gcc warns:
make[5]: Entering directory `./NetworkManager/src/platform/tests'
CC platform.o
platform.c: In function ‘do_ip6_route_add’:
platform.c:696:2: error: ‘plen’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
return nm_platform_ip6_route_add (ifindex, NM_PLATFORM_SOURCE_USER,
^
platform.c: In function ‘do_ip6_route_delete’:
platform.c:724:2: error: ‘plen’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
return nm_platform_ip6_route_delete (ifindex, network, plen, metric);
^
platform.c: In function ‘do_ip4_route_delete’:
[...]
Signed-off-by: Thomas Haller <thaller@redhat.com>
clang warns:
make[5]: Entering directory `./NetworkManager/src/platform/tests'
CC test_link_fake-test-link.o
test-link.c:133:1: error: control may reach end of non-void function [-Werror,-Wreturn-type]
}
^
test-link.c:191:10: error: declaration shadows a variable in the global scope [-Werror,-Wshadow]
char *stdout = NULL;
^
/usr/include/stdio.h:173:16: note: expanded from macro 'stdout'
#define stdout stdout
^
/usr/include/stdio.h:169:25: note: previous declaration is here
extern struct _IO_FILE *stdout; /* Standard output stream. */
^
Signed-off-by: Thomas Haller <thaller@redhat.com>
Since kernel commit 8fe02e167efa8ed4a4503a5eedc0f49fcb7e3eb9,
the value NL80211_FREQUENCY_ATTR_NO_IR replaces PASSIVE_SCAN
and NO_IBSS. Hence their numerical values are identical and
cause the following compiler warning.
clang warns:
make[4]: Entering directory `./NetworkManager/src'
CC wifi-utils-nl80211.lo
platform/wifi/wifi-utils-nl80211.c:683:48: error: initializer overrides prior initialization of this subobject [-Werror,-Winitializer-overrides]
[NL80211_FREQUENCY_ATTR_NO_IBSS] = { .type = NLA_FLAG },
^~~~~~~~
platform/wifi/wifi-utils-nl80211.c:682:53: note: previous initialization is here
[NL80211_FREQUENCY_ATTR_PASSIVE_SCAN] = { .type = NLA_FLAG },
^~~~~~~~
Signed-off-by: Thomas Haller <thaller@redhat.com>
The variable is not actually unused, because it is used
to free the nl_object instance.
clang warns:
make[4]: Entering directory `./NetworkManager/src'
CC nm-linux-platform.lo
platform/nm-linux-platform.c:1746:35: error: unused variable 'obj_cleanup' [-Werror,-Wunused-variable]
auto_nl_object struct nl_object *obj_cleanup = obj;
^
Signed-off-by: Thomas Haller <thaller@redhat.com>
Some subdirectories of src/ encapsulate large chunks of functionality,
but src/config/, src/logging/, and src/posix-signals/ are really only
separated out because they used to be built into separate
sub-libraries that were needed either for test programs, or to prevent
circular dependencies. Since this is no longer relevant, simplify
things by moving their files back into the main source directory.
At some places, we considered a default route to be a route with
destination network 0.0.0.0 (::). This is wrong because a default route
is a route with plen==0.
This is for example relevant for OpenVPN which adds two routes
0.0.0.0/1 and 128.0.0.0/1 to hijack the default route. We should
not treat 0.0.0.0/1 as default route, instead NM should treat
it as any other subnet route (even if it effectively routes large
parts).
Signed-off-by: Thomas Haller <thaller@redhat.com>
Clean up some of the cross-includes between headers (which made it so
that, eg, if you included NetworkManagerUtils.h in a test program, you
would need to build the test with -I$(top_srcdir)/src/platform, and if
you included nm-device.h you'd need $(POLKIT_CFLAGS)) by moving all
GObject struct definitions for src/ and src/settings/ into nm-types.h
(which already existed to solve the NMDevice/NMActRequest circular
references).
Update various .c files to explicitly include the headers they used to
get implicitly, and remove some now-unnecessary -I options from
Makefiles.
'address' was not initialized, so using address.ifindex was not
going to work.
