Cross Reference: /system/connectivity/shill/shill

History log of /system/connectivity/shill/shill_daemon.cc
Revision	Date	Author	Comments (<<< Hide modified files) (Show modified files >>>)
59b397b6ac7fa39edf26d8e0ff2ecac6ac8ebc5b	20-Jan-2016	Samuel Tan <samueltan@google.com>	shill: rename ChromeosDaemon to DaemonTask Rename ChromeosDaemon to DaemonTask since this class contains all the logic for daemon-related tasks (e.g. start/stop, init/shutdown). This will also prevent this class from being confused with ShillDaemon. BUG: 26653263 Change-Id: I9713203cb1f933163d7067a7ae982622bf5b17af TEST: shill compiles and unit tests pass on Android. TEST: shill compiles and unit tests pass on Chrome OS. /system/connectivity/shill/shill_daemon.cc
60ea15d40a3ad77536e01f2bb780181aa9c3c084	20-Jan-2016	Samuel Tan <samueltan@google.com>	shill: refactor ChromeosDaemon Refactor ChromeosDaemon so that it no longer inherits from brillo::Daemon. This is necessary to allow its unit tests to run without the extra base::ExitManager inherited from brillo::Daemon. Do this by introducing a new ShillDaemon class, which inherits the logic in ChromeosDaemon, and overrides brillo::Daemon methods. BUG: 26653263 Change-Id: Ia256bc540d261fb183aae0100946afdcdea80fb3 TEST: shill compiles and unit tests pass on Android. TEST: shill compiles and unit tests pass on Chrome OS. /system/connectivity/shill/shill_daemon.cc
70fc5accfe83fac147ce47e24311d6739edaeedb	18-Jan-2016	Christopher Grant <cjgrant@google.com>	shill: remove obsolete shill daemon source files Bug: None TEST=Shill builds; unit tests pass. Change-Id: I4601af4bcf48311543945e0649f6e16c963914e5 /system/connectivity/shill/shill_daemon.cc
bad1c10ffd2d4ac14f7bd9f4ef6a8982e711f566	19-Sep-2015	mukesh agrawal <quiche@google.com>	shill: remove references to GLib There are a number of references to GLib (the object), even though no one uses GLib in a substantive way. Remove these obsolete references. Bug: 23386647 TEST=unit tests Change-Id: I8d574980bc8e853c8b87ef64b581d64a71491201 /system/connectivity/shill/shill_daemon.cc
c0beca55d290fe0b1c96d78cbbcf94b05c23f5a5	03-Sep-2015	Peter Qiu <zqiu@google.com>	License shill to Apache2 This patch automatically replaced the license on all text files from Chromium OS (BSD style) to AOSP (Apache2), keeping the original year as a reference. The license header was added to .gyp and .gypi, the NOTICE was added with a copy of the Apache2 license and MODULE_LICENSE_* file was added. BUG=23587594 TEST=grep 'Chromium OS Authors' doesn't find anything. Change-Id: If41ede586c2465f5831fb7fee270ff41dbfdb596 /system/connectivity/shill/shill_daemon.cc
608ec29525f553d51f0a92e84176e3d4b45930a9	31-Jul-2015	Peter Qiu <zqiu@chromium.org>	shill: use ControlInterface to create proxy objects So that appropriate proxy objects can be created based on the type of the control interface. While there, remove ProxyFactory related files. BUG=chromium:513791 TEST=USE="asan clang cellular wimax dhcpv6" FEATURES=test emerge-$BOARD shill Change-Id: I465b0807ab4ccacac3891638db41384f43ef83c5 Reviewed-on: https://chromium-review.googlesource.com/289855 Reviewed-by: Zeping Qiu <zqiu@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> Tested-by: Zeping Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
1a212a6b98b22ad1d69652bb26a9e94138635476	16-Jun-2015	Paul Stewart <pstew@chromium.org>	shill: Top-level files n-v: Switch pointer/ref spacing Switch to Chrome style spacing. BUG=chromium:501026 TEST=Unit tests Change-Id: I8bd6f66ea185ca3a2687986f6f2fd99be957dffe Reviewed-on: https://chromium-review.googlesource.com/278070 Reviewed-by: Rebecca Silberstein <silberst@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
1ce231c71932200e4d02c71567f8e93788120781	13-Jun-2015	Paul Stewart <pstew@chromium.org>	shill: Manager: Refactor command line setup In anticipation of a new and different implementation of shill::Daemon, this CL re-factors the interaction between the main() function and shill::Daemon by passing a new Daemon::Settings structure. This way the process of parsing command line parameters is separated from the task of delivering this configuration the Daemon object. While here, the "default_technology_order" field of was unified with the other parameters passed to the Manager from the Daemon. However this change is somewhat dwarfed by the technical debt that needed to be paid off in order to allow this new change to be tested. The new Daemon::ApplySettings method makes no new calls to the Manager object but none of these methods were previously mocked in the MockManager object. More troublingly the side-effects of these