Commit Graph

25044 Commits

Author SHA1 Message Date
Beniamino Galvani
9aa00a8a14 dhcp: nettools: fix parsing of search domains option
'first' was never modified and so the dot was never added.

Fixes: 6adade6f21 ('dhcp: add nettools dhcp4 client')

https://bugzilla.redhat.com/show_bug.cgi?id=1783981
2019-12-17 14:33:25 +01:00
Thomas Haller
bc2ca6e603 release: bump version to 1.23.1-dev after 1.22.0 release
After 1.22.0 is released, merge it back into master so that
1.22.0 is part of the history of master. That means,
  $ git log --first-parent master
will also traverse 1.22.0 and 1.22-rc*.

Also bump the micro version to 1.23.1-dev to indicate that this is
after 1.22.0 is out.
2019-12-17 09:58:12 +01:00
Thomas Haller
7fe734f8bc release: bump version to 1.22.0 2019-12-17 09:24:16 +01:00
Thomas Haller
eae4ef7982 NEWS: update 2019-12-17 09:24:09 +01:00
Thomas Haller
f343f6f31d libnm/secret-agent: remove unused variable
Fixes: f0d3243f2b ('libnm/secret-agent: fix race registering secret agent')
(cherry picked from commit 5f9bcc91c7)
2019-12-17 09:00:00 +01:00
Thomas Haller
5f9bcc91c7 libnm/secret-agent: remove unused variable
Fixes: f0d3243f2b ('libnm/secret-agent: fix race registering secret agent')
2019-12-17 08:57:04 +01:00
Thomas Haller
bfd41a34c0 libnm: merge branch 'th/secret-agent-register-race'
https://bugzilla.redhat.com/show_bug.cgi?id=1781084

https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/364

(cherry picked from commit 85042e7244)
2019-12-16 18:55:26 +01:00
Thomas Haller
793e390c22 libnm/secret-agent: fix reseting timeout when registration completes
Fixes: f0d3243f2b ('libnm/secret-agent: fix race registering secret agent')
(cherry picked from commit 1f9dabcb73)
2019-12-16 18:54:58 +01:00
Thomas Haller
494cbe144d libnm/secret-agent: fix race registering secret agent
When NetworkManager starts, NMSecretAgentOld gets a name-owner changed
signal and registers right away.

Especially since commit ce0e898fb4 ('libnm: refactor caching of D-Bus
objects in NMClient') this hits a race where NetworkManager does not yet
export the org.freedesktop.NetworkManager.AgentManager interface and
the registration fails:

   GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager

Previously, when NMClient recevied a name-owner changed, that would
block the main loop long enough to avoid the race. Note that NMClient
has nothing to do with NMSecretAgentOld, however in practice all
applications that use NMSecretAgentOld also use NMClient.

While we should fix the race server-side, we also need to work around it
in the client. Retry.

Also, make the async request actually cancellable and actually honor the passed
GCancellable.

Check output:

    $ LIBNM_CLIENT_DEBUG=trace ./clients/cli/nmcli agent secret |& grep secret-agent
    libnm-dbus: <trace> [21399.04862] secret-agent[2f2af4ee102d7570]: create new instance
    libnm-dbus: <trace> [21399.04863] secret-agent[2f2af4ee102d7570]: init-sync
    libnm-dbus: <trace> [21404.08147] secret-agent[2f2af4ee102d7570]: name owner changed: (null)
    libnm-dbus: <trace> [21404.09085] secret-agent[2f2af4ee102d7570]: name owner changed: ":1.2504"
    libnm-dbus: <trace> [21404.09085] secret-agent[2f2af4ee102d7570]: register: starting asynchronous registration...
    libnm-dbus: <trace> [21404.09178] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 0 msec...
    libnm-dbus: <trace> [21404.09178] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21404.09195] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 4 msec...
    libnm-dbus: <trace> [21404.09236] secret-agent[2f2af4ee102d7570]: register: retry registration...
    [...]
    libnm-dbus: <trace> [21405.01782] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 128 msec...
    libnm-dbus: <trace> [21405.03063] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21405.03068] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 128 msec...
    libnm-dbus: <trace> [21405.04354] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21406.01097] secret-agent[2f2af4ee102d7570]: register: registration succeeded

(cherry picked from commit f0d3243f2b)
2019-12-16 18:54:33 +01:00
Thomas Haller
37f1c52fa9 libnm/secret-agent: use GTask for nm_secret_agent_old_register*()
This change is of course right and read nicer. Also, the GTask captures the
current g_main_context_get_thread_default(). We will need that next.

