NetworkManager merge requestshttps://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests2021-05-04T06:27:15Zhttps://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/836core/firewalld: use changeZoneOfInterface() instead of deprecated changeZone()2021-05-04T06:27:15ZThomas Hallercore/firewalld: use changeZoneOfInterface() instead of deprecated changeZone()`man firewalld.dbus` comments that changeZone() is deprecated.
Apparently, this is the case since 0.3.3-1 (2013).`man firewalld.dbus` comments that changeZone() is deprecated.
Apparently, this is the case since 0.3.3-1 (2013).https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/835[th/uuid] drop util-linux' libuuid and normalize connection.uuid2021-05-04T13:53:37ZThomas Haller[th/uuid] drop util-linux' libuuid and normalize connection.uuid- drop dependency on util-linux' libuuid. We used it to parse, unparse and generate a random UUID. These are trivial things, let's not drag in a dependency to libnm.
- move UUID helper code from libnm-core-impl to libnm-glib-aux. This i...- drop dependency on util-linux' libuuid. We used it to parse, unparse and generate a random UUID. These are trivial things, let's not drag in a dependency to libnm.
- move UUID helper code from libnm-core-impl to libnm-glib-aux. This is independent code from NetworkManager, so it should not reside in libnm(-core).
- `connection.uuid` accepts certain invalid UUIDs and does not enforce a consistent (lower) case. Add a normalization for that. This changes behavior for existing profiles that used such UUIDs.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/833Fixes and small refactor of wpa_supplicant config generation2022-01-10T18:05:56ZJonas DreßlerFixes and small refactor of wpa_supplicant config generationThis MR is a follow-up of https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/826.
The biggest changes are:
- refactors the way `key_mgmt` of the wpa_supplicant config gets generated (this is kind of optional,...This MR is a follow-up of https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/826.
The biggest changes are:
- refactors the way `key_mgmt` of the wpa_supplicant config gets generated (this is kind of optional, but I think it's way cleaner)
- makes PMF necessary for all WPA3 AKM suites
- Use wpa-psk as key-mgmt instead of sae when AP support both WPA2 and WPA3https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/830[th/platform-ethtool-cleanup] platform: cleanup ethtool calls in "nm-platform...2021-05-03T11:57:44ZThomas Haller[th/platform-ethtool-cleanup] platform: cleanup ethtool calls in "nm-platform-utils.c"- consistently check for success/failure of _ethtool_call_handle()
with "< 0" / ">= 0".
- drop unnecessary memset(). In the past, I argued to add this because
there were obscure cases with valgrind where this made a difference.
As...- consistently check for success/failure of _ethtool_call_handle()
with "< 0" / ">= 0".
- drop unnecessary memset(). In the past, I argued to add this because
there were obscure cases with valgrind where this made a difference.
As it's not clear when/how that is necessary, drop it again.
Also, we want to prefer explicit struct initialization over memset(),
so if memset() would be necessary, those places would be problematic
as well.
- inline unnecessary helper functions. They had only one caller and
only make the code more verbose.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/829Introducing ethtool PAUSE support2021-05-14T11:15:08ZGris GeIntroducing ethtool PAUSE supportIntroducing ethtool PAUSE support with:
* ethtool.pause-autoneg
* ethtool.pause-rx
* ethtool.pause-tx
Limitations:
* When `ethtool.pause-autoneg` is set to true, the `ethtool.pause-rx`
and `ethtool.pause-tx` will be ignored. We ...Introducing ethtool PAUSE support with:
* ethtool.pause-autoneg
* ethtool.pause-rx
* ethtool.pause-tx
Limitations:
* When `ethtool.pause-autoneg` is set to true, the `ethtool.pause-rx`
and `ethtool.pause-tx` will be ignored. We don't have warning for
this.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/828iwd: A few fixes for the iwd backend2021-04-27T17:55:21ZJonas Dreßleriwd: A few fixes for the iwd backendJust a few things I noticed when playing around with the iwd backend.Just a few things I noticed when playing around with the iwd backend.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/825[th/systemd-resolved-resolve] add code for making D-Bus call to systemd-resol...2021-04-28T11:21:57ZThomas Haller[th/systemd-resolved-resolve] add code for making D-Bus call to systemd-resolved's ResolveAddress APIThis will be used later for #601.This will be used later for #601.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/824Replace `./tools/generate-docs-nm-property-infos.pl` with python script2021-05-05T13:36:36ZWen LiangReplace `./tools/generate-docs-nm-property-infos.pl` with python scriptSigned-off-by: Wen Liang <liangwen12year@gmail.com>Signed-off-by: Wen Liang <liangwen12year@gmail.com>https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/822[th/dhcp-request-broadcast] set DHCP request-broadcast based on udev property...2021-04-28T11:22:13ZThomas Haller[th/dhcp-request-broadcast] set DHCP request-broadcast based on udev property ID_NET_DHCP_BROADCASTSee https://github.com/systemd/systemd/pull/19346
