Program received signal SIGSEGV, Segmentation fault.
0x000000000041b3e5 in nmt_page_dsl_constructed (object=0x7fffec0228c0 [NmtPageDsl]) at nmt-page-dsl.c:109
109 priv->ethernet_page = nmt_page_ethernet_new (conn, nmt_editor_page_device_get_device_entry (NMT_EDITOR_PAGE_DEVICE (dsl)));
Missing separate debuginfos, use: debuginfo-install libffi-3.0.10-3.fc18.x86_64 newt-0.52.15-1.fc19.x86_64
(gdb) bt
#0 0x000000000041b3e5 in nmt_page_dsl_constructed (object=0x7fffec0228c0 [NmtPageDsl]) at nmt-page-dsl.c:109
#1 0x0000003e9ac161e3 in g_object_newv (object_type=1, object_type@entry=16203872, n_parameters=n_parameters@entry=2, parameters=parameters@entry=0xf74350)
at gobject.c:1741
#2 0x0000003e9ac16560 in g_object_new_valist (object_type=object_type@entry=16203872, first_property_name=first_property_name@entry=0x42e524 "connection",
var_args=var_args@entry=0x7fffffffcef8) at gobject.c:1830
#3 0x0000003e9ac16894 in g_object_new (object_type=16203872, first_property_name=first_property_name@entry=0x42e524 "connection") at gobject.c:1545
#4 0x000000000041b504 in nmt_page_dsl_new (conn=<optimized out>, deventry=deventry@entry=0x974690 [NmtDeviceEntry]) at nmt-page-dsl.c:47
#5 0x000000000041808e in nmt_editor_constructed (object=<optimized out>) at nmt-editor.c:355
#6 0x0000003e9ac161e3 in g_object_newv (object_type=1, object_type@entry=16174256, n_parameters=n_parameters@entry=4, parameters=parameters@entry=0xf6cec0)
at gobject.c:1741
...
(gdb) p priv
$1 = (NmtPageDslPrivate *) 0x0
We wait for each deletion to complete, so the connections were getting
removed from the connections array as we edited it (unlike with the
old transfer-container GSList-based code). Fix this by copying the
slaves out into their own list first.
Coverity: Defect type: SWAPPED_ARGUMENTS
clients/tui/newt/nmt-newt-section.c:328: swapped_arguments: The positions of arguments in the call to "nmt_newt_widget_size_request" do not match the ordering of the parameters: * "&border_height" is passed to "width" * "&border_width" is passed to "height"
set_connection_path() is almost always called right before enable(),
and it's unclear why it would be called anywhere else. So just
merge the two methods.
nm_secret_agent_simple_*() functions should take an
NMSecretAgentSimple, not an NMSecretAgent.
The type macros were incorrectly validating against
NM_TYPE_SECRET_AGENT rather than NM_TYPE_SECRET_AGENT_SIMPLE.
nmt-editor.c: In function ‘nmt_editor_constructed’:
nmt-editor.c:365:2: error: ‘page’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
add_sections_for_page (editor, grid, page);
^
If an NmtEditorGrid row has a widget, but no label, then we make the
widget span both the label and widget columns. But previously we
weren't doing the same for rows with labels but no widgets. (In fact,
we didn't even allow rows with no widgets; label-only rows had to
specify dummy widgets.)
Fix it so that labels will span into an empty widget column. (This
ensures that a long section name won't force the entire grid to have
an overwide label column).
Also, in both the "no label" and "no widget" cases, still show the
"extra" column if it's present.
Rather than having NmtEditor need to know that DSL requires a DSL
page, and Ethernet page, and a PPP page, just have the DSL page itself
return multiple sections.
Instead of having NmtEditorPage be a widget itself, have it just be an
object that returns a list of NmtEditorSections, where
NmtEditorSection is a subclass of NmtNewtSection.
(This will be important when adding VPN pages, which will be split up
into multiple sections, but with the different sections needing to
cooperate on updating the NMSettingVpn. This reorganization lets us
have an NMPageVpn containing multiple sections, with the NMPageVpn
object handling the coordination between the sections.)
All page types except NmtPageMain are displayed as collapsible
sections. NmtPageMain behaved quite differently from everything else.
So merge it into its container, NmtEditor.
Rename NmtPageDevice to NmtEditorPageDevice, and rename NmtPageGrid to
NmtEditorGrid.
Now all types with names starting with "NmtPage" are instantiable page
types, not abstract classes or low-level widgets.
If we're activating the device without knowing the connection in advance, defer
servicing the requests for the secrets until we set the path.
[lkundrak@fedora20-2 ~]$ nmcli --ask c ifname wlan0
(process:18405): libnm-CRITICAL **: nm_object_get_path: assertion 'NM_IS_OBJECT (object)' failed
Error: Connection activation failed: The device has no connections available.
[lkundrak@fedora20-2 ~]$
We were refreshing the list when the set of available devices changed,
or the set of active connections changed, but not when the set of
available connections changed.
https://bugzilla.gnome.org/show_bug.cgi?id=740333
config.h should be included from every .c file, and it should be
included before any other include. Fix that.
(As a side effect of how I did this, this also changes us to
consistently use "config.h" rather than <config.h>. To the extent that
it matters [which is not much], quotes are more correct anyway, since
we're talking about a file in our own build tree, not a system
include.)
Pop-up menus were slightly mis-aligned in the main window, and even
more mis-aligned in slave-editing windows. Fix that.
Also add a bit of padding to the pop-up window, because it just looks
better that way.