(cherry picked from commit cff4e937ac)
2019-12-16 18:54:12 +01:00
Thomas Haller
eeb3a25ca2 libnm/secret-agent: support debug logging from secret-agent
(cherry picked from commit 392befb5fd)
2019-12-16 18:54:12 +01:00
Thomas Haller
c414173fa9 libnm/secret-agent: drop fallback to Register() method for secret-agent
RegisterWithCapabilities() is supported since NetworkManager 0.9.9.1. Of course,
we don't support such old server anymore (also, because we require the standard
D-Bus interfaces like ObjectManager).

(cherry picked from commit 263aa63caa)
2019-12-16 18:54:12 +01:00
Thomas Haller
2b0533012a libnm: allow using _LOGx() macros in libnm
(cherry picked from commit 1b0f0f8c47)
2019-12-16 18:54:12 +01:00
Thomas Haller
e20c8d8ad5 shared: implement _LOGx() macros using log levels that are themself defines
"nm-glib-aux/nm-logging-fwd.h" provides macros like _LOGD() to be reused
by various parts which implement logging (by defining _NMLOG() accordingly).

libnm also has logging, however it uses different logging levels
aside LOGD_DEBUG.

Instead, implement _LOGD() using a define _LOGL_DEBUG, so that libnm can
redefine thos _LOGL_DEBUG defines and use the _LOGD() macro.

(cherry picked from commit 1b00fd2fd2)
2019-12-16 18:54:12 +01:00
Thomas Haller
452f14216a shared: add nm_dbus_error_is() helper
(cherry picked from commit ce36494c0a)
2019-12-16 18:54:12 +01:00
Thomas Haller
1f9dabcb73 libnm/secret-agent: fix reseting timeout when registration completes
Fixes: f0d3243f2b ('libnm/secret-agent: fix race registering secret agent')
2019-12-16 18:53:23 +01:00
Thomas Haller
85042e7244 libnm: merge branch 'th/secret-agent-register-race'
https://bugzilla.redhat.com/show_bug.cgi?id=1781084

https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/364
2019-12-16 18:44:41 +01:00
Thomas Haller
f0d3243f2b libnm/secret-agent: fix race registering secret agent
When NetworkManager starts, NMSecretAgentOld gets a name-owner changed
signal and registers right away.

Especially since commit ce0e898fb4 ('libnm: refactor caching of D-Bus
objects in NMClient') this hits a race where NetworkManager does not yet
export the org.freedesktop.NetworkManager.AgentManager interface and
the registration fails:

   GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager

Previously, when NMClient recevied a name-owner changed, that would
block the main loop long enough to avoid the race. Note that NMClient
has nothing to do with NMSecretAgentOld, however in practice all
applications that use NMSecretAgentOld also use NMClient.

While we should fix the race server-side, we also need to work around it
in the client. Retry.

Also, make the async request actually cancellable and actually honor the passed
GCancellable.

Check output:

