Jukka Rissanen [Mon, 12 Nov 2012 12:07:28 +0000 (14:07 +0200)]
plugin: Add VPN plugin that will interact with vpnd
Jukka Rissanen [Mon, 12 Nov 2012 12:07:27 +0000 (14:07 +0200)]
provider: Support functions added
Jukka Rissanen [Mon, 12 Nov 2012 12:07:26 +0000 (14:07 +0200)]
provider: Remove functionality that is no longer used
The VPN support functionality is moved into connman-vpnd.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:25 +0000 (14:07 +0200)]
provider: Add more driver interface functions needed by vpnd plugin
Jukka Rissanen [Mon, 12 Nov 2012 12:07:24 +0000 (14:07 +0200)]
provider: Provider disconnect function made available to plugin
The VPN needs to be able to call the provider disconnect.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:23 +0000 (14:07 +0200)]
test: New VPN test script that connect to connman-vpnd
Jukka Rissanen [Mon, 12 Nov 2012 12:07:22 +0000 (14:07 +0200)]
test: Renamed old VPN scripts
Renamed connect-vpn to connect-provider and
disconnect-vpn to remove-provider
In order to not confuse the user, change the name of the old
vpn connect/disconnect script. Now the name of the script
reflects the name of the manager dbus api.
The connect-provider will use the old ConnectProvider
manager dbus interface instead of directly connecting
to connman-vpnd.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:21 +0000 (14:07 +0200)]
vpn: New vpn daemon that handles vpn connections and clients
Jukka Rissanen [Mon, 12 Nov 2012 12:07:20 +0000 (14:07 +0200)]
log: Give program name and version in log init
This allows the correct output when the log functions
are used from separate vpn daemon.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:19 +0000 (14:07 +0200)]
device: Move device creation func into device.c
The device creation function should be part of device.c and
not inet.c. After this change the inet.c can be used by
separate vpn daemon.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:18 +0000 (14:07 +0200)]
ipconfig: Move IP address API into separate ipaddress.c file
Done so that connman_ipaddress_* functions can be used from
separate vpn daemon.
Jukka Rissanen [Mon, 12 Nov 2012 12:07:17 +0000 (14:07 +0200)]
doc: VPN daemon API descriptions
Patrik Flykt [Wed, 21 Nov 2012 14:25:11 +0000 (16:25 +0200)]
core: Use interface index instead of interface name
Service and network code use only interface indexes. Convert the rest of
the code using interface names to use interface indexes instead. The files
affected are:
include/resolver.h
src/connman.h
src/dnsproxy.c
src/resolver.c
src/rtnl.c
src/service.c
src/tethering.c
As the network and service code used interface indexes, the interface
name had to be looked up via a SIOCGIFNAME ioctl when needed. If a service
was removed due to the interface and network being taken down, the lookup
wouldn't work. This caused DNS servers to be left behind with broken sockets
since removing the DNS servers was done using the interface name.
Patrik Flykt [Fri, 16 Nov 2012 13:15:44 +0000 (15:15 +0200)]
TODO: Add WiFi plugin data struct handling task
Patrik Flykt [Fri, 16 Nov 2012 12:32:58 +0000 (14:32 +0200)]
dnsproxy: Use TCP connections only for the intended request
The DNS proxy code creates and uses a TCP connection only for a single
request at a time. A TCP DNS proxy is added to the end of the server
list, but should not be used for generic requests as the connection
will be closed down after the response to the initial request. Any
other requests sent over the connection will silently disappear.
Add a check to ignore TCP DNS servers for generic lookups.
Patrik Flykt [Mon, 19 Nov 2012 08:21:00 +0000 (10:21 +0200)]
dnsproxy: Check the result of socket creation when resolving
If the socket creation failed, continue with the next server.