We don't currently support configuring request-broadcast in NetworkManager's connection profile. But we still can honor the udev property...See https://github.com/systemd/systemd/pull/19346
We don't currently support configuring request-broadcast in NetworkManager's connection profile. But we still can honor the udev property...https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/821[th/cloud-setup-azure-fix-gateway]2021-10-14T17:36:22ZThomas Haller[th/cloud-setup-azure-fix-gateway]Probably fixes https://bugzilla.redhat.com/show_bug.cgi?id=1912236
See the logfile from https://bugzilla.redhat.com/show_bug.cgi?id=1912236#c3, that we add a route
```
platform: (eth0) route: append IPv4 route: type unicast table 3...Probably fixes https://bugzilla.redhat.com/show_bug.cgi?id=1912236
See the logfile from https://bugzilla.redhat.com/show_bug.cgi?id=1912236#c3, that we add a route
```
platform: (eth0) route: append IPv4 route: type unicast table 30400 0.0.0.0/0 via 0.0.0.1 dev 2 metric 10 mss 0 rt-src user
```https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/820[th/minor-cleanups]2021-04-20T15:33:23ZThomas Haller[th/minor-cleanups]Some simple patches, that fell out of `th/l3cfg-18` WIP branch.Some simple patches, that fell out of `th/l3cfg-18` WIP branch.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/819[th/nmclient-context-doc] libnm/doc: clarify GMainContext handling in NMClien...2021-04-19T07:32:16ZThomas Haller[th/nmclient-context-doc] libnm/doc: clarify GMainContext handling in NMClient documentationsome doc improvements(?)some doc improvements(?)https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/818platform/tests: relax unit tests for time values2021-04-21T05:48:37ZAaron Baranyplatform/tests: relax unit tests for time valuesConverting the clock value with clock_t_to_jiffies() can cause rounding
errors depending on the system settings. For example, the default settings
in Arch Linux cause the values to be off by one.Converting the clock value with clock_t_to_jiffies() can cause rounding
errors depending on the system settings. For example, the default settings
in Arch Linux cause the values to be off by one.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/815[th/lldp-listener-cleanup] simplify NMLldpListener2021-04-16T13:23:53ZThomas Haller[th/lldp-listener-cleanup] simplify NMLldpListenerSimplify the NMLldpListener API, so that we enable/destroy the NMLldpListener instance to start/stop LLDP. There is no more further state, like running. If it exists, it's running. The goal is to simplify NMDevice, which uses this API.Simplify the NMLldpListener API, so that we enable/destroy the NMLldpListener instance to start/stop LLDP. There is no more further state, like running. If it exists, it's running. The goal is to simplify NMDevice, which uses this API.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/814nm-settings-connection: Remove the system secrets cache2021-04-23T14:53:51ZAndrew Zaborowskinm-settings-connection: Remove the system secrets cacheAs analyzed in #599 NMSettingsConnection's `priv->system_secrets` and the dance that's done to copy the secrets to and from it seems to be a left-over and not currently needed. After looking at all users of `nm_settings_connection_clear...As analyzed in #599 NMSettingsConnection's `priv->system_secrets` and the dance that's done to copy the secrets to and from it seems to be a left-over and not currently needed. After looking at all users of `nm_settings_connection_clear_secrets` I conclude that nothing depends on it so I'm dropping that function and the system secrets cache as proposed in #599 and simplifying some code paths. My testing was limited to running the unit tests and checking a few Wi-Fi scenarios with wrong secrets and correct secrets provided through the agent UI with the IWD and wpa_supplicant backends.
This MR is based on !813 as that ones is less invasive. There are only two new commits on top of !813.
Note: there's also an agent-owned secrets cache in NMSettingsConnection that is only used once, to preserve the agent-owned secrets over a D-Bus .Update() call.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/813Minor IWD improvments related to secrets and the config file writing2021-04-23T14:53:51ZAndrew ZaborowskiMinor IWD improvments related to secrets and the config file writingA few changes to the new IWD config file generation feature and secret requests.
