<error> level is for something really bad happening. When another party
(iwd in this case) sends a D-Bus request that we cannot meaningfully handle,
that is hardly reason to warn about. <debug> level is enough in this case.
Also, give all messages a common prefix "agent-request" so that we have
something to grep for.
nm_utils_random_bytes() will always try its best to give some
random numbers. A failure only means, that the kernel interfaces
get_random() or /dev/urandom failed to provide good randomness. We
don't really need good random numbers here, so no need to handle
a failure.
Allow the IWD backend to use secrets provided in the connection settings
on initial connection attempt, only require new secrets on subsequent
connections when IWD asks for them -- it only asks if fresh secrets are
required.
The IWD DBus interface currently
(https://git.kernel.org/pub/scm/network/wireless/iwd.git/tree/doc/agent-api.txt?id=38952813dddd776f66d2ed5e88eca9a892964c06)
knows about 3 secret types related to 802.1x authentication in addition
to the PSK secret request. Add support for the new methods and the new
secret types in NM's implementation of the IWD secret agent. Note that
the secret types are mapped to NMSetting8021x property keys and they are
then sent to the NM Secret Agent in the hints parameter to GetSecrets,
this will need support in the NM clients as the exact usage of the
hints parameter is specified a little ambiguously, but this seems to be
one of the permitted usages.
Rework the IWD agent interface info initialization to use NM convenience
macros.
To improve the code logic and reduce space for bugs, don't save the
dbus invocation object as priv->secrets_request, instead move it to
the nm_act_request_get_secrets()'s user_data as we only need the
invocation object for exactly the life time of the request. See
https://github.com/NetworkManager/NetworkManager/pull/139 for
discussion.
- in wake_on_wlan_restore(), if we decide that there is something
to restore, also clear priv->wowlan_restore by setting it to
IGNORE. That way, we are sure to only try resetting the value
once after setting it.
- from nm_platform_wifi_get_wake_on_wlan(), return IGNORE if
the value cannot be read. If we could not read the value
we should not restore NONE, but don't restore.
Instead of returning a boolean @is_user_config value from
get_configured_mtu(), return an mtu-source enum with possible values
NONE,CONNECTION. This enum will be expanded later; for now there is no
change in behavior.
_LOGD() is preferred, because it includes a common prefix depending
on the device. This macro requires, that we have a suitable @self
variable in the local scope.
Coccinelle:
@@
expression a, b;
@@
-a ? a : b
+a ?: b
Applied with:
spatch --sp-file ternary.cocci --in-place --smpl-spacing --dir .
With some manual adjustments on spots that Cocci didn't catch for
reasons unknown.
Thanks to the marvelous effort of the GNU compiler developer we can now
spare a couple of bits that could be used for more important things,
like this commit message. Standards commitees yet have to catch up.
if device *is* a NM_DEVICE_IWD, then make sure to not pass that to
_nm_device_wifi_request_scan (which asserts on anything else than a
NM_DEVICE_WIFI device).
The crash can be triggered by enabling wifi.backend=iwd and clicking
on the 'select network' item in gnome shell for example. The journal
output looks like this:
NetworkManager[1861]: invalid cast from 'NMDeviceIwd' to 'NMDeviceWifi'
NetworkManager[1861]: **
NetworkManager[1861]: NetworkManager:ERROR:src/devices/wifi/nm-device-wifi.c:1127:_nm_device_wifi_request_scan: assertion failed: ((((__extension__ ({ GTypeInstance *__inst = (GTypeInstance*) ((_obj)); GType __t = ((nm_device_wifi_get_type ())); gboolean __r; if (!__inst) __r = (0); else if (__inst->g_class && __inst->g_class->g_type == __t) __r = (!(0)); else __r = g_type_check_instance_is_a (__inst, __t); __r; })))))
systemd[1]: NetworkManager.service: Main process exited, code=dumped, status=6/ABRT
systemd[1]: NetworkManager.service: Failed with result 'core-dump'.
Fixes: 297d4985abhttps://github.com/NetworkManager/NetworkManager/pull/107
There are multiple tests with the same in different directories; add a
unique prefix to test names so that it is clear from the output which
one is running.
Similar cases of such a field are named "manager". Also,
internal names shall not have an "nm" prefix, contrary
to names in a header file, which shall have such a prefix.
