summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* platform: add flags argument to nm_platform_ip4_address_add()th/platform-ipv4-flagsThomas Haller2016-02-299-21/+32
| | | | | The argument is still always unset. We will need it later to set IFA_F_NOPREFIXROUTE.
* platform: handle ifa_flags for IPv4 addressThomas Haller2016-02-293-16/+34
| | | | | | | | | | | It's useful to handle for IPv4 addresses too. - we might want to sort IPv4 addresses according to whether they are permanent or dynamic. - we want to set IFA_F_NOPREFIXROUTE also for IPv4 addresses. While the ability to "set" a flag doesn't necessarily require that we also keep the flags in NMPlatformIP4Address, it is more consistent.
* merge: branch 'bg/vlan-reorder-hdr-bgo762626'Beniamino Galvani2016-02-2912-47/+95
|\ | | | | | | https://bugzilla.gnome.org/show_bug.cgi?id=762626
| * libnm-core: remove special handling of missing VLAN_FLAGSBeniamino Galvani2016-02-292-20/+3
| | | | | | | | | | | | | | | | | | | | On older NM versions the default value for vlan.flags was 0, but then the actual value set on interfaces was REORDER_HDR. In order to maintain backwards compatibility in behavior, remove the special handling of vlan.flags so that a missing key is treated as the default value REORDER_HDR. https://bugzilla.gnome.org/show_bug.cgi?id=762626
| * ifcfg-rh: change the handling of REORDER_HDR flagBeniamino Galvani2016-02-299-27/+79
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On NM 1.0 connections were created by default without the REORDER_HDR flag, but then due to a bug in platform code (fixed in [1]), the kernel interface always had the flag set. Now that the setting is honored, users upgrading to the new version of NM will see a change from the previous behavior, since interfaces will not have REORDER_HDR and this will certainly break functionality. The only solution here seems to be to ignore the REORDER_HDR variable in ifcfg files (since it never had any effect) and introduce a new NO_REORDER_HDR option for the VLAN_FLAGS variable which allows to turn the flag off. The consequence is that the flag will be set for all old connections. This change introduces an incompatibility with initscripts, however is necessary to avoid breaking user functionality upon upgrade. Connections created through NetworkManager will still be parsed correctly by initscripts (since we always write the REORDER_HDR variable). [1] db62fc9d72fa ("platform: fix adding VLAN flags") https://bugzilla.gnome.org/show_bug.cgi?id=762626
| * utils: add nm_utils_strbuf_init() macroThomas Haller2016-02-291-0/+13
|/
* platform/trivial: rename NMPlatformIP6Address's "flags" field to "n_ifa_flags"Thomas Haller2016-02-298-25/+25
| | | | | | There are far too many "flags". Rename the "flags" to "n_ifa_flags" which reminds to "ifa_flags" in 'struct ifaddrmsg', but with a distinctive "n_" prefix.
* platform: adjust type of IPv6 address flags to guintThomas Haller2016-02-299-14/+15
| | | | | | | Address flags inside 'struct ifaddrmsg' are uint8_t and they are extended to uint32_t via IFA_FLAGS. Change it to be guint32.
* platform: streamline fields in platform's structuresThomas Haller2016-02-291-32/+33
| | | | | - sort the fields to better fill gaps due to alignment. - use "bool" bitfield instead of gboolean.
* platform/trivial: rename NMPlatformLink's "flags" field to "n_ifi_flags"Thomas Haller2016-02-297-29/+29
| | | | | | There are far too many "flags". Rename the "flags" to "n_ifi_flags" which reminds to "ifi_flags" in 'struct ifinfomsg', but with a distinctive "n_" prefix.
* platform: adjust type of link flags to guintThomas Haller2016-02-293-6/+3
| | | | The type of ifi_flags is unsigned, not uint32_t.
* systemd,lldp: merge branch 'th/systemd-lldp'Thomas Haller2016-02-2966-162/+88
|\
| * lldp/tests: unref lldp instance in testThomas Haller2016-02-291-1/+1
| |
| * lldp: handle NULL values in lldp_neighbor_id_hash()Thomas Haller2016-02-291-4/+6
| | | | | | | | | | | | g_str_hash() can not be called with NULL. Ensure that we don't crash. Thereby, refactor the hashing algorithm because the chassis-id and port-id are small numbers and xor-ing can cancel them easily.
| * lldp: don't mix malloc() with g_malloc()Thomas Haller2016-02-291-1/+1
| |
| * lldp: refactor logging to use _LOG() macrosThomas Haller2016-02-291-8/+48
| |
| * lldp: fix calling free on memory returned by systemdThomas Haller2016-02-291-3/+1
| | | | | | | | | | systemd's malloc() must be matched by free() and not g_free(). Thereby, use a cleanup attribute.
| * lldp: refactor process_lldp_neighbors()Thomas Haller2016-02-291-16/+15
| |
| * macros: add nm_auto() macro to cleanup functionsThomas Haller2016-02-291-1/+3
| |
| * systemd: don't include "nm-sd-adapt.h" in systemd header filesThomas Haller2016-02-2962-120/+5
| | | | | | | | | | | | The header files are not compiled directly, but always included while compiling other source files. We already require every source file to include "nm-sd-adapt.h" as first, thus we don't need the include in the headers too.
| * build: sort files for libsystemd_nm_la_SOURCES alphabeticallyThomas Haller2016-02-291-8/+8
|/
* cli: fix use-after-free when adding ADSL connectionsBeniamino Galvani2016-02-291-7/+9
| | | | | | | check_adsl_protocol() can free and assign a new string to @protocol_ask, leaving @protocol dangling. Fix this. Fixes: 290c1626b9788aa90861ca423c0dffb59fe29876
* introspection: fix description for vxlan ageing propertyBeniamino Galvani2016-02-291-2/+2
|
* device: fix compile error due to typoThomas Haller2016-02-291-1/+1
| | | | Fixes: 6d409653b1bc9116bfb58b074f0c61407aef1038
* device/trivial: adjust whitespace for property definitionsThomas Haller2016-02-291-191/+191
|
* device: use _notify() instead of g_object_notify() for object propertiesatul2016-02-291-209/+139
| | | | https://github.com/NetworkManager/NetworkManager/pull/3
* device: the veth should have the same priority as ethernetLubomir Rintel2016-02-271-2/+1
| | | | It's supposed to be a virtual ethernet.