Patrik Flykt [Fri, 16 Nov 2012 12:12:46 +0000 (14:12 +0200)]
dnsproxy: Check GIO channel for NULL before use
Check GIO channel for NULL before using it. Also shut down the GIO channel
so that already pending input will not use the free'd structure.
Forest Bond [Thu, 15 Nov 2012 15:23:50 +0000 (10:23 -0500)]
dnsproxy: Close the UDP socket on error condition.
The socket is created again on the next request.
Previously, an error condition only caused the event watch to be removed
and then added again on the next request to avoid processing spurious
error events. But in some network environments, the error condition is
caused by an overly aggressive firewall that rejects our UDP packets and
sends an ICMP error packet indicating an unreachable host and/or port.
In some of these cases we can recover by using a different UDP source
port for outgoing requests (i.e. by closing and re-opening the socket).
Daniel Wagner [Thu, 15 Nov 2012 09:50:15 +0000 (10:50 +0100)]
ofono: Remove network if GPRS context is removed
Whenever the GPRS contextext is removed we need also to remove
the network object. Following sequence was observed:
cm_update_attached() /huawei_1 Attached 1
modem_changed() /huawei_1 Interfaces 0x07
Plugin creates network object and registers it at core
cm_update_attached() /huawei_1 Attached 0
service.c:run_auto_connect()
set_property() /huawei_1 path /huawei_1/context1 org.ofono.ConnectionContext.Active
Obviously, the context is not attached at this point so we get an
Failed to change property: /huawei_1/context1 org.ofono.ConnectionContext.Active: org.ofono.Error.NotAttached GPRS is not
attached
Fixes BMC#25725
Danny Jeongseok Seo [Tue, 13 Nov 2012 01:44:26 +0000 (10:44 +0900)]
iptables: A file descriptor cannot be a nagative integer
When a socket() has failed and a negative file descriptor has been set,
close() cannot accept a negative number as a parameter.
Danny Jeongseok Seo [Mon, 12 Nov 2012 11:34:42 +0000 (20:34 +0900)]
manager: Fix negative error code
Fix error code for __connman_error_failed
Patrik Flykt [Fri, 9 Nov 2012 08:54:21 +0000 (10:54 +0200)]
wifi: Handle D-Bus method call failure
If -ECONNABORTED is received from the gsupplicant network disconnect
callback, it's an indication that the interface has already been
removed from wpa_supplicant side. This can happen due to a race condition
between rtnl and wpa_supplicant interface removal when a USB WiFi dongle
is pulled out.
The pointers given to the callback can no longer be trusted, don't touch
them in the callback. The have been removed by other code at this point.
Patrik Flykt [Fri, 9 Nov 2012 08:54:20 +0000 (10:54 +0200)]
supplicant: Handle race condition on WiFi dongle removal
When a WiFi USB dongle is removed, both rtnl and wpa_supplicant will
be telling that the device is gone. While disconnecting and removing
an interface via gsupplicant, check that D-Bus doesn't give us an
error indicating the interface is already unregistered from D-Bus.
The gdb output looks like:
(gdb) where
#0 0x0042247c in network_connect (network=0x2) at plugins/wifi.c:1197
#1 0x004228f0 in disconnect_callback (result=<value optimized out>,
interface=0x0, user_data=0x99) at plugins/wifi.c:1254
#2 0x0042b910 in network_remove_result (error=0x4d5458
"org.freedesktop.DBus.Error.UnknownMethod", iter=<value optimized
out>,
user_data=<value optimized out>) at gsupplicant/supplicant.c:3593
#3 0x0042ede0 in method_call_reply (call=0x4c6fd0,
user_data=0x4d3be8) at gsupplicant/dbus.c:386
#4 0x2abfddb4 in complete_pending_call_and_unlock () from
/usr/lib/libdbus-1.so.3
Also don't call the same callback twice.
Patrik Flykt [Tue, 13 Nov 2012 12:11:18 +0000 (14:11 +0200)]
bridge: Close socket on error
Danny Jeongseok Seo [Fri, 9 Nov 2012 08:21:24 +0000 (17:21 +0900)]
gdhcp: Fix resource leak
The socket was not closed when the sendto failed.
