Switch platform caching implementation. Instead of caching libnl
objects, cache our own types.
Don't remove yet the now obsolete functions.
Advantage:
* Performance
- as we now cache our native NMPlatformObject instances, we no longer
have to convert libnl objects every time we access the platform
cache.
- for most cases, access is now O(1) because we can lookup the object
in a hash table. Note that ip4_address_get_all() still has to
create a copy of the result (O(n)), but as the caller is about to
use those elements, he cannot do better then O(n) anyway.
* We cache our own native types and have full control over them. We
cannot extend the libnl objects, which has many short-commings:
- _rtnl_addr_hack_lifetimes_rel_to_abs() to convert the timestamps
to absolute values (and back).
- hack_empty_master_iff_lower_up() would modify the internal flag,
but it looses the original value. That means, we can only hack
the state before putting a link into the cache, but we cannot revert
that change, when a slave in the cache changes state.
That was previously solved by always refetching the master when
a slave changed. Now we can re-evaluate the connected state
(DELAYED_ACTION_TYPE_MASTER_CONNECTED).
- we implement functions like equality, to-string as most suitable
for us. Before we needed hacks like nm_nl_object_diff(),
nm_nl_cache_search(), route_search_cache().
- we can extend our objects with exactly those properties we care,
and possibly additional properties that are not representable in
the libnl objects.
- we no longer cache RTM_F_CLONED routes and they get rejected early
on as we receive them.
- In the future, maybe it'd be interesting the make platform objects
immutable (and ref-counted) and expose them directly.
* Previous implementation did not order the refresh of objects but
called check_cache_items(). Now, those actions are delayed and
combined in an attempt to reduce the overall number of reloads.
Realize how expensive a check_cache_items() for addresses and routes
was: it would iterate all addresses/routes and call refresh_object().
The latter obtains a full dump of *all* objects again, and ignores
all but the needle.
Note that we probably still schedule some delayed actions that
are not needed.
Later we can optimize that further (related bug bgo #747985).
While some of these points could also have been implemented with
caching of libnl objects, that would have become hard to maintain.
https://bugzilla.gnome.org/show_bug.cgi?id=747981
NMPObject is a simple "object" implemenation around NMPlatformObject.
They are ref-counted and have a class-pointer. Several basic functions
like equality, hash, to-string are implemented.
NMPCache is can be used to store the NMPObject. Objects are indexed
via their primary id, but there is also multi-lookup via NMCacheId
and NMMultiIndex.
Part of the implementation is inside "nm-linux-platform.c",
because it depends on utility functions from there.
Cache the scope as part of the NMPlatformIP4Route and
no longer read it from libnl object when needed. Later
there will be no more libnl objects around, and we need
to scope when deleting an IPv4 route.
warning: function declaration isn’t a prototype [-Wstrict-prototypes]
In C function() and function(void) are two different prototypes (as opposed to
C++).
function() accepts an arbitrary number of arguments
function(void) accepts zero arguments
We already populate the netlink cache in constructed(). No need
to wait with udev devices until nm_platform_query_devices(). Just
do it right away.
Add a hack to keep 'lo' default-unmanaged. Now that we load
udev devices earlier, we end up clearing the default-unmanged
flag on 'lo', which has bad consequences.
We don't want error logging for nm_platform_link_add() which
tries to load the bonding module. Later we will run tests as non-root,
where modprobe will fail. Logging an error would break the tests.
We'd be able to do so for already existing devices, but not for devices that
are added afterwards, since gudev is hardwired not to listen for events from
kernel.
Ethernet, WiFi, and VLAN used the same implementation for initial address.
Ethernet and WiFi used the same implementation (and duplicated code) for
permanent MAC address, plus they both used ethtool in what should be
generic code, which is better done in the platform.
Throwing away the udev_device instance is wrong. There are two cases:
- the udev-device appears, and the netlink object will never appear
(or is already gone). In this case, keeping the udev-device is ok
because we will eventually get a signal from UDev to cleanup the
device instance.
- the udev-device appears before the netlink object. In this case we
want to keep the udev instance to have it ready.
Fixes: 388b7830f3
Always intern string from udev_get_driver().
We use the result of udev_get_driver() for setting NMPlatformLink.driver.
In all other cases, we already set that value to an interned string,
which simplifies memory handling.
As it was, the lifetime of that string was tied to the lifetime of the
GUdevDevice.
This is not a stelar solution, but we assume that the overall numbers
of different drivers is limited so we don't leak large amounts of
memory.
link_extract_type() would return the NMLinkType and a
@type_name string. If the type was unknown, this string
was rtnl_link_get_type() (IFLA_INFO_KIND).
Split up this behavior and treat those values independently.
link_extract_type() now only detects the NMLinkType. Most users
don't care about unknown types and can just use nm_link_type_to_string()
to get a string represenation.
Only nm_platform_link_get_type_name() (and NMDeviceGeneric:type_description)
cared about a more descriptive type. For that, modify link_get_type_name()
to return nm_link_type_to_string() if NMLinkType could be detected.
As fallback, return rtnl_link_get_type().
Also, rename the field NMPlatformLink:link_type to "kind". For now this
field is mostly unused. It will be used later when refactoring platform
caching.
Given the name nm_link_type_to_string(), we would not expect
to find it in nm-linux-platform.c. It either should be named
nm_platform_link_type_to_string() and be put in a new
nm-platform-utils.c file, or it should be named
nm_utils_link_type_to_string() and be put in NetworkManagerUtils.h.
For now, just leave it here.
link_extract_type() would call tun_get_properties() to determine whether
the link if a TAP or TUN device. The previous implementation would
receive the ifindex, and resolve the ifname via lookup in the platform
cache.
This means, the call on link_extract_type() will only succeed to detect
the TUN/TAP properties, if the libnl object is already in the cache.
Currently that is always the case and there is no problem.
It is desireable, that we can resolve the link type of an object without
consulting the platform cache first.
See "Revert "wireless: Support of IFLA_INFO_KIND rtnl attribute""
http://www.spinics.net/lists/linux-wireless/msg132219.html
The reverted kernel patch caused rtnl_link_get_type() to return "wlan"
for WiFi devices. Since NM depends on this function returning
NULL for WiFi devices so that it goes on to check the sysfs DEVTYPE
attribute, the kernel patch caused WiFi devices to show up as Generic
ones instead. That's wrong, and NM should be able to more easily
handle changes in the kernel drivers from NULL to a more descriptive
rtnl_link_get_type() return, since that's the kernel trend.
What NM should be doing here is to fall back to other detection
schemes if the type is NULL or unrecognized. Make that happen and
clean things up to use a table instead of a giant if(strcmp()) block.
https://bugzilla.gnome.org/show_bug.cgi?id=743209