Manager methods were depended on in MockManager, so a few tests (notably DeviceTest) needed to be re-architected to deal with mock functionality instead of assuming the Mock object did what a real Manager would do. Lastly, both the Device::PrependDNSServers and Manager::FilterPrependDNSServersByFamily methods seemed a bit hard to understand and difficult to test. Minor changes to these APIs allowed the test changes above to work a lot smoother, and also to separate the unit tests to work on individually testing the functionality in each object. BUG=chromium:500991 TEST=Unit tests Change-Id: I04c33c841024c102ff9658e5857236ab2e5bf32f Reviewed-on: https://chromium-review.googlesource.com/278022 Reviewed-by: Garret Kelly <gdk@chromium.org> Reviewed-by: Rebecca Silberstein <silberst@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
d48fa0c5531e2102d4f537e81b9f92afc2d60955	10-Jun-2015	Peter Qiu <zqiu@chromium.org>	shill: command line support for DHCPv6 Add command line option "--dhcpv6-enabled-devices" to enable DHCPv6 for specific devices. The obtained DHCPv6 configuration is for informational purpose only, will not be used to setup connection nor update system configuration by shill. Note: since shill does not use DHCPv6 configuration to setup any connection, other IP configuration (DHCPv4 or IPv6 over RDNSS) is required in order for the device to setup a connection for a given network service. BUG=chromium:492764 TEST=USE="dhcpv6 asan clang" emerge-$BOARD shill TEST=USE="asan clang" emerge-$BOARD shill TEST=Verify with new DHCPv6 client auotest Manual Test: 1. Build/install IPv6 enabled dhcpcd and DHCPv6 enabled shill 2. Setup a virtual ethernet pair using: "ip link add name veth_master type veth peer name slave" 2. Setup a DHCPv6 server on veth_master interface "/usr/sbin/dhcpd -6 -cf /tmp/dhcp6.conf veth_master" 3. Start shill with DHCPv6 enabled for "slave" interface 4. Verify DHCPv6 configuration via "/usr/local/lib/flimflam/test/list-ipconfigs slave" Change-Id: Ic152400f61e5bb4cb02f237b97da971ed24da27f Reviewed-on: https://chromium-review.googlesource.com/276713 Reviewed-by: Paul Stewart <pstew@chromium.org> Tested-by: Zeping Qiu <zqiu@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
675d0b0f04936050a357722f52dc078a3ab671d8	03-Jun-2015	Peter Qiu <zqiu@chromium.org>	shill: move dhcp codes to dhcp directory Number of files related to dhcp will increase significantly after refactoring DHCPConfig and adding support for DHCPv6. So move these files to a dedicated directory. BUG=chromium:496302 TEST=USE="asan clang" FEATURES=test emerge-$BOARD shill Change-Id: I7f47365158fc5e1d902946dc3403720402f9b6ad Reviewed-on: https://chromium-review.googlesource.com/275019 Reviewed-by: Zeping Qiu <zqiu@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> Tested-by: Zeping Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
08add488849f90600a5657a6f54f4dbc34701b8f	20-Apr-2015	Matthew Wein <mattwein@chromium.org>	shill: Allow hostname setting via DHCP Allow shill to optionally allow a Device to (a) request the hostname parameter and (b) accept this hostname from a dhcp config and apply it locally. BUG=chromium:475699 TEST=None CQ-DEPEND=CL:266519 Change-Id: I56140831768255b04c6f5c558feec4a9119f349b Reviewed-on: https://chromium-review.googlesource.com/266647 Reviewed-by: Matthew Wein <mattwein@chromium.org> Tested-by: Matthew Wein <mattwein@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
1a72f5444e077ed21b8e085b17d7f9f1bc44fa5e	15-Apr-2015	Peter Qiu <zqiu@chromium.org>	shill: compiler option support for wifi and wired 802.1x This allows shill to be compiled without wifi and wired 802.1x independently. The supplicant related code will not be compiled when both wifi and wired 802.1x compiler option are disabled. BUG=chromium:475665 TEST=USE="asan clang -wifi" FEATURES=test emerge-$BOARD shill USE="asan clang -wired_8021x" FEATURES=test emerge-$BOARD shill USE="asan clang -wifi -wired_8021x" FEATURES=test emerge-$BOARD shill USE="asan clang" FEATURES=test emerge-$BOARD shill Default shill (including both wifi and wired 802.1x): Passes wifi_matfunc test suite shill without wifi No wifi connectivity Have ethernet connectivity Passes network_8021xWiredAuthentication test shill without wired 802.1x Have wifi and ethernet connectivity Fail network_8021xWiredAuthentication test shill without both wifi and wired 802.1x No wifi connectivity Have ethernet connectivity Fail network_8021xWiredAuthentication test CQ-DEPEND=CL:265614 Change-Id: I6abaa150657f6ca0f6d5c37229fde9a2c0bb36aa Reviewed-on: https://chromium-review.googlesource.com/265774 Tested-by: Zeping Qiu <zqiu@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