==6777== Conditional jump or move depends on uninitialised value(s)
==6777== at 0x44A6DB: check_cache_items (nm-linux-platform.c:1510)
==6777== by 0x44AA4F: announce_object (nm-linux-platform.c:1586)
==6777== by 0x44A59A: refresh_object (nm-linux-platform.c:1666)
==6777== by 0x44A676: check_cache_items (nm-linux-platform.c:1517)
==6777== by 0x44A812: announce_object (nm-linux-platform.c:1566)
==6777== by 0x44D957: handle_udev_event (nm-linux-platform.c:3964)
==6777== by 0x3BAFC125A7: g_closure_invoke (gclosure.c:777)
==6777== by 0x3BAFC2465C: signal_emit_unlocked_R (gsignal.c:3586)
==6777== by 0x3BAFC2C3E1: g_signal_emit_valist (gsignal.c:3330)
==6777== by 0x3BAFC2C69E: g_signal_emit (gsignal.c:3386)
==6777== by 0x3D93A03BE3: ??? (in /usr/lib64/libgudev-1.0.so.0.1.3)
==6777== by 0x3BAF8521D5: g_main_context_dispatch (gmain.c:3066)
nm-version.h was getting disted, making srcdir!=builddir work for
tarball builds, but not for git builds.
Also, remove "-I${top_builddir}/include" from all Makefile.ams, since
there's nothing generated in include/ any more.
The kernel adds these for various operations; they are short-lived,
added often, and not useful to NetworkManager. Ignore them. This
prevents NetworkManager from continuously updating the IPv6 config
and emitting state changes to clients via D-Bus for useless changes.
When starting, the initally loaded addresses were not fixed to have
absolute lifetimes.
This also leads to a crash due to failed assertion[1], when having
already expired addresses during NetworkManager start.
[1] backtrace:
#0 0x00007f39db91ec39 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56
#1 0x00007f39db920348 in __GI_abort () at abort.c:89
#2 0x00007f39dc35a0ed in _g_log_abort () at gmessages.c:255
#3 0x00007f39dc377c97 in g_assertion_message (domain=domain@entry=0x7f39e0085c67 "NetworkManager", file=file@entry=0x7f39e007f070 "platform/nm-linux-platform.c", line=line@entry=1146,
func=func@entry=0x7f39e0081430 <__FUNCTION__.23775> "_init_ip_address_lifetime", message=message@entry=0x7f39e13828b0 "assertion failed: (a_preferred <= a_valid && a_valid > 0 && a_preferred > 0)") at gtestutils.c:2278
#4 0x00007f39dc377cfa in g_assertion_message_expr (domain=domain@entry=0x7f39e0085c67 "NetworkManager", file=file@entry=0x7f39e007f070 "platform/nm-linux-platform.c", line=line@entry=1146,
func=func@entry=0x7f39e0081430 <__FUNCTION__.23775> "_init_ip_address_lifetime", expr=expr@entry=0x7f39e007f4e8 "a_preferred <= a_valid && a_valid > 0 && a_preferred > 0") at gtestutils.c:2293
#5 0x00007f39e0004608 in _init_ip_address_lifetime (address=address@entry=0x7fff7f8a2580, rtnladdr=rtnladdr@entry=0x7f39e12cdb20) at platform/nm-linux-platform.c:1144
#6 0x00007f39e0004ae4 in init_ip6_address (address=address@entry=0x7fff7f8a2580, rtnladdr=rtnladdr@entry=0x7f39e12cdb20) at platform/nm-linux-platform.c:1216
#7 0x00007f39e000a75e in ip6_address_get_all (platform=<optimized out>, ifindex=3) at platform/nm-linux-platform.c:3342
#8 0x00007f39e0032729 in nm_ip6_config_capture (ifindex=ifindex@entry=3, capture_resolv_conf=capture_resolv_conf@entry=0, use_temporary=use_temporary@entry=NM_SETTING_IP6_CONFIG_PRIVACY_UNKNOWN) at nm-ip6-config.c:303
#9 0x00007f39dffee8fa in update_ip_config (self=0x7f39e1380300, initial=<optimized out>) at devices/nm-device.c:5841
#10 0x00007f39dffef52a in nm_device_capture_initial_config (dev=<optimized out>) at devices/nm-device.c:5869
#11 0x00007f39e003e74e in get_existing_connection (manager=manager@entry=0x7f39e132b150, device=device@entry=0x7f39e1380300) at nm-manager.c:1576