* device: when activating without cloned-mac-address, set the permanent oneLubomir Rintel2016-02-267-13/+11
| | | | | | | Don't rely on what's already on the device. It could be that the MAC address set on the device is not meaningful -- the NM crashed while two devices were teamed together and now they have the same hardware address and now it's impossible to bond them with mode=5.
* wifi/test: fix bug in test code wpa_flags_for_idx()Thomas Haller2016-02-261-1/+1
|
* contrib/rpm: print dnf-install command at the end of build.shThomas Haller2016-02-261-1/+3
|
* arping-manager: failure to get a device name should not be fatalLubomir Rintel2016-02-261-2/+11
| | | | The device could just be removed mid-flight.
* device: trivial: fix indentation and whitespacesBeniamino Galvani2016-02-251-10/+10
|
* callouts: downgrade logging severity of messages to g_info() and g_debug()Thomas Haller2016-02-241-5/+14
| | | | | | | | | | | In dispatcher, we install a log-handler which maps G_LOG_LEVEL_MESSAGE to syslog priority LOG_NOTICE, which in turn causes journal to highlight the message. We don't want that so instead use g_info() and g_debug() which maps to lower syslog levels. There is only one problem, in debug-mode, we don't use syslog but the default logging handler from glib. In this case, we have to set G_MESSAGES_DEBUG otherwise g_info()/g_debug() is suppressed.
* nm-glib: backport g_info() macroThomas Haller2016-02-241-0/+7
|
* libnm,core: fix syntax of 'transfer' annotationBeniamino Galvani2016-02-245-7/+7
|
* ifcfg-rh: add test for BOOTPROTO=autoipLubomir Rintel2016-02-243-0/+26
| | | | | Previous commit fixed parsing of these connection; let's check that we don't regress there.
* ifcfg-rh: short circuit the IP configuration parsing on IPv4LLLubomir Rintel2016-02-241-6/+1
| | | | We need the other keys, such as may_fail, too.
* device: check may_fail when progressing to IP_CHECKLubomir Rintel2016-02-241-12/+29
| | | | | Don't let the device progress to check and then in turn to activated unless all the required IP configuration is in place.
* device/trivial: rename nm_device_check_ip_failed()Lubomir Rintel2016-02-241-10/+10
| | | | | | | This saves 110 bytes (which is 1/29986680757 of the size of the books in the Library of Congress). Also, the function is static; not using the nm_device_ prefix is preferred.
* wifi: ignore monitor interfacesDan Williams2016-02-241-0/+11
| | | | | | | If a monitor interface is created, NM will grab that interface and change it to station mode. That's not very nice. https://mail.gnome.org/archives/networkmanager-list/2016-February/msg00068.html
* platform: optimize sysctl_set() to use stack allocated bufferThomas Haller2016-02-241-4/+12
| | | | | | | | | | | The value written to sysctl is usually a short string. It makes sense to optimize for this case and avoid allocating a temporary string on the heap. An alternative would be to use writev(), which effectively does the same and also creates a temporary buffer (preferably stack allocated). https://mail.gnome.org/archives/networkmanager-list/2016-February/msg00070.html
* device: merge branch 'jk/device-types-bgo762538'Thomas Haller2016-02-239-5/+44
|\ | | | | | | https://bugzilla.gnome.org/show_bug.cgi?id=762538
| * device: add new NMDeviceType NM_DEVICE_TYPE_VETHThomas Haller2016-02-239-1/+16
| | | | | | | | Based-on-patch-by: Jiří Klimeš <jklimes@redhat.com>
| * libnm: add missing device types to get_type_name()Thomas Haller2016-02-232-4/+28
|/
* all: merge branch 'jk/device-types'Thomas Haller2016-02-235-10/+31
|\ | | | | | | | | Resurrect old branch by Jirka, rebase it and merge some parts.
| * libnm-util: add missing device types to NMDeviceTypeThomas Haller2016-02-231-0/+4
| | | | | | | | | | | | | | Althoug we don't fully backport the new device types, at least add the types to the NMDeviceType enum. Based-on-patch-by: Jiří Klimeš <jklimes@redhat.com>
| * examples: add missing device-types to examplesThomas Haller2016-02-233-1/+13
| | | | | | | | Based-on-patch-by: Jiří Klimeš <jklimes@redhat.com>
| * doc: update introspection/all.xml to include missing nm-device xmlsThomas Haller2016-02-231-9/+14
|/
* man: clarify DEVICE_IFACE vs. DEVICE_IP_IFACE in `man NetworkManager`Thomas Haller2016-02-221-2/+7
|
* wifi: avoid autoconnecting Ad-Hoc networks with method=autoLubomir Rintel2016-02-221-4/+5
| | | | | | | This fixes the issue where all Ad-Hoc networks try to connect one after one on NM startup instead of the managed network that has AP available. Fixes: e2637760f160f8d790438f3ca26df1b888de7909
View Lorry Controller Status