da86578bc413a76eca81903c42e0123420df4caa	13-Apr-2015	Nathan Bullock <nathanbullock@google.com>	shill: Pass groups to RtnlHandler.Start Allow clients of the RtnlHandler to explicitly define which groups they want to listen to. This will allow more flexibility in the clients without causing everyone to have to listen to all groups. BUG=none TEST=FEATURES=test emerge-$BOARD apmanager shill Change-Id: I2682f6bcfda3a50dade2d1a2b34a2bc1254d7fde Reviewed-on: https://chromium-review.googlesource.com/265460 Reviewed-by: Nathan Bullock <nathanbullock@google.com> Commit-Queue: Nathan Bullock <nathanbullock@google.com> Tested-by: Nathan Bullock <nathanbullock@google.com> /system/connectivity/shill/shill_daemon.cc
782cdcea789d76a7ad7802cf9410b14a2399c0a9	01-Apr-2015	Garret Kelly <gdk@chromium.org>	shill: add minimum-mtu command-line parameter Add the --minimum-mtu option to shill to allow clients to override small MTU values set by DHCP servers. Any MTU value smaller than the provided one is dropped. BUG=chrome-os-partner:36886 TEST=FEATURES=test emerge-amd64-generic shill, ASAN profile Change-Id: I8e79a18f335705887568e04050712d565d53c4f0 Reviewed-on: https://chromium-review.googlesource.com/263522 Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Garret Kelly <gdk@chromium.org> Tested-by: Garret Kelly <gdk@chromium.org> /system/connectivity/shill/shill_daemon.cc
c5f89d131e5b03f448b73fd02d16cab30e438521	18-Feb-2015	Garret Kelly <gdk@chromium.org>	shill: add new --prepend-dns-servers flag The --prepend-dns-servers flag causes every Device to prepend resolvers of the same address family as their local address to their list of resolvers before setting it. It is guaranteed that the same resolver will not appear more than once in the output list. BUG=chrome-os-partner:36696 TEST=amd64-generic unit tests and ASAN profile unit tests CQ-DEPEND=CL:250873 Change-Id: I052f6bb8bd222880e1b18097c72a96f11cd50559 Reviewed-on: https://chromium-review.googlesource.com/250877 Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Garret Kelly <gdk@chromium.org> Tested-by: Garret Kelly <gdk@chromium.org> /system/connectivity/shill/shill_daemon.cc
5a2097c0f4db1704dd5b7b0b874b669bb21c42ad	11-Feb-2015	Jason Simmons <jsimmons@google.com>	shill: Add a command line option that controls the default technology order This can be used to configure a preference for a particular network on a given device. For example, a device may want to use the mobile network for all traffic even if an open wifi network is available. BUG=chromium:453607 TEST=Added the flag via /etc/init/shill.override and confirmed that dbus GetServiceOrder returned the new order Change-Id: Ifac0baa4feeaae4a780d4d74f2bd195b4e4dd31d Reviewed-on: https://chromium-review.googlesource.com/248610 Reviewed-by: Paul Stewart <pstew@chromium.org> Tested-by: Jason Simmons <jsimmons@chromium.org> Commit-Queue: Jason Simmons <jsimmons@chromium.org> /system/connectivity/shill/shill_daemon.cc
83d625655edbc7c35fc436b8b8fe744886d0aae1	10-Feb-2015	Paul Stewart <pstew@chromium.org>	shill: DeviceInfo: Optionally ignore unknown Ethernet Before CL:224422, any unknown network devices (ones that did not report a driver symlink) were treated as "Unknown" and were not managed by shill. However some broken Ethernet drivers do not report this. This change allows a command-line option to revert this workaround so virtual network interfaces and other such non-reported devices can be ignored by the manager again. Additionally, explicitly ignore all devices that appear with a "veth" prefix. BUG=chromium:457043,chromium:456104,chromium:451714 TEST=Unit tests Change-Id: I513c7014c2159d7fea608eed85c0e33b02a893ba Reviewed-on: https://chromium-review.googlesource.com/247904 Reviewed-by: Zeping Qiu <zqiu@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
1d499ed2d5eede24f55c7f6b8aaba06136ef4f22	31-Jan-2015	Peter Qiu <zqiu@chromium.org>	shill: add passive mode support Add a command line option to enable passive-mode. When this option is enabled, shill will not manage any devices by default. Remote application can specify the devices for shill to manage through org.chromium.flimflam.Manager::ReleaseInterface/ClaimInterface APIs. All devices claims will go through the default device claimer, with claimer name of empty string (""). While there, update the ReleaseInterface API to also take claimer name as an argument, so that the claimer name will be check before releasing a device. This is to avoid devices being released by other unknown claimers. BUG=chromium:454007 TEST=USE="asan clang" FEATURES=test emerge-$BOARD shill apmanager Manual Test: 1. Start shill with passive mode enabled, verify "/usr/local/lib/flimflam/list-device" does not return any device. 