In kernel, the onlink flag (RTNH_F_ONLINK) is associated with each
nexthop (rtnh_flags) rather than the route as a whole. NM previously
stored it only per-route in NMPlatformIPRoute.r_rtm_flags, which meant
that two nexthops only differing with the onlink flag were combined
as one entry in the platform cache.
Fix this by tracking the onlink flag per-nexthop.
Resolves: https://issues.redhat.com/browse/NMT-1486
The current implementation returns IP addresses obtained by adding a
counter to a base address. For CLAT we want to return all the 8
addresses in the 192.0.0.0/29 range, but not starting from 192.0.0.0
because that looks more like a network address. Slightly tweak the
algorithm so that addresses can wrap around.
The name suggests that the function always removes all the watchers
with the given tag; instead it removes only "dirty" ones when the
"all" parameter is FALSE. Split the function in two variants.
(cherry picked from commit b6e67c6abc)
When a single-hop route is merged with another single-hop route from a
different interface creating a new ECMP route, NetworkManager needs to
schedule a commit on the l3cfg that is managing the first single-hop
route. Otherwise, the single-hop will continue to be there until a new
commit is scheduled.
Fixes: d9d33e2acc ('netns: fix configuring onlink routes for ECMP routes')
nm_netns_ip_route_ecmp_commit() returns the ECMP routes that didn't find
a merge-partner. It expects NML3Cfg to configure such routes in
platform.
Note that these routes have a positive "weight", which was used for
tracking the ECMP information in NML3Cfg and NMNetns.
But in kernel, single-hop routes don't have a weight. All routes in
NMPlatform will have a zero weight. While it somewhat works to call
nm_platform_ip_route_add() with a single-hop route of non-zero weight,
the result will be a different(!) route. This means for example that
nm_platform_ip_route_sync() will search the NMPlatform cache for whether
the routes exist, but in the cache single-hop routes with positive weight
never exist. Previously this happened and nm_platform_ip_route_sync()
would not delete those routes when it should.
It's also important because NML3Cfg tracks the object states of routes
that it wants to configure, vs routes in kernel (NMPlatform cache). The
route in kernel has a weight of zero. The route we want to configure
must also have a weight of zero.
The solution is that nm_netns_ip_route_ecmp_commit() does not tell
NML3Cfg to add a route, that cannot be added. Instead, it must first
normalize the weight to zero, to have a "regular" single-hop route.
Fixes: 5b5ce42682 ('nm-netns: track ECMP routes')
The implementation came with two flavors, where watcher could either
specify a tag or no tag. That resulted in different usage patterns and
behavior.
Handles with tag are indexed by a dictionary and de-duplicated. Also the intended
pattern is to delete them with nm_netns_watcher_remove_all(),
Currently, nm_netns_watcher_remove_handle() was not permissible to tag-full handles,
because of the de-duplication and because handles had no ref-counting
implemented (the latter would be fixable, so
nm_netns_watcher_remove_handle() would be made to work).
On the other hand, handles without tag are never de-duplicated. They are
also not indexed, so nm_netns_watcher_remove_all() doesn't work for
them. They could only be removed via nm_netns_watcher_remove_handle().
Currently, the only user of the API will use tag-full handles. Drop the
unused API. This is done as a separate commit, to potentially revert and
restore tag-less handles (after they were already implemented).
NML3Cfg will want to know when an address changes -- on any interface.
We want to support gazillion of interfaces, a naive approach is not
going to scale. Instead, NMNetns already subscribes to all platform
signals, it should dispatch events for address changes.
Add a mechanism how users (NML3Cfg) can register watches, and get called
back when the event happens.
The onlink flag is part of each next hop.
When NetworkManager configures ECMP routes, we won't support that. All
next hops of an ECMP route must share the same onlink flag. That is fine
and fixed by this commit.
What is not fine, is that we don't track the rtnh_flags flags in
NMPlatformIP4RtNextHop, and consequently our nmp_object_id_cmp() is
wrong.
Fixes: 5b5ce42682 ('nm-netns: track ECMP routes')
(cherry picked from commit 6ed966258c)
Kernel enforces that all nexthops must be reachable through a route.