    $ LIBNM_CLIENT_DEBUG=trace ./clients/cli/nmcli agent secret |& grep secret-agent
    libnm-dbus: <trace> [21399.04862] secret-agent[2f2af4ee102d7570]: create new instance
    libnm-dbus: <trace> [21399.04863] secret-agent[2f2af4ee102d7570]: init-sync
    libnm-dbus: <trace> [21404.08147] secret-agent[2f2af4ee102d7570]: name owner changed: (null)
    libnm-dbus: <trace> [21404.09085] secret-agent[2f2af4ee102d7570]: name owner changed: ":1.2504"
    libnm-dbus: <trace> [21404.09085] secret-agent[2f2af4ee102d7570]: register: starting asynchronous registration...
    libnm-dbus: <trace> [21404.09178] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 0 msec...
    libnm-dbus: <trace> [21404.09178] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21404.09195] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 4 msec...
    libnm-dbus: <trace> [21404.09236] secret-agent[2f2af4ee102d7570]: register: retry registration...
    [...]
    libnm-dbus: <trace> [21405.01782] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 128 msec...
    libnm-dbus: <trace> [21405.03063] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21405.03068] secret-agent[2f2af4ee102d7570]: register: registration failed with error "GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.NetworkManager.AgentManager” on object at path /org/freedesktop/NetworkManager/AgentManager". Retry in 128 msec...
    libnm-dbus: <trace> [21405.04354] secret-agent[2f2af4ee102d7570]: register: retry registration...
    libnm-dbus: <trace> [21406.01097] secret-agent[2f2af4ee102d7570]: register: registration succeeded
2019-12-16 18:38:49 +01:00
Thomas Haller
cff4e937ac libnm/secret-agent: use GTask for nm_secret_agent_old_register*()
This change is of course right and read nicer. Also, the GTask captures the
current g_main_context_get_thread_default(). We will need that next.
2019-12-16 18:38:49 +01:00
Thomas Haller
392befb5fd libnm/secret-agent: support debug logging from secret-agent 2019-12-16 18:38:49 +01:00
Thomas Haller
263aa63caa libnm/secret-agent: drop fallback to Register() method for secret-agent
RegisterWithCapabilities() is supported since NetworkManager 0.9.9.1. Of course,
we don't support such old server anymore (also, because we require the standard
D-Bus interfaces like ObjectManager).
2019-12-16 18:38:49 +01:00
Thomas Haller
1b0f0f8c47 libnm: allow using _LOGx() macros in libnm 2019-12-16 18:38:49 +01:00
Thomas Haller
1b00fd2fd2 shared: implement _LOGx() macros using log levels that are themself defines
"nm-glib-aux/nm-logging-fwd.h" provides macros like _LOGD() to be reused
by various parts which implement logging (by defining _NMLOG() accordingly).

libnm also has logging, however it uses different logging levels
aside LOGD_DEBUG.

Instead, implement _LOGD() using a define _LOGL_DEBUG, so that libnm can
redefine thos _LOGL_DEBUG defines and use the _LOGD() macro.
2019-12-16 18:37:47 +01:00
Thomas Haller
ce36494c0a shared: add nm_dbus_error_is() helper 2019-12-16 18:16:12 +01:00
Thomas Haller
809d70ee64 all: fix wrong "gs_free GError *" declarations
This is a bug and leads either to a leak or a crash.

(cherry picked from commit 4a3ca7115a)
2019-12-16 17:45:18 +01:00
Thomas Haller
2fa4827eb9 checkpatch: catch "gs_free GError *" declations
(cherry picked from commit ec0adbfaf0)
2019-12-16 17:45:16 +01:00
Thomas Haller
4a3ca7115a all: fix wrong "gs_free GError *" declarations
This is a bug and leads either to a leak or a crash.
2019-12-16 17:42:23 +01:00
Thomas Haller
ec0adbfaf0 checkpatch: catch "gs_free GError *" declations 2019-12-16 17:42:23 +01:00
Thomas Haller
9e02a67619 systemd: merge branch systemd into master 2019-12-16 10:22:09 +01:00
Thomas Haller
0d155d1821 systemd: update code from upstream (2019-12-13)
This is a direct dump from systemd git.

======

SYSTEMD_DIR=../systemd
COMMIT=c8bf87b3399a3dd0b17fd0003b9797635b161ee0

(
  cd "$SYSTEMD_DIR"
  git checkout "$COMMIT"
  git reset --hard
  git clean -fdx
)

git ls-files -z :/src/systemd/src/ \
                :/shared/systemd/src/ \
                :/shared/nm-std-aux/unaligned.h | \
  xargs -0 rm -f

nm_copy_sd_shared() {
    mkdir -p "./shared/systemd/$(dirname "$1")"
    cp "$SYSTEMD_DIR/$1" "./shared/systemd/$1"
}

nm_copy_sd_core() {
    mkdir -p "./src/systemd/$(dirname "$1")"
    cp "$SYSTEMD_DIR/$1" "./src/systemd/$1"
}

nm_copy_sd_stdaux() {
    mkdir -p "./shared/nm-std-aux/"
    cp "$SYSTEMD_DIR/$1" "./shared/nm-std-aux/${1##*/}"
}