2. Use dbus-send command to release "eth0" device from default claimer, and verify "/usr/local/lib/flimflam/list-device" does return eth0 device. 3. Use dbus-send command to claim "eth0" device from default claimer, and verify "/usr/local/lib/flimflam/list-device" does not return any device. 4. Regression test with apmanager. Change-Id: I2b79a13e8aea0d07e9d7923ba3c26adb19237e1a Reviewed-on: https://chromium-review.googlesource.com/245123 Reviewed-by: Zeping Qiu <zqiu@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> Trybot-Ready: Zeping Qiu <zqiu@chromium.org> Tested-by: Zeping Qiu <zqiu@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
6d856a6df7f112a84a5799d66a36b7b5d6bb3e50	14-Jan-2015	Peter Qiu <zqiu@chromium.org>	shill: create Sockets object in RTNLHandler Create and maintain Sockets object in RTNLHandler to prevent the possibility of Sockets object being destructed before RTNLHandler. BUG=chromium:448823 TEST=USE="asan clang" FEATURES=test emerge-$BOARD shill apmanager Verify we still have network connectivity through shill Verify apmanager doesn't seg fault when exiting Change-Id: I3a60068718ba6c547cef1437aa1fec22bd27e3eb Reviewed-on: https://chromium-review.googlesource.com/240801 Trybot-Ready: Zeping Qiu <zqiu@chromium.org> Tested-by: Zeping Qiu <zqiu@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Zeping Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
8dc5e7b99920ab79d2bee7f99fec9fa5765fda68	12-Dec-2014	Paul Stewart <pstew@chromium.org>	shill: DBus: Fix issues with the exit path Fix a short list of issues that prevented shill from exiting cleanly: - In daemon mode stderr is redirected to a child process running "logger". When the shill process receives a signal, the logger process is also killed. Catch and ignore SIGPIPE errors that occur as a result of this defunct process. - Both DHCPControl and ProxyFactory created separate DBus::Connection instances. Either dbus-c++ or the dbus library has a double free problem when the second instance is freed. Instead, create a new singleton SharedDBusConneciton so both these modules can share the DBus connection. - Ownership of dbus_control shill_main is passed to the ShillDaemon instance. As such is an error to create an at-exit handler to explicitly delete it as this would constitute a double-free. - The DHCP provider singleton held a reference to a DBus proxy which was only removed in the LazyInstance at-exit-handler. Release this instance explicitly in Daemon::Stop() so it is freed while the DBus connection is still active. - Explicitly destroy manager_ in Daemon::Stop() so that its adaptor is freed while the DBus connection is still active. BUG=chromium:441413 TEST=Unit tests + "stop shill" and ensure "PIPE" message does not appear in logs. Change-Id: I0239ccc24fee05ef9b2c953ef04fa64acd817349 Reviewed-on: https://chromium-review.googlesource.com/234902 Reviewed-by: Ben Chan <benchan@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
d6a8b519e3057f52313e502f502a4819a942e089	18-Nov-2014	Ben Chan <benchan@chromium.org>	shill: wifi: Move wifi code to its own sub-directory. This CL is part of a series of CLs, which aim to reorganzie shill's flat source code structure into a more modular form by moving technology specific code into their own sub-directories. BUG=chromium:433419 TEST=`FEATURES=test emerge-$BOARD shill` TEST=`USE='clang asan' FEATURES=test emerge-$BOARD shill` Change-Id: Ic6bf64e7fe263fd0b2277877d78cf776f179e7d7 Reviewed-on: https://chromium-review.googlesource.com/230460 Reviewed-by: Paul Stewart <pstew@chromium.org> Tested-by: Ben Chan <benchan@chromium.org> Commit-Queue: Ben Chan <benchan@chromium.org> /system/connectivity/shill/shill_daemon.cc
c9c31d8497c3f053c2160408cc386010fc125fad	22-Oct-2014	Rebecca Silberstein <silberst@chromium.org>	shill: add object id to scoped log messages Add object identifier (DBus::Path) string to scoped logging messages. If the logging message is called from a static method, SLOG will be called with nullptr and the log will use (anon) for the object id. Objects without identifying information will use the (class_name) as their logged identifier. BUG=chromium:403996 TEST=ran unit tests and manually inspected net.log Change-Id: Idf23911a303f5edc4b82917bf1e2cea3f8e44e60 Reviewed-on: https://chromium-review.googlesource.com/224812 Tested-by: Rebecca Silberstein <silberst@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Rebecca Silberstein <silberst@chromium.org> /system/connectivity/shill/shill_daemon.cc