The one change that touches core code is the addition of the `NM_SETTINGS_CONNECTION_UPDATE_REASON_UPDATE_NON_SECRET` flag that, like `NM_SETTINGS_CONNECTI...A few changes to the new IWD config file generation feature and secret requests.
The one change that touches core code is the addition of the `NM_SETTINGS_CONNECTION_UPDATE_REASON_UPDATE_NON_SECRET` flag that, like `NM_SETTINGS_CONNECTION_UPDATE_REASON_RESET_SYSTEM_SECRETS`
and `NM_SETTINGS_CONNECTION_UPDATE_REASON_RESET_AGENT_SECRETS` tells the NMSettingConnection's `updated` signal handlers what has changed in the connection.https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/811dhcp-client: Support using valid lease configuration on timeout2021-06-27T04:51:30ZJonathan Kangdhcp-client: Support using valid lease configuration on timeoutdhclient checks lease file on timeout. If they are still valid, the ip
configs will be used. While currently NetworkManager directly fails
the dhcp configuration once timeout is reached.
Fix that by using ip configs found in the valid l...dhclient checks lease file on timeout. If they are still valid, the ip
configs will be used. While currently NetworkManager directly fails
the dhcp configuration once timeout is reached.
Fix that by using ip configs found in the valid lease file on timeout,
when dhclient is the dhcp backend.
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/issues/682https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/809[th/dns-minor-cleanup] minor cleanups for nm-dns-manager2021-04-07T05:58:56ZThomas Haller[th/dns-minor-cleanup] minor cleanups for nm-dns-managerhttps://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/808[th/systemd] systemd: update code from upstream (2021-04-06)2021-04-06T17:49:36ZThomas Haller[th/systemd] systemd: update code from upstream (2021-04-06)https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/807Draft: cli: optionally support libedit instead of readline2021-07-05T09:02:25ZAdrian FreihoferDraft: cli: optionally support libedit instead of readlineFor preconfigured, closed embedded devices, the inclusion of GPLv3 software
components is not possible. For NetworkManager this affects libreadline, which
is used by nmcli. This change allows compiling with libedit instead of
libreadline...For preconfigured, closed embedded devices, the inclusion of GPLv3 software
components is not possible. For NetworkManager this affects libreadline, which
is used by nmcli. This change allows compiling with libedit instead of
libreadline.
Support for editline seams to be broken. Therefore it gets removed.
* [x] Test build with autotools (readline / libedit)
* [x] Test build with meson
* [ ] Test libedit linked nmcli
* [ ] Fix incompatible pointer types
```
[490/872] Compiling C object src/nmcli/nmcli.p/agent.c.o
../src/nmcli/agent.c: In function ‘get_secrets_from_user’:
../src/nmcli/agent.c:92:31: warning: assignment to ‘int (*)(const char *, int)’ from incompatible pointer type ‘int (*)(void)’ [-Wincompatible-pointer-types]
92 | rl_startup_hook = set_deftext;
| ^
[493/872] Compiling C object src/nmcli/nmcli.p/common.c.o
../src/nmcli/common.c: In function ‘get_secrets_from_user’:
../src/nmcli/common.c:705:50: warning: assignment to ‘int (*)(const char *, int)’ from incompatible pointer type ‘int (*)(void)’ [-Wincompatible-pointer-types]
705 | rl_startup_hook = nmc_rl_set_deftext;
| ^
[494/872] Compiling C object src/nmcli/nmcli.p/connections.c.o
../src/nmcli/connections.c: In function ‘nmcli_editor_tab_completion’:
../src/nmcli/connections.c:6517:64: warning: assignment to ‘void (*)(void)’ from incompatible pointer type ‘void (*)(char **, int, int)’ [-Wincompatible-pointer-types]
6517 | rl_completion_display_matches_hook = uuid_display_hook;
| ^
../src/nmcli/connections.c:6564:60: warning: assignment to ‘void (*)(void)’ from incompatible pointer type ‘void (*)(char **, int, int)’ [-Wincompatible-pointer-types]
6564 | rl_completion_display_matches_hook = uuid_display_hook;
| ^
../src/nmcli/connections.c: In function ‘property_edit_submenu’:
../src/nmcli/connections.c:7358:29: warning: assignment to ‘int (*)(const char *, int)’ from incompatible pointer type ‘int (*)(void)’ [-Wincompatible-pointer-types]
7358 | rl_startup_hook = nmc_rl_set_deftext;
| ^
```