L3Cfg is generating dependent onlink routes to solve this problem but
the IPv4 ECMP commit is happening before that.
To solve this we introduce two boolean fields "is_new" and "is_ready" to
know in which state is the L3Cfg affected. Initially, "is_new" is TRUE
and "is_ready" is FALSE. Here we schedule a commit on idle and we set
"is_new" to FALSE. When revisiting, we set "is_ready" to TRUE and then
we set the ECMP IPv4 routes.
When a reapply kicks in we reset the L3Cfg state by setting "is_new" to
TRUE.
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/1520
(cherry picked from commit 7a844ecba9)
ECMP IPv4 route nexthops requires an onlink route but we should trust
l3cfg when generating and managing such routes.
This reverts commit 737cb5d424.
(cherry picked from commit cbf70b4dca)
When we register/unregister a commit-type or when we add/remove a
config-data to NML3Cfg, that act only does the registration/addition.
Only on the next commit, are the changes actually done. The purpose
of that is to add/register multiple configurations and commit them later
when ready.
However, it would be wrong to not do the commit a short time after. The
configuration state is dirty with need to be committed, and that should
happen soon.
Worse, when a interface disappears, NMDevice will clear the ifindex and
the NML3Cfg instance, thereby unregistering all config data and commit
type. If we previously commited something, we need to do another follow-up
commit to cleanup that state.
That is for example important with ECMP routes, which are registered in
NMNetns. When NML3Cfg goes down, it always must unregister to properly
cleanup. Failure to do so, causes an assertion failure and crash. This
change fixes that.
Fix that by automatically schedule and idle commit on
register/unregister/add/remove of commit-type/config-data.
It should *always* be permissible to call a AUTO commit from
an idle handler, because various parties cannot use NML3Cfg
independently, and they cannot know when somebody else does a
commit.
Note that NML3Cfg remembers if it presiouvly did a commit
("commit_type_update_sticky"), so even if the last commit-type gets
unregistered, the next commit will still do a sticky update (one more
time).
The only remaining question is what happens during quitting. When
quitting, NetworkManager we may want to leave some interfaces up and
configured. If we were to properly cleanup the NML3Cfg we might need a
mechanism to handle that. However, currently we just leak everything
during quit, so that is not a concern now. It is something that needs
to be addressed in the future.
https://bugzilla.redhat.com/show_bug.cgi?id=2158394https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/1505
When committing ECMP routes we are cleaning up dirty routes and freeing
the EcmpTrackObj. We need to free EcmpTrackObj only when it is not
needed anymore so it is the last thing we do when cleaning up the
routes.
If an L3Cfg try to register an ECMP route that we are tracking we must
mark it as needs_update because it means it could be dropped from
kernel. When merging them if the merged route didn't change we should
commit the route anyway because we know it needed update.
When adding a static route, kernel enforce that the gateway is
reachable. To solve this, NetworkManager generates onlink routes for
each static route. As ECMP routes does not follow the same logic than
singlehop routes, we need to add the onlink route for each hop of ECMP
routes once merged.
NML3Cfg will take ownership of the onlink route added and will purge it
when it is not needed anymore.
NML3Cfg and NMNetns are already strongly related and cooperate.
An NML3Cfg instance is created via NMNetns, which is necessary
because NMNetns ensures that there is only one NML3Cfg instance per
ifindex and it won't ever make sense to have multiple NML3Cfg instances
per namespace.
Note that NMNetns tracks additional information for each NML3Cfg.
Previously, in a pointless attempt to separate code, it did so
by putting that information in another struct (L3CfgData).
But as the classes are strongly related, there really is no
reason why we cannot just attach this information to NML3Cfg
directly. Sure, we want that code has low coupling, high cohesion
but that doesn't mean we gain anything by putting data that is
strongly related to the NML3Cfg to another struct L3CfgData.
The advantage is we save some redundant data and an additional
L3CfgData. But the bigger reason is that with this change, it
will be possible to access the NMNetns specific data directly from
an NML3Cfg instance, without another dictionary lookup. Currently
such a lookup is never used, but it will be.
Basically, NML3Cfg and NMNetns shares some state. It is now in the
"internal_netns" field of the NML3Cfg instead of L3CfgData.
