The wpas-helpers provide easy _asynchronous_ access to wpa_supplicant
control interfaces. Compared to the old wpa_ctrl_* stuff it's no longer
synchronous. Thus, it doesn't block our daemon while wpa_supplicant runs
some heavy work again..
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Use a common subdirectory for gdhcp so we can easily replace it once
sd-dhcp has server-side support.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If we get ADD events from udev for a new link but run in unmanaged mode,
skip the event. It is fairly unlikely that another manager has already
started wpa_supplicant on it (and even if, it would be racy).
Hence, lets not support hotplugging in these cases until external managers
provide us a notification API.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If, after 5 times (which is 1s), wpa_supplicant is still not reachable, we
now drop the link. This avoids hangup issues during wpas startup that
might happen from time to time.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If miracled is spawned with --manage-wifi, we run our own copy of
wpa_supplicant on managed devices. This is meant for testing purposes
only. Hopefully, network-manages like NM or connman will provide a P2P API
at some time so we can switch. Until then, we allow the user to control
who is in charge of wpa_supplicant.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
This is very handy to catch wrong usage of shl_strjoin(). We now get a gcc
warning if the sentinel (NULL) is omitted.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Allow users to tag devices as "miracle" via udev so they get picked up
automatically. We also add a generic "remove"-monitor so we correctly drop
links as they get unplugged. The wpa_supplicant notification is unreliable
so we want the proper udev-events here.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
We should notify miraclectl about ScanStopped events so we can correctly
track active scans. If we don't do that, we would have to explicitly track
it on connect/disconnect/etc. commands, which seems cumbersome given the
undeterminisc actions they cause.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If we run as GO we need to support local AP events and dhcp-server
handling. Both was already supported by our code-base so simply hook it up
with miracled-wifi.
Note that this adds a severe restriction: we can only have a single
p2p-connection per interface. The wpa_supplicant events don't provide
enough information to associate the ifnames to devices.
But most drivers only support a single STA, anyway, so that's fine.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Add a gdhcp callback for server-side lease events. Then raise L: and R:
lines for each event we get so miracled can react to it.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
We use "miracle" as project-name and for all links and tokens. The fancy
name stays "MiracleCast".
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Add commands for allow/reject/connect/disconnect and parse all events
properly. Notify the user of any important changes so they can react to
it.
Note that Peer.Connect() is currently not implemented and will stay this
way until we have the dhcp-server and wfd-source ready (or at least
planned).
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If we explicitly call wifi_dev_disconnect(), no WIFI event is broadcasted,
so handle it manually and send a dbus event if it changed.
Note that we explicitly call into wifi even if the device is not connected
to kill any ongoing connection attempts.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Hook up the peer-dbus API with the underlying peer objects. This allows to
handle incoming requests and deal with connection issues.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If we stop a scan in an async-manner and the object is no longer around,
don't print an error. Only do that if the scan is stopped explicitly by
the user.
This allows us to correctly kill scans during link-removal in a follow-up
patch. In such cases, the scan-stop may fail of the dbus object has
already been removed.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
We currently suppress all log-messages except FATAL, which is not really
what we want. Initialize the log-level to NOTICE like the default for
shl_log is, too.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Add readline support to miraclectl so we can have interactive commands.
This rewrites huge parts of miraclectl but hopefully makes it future proof
so no such rewrite will be needed again.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Add few small helpers to shl-macro/util. This includes some _cleanup_
helpers and the qstr suite for quoted/escaped strings.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Print more information now that we have property-support in miracled. Also
split link and peer information, otherwise the output looks misleading.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
The add-link argument turned out to be confusing so avoid the split input
and use the same name as all the other commands. We now split it
internally so the DBus call will work properly.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Three new commands that control scans on a given link are added:
1) start-scan: Starts scanning in background for a given link
2) stop-scan: Stops scanning in background for a given link
3) scan: Does interactive scanning for a given link
The interactive scanning basically calls start-scan and stop-scan. In
between, it listens for dbus-signals for new peers and displays them if
they are assigned to the given link. This should be used for basic
p2p-device scanning on a link.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
If we don't split peer linking off, the peer might get advertised before
the sub-device related data is set. With dbus, this might trigger
callbacks to not return attributes as the peer-type is unknown. Therefore,
properly initialize the peer before linking/advertising it.
This fixes a bug where the initial InterfacesAdded signal for new peers
does not include all properties of a peer.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Both dbus commands allow external programs to make our device discoverable
for P2P devices and also start scanning for remote devices.
The commands are *not* ref-counted right now, so parallel use is
discouraged. However, these commands can interrupt normal operations on a
wifi-device anyway, so if used in parallel, it is very like to break.
Thus, keep the interface simple and require callers to do
access-management (it's root-only, anyway..).
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Send ObjectManager.Interfaces{Added,Removed} signals whenever new objects
appear. Note that we cannot set DBus interfaces there as sd-bus ignores
them. We'll try to fix that upstream and until then just not expose any of
those.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
We need to send PropertiesChanged signals whenever we change properties
that are annotated as such.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
This command prints detailed information on a given peer. This is exactly
the same as 'show-link' but for peers.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Implement basic peer properties and hook them up with dbus. Also add a
device_name parser to the wifi implementation so we correctly read the
friendly-name of remote devices.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Move peer_make_name() to the top to keep consistency with miracled-link.c
and move non-object-based helpers to the top.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
The new show-link command lists detailed information of a given link.
Useful for debugging and introspection.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