I find it slightly nicer and explict. Also, the list elements
are strictly speaking private, we should better not explicitly
use them outside of NMManager/NMDevice. The macro hides this.
- no longer track APs in a hash table with their exported path
as key. The exported path is already tracked by NMDBusManager's
lookup index, so we can reuse that for fast lookup by path. Otherwise,
track the APs in a CList per device.
- as we now track APs in a CList, their order is well defined.
We no longer need to sort APs and obsoletes nm_wifi_aps_get_sorted()
and simplifies nm_wifi_aps_find_first_compatible().
Instead of using a GSList for tracking the devices, use a CList.
I think a CList is in most cases the more suitable data structure
then GSList:
- you can find out in O(1) whether the object is linked. That
is nice, for example to assert in NMDevice's destructor that
the object was unlinked, and we will use that later in
nm_manager_get_device_by_path().
- you can unlink the element in O(1) and you can unlink the
element without having access to the link's head
- Contrary to GSList, this does not require an extra slice
allocation for the link node. It quite possibliy consumes
slightly less memory because the CList structure is embedded
in a struct that we already allocate. Even if slice allocation
would be perfect to only consume 2*sizeof(gpointer) for the link
note, it would at most be as-good as CList. Quite possibly,
there is an overhead though.
- CList possibly has better memory locality, because the link
structure and the data are close to each other.
Something which could be seen as disavantage, is that with CList
one device can only be tracked in one NMManager instance at a time.
But that is fine. There exists only one NMManager instance for now,
and even if we would ever introduce multiple managers, we probably
would not associate one NMDevice instance with multiple managers.
The advantages are arguably not huge, but CList is IMHO clearly the
more suited data structure. No need to stick to a suboptimal data
structure for the job. Refactor it.
NMSettings exposes a cached list of all connection. We don't need
to clone it. Note that this is not save against concurrent modification,
meaning, add/remove of connections in NMSettings will invalidate the
list.
However, it wasn't save against that previously either, because
altough we cloned the container (GSList), we didn't take an additional
reference to the elements.
This is purely a performance optimization, we don't need to clone the
list. Also, since the original list is of type "NMConnection *const*",
use that type insistently, instead of dependent API requiring GSList.
IMO, GSList is anyway not a very nice API for many use cases because
it requires an additional slice allocation for each element. It's
slower, and often less convenient to use.
I dislike the static hash table to cache the integer counter for
numbered paths. Let's instead cache the counter at the class instance
itself -- since the class contains the information how the export
path should be exported.
However, we cannot use a plain integer field inside the class structure,
because the class is copied between derived classes. For example,
NMDeviceEthernet and NMDeviceBridge both get a copy of the NMDeviceClass
instance. Hence, the class doesn't contain the counter directly, but
a pointer to one counter that can be shared between sibling classes.
Previously, we used the generated GDBusInterfaceSkeleton types and glued
them via the NMExportedObject base class to our NM types. We also used
GDBusObjectManagerServer.
Don't do that anymore. The resulting code was more complicated despite (or
because?) using generated classes. It was hard to understand, complex, had
ordering-issues, and had a runtime and memory overhead.
This patch refactors this entirely and uses the lower layer API GDBusConnection
directly. It replaces the generated code, GDBusInterfaceSkeleton, and
GDBusObjectManagerServer. All this is now done by NMDbusObject and NMDBusManager
and static descriptor instances of type GDBusInterfaceInfo.
This adds a net plus of more then 1300 lines of hand written code. I claim
that this implementation is easier to understand. Note that previously we
also required extensive and complex glue code to bind our objects to the
generated skeleton objects. Instead, now glue our objects directly to
GDBusConnection. The result is more immediate and gets rid of layers of
code in between.
Now that the D-Bus glue us more under our control, we can address issus and
bottlenecks better, instead of adding code to bend the generated skeletons
to our needs.
Note that the current implementation now only supports one D-Bus connection.
That was effectively the case already, although there were places (and still are)
where the code pretends it could also support connections from a private socket.
We dropped private socket support mainly because it was unused, untested and
buggy, but also because GDBusObjectManagerServer could not export the same
objects on multiple connections. Now, it would be rather straight forward to
fix that and re-introduce ObjectManager on each private connection. But this
commit doesn't do that yet, and the new code intentionally supports only one
D-Bus connection.