02e3dc3bc6e9fdb475bd13944f1c6764c921abbb	31-Oct-2014	Peter Qiu <zqiu@chromium.org>	shill: Move netlink functionality to libshill-net shared library By exporting netlink functionality to a shared library, other applications can use the interfaces provided by this library to communicate with the cfg80211 kernel module and mac80211 drivers. To use this library, add dependency for libshill-net-<(libbase_ver) in your gyp file, and include the headers from <shill/net/*> BUG=chromium:427982 TEST=unittests, run wifi_matfunc suite for shill regression test Change-Id: Ia8ea063daaa63de27fe81e153458e149586086ec Reviewed-on: https://chromium-review.googlesource.com/226830 Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Peter Qiu <zqiu@chromium.org> Tested-by: Peter Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
8d6b59704591ba9fad57751858835dc332dbdd37	28-Oct-2014	Peter Qiu <zqiu@chromium.org>	shill: export a shared library for RTNL interface. Export a shared library "libshill-net-<(libbase-ver)" from shill, which currently includes the RTNL interface and other codes that it depends on (mainly ByteString, IPAddress, IOHandler, TimeStamp, and Sockets). All the library codes are moved to under "shill/net/". Upcoming changes: - Remove dependency for "shill/error.h" which depends on dbus-c++. - Add Netlink interface to the shared library. BUG=chromium:427982 TEST=unittests CQ-DEPEND=CL:226103 Change-Id: I7f5c183b8a347c02e80ffed5e3497f691dde517c Reviewed-on: https://chromium-review.googlesource.com/226110 Reviewed-by: Peter Qiu <zqiu@chromium.org> Tested-by: Peter Qiu <zqiu@chromium.org> Commit-Queue: Peter Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
37a2b60098f7404a0faa918694072670de1a0a97	27-Oct-2014	Peter Qiu <zqiu@chromium.org>	shill: IO handler support for IO message loop Added IO handler factory for creating IO handlers based on the current message loop type. For IO message loop, IO handlers will be based on the libbase's FileDescriptorWatcher, while UI message loop will use the glib's IO handlers (since libbase doesn't provide FileDescriptorWatcher for UI message loop that uses glib). Also remove the dependency of EventDispatcher from RTNLHandler and NetlinkManager. And redirect the IO handler creation calls in EventDispatcher to IOHandlerFactory. Eventually would delete the IO handler creation calls from EventDispatcher once all references to them are removed. This is a part of an effort to export rtnl/nl80211 functionality as a shared library, so that they can be exported without tying to a specific message loop type. BUG=chromium:427534 TEST=unittests, manual regression test (run wifi_matfunc suite) Change-Id: Ibb3e11ef381d1d93dcae20a697cee310f0a7703f Reviewed-on: https://chromium-review.googlesource.com/225651 Tested-by: Peter Qiu <zqiu@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Peter Qiu <zqiu@chromium.org> /system/connectivity/shill/shill_daemon.cc
fbe8d2b361a51e89fcf1207050729be86a961f5f	16-Sep-2014	Samuel Tan <samueltan@chromium.org>	shill: use ResultAggregator for device suspend actions Use ResultAggregator to have manager asynchronously execute each device's suspend actions before a system suspend takes place. The previous implementation executed these suspend actions synchronously and used a HookTable for other termination/suspend actions. This implementation does not change the use of the HookTable for termination action, but uses only a ResultAggregator to wait for asynchronous suspend actions. Manager now creates a ResultAggregator in OnSuspendImminent() and passes a callback to every device as an argument to OnBeforeSuspend(). Each device then performs synchronous actions and/or posts asychronous tasks in OnBeforeSuspend(), and is responsible for running the received callback after completing these actions. Also: - rename RetrySetWakeOnPacketConnection to RetryApplyWakeOnWiFiSettings for consistency. - add detailed documentation to result_aggregator.h explaining and stating the conventions to be followed when using ResultAggregator. - remove TerminationActionReason from metrics and add separate Termination action and Suspend action metrics in preparation for complete splitting of termination and suspend action codepaths. BUG=chromium:390668 CQ-DEPEND=CL:223098 TEST='P2_TEST_FILTER="shill::*" FEATURES="test" USE="clang asan" emerge-samus shill' succeeds. Change-Id: I5a9d9039669a822e0cb82bf6784522f1ce86a966 Reviewed-on: https://chromium-review.googlesource.com/218551 Reviewed-by: Samuel Tan <samueltan@chromium.org> Tested-by: Samuel Tan <samueltan@chromium.org> Commit-Queue: Samuel Tan <samueltan@chromium.org> /system/connectivity/shill/shill_daemon.cc