nm_copy_sd_core "src/libsystemd-network/arp-util.c"
nm_copy_sd_core "src/libsystemd-network/arp-util.h"
nm_copy_sd_core "src/libsystemd-network/dhcp-identifier.c"
nm_copy_sd_core "src/libsystemd-network/dhcp-identifier.h"
nm_copy_sd_core "src/libsystemd-network/dhcp-internal.h"
nm_copy_sd_core "src/libsystemd-network/dhcp-lease-internal.h"
nm_copy_sd_core "src/libsystemd-network/dhcp-network.c"
nm_copy_sd_core "src/libsystemd-network/dhcp-option.c"
nm_copy_sd_core "src/libsystemd-network/dhcp-packet.c"
nm_copy_sd_core "src/libsystemd-network/dhcp-protocol.h"
nm_copy_sd_core "src/libsystemd-network/dhcp6-internal.h"
nm_copy_sd_core "src/libsystemd-network/dhcp6-lease-internal.h"
nm_copy_sd_core "src/libsystemd-network/dhcp6-network.c"
nm_copy_sd_core "src/libsystemd-network/dhcp6-option.c"
nm_copy_sd_core "src/libsystemd-network/dhcp6-protocol.h"
nm_copy_sd_core "src/libsystemd-network/lldp-internal.h"
nm_copy_sd_core "src/libsystemd-network/lldp-neighbor.c"
nm_copy_sd_core "src/libsystemd-network/lldp-neighbor.h"
nm_copy_sd_core "src/libsystemd-network/lldp-network.c"
nm_copy_sd_core "src/libsystemd-network/lldp-network.h"
nm_copy_sd_core "src/libsystemd-network/network-internal.c"
nm_copy_sd_core "src/libsystemd-network/network-internal.h"
nm_copy_sd_core "src/libsystemd-network/sd-dhcp-client.c"
nm_copy_sd_core "src/libsystemd-network/sd-dhcp-lease.c"
nm_copy_sd_core "src/libsystemd-network/sd-dhcp6-client.c"
nm_copy_sd_core "src/libsystemd-network/sd-dhcp6-lease.c"
nm_copy_sd_core "src/libsystemd-network/sd-ipv4acd.c"
nm_copy_sd_core "src/libsystemd-network/sd-ipv4ll.c"
nm_copy_sd_core "src/libsystemd-network/sd-lldp.c"
nm_copy_sd_core "src/libsystemd/sd-event/event-source.h"
nm_copy_sd_core "src/libsystemd/sd-event/event-util.c"
nm_copy_sd_core "src/libsystemd/sd-event/event-util.h"
nm_copy_sd_core "src/libsystemd/sd-event/sd-event.c"
nm_copy_sd_core "src/libsystemd/sd-id128/id128-util.c"
nm_copy_sd_core "src/libsystemd/sd-id128/id128-util.h"
nm_copy_sd_core "src/libsystemd/sd-id128/sd-id128.c"
nm_copy_sd_core "src/systemd/_sd-common.h"
nm_copy_sd_core "src/systemd/sd-dhcp-client.h"
nm_copy_sd_core "src/systemd/sd-dhcp-lease.h"
nm_copy_sd_core "src/systemd/sd-dhcp-option.h"
nm_copy_sd_core "src/systemd/sd-dhcp6-client.h"
nm_copy_sd_core "src/systemd/sd-dhcp6-lease.h"
nm_copy_sd_core "src/systemd/sd-event.h"
nm_copy_sd_core "src/systemd/sd-id128.h"
nm_copy_sd_core "src/systemd/sd-ipv4acd.h"
nm_copy_sd_core "src/systemd/sd-ipv4ll.h"
nm_copy_sd_core "src/systemd/sd-lldp.h"
nm_copy_sd_core "src/systemd/sd-ndisc.h"
nm_copy_sd_shared "src/basic/alloc-util.c"
nm_copy_sd_shared "src/basic/alloc-util.h"
nm_copy_sd_shared "src/basic/async.h"
nm_copy_sd_shared "src/basic/env-file.c"
nm_copy_sd_shared "src/basic/env-file.h"
nm_copy_sd_shared "src/basic/env-util.c"
nm_copy_sd_shared "src/basic/env-util.h"
nm_copy_sd_shared "src/basic/errno-util.h"
nm_copy_sd_shared "src/basic/escape.c"
nm_copy_sd_shared "src/basic/escape.h"
nm_copy_sd_shared "src/basic/ether-addr-util.c"
nm_copy_sd_shared "src/basic/ether-addr-util.h"
nm_copy_sd_shared "src/basic/extract-word.c"
nm_copy_sd_shared "src/basic/extract-word.h"
nm_copy_sd_shared "src/basic/fd-util.c"
nm_copy_sd_shared "src/basic/fd-util.h"
nm_copy_sd_shared "src/basic/fileio.c"