Danny Jeongseok Seo [Fri, 9 Nov 2012 05:44:12 +0000 (14:44 +0900)]
gweb: Check null before reference
Danny Jeongseok Seo [Fri, 9 Nov 2012 05:46:27 +0000 (14:46 +0900)]
inet: Fix resource leak
The socket was not closed when err < 0 is true.
Jukka Rissanen [Wed, 7 Nov 2012 08:14:05 +0000 (10:14 +0200)]
wifi: Skip service if there is error while loading its settings
If service settings cannot be loaded, then skip that service.
Thanks for Felipe Tonello for reporting the issue.
Jukka Rissanen [Thu, 1 Nov 2012 12:41:55 +0000 (14:41 +0200)]
README: Add information about network interface clearing at startup
Daniel Wagner [Wed, 31 Oct 2012 09:33:39 +0000 (10:33 +0100)]
session: Register session after policy plugin returned config
Move the configuration part of __connman_session_create() into
session_create_cb(). With this change the policy plugin is able
to do async work to retrieve a configuration.
Daniel Wagner [Wed, 31 Oct 2012 09:33:38 +0000 (10:33 +0100)]
session: Factor out user settings in __connman_session_create()
In order to be able to pass the user configuration provided through from
the D-Bus Manager.SessionCreate() call to the callback we need to store
the configuration into a local data data structure. This data structure
can then be passed into the callback introduced later on.
Daniel Wagner [Wed, 31 Oct 2012 09:33:37 +0000 (10:33 +0100)]
session: Remove struct connman_session_bearer
Instead start using enum connman_service_type directly.
Daniel Wagner [Wed, 31 Oct 2012 09:33:36 +0000 (10:33 +0100)]
session: Handle no match case in parse_bearers()
Daniel Wagner [Wed, 31 Oct 2012 09:33:35 +0000 (10:33 +0100)]
session: parse_bearers() should check for invalid types
Daniel Wagner [Wed, 31 Oct 2012 09:33:34 +0000 (10:33 +0100)]
session: Rename session_parse_allowed_bearers() to parse_bearers()
This function name was a source of confusion because in a later patch
we introduce connman_session_parse_allowed_bearers() which will
call parse_bearers(). With this change it should be more readable.
Daniel Wagner [Wed, 31 Oct 2012 09:33:33 +0000 (10:33 +0100)]
session: Remove match_all from connman_session_bearer
The match_all will be expressed through CONNMAN_SESSION_TYPE_UNKNOWN.
The 'no match' case happens when allowed_bearers is NULL.
Daniel Wagner [Wed, 31 Oct 2012 09:33:32 +0000 (10:33 +0100)]
session: Remove name variable from connman_session_bearer
The string is only used when appending the bearer to the D-Bus
message in append_allowed_bearers(). Let's use
__connman_session_type2string() in append_allowed_bearers(). This
saves a bit of memory.
Daniel Wagner [Wed, 31 Oct 2012 09:33:31 +0000 (10:33 +0100)]
session: Add callback to policy create()
Instead returning directly a config when create() is called
in policy plugin, use a callback function for handing over a valid
configuration from the plugin to the session core. This prepares
support for asynchronous create call.
Daniel Wagner [Wed, 31 Oct 2012 09:33:30 +0000 (10:33 +0100)]
session: Policy register should check callback existence
Let's ensure that the policy plugin has all necessary callbacks
installed when connman_session_policy_register() is called. The rest
of the code expects that the create() and destroy() callbacks exist
whenever a plugin is used.
Daniel Wagner [Wed, 31 Oct 2012 09:33:29 +0000 (10:33 +0100)]
session: Pick directly first policy in list
There is no point in iterating over the list when we always
pick the first element in the list.