ca324740fb53093c4bd4ffa73cd4b9ebe09a01be	29-Aug-2014	Paul Stewart <pstew@chromium.org>	shill: ShillDaemonTest: Fix ASAN issues ShillDaemon had a raw pointer to the ControlInterface. Change this to a scoped_ptr so it is freed when the Daemon instance destructs. BUG=chromium:408818 TEST=P2_TEST_FILTER="shill::ShillDaemonTest.*" FEATURES=test \ USE="clang asan" emerge-falco platform2 Change-Id: Ia8e4ed773cacb6a5906c2188a6b6bc3e7c20716a Reviewed-on: https://chromium-review.googlesource.com/215669 Reviewed-by: Ben Chan <benchan@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
3bdf1abfba480415a4714108e615649f422197e0	18-Jul-2014	Paul Stewart <pstew@chromium.org>	shill: DHCPClient: Output status metrics Output metrics about the DHCP client status whenever this state changes. This will allow us track aggregate activity so we can track the efficacy of the various DHCP client features. BUG=chromium:384897 TEST=Unit tests + manual: restart shill and browse to chrome://histograms Change-Id: I8e9628243c8b9b6e8bd0270e86795737a2f75846 Reviewed-on: https://chromium-review.googlesource.com/208825 Reviewed-by: Peter Qiu <zqiu@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
c350e68360d08626cff8c4020c743b61d7da5a2b	20-Jun-2014	Paul Stewart <pstew@chromium.org>	shill: Further deprecate CaCertNSS Chrome has been updated to no longer set the CaCertNSS property for OpenVPN, L2TP/IPSec and 802.1x connections. It also has code to migrate such properties over to the equivalent CaCertPEM property. This CL rips out the underpinnings in shill that used to convert NSS properties by reading the user NSS database. It still supports the various NSS service properties for read (to allow Chrome to detect and migrate away from the NSS property) and write (so Chrome can clear the vestigial property) in all the places where that was previously supported. CQ-DEPEND=CL:205152 BUG=chromium:385401 TEST=Unit tests Change-Id: Id3df385148acb8e9c4240bbaa980e4118aa6b088 Reviewed-on: https://chromium-review.googlesource.com/204763 Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
b7bf216f6f8b482123820afcfe22a7c2d3b19c9f	12-Jun-2014	Ben Chan <benchan@chromium.org>	shill: Fix variable declaration order in Daemon. This CL fixes the declaration order of member variables of Daemon such that variables being depended on are declared first. This CL also removes unused header inclusions and fixes some formatting issues. BUG=None TEST=`FEATURES=test emerge-$BOARD platform2` Change-Id: I99d40638011efe24cab8e519619931f5c9a5c904 Reviewed-on: https://chromium-review.googlesource.com/203562 Reviewed-by: Paul Stewart <pstew@chromium.org> Tested-by: Ben Chan <benchan@chromium.org> Commit-Queue: Ben Chan <benchan@chromium.org> /system/connectivity/shill/shill_daemon.cc
a3ac09632c45a9dce9ba21bb6437f2858336943c	12-Jun-2014	Ben Chan <benchan@chromium.org>	shill: Fix a crash when terminating daemon in presence of cellular device This CL fixes a crash caused by destroying a Cellular object in midst of executing the completion callback of its termination action. BUG=chromium:383767 TEST=Tested the following: 1. `FEATURES=test emerge-$BOARD platform2` 2. On a DUT with a cellular device in the enabled state, verify that `restart shill` will cause shill to execute the termination action without crashing. Change-Id: Ia521b00688816961e25f74dd87d81176220e7ea9 Reviewed-on: https://chromium-review.googlesource.com/203561 Reviewed-by: Paul Stewart <pstew@chromium.org> Tested-by: Ben Chan <benchan@chromium.org> Commit-Queue: Ben Chan <benchan@chromium.org> /system/connectivity/shill/shill_daemon.cc
0a59a5a484634f798222ad077dbbfa7faa76651d	25-Apr-2014	mukesh agrawal <quiche@chromium.org>	shill: stop passing around format strings for profiles Prior to this CL, shill passed around format strings, to specify where the user profiles should be stored. While this makes things more configurable, it also a) makes things more complex, and b) causes g++ to emit warnings that some format strings can't be checked statically. The location of user profiles doesn't need the full power of printf-format style configurability. We really only need the ability to specify which directory the user profiles are stored in. Note that this changes requires updating some unit tests, which actually did take advantage of the flexibility. In particular, some of the Manager and Profile unit tests placed user profiles directly in a user profile directory, rather than creating a user-specific sub-directory first. Creating the user-specific directories seems like the right thing to do, since that it what the normal (non-test) code does. BUG=chromium:293668 TEST=unit tests Change-Id: Ic1afeec84a7797105c9a49b8261a9677e17d91ee Reviewed-on: https://chromium-review.googlesource.com/197061 Tested-by: mukesh agrawal <quiche@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: mukesh agrawal <quiche@chromium.org> /system/connectivity/shill/shill_daemon.cc
