If the tracked device is a control device only (has no network interface)
like in the case of a cdc-wdm device, get the mtu from the ip interface
(the exposed wwan network interface in this case).
https://bugzilla.redhat.com/show_bug.cgi?id=1460217
/proc/sys might be read-only but we want to set it for
enabling shared mode.
Check first if the sysctl already has the expected value,
and if so, do nothing.
https://bugzilla.gnome.org/show_bug.cgi?id=790726
Also downgrade a few intermediate error logging messages
for failures that happen while start_sharing(). A debug
message is enough in this case, because we propagate now
the error to the caller, which logs a warning anyway.
check_and_add_ipv6ll_addr() checks whether a link-local address is
already present in priv->ip6_config and if so, it returns with no
action.
priv->ip6_config is only updated after a merge-and-apply or (in an
idle source) when the external configuration changes and so there is
no guarantee that the addresses there are up-to-date.
priv->ext_ip6_config_captured should be checked instead, because it is
updated from platform right before starting the generation of a
link-local address. Note that also linklocal6_start() already checks
the captured external configuration rather than priv->ip6_config.
https://bugzilla.redhat.com/show_bug.cgi?id=1500350
If a device is 'external' (which means that NM generated an in-memory
connection to only to track the device state) we should not change its
IP configuration.
https://bugzilla.redhat.com/show_bug.cgi?id=1512316
In the next commit we will modify ipX_config_merge_and_apply to never
touch external devices. When a "reapply" call is issued on an external
device we are no longer simply tracking its state but we are actively
managing it and so its sys-iface-state must be promoted to managed.
https://bugzilla.redhat.com/show_bug.cgi?id=1512316
We also do this for libnm, where it causes visible changes
in behavior. But if somebody would rely on the hashing implementation
for hash tables, it would be seriously flawed.
GHashTable optimizes a NULL equality function to use direct pointer
comparison. That saves the overhead of calling g_direct_equal().
This is also documented behavior for g_hash_table_new().
While at it, also don't pass g_direct_hash() but use the default
of %NULL. The behavior is the same, but consistently don't use
g_direct_hash().
When a master connection is deactivated by user, we set the
autoconnect-blocked reason 'user-request' for the connection and we
propagate the same reason to slaves. Doing so prevents the
autoactivation of slaves when the master is manually activated again,
because the only way to override the 'user-request' blocked reason is
through manual activation of slaves.
Instead what should happen is that the manual deactivation of a master
marks slaves as blocked for failed dependencies. When the master
becomes available again, slaves can autoactivate if the profile allows
it.
https://bugzilla.redhat.com/show_bug.cgi?id=1437598
The _NM_GET_PRIVATE() macro already preserved and propagated
the constness of @self to the resulting private pointer.
_NM_GET_PRIVATE_PTR() didn't do that. Extend the macro,
to make that possible.
Fix the following warning:
src/devices/nm-device.c: In function ‘activation_source_schedule’:
src/devices/nm-device.c:4995:9: error: ‘source_func’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
new_id = g_idle_add (source_func, self);
~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Don't necessarily fail the entire connection if a duplicate IPv4
address is detected, but instead look at the may-fail property and at
the outcome of IPv6.
https://bugzilla.redhat.com/show_bug.cgi?id=1508001
- split NM_DEVICE_AUTOCONNECT_BLOCKED_INTERN in two parts:
"wrong-pin" and "manual-disconnect". Setting/unsetting them
should be tracked differently, as their reason differs.
- no longer initialize/clear the autoconnect-blocked reasons
during realize/unrealize of the device. Instead, initialize
it once when the object gets created (nm_device_init()), and
keep the settings beyond unrealize/realize cycles. This only
matters for software devices, as regular devices get deleted
after unrealizing once. But for software devices it is essential,
because we don't want to forget the autoconnect settings of
the device instance.
- drop verbose logging about blocking autoconnect due to failed
pin. We already log changes to autoconnect-blocked flags with
TRACE level. An additional message about this particular issue
seems not necessary at INFO level.
- in NMManager's do_sleep_wake(), no longer block autoconnect
for devices during sleep. We already unmanage the device, which
is a far more effective measure to prevent activation. We should
not also block autoconnect.
The flags allow for more then two reasons. Currently the only reasons
for allowing or disallowing autoconnect are "user" and "intern".
It's a bit odd, that NMDeviceAutoconnectBlockedFlags has a negative
meaning. So
nm_device_set_autoconnect_intern (device, FALSE);
gets replaced by
nm_device_set_autoconnect_blocked_set (device, NM_DEVICE_AUTOCONNECT_BLOCKED_INTERN);
and so on.
However, it's chosen this way, because autoconnect shall be allowed,
unless any blocked-reason is set. That is, to check whether autoconnect
is allowed, we do
if (!nm_device_get_autoconnect_blocked (device, NM_DEVICE_AUTOCONNECT_BLOCKED_ALL))
The alternative check would be
if (nm_device_get_autoconnect_allowed (device, NM_DEVICE_AUTOCONNECT_ALLOWED_ALL) == NM_DEVICE_AUTOCONNECT_ALLOWED_ALL)
which seems odd too.
So, add the inverse flags to block autoconnect.
Beside refactoring and inverting the meaning of the autoconnect
settings, there is no change in behavior.
nm_device_can_auto_connect() only has one caller, auto_activate_device()
in NMPolicy.