Daniel Wagner [Wed, 31 Oct 2012 09:33:28 +0000 (10:33 +0100)]
session: Factor out memory release part of cleanup_session
We want to reuse this code snippet for the error case in
__connman_session_create() too.
Daniel Wagner [Wed, 31 Oct 2012 09:33:27 +0000 (10:33 +0100)]
manager: Allow async CreateSession method call
The CreateSession D-Bus call should be marked as async call in order
to allow the session core to defer the response.
Daniel Wagner [Wed, 31 Oct 2012 09:33:26 +0000 (10:33 +0100)]
storage: Add void if function does not have any arguments
Jukka Rissanen [Thu, 1 Nov 2012 12:16:24 +0000 (14:16 +0200)]
dhcp: Clear ipaddresses from ipconfig before starting DHCP
This is needed if manually configured addresses were used and later
DHCP was taken into use. If the manually configured IP information
(address, netmask and gateway) and the information given by DHCP is the
same, DHCP will not set the IP address to the interface.
Jukka Rissanen [Thu, 1 Nov 2012 12:16:23 +0000 (14:16 +0200)]
ipconfig: Add function to clear ipaddress information
Emil Renner Berthing [Tue, 23 Oct 2012 10:29:24 +0000 (13:29 +0300)]
openvpn: Support '--tls-auth <file>' with optional '<direction>' argument
Connman OpenVPN
OpenVPN.TLSAuth --tls-auth <file> [<direction>]
OpenVPN.TLSAuthDir optional <direction> argument to '--tls-auth'
Since 'OpenVPN.TLSAuthDir' contains the optional <direction> argument for
'OpenVPN.TLSAuth', handle them separately. The special handling is
identified by setting openvpn command line argument to NULL.
Emil Renner Berthing [Tue, 23 Oct 2012 10:29:23 +0000 (13:29 +0300)]
openvpn: Add more options
Connman OpenVPN
OpenVPN.NSCertType --ns-cert-type <client|server>
OpenVPN.AskPass --askpass <file>
OpenVPN.AuthNoCache --auth-nocache
Marcel Holtmann [Fri, 26 Oct 2012 08:48:16 +0000 (01:48 -0700)]
Release 1.9
Tomasz Bursztyka [Fri, 26 Oct 2012 07:33:11 +0000 (10:33 +0300)]
technology: Do not proceed with tethering if technology is not enabled
Fixes BMC#25846
Marti Raudsepp [Thu, 25 Oct 2012 18:30:44 +0000 (21:30 +0300)]
systemd: prevent duplicate logging messages in journal
By default, both stdout and syslog messages go to the systemd journal,
which results in duplicate messages being logged.
Thanks to Vinicius Costa Gomes for pointing out this problem.
Tomasz Bursztyka [Thu, 25 Oct 2012 07:54:57 +0000 (10:54 +0300)]
technology: Do not presume a technology being hardblocked when created
This fixes an issue with the TechnologyAdded signal when the first
un-hardblock event occurs:
- when a technology was created, D-Bus registration was done and hardblock
was set to TRUE even if there was no evidence that the technology was
rfkill driven
- when the technology was updated to be rfkill driven, hardblock was already
set to TRUE and thus the technology was not unregistered
- when an rfkill event un-hardblocks the technology, the TechnologyAdded
signal was not sent since the technology was already registered to D-Bus
Patrik Flykt [Thu, 25 Oct 2012 10:52:22 +0000 (13:52 +0300)]
dnsproxy: Add reply host part length checking
Check that the first part of the name is not of zero length before
attempting to calculate the length of the domain part. Also ensure
the domain lenght checking does not run outside of the receive
buffer.
Also add debug messages for ids and lengths in order to pinpoint
any possible problems.
Tomasz Bursztyka [Tue, 23 Oct 2012 07:26:44 +0000 (10:26 +0300)]
wispr: Free wispr portal context in the relevant place
This fixes 2 issues:
- calling __connman_service_ipconfig_indicate_state() might lead to
restart wispr_portal check, so context should not be freed afterward
but beforehand.