- name things related to `in_addr_t`, `struct in6_addr`, `NMIPAddr` as
`nm_ip4_addr_*()`, `nm_ip6_addr_*()`, `nm_ip_addr_*()`, respectively.
- we have a wrapper `nm_inet_ntop()` for `inet_ntop()`. This name
of our wrapper is chosen to be familiar with the libc underlying
function. With this, also name functions that are about string
representations of addresses `nm_inet_*()`, `nm_inet4_*()`,
`nm_inet6_*()`. For example, `nm_inet_parse_str()`,
`nm_inet_is_normalized()`.
<<<<
R() {
git grep -l "$1" | xargs sed -i "s/\<$1\>/$2/g"
}
R NM_CMP_DIRECT_IN4ADDR_SAME_PREFIX NM_CMP_DIRECT_IP4_ADDR_SAME_PREFIX
R NM_CMP_DIRECT_IN6ADDR_SAME_PREFIX NM_CMP_DIRECT_IP6_ADDR_SAME_PREFIX
R NM_UTILS_INET_ADDRSTRLEN NM_INET_ADDRSTRLEN
R _nm_utils_inet4_ntop nm_inet4_ntop
R _nm_utils_inet6_ntop nm_inet6_ntop
R _nm_utils_ip4_get_default_prefix nm_ip4_addr_get_default_prefix
R _nm_utils_ip4_get_default_prefix0 nm_ip4_addr_get_default_prefix0
R _nm_utils_ip4_netmask_to_prefix nm_ip4_addr_netmask_to_prefix
R _nm_utils_ip4_prefix_to_netmask nm_ip4_addr_netmask_from_prefix
R nm_utils_inet4_ntop_dup nm_inet4_ntop_dup
R nm_utils_inet6_ntop_dup nm_inet6_ntop_dup
R nm_utils_inet_ntop nm_inet_ntop
R nm_utils_inet_ntop_dup nm_inet_ntop_dup
R nm_utils_ip4_address_clear_host_address nm_ip4_addr_clear_host_address
R nm_utils_ip4_address_is_link_local nm_ip4_addr_is_link_local
R nm_utils_ip4_address_is_loopback nm_ip4_addr_is_loopback
R nm_utils_ip4_address_is_zeronet nm_ip4_addr_is_zeronet
R nm_utils_ip4_address_same_prefix nm_ip4_addr_same_prefix
R nm_utils_ip4_address_same_prefix_cmp nm_ip4_addr_same_prefix_cmp
R nm_utils_ip6_address_clear_host_address nm_ip6_addr_clear_host_address
R nm_utils_ip6_address_same_prefix nm_ip6_addr_same_prefix
R nm_utils_ip6_address_same_prefix_cmp nm_ip6_addr_same_prefix_cmp
R nm_utils_ip6_is_ula nm_ip6_addr_is_ula
R nm_utils_ip_address_same_prefix nm_ip_addr_same_prefix
R nm_utils_ip_address_same_prefix_cmp nm_ip_addr_same_prefix_cmp
R nm_utils_ip_is_site_local nm_ip_addr_is_site_local
R nm_utils_ipaddr_is_normalized nm_inet_is_normalized
R nm_utils_ipaddr_is_valid nm_inet_is_valid
R nm_utils_ipx_address_clear_host_address nm_ip_addr_clear_host_address
R nm_utils_parse_inaddr nm_inet_parse_str
R nm_utils_parse_inaddr_bin nm_inet_parse_bin
R nm_utils_parse_inaddr_bin_full nm_inet_parse_bin_full
R nm_utils_parse_inaddr_prefix nm_inet_parse_with_prefix_str
R nm_utils_parse_inaddr_prefix_bin nm_inet_parse_with_prefix_bin
R test_nm_utils_ip6_address_same_prefix test_nm_ip_addr_same_prefix
./contrib/scripts/nm-code-format.sh -F
NetworkManager primarily manages interfaces in an independent fashion.
That means, whenever possible, we want to have a interface specific
view. In many cases, the underlying kernel API also supports that view.
For example, when configuring IP addresses or unicast routes, we do so
per interfaces and don't need a holistic view.
However, that is not always sufficient. For routing rules and certain
route types (blackhole, unreachable, etc), we need a system wide view
of all the objects in the network namespace.