a0ddf46e466bd4ba3d20952f0a6988c680c1af14	06-Feb-2014	Ben Chan <benchan@chromium.org>	shill: Update to build against libchrome-242728. BUG=chromium:341521 CQ-DEPEND=CL:193660 TEST=Tested the following: 1. `FEATURES=test USE='cellular wimax' emerge-$BOARD platform2` 2. Run the following tests: - network_3GSmokeTest - network_VPNConnect.* - network_WiMaxSmoke - wifi_matfunc Change-Id: Ic1553c182ab7a833a68c45f012f646b8930cb095 Reviewed-on: https://chromium-review.googlesource.com/193606 Reviewed-by: Ben Chan <benchan@chromium.org> Tested-by: Ben Chan <benchan@chromium.org> Commit-Queue: Ben Chan <benchan@chromium.org> /system/connectivity/shill/shill_daemon.cc
c272896939abdb97c686df48345c7d6651337736	10-Jul-2013	Wade Guthrie <wdg@chromium.org>	shill: Fixes disconnect UMA stat To determine the kernel's disconnect notification behaviour for these UMA stats, I originally reverse-engineered the disconnect netlink messages that we get from the kernel (should have looked at the kernel code which, eventually, I did). I found, erroneously, that all of the disconnect reasons were found in the disassociate message. It turns out that the disconnect reasons are from the disassociate message only for (some!) disconnections instigated by the station. When the AP is the source of the disconnect, the reason is stored in a disconnect message. This code, then, extracts the reason code from whatever message type (disconnect or disassociate) that contains one. The message will still be a little noisy since some station-started disconnections don't provide a disassociate message and I've noticed that some AP-started disconnections seem to send 2 messages. I haven't tracked any of this down. BUG=chromium:215808 TEST=unittest (new) and manual - I test using the AP and description found in chromium:294315. CQ-DEPEND=Icf79aa729b1ed125743686c4536fe1b59183fed2 Change-Id: If648d530c613f485c074acf58ddb0bca4de22084 Reviewed-on: https://chromium-review.googlesource.com/170926 Reviewed-by: Wade Guthrie <wdg@chromium.org> Commit-Queue: Wade Guthrie <wdg@chromium.org> Tested-by: Wade Guthrie <wdg@chromium.org> /system/connectivity/shill/shill_daemon.cc
13804a1acd238a3c7ba8f4f44db96a97fa04fbda	09-Oct-2013	Ben Chan <benchan@chromium.org>	shill: Run termination actions before stopping DBus proxies. shill uses a termination action to disable the cellular device upon termination (CL:171374). It needs to execute the termination action before calling Daemon::Stop(), which stops the DBus proxies to ModemManager. Otherwise, the termination action will fail to initiate the disable operation over ModemManager's DBus interface. shill previously disconnected but didn't disable the cellular device upon termination, and thus called Daemon::Stop() before running the termination action in order to prevent the autoconnect code from trying to reconnect the disconnected cellular service. However, that's not necessary as Manager::AutoConnect() actually checks if the manager is still running. BUG=chromium:305445 TEST=Manually test the following with E396, E362, ALT3100, MU736 modem: 1. Enable shill verbose logging under a root shell: ff_debug manager+device+service+cellular+modem ff_debug --level -5 2. Connect to Ethernet, WiFi, and cellular. 3. Run `stop shill` under a root shell to terminate the shill process. 4. Verify from /var/log/net.log that shill starts the termination actions and completes them before destroying DBus proxies to ModemManager. WiFi is disconnected and cellular is disabled. shill: [1008/170201:INFO:shill_main.cc(122)] Shutting down due to received signal. shill: [1008/170201:INFO:manager.cc(1113)] Running termination actions. shill: [1008/170201:INFO:cellular.cc(802)] StartTermination ... shill: [1008/170204:INFO:cellular.cc(819)] OnTerminationCompleted: org.chromium.flimflam.Error.Success:Success (no error) ... shill: [1008/170211:INFO:modem_manager.cc(50)] Stop watching modem manager service: org.chromium.ModemManager shill: [1008/170211:INFO:modem_manager.cc(50)] Stop watching modem manager service: org.freedesktop.ModemManager1 ... shill: [1008/170211:VERBOSE2:shill_main.cc(115)] DeleteDBusControl Change-Id: I424f79036f07b7aacf795389dd06654c7ece4606 Reviewed-on: https://chromium-review.googlesource.com/172330 Tested-by: Ben Chan <benchan@chromium.org> Reviewed-by: Thieu Le <thieule@chromium.org> Reviewed-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Ben Chan <benchan@chromium.org> /system/connectivity/shill/shill_daemon.cc