That caller already checks whether the connection has autoconnect
enabled, so drop the duplicate check.
This saves some duplication, but it also makes some sense:
NMSettingsConnection has a complex blocking of autoconnect,
so just looking at connection.autoconnect is not enough in
any case to determine whether the connection should autoconnect.
We move thus more handling of autoconnect to NMPolicy, where
it belongs.
The number of authentication retires is useful also for passwords aside
802-1x settings. For example, src/devices/wifi/nm-device-wifi.c also has
a retry counter and uses a hard-coded value of 3.
Move the setting, so that it can be used in general. Although it is still
not implemented for other settings.
This is an API and ABI break.
Since commit 4a6fd0e83e (device: honor the
connection.autoconnect-retries for 802.1X) and the related bug bgo#723084,
we reuse the autoconnect-retries setting to control the retry count
for requesting passwords.
I think that is wrong. These are two different settings, we should not
reuse the autoconnect retry counter while the device is still active.
For example, the user might wish to set autoconnect-retries to infinity
(zero). In that case, we would retry indefinitly to request a password.
That could be problematic, if there is a different issue with the
connection, that makes it appear tha the password is wrong.
A full re-activation might succeed, but we would never stop retrying
to authenticate. Instead, we should have two different settings for
retrying to authenticate and to autoconnect.
This is a change in behavior compared to 1.8.
Since 32b3eb1181 [core: merge IPv4 and IPv6 implementation of
nm_utils_ip4_property_path()], nm_utils_sysctl_ip_conf_path() introduced
in cd271d5cb1 [core: add nm_utils_sysctl_ip_conf_is_path() util] is used to
cunstruct sysctl paths and it is way less tolerant towards using something
that is not an interface name in the path.
It's always been incorrect to assume the ifname is a linux link name and
it resulted it ugly, if benign, sysctl access attempts such as
"/sys/class/net/28:B2:BD:5D:23:AB/phys_port_id" etc.
Now it triggers an assertion failure. Let's guard all such accesses.
Fixes: 32b3eb1181
Fixes: cd271d5cb1
For a while now, all NMPObject instances are not modified after
being cached. They are immutable, and can be passed around by keeping
a reference to them.
No longer copy the NMPlatformLink data to a @info variable. Instead,
take a reference (which ensures that the instance stays alive). It
won't change, as it's immutable.
The advantage is, that whenever you see a NMPlatformLink pointer,
for exmple in device_recheck_slave_status(), you can be sure that
it's actually a NMPObect, and NMP_OBJECT_UP_CAST() will work.
We now can be enslaved and have L3 configuration at the same time.
This also reduces some unnecessary complexity, because the decision to
progress to IP_CHECK or SECONDARIES now happens in a single place, in
the check_ip_state() routine.
The OpenVSwitch interfaces come into existence by their enslavement to a port.
They can also bear an IP4 or IP6 configuration -- waiting on a carrier would
deadlock the acitvation.
That one is special. All interfaces that are attached to OpenVSwitch
ports appear as slaves to that one even for our purposes we like to
pretend they're slaves to the actual OpenVSwitch bridges.
For some software devices, the platform link appears only after they've been
realized. Update their properties and let them know that the link has changed
so they can eventually proceed with activation.
Also, reset the properties (udi, iface, driver) that are set from the platform
link when the link goes away. At that point they don't reflect reality anymore.
Removes some code duplication too.
Coverity doesn't like this. Refactor a bit, hoping that it fares better.
1. Defect type: ASSERT_SIDE_EFFECT
1. NetworkManager-1.9.2/src/devices/nm-device.c:10226: assignment_where_comparison_intended: Assignment "ip_ifindex = nm_device_get_ip_ifindex(self)" has a side effect. This code will work differently in a non-debug build.
2. NetworkManager-1.9.2/src/devices/nm-device.c:10226: remediation: Did you intend to use a comparison ("==") instead?
Add a proxy setting to generated connections because after commit
6f94b16507 ("libnm: fix nm_connection_diff() for settings without
properties") the matching between a connection having the setting and
another connection without it fails. Before the commit, since no proxy
property is marked as inferrable, such comparison succeeded.
Since commit 5c299454b4 we can configure
multiple default-routes.
That is especially useful with IPv6 to configure multiple routers.
It will also be useful, once we allow configuring manual default-routes,
like regular static routes.
However the problem is, that the default-route for the manual gateway
and the gateway from DHCP both get the same metric. So it's undefined
which route is used. To avoid that problem, and to restore previous
behavior, don't accept any default-routes if a gateway is set.
Fixes: 5c299454b4
Setting the MTU might fail when the underlying device's MTU
is not set.
Detect that case, and log a better warning message.
Unfortunately, it's tricky to detect whether this is a complete
failure, or whether we will later try again to change the MTU.
So, we log a failure, altough later we might fix it. It would
be better not to warn about non-errors.
and nm_utils_ip6_property_path(). The API with static buffers
looks a bit nicer. But I think they are dangerous, because
we tend to pass the buffer down several layers of the stack, and
it's not immediately clear, that we don't overwrite the static
buffer again (which we probably did not, but it's hard to verify
that there is no bug there).
Kernel does not allow setting the MTU of a VLAN larger
then the MTU of the underlying device. Hence, we might
initially fail to set a large MTU of the VLAN, but we
have to retry when the MTU of the parent changes.
https://bugzilla.redhat.com/show_bug.cgi?id=1414901