Originally, NMPRulesManager was added to track routing rules. Then, it
was extended to also track certain route types, and the API was renamed to
NMPRouteManager.
This will also be used to track MPTCP addresses.
So rename again, to give it a general name that is suitable for what it
does. Still, the name is not great (suggestion welcome), but it should
cover the purpose of the API well enough. And it's the best I came
up with.
Rename.
Routes of type blackhole, unreachable, prohibit don't have an
ifindex/device. They are thus in many ways similar to routing rules,
as they are global. We need a mediator to keep track which routes
to configure.
This will be very similar to what NMPRulesManager already does for
routing rules. Rename the API, so that it also can be used for routes.
Renaming the file will be done next, so that git's rename detection
doesn't get too confused.
(cherry picked from commit ea4f6d7994)
We use clang-format for automatic formatting of our source files.
Since clang-format is actively maintained software, the actual
formatting depends on the used version of clang-format. That is
unfortunate and painful, but really unavoidable unless clang-format
would be strictly bug-compatible.
So the version that we must use is from the current Fedora release, which
is also tested by our gitlab-ci. Previously, we were using Fedora 34 with
clang-tools-extra-12.0.1-1.fc34.x86_64.
As Fedora 35 comes along, we need to update our formatting as Fedora 35
comes with version "13.0.0~rc1-1.fc35".
An alternative would be to freeze on version 12, but that has different
problems (like, it's cumbersome to rebuild clang 12 on Fedora 35 and it
would be cumbersome for our developers which are on Fedora 35 to use a
clang that they cannot easily install).
The (differently painful) solution is to reformat from time to time, as we
switch to a new Fedora (and thus clang) version.
Usually we would expect that such a reformatting brings minor changes.
But this time, the changes are huge. That is mentioned in the release
notes [1] as
Makes PointerAligment: Right working with AlignConsecutiveDeclarations. (Fixes https://llvm.org/PR27353)
[1] https://releases.llvm.org/13.0.0/tools/clang/docs/ReleaseNotes.html#clang-format
Completely rework IP configuration in the daemon. Use NML3Cfg as layer 3
manager for the IP configuration of an interface. Use NML3ConfigData as
pieces of configuration that the various components collect and
configure. NMDevice is managing most of the IP configuration at a higher
level, that is, it starts DHCP and other IP methods. Rework the state
handling there.
This is a huge rework of how NetworkManager daemon handles IP
configuration. Some fallout is to be expected.
It appears the patch deletes many lines of code. That is not accurate, because
you also have to count the files `src/core/nm-l3*`, which were unused previously.
Co-authored-by: Beniamino Galvani <bgalvani@redhat.com>
"libnm-core/" is rather complicated. It provides a static library that
is linked into libnm.so and NetworkManager. It also contains public
headers (like "nm-setting.h") which are part of public libnm API.
Then we have helper libraries ("libnm-core/nm-libnm-core-*/") which
only rely on public API of libnm-core, but are themself static
libraries that can be used by anybody who uses libnm-core. And
"libnm-core/nm-libnm-core-intern" is used by libnm-core itself.
Move "libnm-core/" to "src/". But also split it in different
directories so that they have a clearer purpose.
The goal is to have a flat directory hierarchy. The "src/libnm-core*/"
directories correspond to the different modules (static libraries and set
of headers that we have). We have different kinds of such modules because
of how we combine various code together. The directory layout now reflects
this.
Currently "src/" mostly contains the source code of the daemon.
I say mostly, because that is not true, there are also the device,
settings, wwan, ppp plugins, the initrd generator, the pppd and dhcp
helper, and probably more.
Also we have source code under libnm-core/, libnm/, clients/, and
shared/ directories. That is all confusing.
We should have one "src" directory, that contains subdirectories. Those
subdirectories should contain individual parts (libraries or
applications), that possibly have dependencies on other subdirectories.
There should be a flat hierarchy of directories under src/, which
contains individual modules.
As the name "src/" is already taken, that prevents any sensible
restructuring of the code.
As a first step, move "src/" to "src/core/". This gives space to
reorganize the code better by moving individual components into "src/".
For inspiration, look at systemd's "src/" directory.
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/743