#12 0x00007f39e003edb2 in add_device (self=self@entry=0x7f39e132b150, device=device@entry=0x7f39e1380300, generate_con=<optimized out>) at nm-manager.c:1826
#13 0x00007f39e003f503 in platform_link_added (reason=<optimized out>, plink=0x7f39e1372340, ifindex=3, self=0x7f39e132b150) at nm-manager.c:2205
#14 platform_link_cb (platform=<optimized out>, ifindex=3, plink=0x7f39e1372340, change_type=<optimized out>, reason=<optimized out>, user_data=<optimized out>) at nm-manager.c:2220
#15 0x00007f39da868d8c in ffi_call_unix64 () at ../src/x86/unix64.S:76
#16 0x00007f39da8686bc in ffi_call (cif=cif@entry=0x7fff7f8a2a40, fn=0x7f39e003f360 <platform_link_cb>, rvalue=0x7fff7f8a29b0, avalue=avalue@entry=0x7fff7f8a2930) at ../src/x86/ffi64.c:522
#17 0x00007f39dc84aad8 in g_cclosure_marshal_generic (closure=0x7f39e1368080, return_gvalue=0x0, n_param_values=<optimized out>, param_values=<optimized out>, invocation_hint=<optimized out>, marshal_data=0x0) at gclosure.c:1454
#18 0x00007f39dc84a298 in g_closure_invoke (closure=0x7f39e1368080, return_value=return_value@entry=0x0, n_param_values=5, param_values=param_values@entry=0x7fff7f8a2c40, invocation_hint=invocation_hint@entry=0x7fff7f8a2be0)
at gclosure.c:777
#19 0x00007f39dc85c35d in signal_emit_unlocked_R (node=node@entry=0x7f39e12c16c0, detail=detail@entry=0, instance=instance@entry=0x7f39e12c30a0, emission_return=emission_return@entry=0x0,
instance_and_params=instance_and_params@entry=0x7fff7f8a2c40) at gsignal.c:3586
#20 0x00007f39dc8640f2 in g_signal_emit_valist (instance=<optimized out>, signal_id=<optimized out>, detail=<optimized out>, var_args=var_args@entry=0x7fff7f8a2e20) at gsignal.c:3330
#21 0x00007f39dc8643af in g_signal_emit (instance=<optimized out>, signal_id=<optimized out>, detail=detail@entry=0) at gsignal.c:3386
#22 0x00007f39e000e24c in nm_platform_query_devices () at platform/nm-platform.c:330
#23 0x00007f39e0040f30 in nm_manager_start (self=0x7f39e132b150) at nm-manager.c:4207
#24 0x00007f39dffe53ab in main (argc=1, argv=0x7fff7f8a3468) at main.c:657
Regression introduced by commit 8310a039d8.
https://bugzilla.gnome.org/show_bug.cgi?id=732472
Signed-off-by: Thomas Haller <thaller@redhat.com>
There is no point in raising events that a route/address was added, if
a subsequent call to ip4_address_get_all() et.al don't show these
objects.
Signed-off-by: Thomas Haller <thaller@redhat.com>
If the timestamp is set to zero, the to_string() functions treat the lifetime
as based on *now*. For nm_platform_ip_address_cmp_expiry() this makes no
sense, because there is no absolute exiry to compare. Instead compare
them as expire earlier then the other address.
Signed-off-by: Thomas Haller <thaller@redhat.com>
The "lifetime" part when printing an address in nm_platform_ip[46]_address_to_string()
is supposed to show the raw, internal values of the address.
We already have the "lft" and "pref" output that presents the expiries based on now.
These fields are already crafted to show what the user probably wants
to see when looking at debugging log. "lifetime" should not do any
special casing and just print the raw values.
Signed-off-by: Thomas Haller <thaller@redhat.com>
When printing an address in nm_platform_ip4_address_to_string()
and nm_platform_ip6_address_to_string() treat an unset @timestamp
as counting from @now.
This is useful, if you just have the remaining lifetime at hand
and want to print an address. In general it is not a good idea to
leave the timestamp not anchored to an absolute @timestamp.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Previous patch 8310a039d8 modified
platform to set the timestamp of addresses always to 1.