- freeing the context in wispr_manage_message() should not happen since
wispr_manage_message() will return back to wispr_portal_web_result()
where we can still use the context.
Thanks to Julien Massot <jmassot@aldebaran-robotics.com> who reported
the issue and provided logs.
Tomasz Bursztyka [Mon, 22 Oct 2012 07:07:49 +0000 (10:07 +0300)]
device: Do not mark connecting networks unavailable
When autoscan fallback code is started, it sets scanning to true which
in turn marks all networks unavailable except for the ones that are
already connected. When connecting during an ongoing autoscan, the
connection attempt stops autoscan and all unavailable networks are
removed, also the one to be connected.
The fix is to ignore both connected and connecting networks when
marking networks unavailable.
Jukka Rissanen [Fri, 19 Oct 2012 07:44:03 +0000 (10:44 +0300)]
network: Fix typo in debug print
Jukka Rissanen [Thu, 18 Oct 2012 11:35:19 +0000 (14:35 +0300)]
web: Make debug func print more useful information
The file and function name are printed in debug prints.
Jukka Rissanen [Thu, 18 Oct 2012 11:35:18 +0000 (14:35 +0300)]
gresolv: Remove query from queue before destroying the query
The order of actions is important here.
Jukka Rissanen [Thu, 18 Oct 2012 11:35:17 +0000 (14:35 +0300)]
gresolv: Make sure we will not receive DNS data after closing
We must close the channel when freeing the resolver object,
otherwise we might still receive data when the resolver has been
freed already.
Fixes BMC#25757
Jukka Rissanen [Thu, 18 Oct 2012 11:35:16 +0000 (14:35 +0300)]
gresolv: Add more debug prints
Jukka Rissanen [Thu, 18 Oct 2012 11:35:15 +0000 (14:35 +0300)]
gresolv: Make debug func print more useful information
The file and function names are printed in debug prints.
Patrik Flykt [Wed, 17 Oct 2012 10:34:43 +0000 (13:34 +0300)]
service: Prefer user connected services with SingleConnectedTechnology
Don't override user connected services with the ones selected by the
preferred technology list when SingleConnectedTechnology is enabled.
Do this by checking each connected service sorted in the beginning of
the service list for the userconnect flag.
Patrik Flykt [Wed, 17 Oct 2012 10:34:42 +0000 (13:34 +0300)]
service: Clear user connected flag on disconnect
Remember whether the service was connected by the user via D-Bus
until the service gets disconnected.
Patrik Flykt [Wed, 17 Oct 2012 10:34:41 +0000 (13:34 +0300)]
service: Keep only a single connected technology if configured
If SingleConnectedTechnology is enabled in main.conf, disconnect any
previously connected services when the new service enters ready state.
Patrik Flykt [Wed, 17 Oct 2012 10:34:40 +0000 (13:34 +0300)]
main: Add SingleConnectedTechnology main.conf variable
Patrik Flykt [Wed, 17 Oct 2012 10:34:39 +0000 (13:34 +0300)]
main.conf: Add SingleConnectedTechnology description
Patrik Flykt [Wed, 17 Oct 2012 10:34:23 +0000 (13:34 +0300)]
service: Fix default service switching and setting gateway
Calling switch_default_service() didn't change the service order since
the services were already sorted that way. Also update the gateway
immediately.
Patrik Flykt [Wed, 17 Oct 2012 10:34:22 +0000 (13:34 +0300)]
main.conf: Update preferred technology description
Patrik Flykt [Wed, 17 Oct 2012 10:34:21 +0000 (13:34 +0300)]
service: A preferred service in state ready is good enough
Simplify the preferred service selection such that a connected
service is good enough, especially since a connecting service
will also terminate the search for the current preferred one.