nm_copy_sd_shared "src/basic/fileio.h"
nm_copy_sd_shared "src/basic/format-util.c"
nm_copy_sd_shared "src/basic/format-util.h"
nm_copy_sd_shared "src/basic/fs-util.c"
nm_copy_sd_shared "src/basic/fs-util.h"
nm_copy_sd_shared "src/basic/hash-funcs.c"
nm_copy_sd_shared "src/basic/hash-funcs.h"
nm_copy_sd_shared "src/basic/hashmap.c"
nm_copy_sd_shared "src/basic/hashmap.h"
nm_copy_sd_shared "src/basic/hexdecoct.c"
nm_copy_sd_shared "src/basic/hexdecoct.h"
nm_copy_sd_shared "src/basic/hostname-util.c"
nm_copy_sd_shared "src/basic/hostname-util.h"
nm_copy_sd_shared "src/basic/in-addr-util.c"
nm_copy_sd_shared "src/basic/in-addr-util.h"
nm_copy_sd_shared "src/basic/io-util.c"
nm_copy_sd_shared "src/basic/io-util.h"
nm_copy_sd_shared "src/basic/list.h"
nm_copy_sd_shared "src/basic/log.h"
nm_copy_sd_shared "src/basic/macro.h"
nm_copy_sd_shared "src/basic/memory-util.c"
nm_copy_sd_shared "src/basic/memory-util.h"
nm_copy_sd_shared "src/basic/mempool.c"
nm_copy_sd_shared "src/basic/mempool.h"
nm_copy_sd_shared "src/basic/missing_fcntl.h"
nm_copy_sd_shared "src/basic/missing_random.h"
nm_copy_sd_shared "src/basic/missing_socket.h"
nm_copy_sd_shared "src/basic/missing_stat.h"
nm_copy_sd_shared "src/basic/missing_syscall.h"
nm_copy_sd_shared "src/basic/missing_type.h"
nm_copy_sd_shared "src/basic/parse-util.c"
nm_copy_sd_shared "src/basic/parse-util.h"
nm_copy_sd_shared "src/basic/path-util.c"
nm_copy_sd_shared "src/basic/path-util.h"
nm_copy_sd_shared "src/basic/prioq.c"
nm_copy_sd_shared "src/basic/prioq.h"
nm_copy_sd_shared "src/basic/process-util.c"
nm_copy_sd_shared "src/basic/process-util.h"
nm_copy_sd_shared "src/basic/random-util.c"
nm_copy_sd_shared "src/basic/random-util.h"
nm_copy_sd_shared "src/basic/set.h"
nm_copy_sd_shared "src/basic/signal-util.c"
nm_copy_sd_shared "src/basic/signal-util.h"
nm_copy_sd_shared "src/basic/siphash24.h"
nm_copy_sd_shared "src/basic/socket-util.c"
nm_copy_sd_shared "src/basic/socket-util.h"
nm_copy_sd_shared "src/basic/sort-util.h"
nm_copy_sd_shared "src/basic/sparse-endian.h"
nm_copy_sd_shared "src/basic/stat-util.c"
nm_copy_sd_shared "src/basic/stat-util.h"
nm_copy_sd_shared "src/basic/stdio-util.h"
nm_copy_sd_shared "src/basic/string-table.c"
nm_copy_sd_shared "src/basic/string-table.h"
nm_copy_sd_shared "src/basic/string-util.c"
nm_copy_sd_shared "src/basic/string-util.h"
nm_copy_sd_shared "src/basic/strv.c"
nm_copy_sd_shared "src/basic/strv.h"
nm_copy_sd_shared "src/basic/strxcpyx.c"
nm_copy_sd_shared "src/basic/strxcpyx.h"
nm_copy_sd_shared "src/basic/time-util.c"
nm_copy_sd_shared "src/basic/time-util.h"
nm_copy_sd_shared "src/basic/tmpfile-util.c"
nm_copy_sd_shared "src/basic/tmpfile-util.h"
nm_copy_sd_shared "src/basic/umask-util.h"
nm_copy_sd_shared "src/basic/utf8.c"
nm_copy_sd_shared "src/basic/utf8.h"
nm_copy_sd_shared "src/basic/util.c"
nm_copy_sd_shared "src/basic/util.h"
nm_copy_sd_shared "src/shared/dns-domain.c"
nm_copy_sd_shared "src/shared/dns-domain.h"
nm_copy_sd_stdaux "src/basic/unaligned.h"
2019-12-15 15:14:48 +01:00
Thomas Haller
f49ce41214 client: fallback to CLOCK_MONOTONIC for timerfd
RHEL7 supports clock_gettime(CLOCK_BOOTIME), but it does not support
timerfd_create(CLOCK_BOOTIME). Creating a timerfd will fail with EINVAL.
Fallback to CLOCK_MONOTONIC.