So, when adding an address platform logging looked like:
signal: address 4 added: 192.168.232.3/24 lft 2000sec pref 1000sec lifetime 12345-1[13344,14344] dev em1 src kernel
This is confusing in the log file and during debugging. Instead set the
timestamp to the last modification time of the address so that it will
look like:
signal: address 4 added: 192.168.232.3/24 lft 2000sec pref 1000sec lifetime 12345-12345[1000,2000] dev em1 src kernel
Signed-off-by: Thomas Haller <thaller@redhat.com>
When setting the timestamp to 1, we have to subtract(!) one second
from a_valid and a_preferred.
Due to this error, NM saw the lifetimes of addresses from system as two
seconds larger then the actual value.
Signed-off-by: Thomas Haller <thaller@redhat.com>
_rebase_relative_time_on_now() is used both by _address_get_lifetime()/nm_platform_ip[46]_address_sync()
and the to_string() functions.
In the latter case, we want to print the original value, without padding. Otherwise in
the addresses are printed in the logs with an additional 5 seconds
padding, which is confusing.
For adding addresses in platform however, we still want to keep the
padding. So pass it on as additional parameter.
Signed-off-by: Thomas Haller <thaller@redhat.com>
CC nm-linux-platform.lo
platform/nm-linux-platform.c: In function '_nm_platform_link_get':
platform/nm-linux-platform.c:161:17: error: 'rtnllink' may be used uninitialized in this function [-Werror=maybe-uninitialized]
nl_object_put (*object);
^
platform/nm-linux-platform.c:1923:35: note: 'rtnllink' was declared here
auto_nl_object struct rtnl_link *rtnllink;
^
cc1: all warnings being treated as errors
Signed-off-by: Thomas Haller <thaller@redhat.com>
When only the lifetime of an address changes, we did not get a platform signal
as libnl does not consider the time fields in nl_object_diff().
Workaround by comparing the timestamps manually.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Moving setting the source of the address to the init_* functions.
This also has the advantage, that the platform internal to_string functions have the proper
source set.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Change the to_string functions to convert the lifetime/preferred values
to the time remaining when the function is evaluated. These functions
are used for printing/debugging, so it's more sensible to show the
remaining time.
On the other hand, for debugging, it's better to see the raw values (also).
In addition to the remaining time we keep to print the timestamps+now if the
address is not permanent. So when inspecting the logs it is possible to figure
out the real values.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Replace the calls to subtract_guint32() by _rebase_relative_time_on_now()
and _address_get_lifetime().
Signed-off-by: Thomas Haller <thaller@redhat.com>
Before nm_platform_ip4_address_sync() set the preferred time to the same value
as the address lifetime. The result was that the preferred time was
always identical to valid lifetime.
This will lead to the kernel using the address longer then the desired
preferred time (until validity of the address expires).
https://bugzilla.redhat.com/show_bug.cgi?id=1082041https://bugzilla.redhat.com/show_bug.cgi?id=1083283
Reported-by: Kai Engert <kengert@redhat.com>
Signed-off-by: Thomas Haller <thaller@redhat.com>
The kernel tells the address lifetimes in the 'struct ifa_cacheinfo'
attribute. This contains two timestamps (cstamp and tstamp) and two
relative lifetimes (ifa_prefered and ifa_valid).
The timestamps are equal to clock_gettime(CLOCK_MONOTONIC) scale in
1/100th of a second (wrapping every 497 days).
The preferred/valid times are re-adjusted everytime when sending the
message and count down as the time goes by. In other words, they are
anchored relatively to the moment of when kernel creates the netlink
message.
As platform is caching the rtnl_addr object, the information of *when* the
lifetimes started counting is not available.
This patch fixes reading these values by hacking the libnl object
when it gets received, so that valid and preferred are instead absolute
expiration timestamps in scale nm_utils_get_monotonic_timestamp_s() --
which NM internally is used for address timestamps.
There are two minor downsides to this hack:
- the valid and preferred properties of a cached rtnl_addr object have
an unexpected meaning, i.e. they are absolute and in a different time
scale.
- later when converting rtnl_addr to NMPlatformIPAddress, the base
timestamp is set to "1", i.e. an NMPlatformIPAddress has no knowledge
of when the address was created or last modified. The timestamp
property of NMPlatformIPAddress is solely there to anchor the relative
timestamps lifetime and preferred. Do not use it for anything
else.