Patrik Flykt [Wed, 17 Oct 2012 10:34:20 +0000 (13:34 +0300)]
service: Don't trigger autoconnect when neither service is preferred
Don't trigger a new autoconnect when neither the default nor the new
service is preferred. Rely on the fact that normal autoconnect selection
mechanism has done the work for us already.
Tomasz Bursztyka [Wed, 17 Oct 2012 07:46:15 +0000 (10:46 +0300)]
client: Add Remove() method support for service
Tomasz Bursztyka [Wed, 17 Oct 2012 07:46:14 +0000 (10:46 +0300)]
client: Fix various memory leaks in monitor and services parts
Tomasz Bursztyka [Wed, 17 Oct 2012 07:46:13 +0000 (10:46 +0300)]
client: Fix some memory leaks in data_manager part
Tomasz Bursztyka [Wed, 17 Oct 2012 07:46:12 +0000 (10:46 +0300)]
client: Fix some memory leaks in commands part
Jukka Rissanen [Tue, 16 Oct 2012 14:19:25 +0000 (17:19 +0300)]
gresolv: Remove the lookup for real
The call to g_resolv_cancel_lookup() will do nothing
because we just removed the lookup from the queue.
The fix is to remove the lookup directly and not call
the cancel function.
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:23 +0000 (15:35 +0300)]
technology: Fix hardblocked state on non hw rfkillable switch added
If hardblock is on and a new device is inserted and detected as not
hardblocked, then it will be possible to enable/disable it (soft rfkill)
independently to the main hw rfkill switch.
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:22 +0000 (15:35 +0300)]
technology: Recompute hardblocked state on rfkill remove event
This fixes the case of cascading rfkill switches: if enabled, hard rfkilling
such technology might generate contradictory events.
1 - first all switches are hardblocked
2 - then one of these switch (usually: device's switch) gets fully unblocked
3 - then this same switch gets removed
Step 2 is in contradiction with step 1, so we need to care about such switch
getting removed by recomputing the hardblocked state.
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:21 +0000 (15:35 +0300)]
technology: Properly handle rfkill driven state
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:20 +0000 (15:35 +0300)]
device: Add a getter for powered property
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:19 +0000 (15:35 +0300)]
technology: Use the right method to enable or disable
If rfkill driven, use rfkill soft block/unblock. If not, request the
device to be enabled or disabled.
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:18 +0000 (15:35 +0300)]
technology: Change enabled to a boolean and refactor accordingly
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:17 +0000 (15:35 +0300)]
technology: Refactor and split functions setting enabled state
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:16 +0000 (15:35 +0300)]
technology: Refactor rfkill event handling according to soft/hard block
No need to proceed with softblocked if technology is already hardblocked.
Apply offlinemode and persistant state according to softblocked state.
(saner logic and helped to cleanup code from style point of view too)
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:15 +0000 (15:35 +0300)]
technology: Add and handle a marker for softblock status
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:14 +0000 (15:35 +0300)]
technology: Add a marker to know if a technology is rfkill driven
Useful for coming patches: enabling/disabling technologies will be done
differently whether technology is rfkill driven or not:
- if rfkill driven -> enabled will rely on rfkill states
- if not -> enabled will rely on driver/devices states
Tomasz Bursztyka [Mon, 15 Oct 2012 12:35:13 +0000 (15:35 +0300)]
technology: Refactor how a technology is enabled or disabled
Refactor how a device list is enabled/disabled: this will be useful
for coming patches. Simplify also the code, and remove useless gotos.
Patrik Flykt [Tue, 16 Oct 2012 10:56:32 +0000 (13:56 +0300)]
wispr: Keep track of proxy callback timeout
Keep track of the zero-second no proxy callback timeout and remove
it when freeing up the WISPr context.