Also, the D-Bus startup was simplified. There is no retry, either nm_dbus_manager_start()
succeeds, or it detects the initrd case. In the initrd case, bus manager never tries to
connect to D-Bus. Since the initrd scenario is not yet used/tested, this is good enough
for the moment. It could be easily extended later, for example with polling whether the
system bus appears (like was done previously). Also, restart of D-Bus daemon isn't
supported either -- just like before.
Note how NMDBusManager now implements the ObjectManager D-Bus interface
directly.
Also, this fixes race issues in the server, by no longer delaying
PropertiesChanged signals. NMExportedObject would collect changed
properties and send the signal out in idle_emit_properties_changed()
on idle. This messes up the ordering of change events w.r.t. other
signals and events on the bus. Note that not only NMExportedObject
messed up the ordering. Also the generated code would hook into
notify() and process change events in and idle handle, exhibiting the
same ordering issue too.
No longer do that. PropertiesChanged signals will be sent right away
by hooking into dispatch_properties_changed(). This means, changing
a property in quick succession will no longer be combined and is
guaranteed to emit signals for each individual state. Quite possibly
we emit now more PropertiesChanged signals then before.
However, we are now able to group a set of changes by using standard
g_object_freeze_notify()/g_object_thaw_notify(). We probably should
make more use of that.
Also, now that our signals are all handled in the right order, we
might find places where we still emit them in the wrong order. But that
is then due to the order in which our GObjects emit signals, not due
to an ill behavior of the D-Bus glue. Possibly we need to identify
such ordering issues and fix them.
Numbers (for contrib/rpm --without debug on x86_64):
- the patch changes the code size of NetworkManager by
- 2809360 bytes
+ 2537528 bytes (-9.7%)
- Runtime measurements are harder because there is a large variance
during testing. In other words, the numbers are not reproducible.
Currently, the implementation performs no caching of GVariants at all,
but it would be rather simple to add it, if that turns out to be
useful.
Anyway, without strong claim, it seems that the new form tends to
perform slightly better. That would be no surprise.
$ time (for i in {1..1000}; do nmcli >/dev/null || break; echo -n .; done)
- real 1m39.355s
+ real 1m37.432s
$ time (for i in {1..2000}; do busctl call org.freedesktop.NetworkManager /org/freedesktop org.freedesktop.DBus.ObjectManager GetManagedObjects > /dev/null || break; echo -n .; done)
- real 0m26.843s
+ real 0m25.281s
- Regarding RSS size, just looking at the processes in similar
conditions, doesn't give a large difference. On my system they
consume about 19MB RSS. It seems that the new version has a
slightly smaller RSS size.
- 19356 RSS
+ 18660 RSS
If SSID is an empty string there's no need to call nm_wifi_ap_set_ssid
as it won't do anything. It also has an assert checking that NULL is
passed for an empty SSID and we were passing a non-NULL pointer.
This bug was not causing a crash for me because of the !IS_NM_DEVICE_IWD
check and because my glib version probably had the assertion within
NM_IWD_MANAGER_GET_PRIVATE disabled.
While there, change the g_signal_connect line to use the macro for the
signal name.
Make sure .set_enabled uses the Device.Powered property to basically
bring the netdev UP and DOWN as I understand is expected by the
nm_device logic.
Device.Powered should generally reflect the UP state immediately but
just to avoid possible race conditions .is_available() will now return
a value that is an AND of the local "enabled" state and IWD's Powered
property.
Change from the default dbus call timeout (-1) to infinite (G_MAXINT)
because the call may now include the secret requests which have their
own timeout policies.
Remove the code (mostly copied from nm-device-wifi.c) that handles
checking if the secrets were provided and requesting missing secrets
before starting a connection attempt. Instead, request secrets when
we're asked for them by IWD through its agent interface. This happens
while the dbus Connect call is running. We change the NMDevice from
the CONFIG state to NEED_AUTH and then change back to CONFIG once we
sent the secrets back to IWD.
The current code would require the secrets only based on whether a
network is a KnownNetwork but IWD may need a new passwords even for
KnownNetworks if the last connection attempt has failed.