Compare this to n-acd which also has compatibility code to fallback to
CLOCK_MONOTONIC. However when n-acd falls back to CLOCK_MONOTONIC, it uses
monotonic clock also for clock_gettime().

For n-dhcp4, the timestamps are also exposed in the public API
(n_dhcp4_client_lease_get_lifetime()). Hence, for timestamps n-dhcp4
still uses and requires clock_gettime(CLOCK_BOOTIME). Only the internal
timeout handling with the timerfd falls back to CLOCK_MONOTONIC.

https://github.com/nettools/n-dhcp4/pull/13
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/362
(cherry picked from commit a1771c738d)
2019-12-15 13:51:44 +01:00
Thomas Haller
a1771c738d client: fallback to CLOCK_MONOTONIC for timerfd
RHEL7 supports clock_gettime(CLOCK_BOOTIME), but it does not support
timerfd_create(CLOCK_BOOTIME). Creating a timerfd will fail with EINVAL.
Fallback to CLOCK_MONOTONIC.

Compare this to n-acd which also has compatibility code to fallback to
CLOCK_MONOTONIC. However when n-acd falls back to CLOCK_MONOTONIC, it uses
monotonic clock also for clock_gettime().

For n-dhcp4, the timestamps are also exposed in the public API
(n_dhcp4_client_lease_get_lifetime()). Hence, for timestamps n-dhcp4
still uses and requires clock_gettime(CLOCK_BOOTIME). Only the internal
timeout handling with the timerfd falls back to CLOCK_MONOTONIC.

https://github.com/nettools/n-dhcp4/pull/13
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/362
2019-12-15 13:50:45 +01:00
Beniamino Galvani
944ff9f9dc manager: create a virtual device only if the connection can autoconnect
The autoconnection for virtual devices currently works in two
phases. First we detect that there is suitable profile that can
autoconnect and we realize the device. Then, when the device becomes
'disconnected', autoconnect kicks in and starts the activation.

However, if autoconnect is blocked for a device, currently we do step
1 without step 2, leaving a stale interface around. Fix this by also
checking that autoconnect is not blocked during step 1.

https://bugzilla.redhat.com/show_bug.cgi?id=1765047

https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/360
(cherry picked from commit 6c716912eb)
2019-12-15 09:37:25 +01:00
Beniamino Galvani
6c716912eb manager: create a virtual device only if the connection can autoconnect
The autoconnection for virtual devices currently works in two
phases. First we detect that there is suitable profile that can
autoconnect and we realize the device. Then, when the device becomes
'disconnected', autoconnect kicks in and starts the activation.

However, if autoconnect is blocked for a device, currently we do step
1 without step 2, leaving a stale interface around. Fix this by also
checking that autoconnect is not blocked during step 1.

https://bugzilla.redhat.com/show_bug.cgi?id=1765047

https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/360
2019-12-15 09:36:27 +01:00
Beniamino Galvani
f1d4853927 device: don't reapply IP configuration if the ifindex is missing
Assertions will fail in ip_config_merge_and_apply() if the device
doesn't have an ifindex. Reproducible with:

 $ nmcli connection add type ovs-bridge ifname ovs0 ipv4.method disabled ipv6.method disabled
 Connection 'ovs-bridge-ovs0' (1d5e794b-10ad-4b2b-aa7c-5ca7e34b0a55) successfully added

 $ nmcli device reapply ovs0
 Error: Reapplying connection to device '(null)' (/org/freedesktop/NetworkManager/Devices/16) failed: Remote peer disconnected

 $ journalctl -u NetworkManager -e
 ...
 NetworkManager[73824]: nm_ip4_config_add_dependent_routes: assertion 'ifindex > 0' failed
 systemd[1]: NetworkManager.service: Main process exited, code=dumped, status=5/TRAP
 ...