Tomasz Bursztyka [Tue, 16 Oct 2012 05:35:17 +0000 (08:35 +0300)]
wifi: Do not start autoscan if interface was not ready
Tomasz Bursztyka [Tue, 16 Oct 2012 05:35:16 +0000 (08:35 +0300)]
gsupplicant: Send -ENOLINK error on scan reply if interface was not ready
Tomasz Bursztyka [Wed, 10 Oct 2012 08:18:45 +0000 (11:18 +0300)]
gsupplicant: Add a new state according to 'interface_disabled'
When soft rfkill is on, wpa_supplicant sets the interface disabled and
sends a state named 'interface_disabled'. Taking this information into
account fixes the following issue:
- disable wifi (user setting) and hard rfkill it
- then un-hard rfkill it, whereafter rfkill states (soft/hard) will go
like this:
* from 1/1 to 0/0
* from 0/0 to 1/0
when 0/0 occurs, connman will request to enable wifi
The problem with this is that enabling wifi takes quite some time and
in between ConnMan will soft block wifi to disable it (according to
previous user setting). Thus it will request to disable wifi but since
enabling is still going on, this request won't do anything. Meanwhile
wpa_supplicant will also catch the soft rfkill event and wpa_supplicant
will set the state to 'interface_disabled', but since it's not handled
properly by ConnMan, the wifi_enable() callback will be called and the
function will assume wifi got enabled.
Tomasz Bursztyka [Tue, 9 Oct 2012 10:35:16 +0000 (13:35 +0300)]
rfkill: Return the error when write() fails
Jukka Rissanen [Mon, 15 Oct 2012 10:36:04 +0000 (13:36 +0300)]
gresolv: Remove all pending lookups when resolver is removed
Remove all lookups found in queue when GResolv object is removed.
Jukka Rissanen [Mon, 15 Oct 2012 10:36:03 +0000 (13:36 +0300)]
gresolv: Avoid accessing already freed memory
We must remove the lookup from lookup queue and query from query queue
before calling user callback. The callback might unref the GResolv which in
turn would remove the lookup/query what we are trying to access after
the callback is returned.
So it is enough to remove the lookup or query entry from queue before
cb is called and then manually remove it after the callback has returned.
Jukka Rissanen [Fri, 12 Oct 2012 11:01:31 +0000 (14:01 +0300)]
iptables: Ignore module loading error
In iptables 1.4.9 module loading gives an error even if the module
is built in. Ignore the loading errors because the missing iptables
support is noticed when trying to get the iptables socket options.
Jukka Rissanen [Fri, 12 Oct 2012 17:13:22 +0000 (20:13 +0300)]
gdhcp: Returned IP address is already in host byte order
Jukka Rissanen [Thu, 11 Oct 2012 10:45:41 +0000 (13:45 +0300)]
gresolv: Remove lookup from correct queue
The lookup must be removed from lookup queue and not from query
queue when cancelling the lookup. Otherwise it is possible that
we might access an already removed lookup that is still found in
lookup queue.
Fixes BMC#25728
Jukka Rissanen [Thu, 11 Oct 2012 08:21:57 +0000 (11:21 +0300)]
nat: Memory leak when cleaning up
The nat struct was not freed when it is was removed from the hash.
Jukka Rissanen [Thu, 11 Oct 2012 07:11:52 +0000 (10:11 +0300)]
gdhcp: Use data size specific option setting function
Instead of using dhcp_add_simple_option() the gdhcp now uses
three functions for setting uint8, uint16 and uint32 values.
Jukka Rissanen [Thu, 11 Oct 2012 07:11:51 +0000 (10:11 +0300)]
gdhcp: Use host byte order internally
Convert to network byte order just before sending the packet.
Jukka Rissanen [Thu, 11 Oct 2012 07:11:50 +0000 (10:11 +0300)]
gdhcp: Add unalignment macros
The alignment macros are copied from Bluez.
Jukka Rissanen [Thu, 11 Oct 2012 07:11:49 +0000 (10:11 +0300)]
gdhcp: Remove obsolete load lease function