318445d33275a29a5d5f50a507b138a28b106d1b	16-May-2013	Wade Guthrie <wdg@chromium.org>	shill: Cleanup: Removes NetlinkManager::GetMessageType Code could get the netlink message type from the NetlinkManager or from the specific message class (e.g., Nl80211Message::GetMessageType). This CL removes the former in favor of the latter. BUG=None TEST=unittest Change-Id: Ib8d280dff96f4a87cdde20a7b7529280e8414011 Reviewed-on: https://gerrit.chromium.org/gerrit/51501 Reviewed-by: Darin Petkov <petkov@chromium.org> Commit-Queue: Wade Guthrie <wdg@chromium.org> Reviewed-by: Wade Guthrie <wdg@chromium.org> Tested-by: Wade Guthrie <wdg@chromium.org> /system/connectivity/shill/shill_daemon.cc
e6c44440f5cf102c14581c961028a7830cc6448f	19-Apr-2013	Paul Stewart <pstew@chromium.org>	shill: Remove "--push" argument parsing Now that init no longer uses this argument in vain, remove the "--push" argument from shill's command-line parsing. CQ-DEPEND=CL:48604 BUG=chromium:231858 TEST=Unit tests Change-Id: I3838057b2127f271a23ec40d27b248a91a840bcf Reviewed-on: https://gerrit.chromium.org/gerrit/48607 Reviewed-by: Christopher Wiley <wiley@chromium.org> Tested-by: Paul Stewart <pstew@chromium.org> Commit-Queue: Paul Stewart <pstew@chromium.org> /system/connectivity/shill/shill_daemon.cc
bb9fca2e550f92d694867f0bcf0efe9426dfc2bb	11-Apr-2013	Wade Guthrie <wdg@chromium.org>	shill: Renames Config80211 to NetlinkManager. The domain of Config80211 has extended to different types of netlink message communication. That warrants a name change. This CL changes the name of the class plus a bunch of objects of that class. This required changing the names of some files. In addition, a few (half a dozen, maybe) readability issues were addressed. These include inaccurate header guard comments and bad indentation. BUG=None TEST=unittest Change-Id: Id0006d5d37a98ed6c138f9982a2aa3624b308757 Reviewed-on: https://gerrit.chromium.org/gerrit/47860 Reviewed-by: mukesh agrawal <quiche@chromium.org> Commit-Queue: Wade Guthrie <wdg@chromium.org> Reviewed-by: Wade Guthrie <wdg@chromium.org> Tested-by: Wade Guthrie <wdg@chromium.org> /system/connectivity/shill/shill_daemon.cc
12f113aa68bcb9084e342ab1b985caf1aded6f1c	13-Mar-2013	Wade Guthrie <wdg@chromium.org>	shill: Splits the netlink message factory between message types. Provides a mechanism to register message factories with \|NetlinkMessageFactory\|. Splits \|NetlinkMessageFactory\| into a generic factory and one for \|ControlNetlinkMessage\| and \|Nl80211Message\|. Uses the registration mechanism to install the control and nl80211 message factories. This CL also moves \|UnknownMessage\| from being a subtype of \|Nl80211Message\| to being a subtype of \|NetlinkMessage\|. Lucy has some 'splaining to do: At startup time, the message type isn't known for some types of netlink message (notably, the nl80211 class of messages). This message type is discovered, in the code, through the \|Config80211::AddFamilyByString\| call. This CL adds a message factory closure to that call so that Config80211 can install the factory when the code knows the message type against which to register the factory. BUG=chromium:220372 TEST=unittest and manual (restarted shill with log=-10,wifi and observed messages of all types being created for netlink messages coming from the kernel). Change-Id: I496b405c20ef98b6e23e96e8088fc542ff29264c Reviewed-on: https://gerrit.chromium.org/gerrit/45314 Commit-Queue: Wade Guthrie <wdg@chromium.org> Reviewed-by: Wade Guthrie <wdg@chromium.org> Tested-by: Wade Guthrie <wdg@chromium.org> /system/connectivity/shill/shill_daemon.cc
c6c8196e736baea5269cd8a716ddd455b515cac9	07-Mar-2013	Wade Guthrie <wdg@chromium.org>	shill: Simplifies netlink message callbacks. Config80211 had required broadcast callbacks to be child classes of some special (and, as it turns out, completely unnecessary) superclass. This CL a) replaces the superclass with \|base::Callback\|, b) updates callback_metrics to use this new mechanism, and c) removes callback_object (which was really just a template for future callback and is no longer necessary). BUG=None. TEST=unittests. Change-Id: I8f080521ba17e45e678289d10060f8e8927c0496 Reviewed-on: https://gerrit.chromium.org/gerrit/44779 Commit-Queue: Wade Guthrie <wdg@chromium.org> Reviewed-by: Wade Guthrie <wdg@chromium.org> Tested-by: Wade Guthrie <wdg@chromium.org> /system/connectivity/shill/shill_daemon.cc