(cherry picked from commit 6d6e1402dc)
2019-12-14 21:05:54 +01:00
Beniamino Galvani
3381299562 ovs: check state before starting ip configuration after link change
When the link becomes available, check that the device is in the
ip-config state before starting ip configuration. Also, reset the
'waiting_for_interface' flag when the device deactivates.

https://bugzilla.redhat.com/show_bug.cgi?id=1781165
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/358
(cherry picked from commit 0738c10445)
2019-12-14 21:05:53 +01:00
Beniamino Galvani
6d6e1402dc device: don't reapply IP configuration if the ifindex is missing
Assertions will fail in ip_config_merge_and_apply() if the device
doesn't have an ifindex. Reproducible with:

 $ nmcli connection add type ovs-bridge ifname ovs0 ipv4.method disabled ipv6.method disabled
 Connection 'ovs-bridge-ovs0' (1d5e794b-10ad-4b2b-aa7c-5ca7e34b0a55) successfully added

 $ nmcli device reapply ovs0
 Error: Reapplying connection to device '(null)' (/org/freedesktop/NetworkManager/Devices/16) failed: Remote peer disconnected

 $ journalctl -u NetworkManager -e
 ...
 NetworkManager[73824]: nm_ip4_config_add_dependent_routes: assertion 'ifindex > 0' failed
 systemd[1]: NetworkManager.service: Main process exited, code=dumped, status=5/TRAP
 ...
2019-12-14 21:03:34 +01:00
Beniamino Galvani
0738c10445 ovs: check state before starting ip configuration after link change
When the link becomes available, check that the device is in the
ip-config state before starting ip configuration. Also, reset the
'waiting_for_interface' flag when the device deactivates.

https://bugzilla.redhat.com/show_bug.cgi?id=1781165
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/merge_requests/358
2019-12-14 21:02:31 +01:00
Thomas Haller
3e77f0a5a8 cloud-setup/trivial: rename timeout_ms to timeout_msec
I think "ms" as abbreviation is too short and unclear. Also, it's
not used consistently. Rename variable.

(cherry picked from commit 13741aa952)
2019-12-13 17:47:05 +01:00
Thomas Haller
bd9b253540 all: rename time related function to spell out nsec/usec/msec/sec
The abbreviations "ns" and "ms" seem not very clear to me. Spell them
out to nsec/msec. Also, in parts we already used the longer abbreviations,
so it wasn't consistent.
2019-12-13 16:54:40 +01:00
Thomas Haller
13741aa952 cloud-setup/trivial: rename timeout_ms to timeout_msec
I think "ms" as abbreviation is too short and unclear. Also, it's
not used consistently. Rename variable.
2019-12-13 16:54:40 +01:00
Thomas Haller
490c46e527 meson/build: avoid using ternary as it breaks older meson
See mesonbuild issue 5003.

Oddly, this isn't inside a method call, while we use ternaries
inside a method calls otherwise. Anyway, workaround this and
avoid the crash.

Fixes: c21c6bc0be ('build/meson: allow configuring default for main.auth-polkit setting')
(cherry picked from commit 0e3400bef7)
2019-12-13 14:58:49 +01:00
Thomas Haller
0e3400bef7 meson/build: avoid using ternary as it breaks older meson
See mesonbuild issue 5003.

Oddly, this isn't inside a method call, while we use ternaries
inside a method calls otherwise. Anyway, workaround this and
avoid the crash.

Fixes: c21c6bc0be ('build/meson: allow configuring default for main.auth-polkit setting')
2019-12-13 13:48:25 +01:00
Thomas Haller
d8cc6af058 shared/glib: reimplement g_atomic_pointer_compare_and_exchange() macro
With glib 2.63.2 and clang 9.0.0 (Fedora 32) we get compile errors:

  ../clients/cloud-setup/nmcs-provider-ec2.c:51:8: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'const char **' [-Werror,-Wincompatible-pointer-types]
                  if (!g_atomic_pointer_compare_and_exchange (&base_cached, NULL, base))
                       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/bluetooth/nm-bluez-manager.c:2836:2: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'const NMBtVTableNetworkServer **' (aka 'const struct _NMBtVTableNetworkServer **') [-Werror,-Wincompatible-pointer-types]
          g_atomic_pointer_compare_and_exchange (&nm_bt_vtable_network_server, NULL, &priv->vtable_network_server);
          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/bluetooth/nm-bluez-manager.c:2853:2: error: passing 'typeof ((&priv->vtable_network_server)) *' (aka 'struct _NMBtVTableNetworkServer **') to parameter of type 'const NMBtVTableNetworkServer **' (aka 'const struct _NMBtVTableNetworkServer **') discards qualifiers in nested pointer types [-Werror,-Wincompatible-pointer-types-discards-qualifiers]
          g_atomic_pointer_compare_and_exchange (&nm_bt_vtable_network_server, &priv->vtable_network_server, NULL);
          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/nm-device.c:8857:8: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'GBytes **' (aka 'struct _GBytes **') [-Werror,-Wincompatible-pointer-types]
                  if (!g_atomic_pointer_compare_and_exchange (&global_duid, NULL, p)) {
                       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~

The issue happens because glib passes the "atomic" argument to

  __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)

without cast, and clang 9 seems to be picky about const pointers.
Add our own version of the macro that does better casts while also having
better compile time checks for valid arguments.

(cherry picked from commit f5b0713651)
2019-12-13 13:48:00 +01:00
Thomas Haller
f5b0713651 shared/glib: reimplement g_atomic_pointer_compare_and_exchange() macro
With glib 2.63.2 and clang 9.0.0 (Fedora 32) we get compile errors:

  ../clients/cloud-setup/nmcs-provider-ec2.c:51:8: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'const char **' [-Werror,-Wincompatible-pointer-types]
                  if (!g_atomic_pointer_compare_and_exchange (&base_cached, NULL, base))
                       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/bluetooth/nm-bluez-manager.c:2836:2: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'const NMBtVTableNetworkServer **' (aka 'const struct _NMBtVTableNetworkServer **') [-Werror,-Wincompatible-pointer-types]
          g_atomic_pointer_compare_and_exchange (&nm_bt_vtable_network_server, NULL, &priv->vtable_network_server);
          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/bluetooth/nm-bluez-manager.c:2853:2: error: passing 'typeof ((&priv->vtable_network_server)) *' (aka 'struct _NMBtVTableNetworkServer **') to parameter of type 'const NMBtVTableNetworkServer **' (aka 'const struct _NMBtVTableNetworkServer **') discards qualifiers in nested pointer types [-Werror,-Wincompatible-pointer-types-discards-qualifiers]
          g_atomic_pointer_compare_and_exchange (&nm_bt_vtable_network_server, &priv->vtable_network_server, NULL);
          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~
  ../src/devices/nm-device.c:8857:8: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'GBytes **' (aka 'struct _GBytes **') [-Werror,-Wincompatible-pointer-types]
                  if (!g_atomic_pointer_compare_and_exchange (&global_duid, NULL, p)) {
                       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
      __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                             ^~~~~~~~~~~~~~

The issue happens because glib passes the "atomic" argument to

  __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)

without cast, and clang 9 seems to be picky about const pointers.
Add our own version of the macro that does better casts while also having
better compile time checks for valid arguments.
2019-12-13 13:46:37 +01:00
Thomas Haller
dc981b12a9 clients/tests: merge branch 'th/clients-test-select-fields'
(cherry picked from commit 4711019d34)
2019-12-13 13:20:58 +01:00
Thomas Haller
30b7ab6cb9 clients/tests: add test for selecting fields from multiple connections with nmcli con show
(cherry picked from commit f230342478)
2019-12-13 13:20:57 +01:00
Thomas Haller
64f103416f clients/tests: support Util.replace_text_sort_list() helper
This will be necessary to sort a list of text that contains text to
replace. The use case is to get all UUIDs from the server, but some
of these UUIDs are dynamic. So, in order to have a fixed sort order,
we need to handle sorting the list, based on the replacement.

(cherry picked from commit 347253be2f)
2019-12-13 13:20:57 +01:00
Thomas Haller
1a20c9b2b5 clients/tests: add NMStubServer.findConnections() helper function
Will be used to get all connection from the test stub server.

(cherry picked from commit 0d24ac4996)
2019-12-13 13:20:57 +01:00
Thomas Haller
50d7b1cd7c clients/tests: support replacing UUIDs in command text
We want to issue a nmcli command with the connection's UUID in the command
line. As the command line is stored to disk and compared, we cannot use UUIDs
that are randomly generated. Instead, we need to support replacing the text
in the command line.

(cherry picked from commit 7d048f1a5b)
2019-12-13 13:20:57 +01:00