Another reason the timestamp property is meaningless is that
its scale nm_utils_get_monotonic_timestamp_s() starts counting at
process start. So addresses that existed before would have a negative
or zero timestamp, which we avoid. This in turn could be solved by either
allowing negative timestamps or by shifting
nm_utils_get_monotonic_timestamp_*(). Both is viable, but not
necessary (ATM), because the age of an address has no other apparent
use then to anchor the relative timestamps.
Another implication is, that we potentially could get rid of the
timestamp completely, and insteat make preferred and lifetime be
absolute expiries.
This will be fixed properly later, by not caching libnl objects but instead
native NMPlatform objects. For those we have full control over their properties.
https://bugzilla.gnome.org/show_bug.cgi?id=727382
Signed-off-by: Thomas Haller <thaller@redhat.com>
NMIP[46]Route had a "source" field, but it was always set to KERNEL
for routes read from the kernel (even if they were originally added by
NM).
Fix things a bit by translating between our "source" field and the
kernel's "protocol" field.
https://bugzilla.gnome.org/show_bug.cgi?id=729203
In this mode, nmtst itself will not log anything and not set the logging
level. Also, it will set g_log_set_always_fatal().
This is for tests that want to assert against all logged messages via
g_test_expect_message().
In this mode also setting the logging level via NMTST_DEBUG variable has
no effect. The test is expected to manage the logging level itself and
changing the logging level might interfere with the test.
As a showcase, move keyfile/tests/test-keyfile.c to nmtst.
Signed-off-by: Thomas Haller <thaller@redhat.com>
Call to nmtst_reexec_sudo(), which allows you to specify a program
via environment variable to exec the test.
This is useful to exec the test program with sudo.
NMTST_DEBUG="no-debug,sudo-cmd=$PWD/tools/test-sudo-wrapper.sh" make -C src/platform/tests/ check
Signed-off-by: Thomas Haller <thaller@redhat.com>
Always run the linux platform tests, even if called as non-root user.
In such a case, print a message and return 77 (signalizing that the test
was skipped).
Only if we configured with --enable-test=root, we enforce that the
user executes the tests as root.
Co-Authored-By: Pavel Šimerda <psimerda@redhat.com>
Signed-off-by: Thomas Haller <thaller@redhat.com>
In older versions of team (e.g. Fedora 17), the master team device
stays up, even if no slaves are IFF_LOWER_UP. Workaround this bug.
Signed-off-by: Thomas Haller <thaller@redhat.com>
In this case, the fake platform implementation was wrong in that it did
not set the source property of the route/address objects like linux
platform does. Fix the test and the fake platform.
https://bugzilla.gnome.org/show_bug.cgi?id=706293
Signed-off-by: Thomas Haller <thaller@redhat.com>
The handling for announcing links was broken resulting in
duplicate link-added signals from platform.
Co-Authored-By: Thomas Haller <thaller@redhat.com>
Signed-off-by: Thomas Haller <thaller@redhat.com>
When adding a link, the Linux platform implementation raises the
link-changed signal synchronously. Fix the fake platform to behave identically
and also fix all the tests.
This also fixes the Linux platform tests for the most part because now the
test functions (and fake platform) behave like the Linux system
implementation.
https://bugzilla.gnome.org/show_bug.cgi?id=706293
Co-Authored-By: Thomas Haller <thaller@redhat.com>
Signed-off-by: Thomas Haller <thaller@redhat.com>
When doing a lookup for an libnl route, the cache comparison function
for routes takes into account 'family', 'tos', 'table', 'dst', and 'prio'.
In NetworkManager we don't use all of these properties for a route, so
at several places when doing a cache lookup we don't have all identifying
properties. Usually we only have 'family' and 'dst' ('table' is
implicit 0, because NM does currently not care about any other tables).
The problem is that NM sees routes with different 'tos', 'prio', but it
cannot look them up in the cache. Add a hack to search the cache
fuzzy.
This is similar to the hack for link, where the identifying properties
are 'family' and 'ifindex', but we only have 'ifindex' at hand. However,
contrary to this hack, we coerce the 'family' to AF_UNSPEC for every link cache
operation. This is not viable in this case, because we internally need
the 'tos' field.
We need the 'tos' field because when deleting an IPv4 route, the 'tos' field must
match. See fib_table_delete(). This was already partially fixed by commit
f0daf90298, but before the lookup to the
cached object would fail for any non-zero 'tos'.
Signed-off-by: Thomas Haller <thaller@redhat.com>
