.fd.NM.Settings that can be used to manually reload connections, and
add an nm-cli command to call it.
Some plugins may emit :new-connection or :unmanaged-specs-changed
while reading connections, so don't connect to those signals until
after the initial load_connections() (and just unconditionally emit
:unmanaged-specs-changed at that point).
In ifcfg-rh's get_unmanaged_specs(), don't bother to try to read the
connections first; if they haven't been read yet, just return NULL;
NMSettings will call it again after the connections have been read.
ifcfg-rh didn't let you unmanage an InfiniBand device by hardware
address because it was recording the hardware address with uppercase
letters, while nm_match_spec_hwaddr() required lowercase. Fix this by
making nm_match_spec_hwaddr() match case-insensitively (and remove the
manual lowercasing that several other places were doing to work around
this.)
keyfile didn't let you unmanage an InfiniBand device by hardware
address because it only accepted ARPHRD_ETHER hardware addresses. Fix
that by using nm_utils_hwaddr_valid() instead.
We don't always want to immediately write new connections to disk, to
facilitate "runtime" or "temporary" connections where an interface's
runtime config isn't backed by on-disk config. Also, just because
an interface's configuration is changed doesn't necessarily mean
that new configuration should be written to disk either.
Add D-Bus methods for adding new connections and for updating existing
connections that don't immediately save the connection to disk.
Also add infrastructure to indicate to plugins that the new connection
shouldn't be immediately saved if the connection was added with the
new method.
We'll need this for later with unsaved connections. The ifnet
plugin previously tracked connections by the "conn_name" which
was derived from keys in the /etc/conf.d/net file. These keys
take two forms:
1) interface name
config_eth0=(
"192.168.4.121/24"
"dhcp6"
)
2) wifi SSID, either text or hex-encoded:
config_myssid=("dhcp")
config_0xab3ace=("dhcp")
The conf.d net connection name is apparently usually an interface
name, so when writing to /etc/conf.d/net the NM connection name is
changed from eg "Ethernet connection 1" to the next available
interface name based on the type of connection, eg "eth0".
The ifnet plugin actively removed connections that were not present
in /etc/conf.d/net during the reload_connections() call, but in the
future we'll want to allow unsaved connections which in the case of
ifnet clearly won't yet be written to the file. Since only
connections written to the file have a "conn_name", tracking
connections by conn_name no longer works.
Use the new NMConnection 'changed' signal to mark connections
as dirty/unsaved, and reset that when they get flushed to disk.
Previously, the 'Updated' signal was emitted only when the
connection was changed and flushed to disk, but now we have
more granular needs, and the signal is emitted whenever the
connection actually *is* changed, regardless of whether its
flushed to disk or not.
If a kernel interface changes its MAC address, and NM is not
supposed to manage that interface, ifupdown needs to notice
that MAC address change and tell NM that the unmanaged devices
have changed, so that NM continues to not touch the device
after the MAC has changed.
well_known_interfaces -> eni_ifaces, since it's a hash of any
interfaces read from /etc/network/interfaces.
well_known_ifaces -> kernel_ifaces, since it's a hash of any
network subsystem interface the kernel knows about
'iface_connections' is really the list of all NMIfupdownConnections
known to the plugin, read from /e/n/i and hashed by block name. Since
ifupdown doesn't store anything *except* connections from /e/n/i,
just rename it to 'connections' to reduce confusion with the
well_known_interfaces and well_known_ifaces hashes.
Although having different parts of NM in different subdirectories
keeps the source tree neat, it has made the build messy, particularly
because of cross-dependencies between the subdirs.
Reorganize to build all of the pieces of the NetworkManager binary
from src/Makefile, and only use recursive make for test programs,
helper binaries, and plugins.
As part of this, get rid of all the per-directory convenience
libraries, and switch to building a single top-level
libNetworkManager.la, containing everything except main.c, which all
of the test programs can then link against.
Settings with all-default values are not written to reduce
complexity of the keyfile (and be more human-readable friendly)
and that includes VLAN settings with a VLAN ID of zero. So
when reading this file back, if there is no 'base type' setting
(eg, the setting specified by the connection::type property)
then just add that setting. nm_connection_verify() will catch
cases where an empty 'base type' setting is invalid.
Add these aliases for the setting names '802-3-ethernet',
'802-11-wireless', and '802-11-wireless-security' and write them by
default. It's much friendlier for administrators to type, and a lot
less ugly.
Also works for:
[connection]
type=ethernet
test-keyfile.c: In function 'test_read_string_ssid':
test-keyfile.c:1154:51: error: argument to 'sizeof' in 'memcmp' call is the
same expression as the second source; did you mean to provide an explicit
length? [-Werror=sizeof-pointer-memaccess]
ASSERT (memcmp (array->data, expected_ssid, sizeof (expected_ssid)) == 0,
Add some new API to NMConfig so that NMSettings and its plugins can
use NMConfig to look up values rather than reparsing the config file
themselves.
Also, move the no-auto-default cache from NetworkManager.conf to
$NMSTATEDIR/no-auto-default.state, so NM isn't rewriting its own
config file at runtime.
STP defaults to yes in NetworkManager (and the initscripts), so a missing
STP value in an ifcfg file means yes/on. Calling svSetValue(STP, NULL)
clears that line from the ifcfg, and thus STP gets interpreted as yes.
Explicitly set stp to "no" so that the value actually gets saved.
Just code cleanup: This is much less error-prone than manual nesting,
and will mesh very well with future changes to use the libgsystem
cleanup macros.
Avoid warnings about GValueArray being deprecated by adding macros
that wrap G_GNUC_BEGIN_IGNORE_DEPRECATIONS /
G_GNUC_END_IGNORE_DEPRECATIONS around the GValueArray calls.
For settings corresponding to devices that have a :carrier property
(ie bond, bridge, infiniband, vlan, and wired), add a :carrier-detect
property specifying how that affects the connection:
yes: The connection can only be activated when the device
has carrier, and will be deactivated if the device loses
carrier (for more than 4 seconds).
no: The connection ignores carrier on the device; it can be
activated when there is no carrier, and stays activated
when carrier is lost.
on-activate: The connection can only be activated when the
device has carrier, but it will not be deactivated if the
device loses carrier.
https://bugzilla.gnome.org/show_bug.cgi?id=688284
GObject creation cannot normally fail, except for types that implement
GInitable and take a GError in their _new() method. Some NM types
override constructor() and return NULL in some cases, but these
generally only happen in the case of programmer error (eg, failing to
set a mandatory property), and so crashing is reasonable (and most
likely inevitable anyway).
So, remove all NULL checks after calls to g_object_new() and its
myriad wrappers.
https://bugzilla.gnome.org/show_bug.cgi?id=693678
g_malloc(), etc, never return NULL, by API contract. Likewise, by
extension, no other glib function ever returns NULL due to lack of
memory. So remove lots of unnecessary checks (the vast majority of
which would have immediately crashed had they ever run anyway, since
g_set_error(), g_warning(), and nm_log_*() all need to allocate
memory).
https://bugzilla.gnome.org/show_bug.cgi?id=693678
Allows to attach any connection to a bridge using the BRIDGE= key.
IP configuration is optional for bridge components but not
prohibited. Test case included.