Avoid calling nm_device_iwd_set_dbus_object (device, NULL) if the
dbus_object was NULL already. Apparently gdbus guarantees that a
name-owner notification either has a NULL old owner or a NULL new owner
but can also have both old and new owner NULL.
Reuse the apparent workaround from libnm/nm-client.c in which the
GDbusObjectManagerClient is recreated every time the name owner
pops up, instead of creating it once and using that object forever.
Resubscribe to all the signals on the new object. The initial
GDbusObjectManager we create is only used to listed for the name-owner
changes.
There's nothing in gdbus docs that justifies doing that but there
doesn't seem to be any way to reliably receive all the signals from
the dbus service the normal way. The signals do appear on dbus-monitor
and the gdbus apparently subscribes to those signals with AddMatch()
correctly but they sometimes won't be received by the client code,
unless this workaround is applied.
While making changes to got_object_manager, don't destroy the
cancellable there as it is supposed to be used throughout the
NMIwdManager life.
Disconnect from NMManager signals in our cleanup, make sure the
NMManager singleton is not destroyed before we are by keeping a
reference until we've disconnected from its signals.
Add very simple periodic scanning because IWD itself only does periodic
scanning when it is in charge of autoconnecting (by policy). Since we
keep IWD out of the autoconnect state in order to use NM's autoconnect
logic, we need to request the scanning. The policy in this patch is to
use a simple 10s period between the end of one scan the requesting of
another while not connected, and 20s when connected. This is so that
users can expect similar results from both wifi backends but without
duplicating the more elaborate code in the wpa_supplicant backend which
can potentially be moved to a common superclass.
Must always initialize cleanup variable, to be able to build with
"-fexceptions".
make[2]: Entering directory './contrib/fedora/rpm/NetworkManager.20180124-060444.C5tHCi/BUILD/NetworkManager-1.11.1'
CC src/devices/wifi/src_devices_wifi_libnm_device_plugin_wifi_la-nm-device-iwd.lo
In file included from ./shared/nm-utils/nm-glib.h:27:0,
from ./shared/nm-utils/nm-macros-internal.h:60,
from ./shared/nm-default.h:257,
from src/devices/wifi/nm-device-iwd.c:21:
src/devices/wifi/nm-device-iwd.c: In function ‘deactivate_async_finish’:
./shared/nm-utils/gsystem-local-alloc.h:37:8: error: ‘variant’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
if (*(Type*)v) \
^
src/devices/wifi/nm-device-iwd.c:405:29: note: ‘variant’ was declared here
gs_unref_variant GVariant *variant;
^~~~~~~
Fixes: d0c1e1a62a
In a previous patch I added deactivate_async to make sure that NM
auto re-connect waits for the IWD state to changed from "disconnecting"
to "disconnected" before starting a new activation when the user wants
to switch from one profile to another. This doesn't account for when
IWD itself goes into "disconnecting" because of a connect failure.
When IWD goes into the "disconnecting" state we call
nm_device_state_changed (NM_DEVICE_STATE_FAILED,
NM_DEVICE_STATE_REASON_SUPPLICANT_DISCONNECT)
immediately to give feedback to user as soon as possible. We will
return FALSE from get_autoconnect_allowed for the period the
"disconnecting" state.
Deactivating the connection translates to a Device.Disconnect dbus call
to IWD. The dbus call normally returns immediately because the
corresponding nl80211 command returns immediately but we can't rely on
that. Make sure that NMDevice waits for the IWD method return before
continuing with the new activation request when switching wifi networks.
The handler would until now check if nm_device_is_activating() was true
or the NMDevice state was "activated" to decide whether to report to
NMDevice that we'd been unexpectedly disconnected (i.e. connection
failed). However NM's "prepare" and "need auth" states correspond to
IWD's "disconnected" state because they don't involve Wifi
authentication/association.
Additionally nm_device_is_activating() returns TRUE even when NMDevice
state is "disconnected" but an activation request is pending. As a
result when switching networks, NMDevice would first save the activation
request and go into the "disconnected" state, we'd then call the IWD's
Disconnect method and when we received the IWD state change notification
to "disconnected", we'd cause the pending activation request to be
considered a failure. The handler shouldn't report a failed
connection when the NMDevice state is "disconnected".