Since adding NMSettingIPConfig:gateway, we were just binding that
property to the Gateway entry as a string. But this caused two
different problems: first, we were trying to set the :gateway property
from the entry even when the IP address in the entry was incomplete
(causing warnings), and second, we were no longer enforcing the rule
that the gateway can only be set when there are static addresses
configured.
Fix this by adding back nm_editor_bind_ip_gateway_to_string(), but
with new semantics reflecting the new way NMSettingIPConfig:addresses
and :gateway work. (Besides just fixing the new bugs, this also makes
the Gateway entry insensitive when there are no addresses; before,
nmtui would allow you to type there, but the value would not be
saved.)
Fixes: Test263_nmtui_ipv4_addresses_delete_ip_and_back_to_auto
https://bugzilla.gnome.org/show_bug.cgi?id=740017
When a connection is being activated, nmcli could ask for secrets for another
connection, which might confuse users. We check the request now and only ask
for secrets of connection being activated.
Test case:
$ nmcli con up my-ethernet0
Passwords or encryption keys are required to access the wireless network 'Red Hat'.
Warning: password for '802-1x.identity' not given in 'passwd-file' and nmcli cannot ask without '--ask' option.
The gateway is a global property of the IPv4/IPv6 configuration, not
an attribute of any particular address. So represent it as such in the
API; remove the gateway from NMIPAddress, and add it to
NMSettingIPConfig.
Behind the scenes, the gateway is still serialized along with the
first address in NMSettingIPConfig:addresses, and is deserialized from
that if the settings dictionary doesn't contain a 'gateway' key.
Adjust nmcli's interactive mode to prompt for IP addresses and gateway
separately. (Patch partly from Jirka Klimeš.)
Split a base NMSettingIPConfig class out of NMSettingIP4Config and
NMSettingIP6Config, and update things accordingly.
Further simplifications of now-redundant IPv4-vs-IPv6 code are
possible, and should happen in the future.
Merge NMIP4Address and NMIP6Address into NMIPAddress, and NMIP4Route
and NMIP6Route into NMIPRoute. The new types represent IP addresses as
strings, rather than in binary, and so are address-family agnostic.
libnm mostly used GPtrArrays in its APIs, except that arrays of
connections were usually GSLists. Fix this and make them GPtrArrays
too (and rename nm_client_list_connections() to
nm_client_get_connections() to match everything else).
Consolidate NMClientError and NMObjectError (such that there is now
only one libnm-API-specific error domain). In particular, merge
NM_CONNECTION_ERROR_CONNECTION_REMOVED with
NM_OBJECT_ERROR_OBJECT_CREATION_FAILURE as the new
NM_CONNECTION_ERROR_OBJECT_CREATION_FAILED.
Also make object_creation_failed() be a plain method rather than a
signal, since there's no reason for anyone to be connecting to it on
another object. And remove its GError argument because the subclass
can just create its own more-specific error.
Move the definition of NMSecretAgentError to nm-errors, register it
with D-Bus, and verify in the tests that it maps correctly.
NM_SECRET_AGENT_ERROR_INTERNAL_ERROR is renamed to
NM_SECRET_AGENT_ERROR_FAILED, and NM_SECRET_AGENT_ERROR_NOT_AUTHORIZED
to NM_SECRET_AGENT_ERROR_PERMISSION_DENIED, for consistency with other
error domains. While NMSecretAgentError, unlike most other error
domains, has always been correctly mapped across D-Bus, the renaming
is not an ABI break, because the daemon never checks for either of
those values, so all versions of the daemon will treat
"org.freedesktop.NetworkManager.SecretAgent.InternalError" and
"org.freedesktop.NetworkManager.SecretAgent.Failed" the same (by just
ignoring the error name and keeping only the error message).
PPPoE connections involve two different network connections, making it
ambiguous what the "Device" field refers to. Clarify that it refers to
the Ethernet device, not the PPP device.
NMActiveConnection:connection was an object path rather than an
NMRemoteConnection because in the past the NMObject property system
wasn't capable of handling NMRemoteConnection-valued properties.
NMActiveConnection:master was an object path rather than an NMDevice
entirely by accident. Fix both.
NMActiveConnection:specific-object can't currently be converted to an
object, because we don't know ahead of time what object type it is,
and NMObject can't deal with that. Instead, we rename it to
:specific-object-path (and likewise for its get function), both to
emphasize that it doesn't behave like other properties, and to leave
the old name open for an actual object-valued property later on.
If a master is deleted but its slaves are left behind, show those
slaves in the appropriate part of the connection list, so they can be
deleted. (This code is just copied from nm-connection-editor.)
nmt_newt_grid_size_allocate() depends on nmt_newt_grid_size_request()
having been called immediately prior, which would normally be true,
except that NmtNewtSection adjusts the label widgets in its border to
match its allocation, so when its size changes, it will end up calling
size_allocate() on the border with out-of-date requisition data. Fix
that by re-size_requesting the border after modifying it.
https://bugzilla.gnome.org/show_bug.cgi?id=738010
Add nm_utils_wifi_strength_bars(), which figures out whether the
terminal can display graphical wifi strength bars, and converts a
numerical value to the appropriate Unicode or ASCII characters.
This also now takes into consideration the fact that the console font
doesn't contain all of the necessary characters, so we can't display
the graphical bars there. (rh #1131491)
Make synchronous APIs take GCancellables, and make asynchronous APIs
use GAsyncReadyCallbacks and have names ending in "_async", with
"_finish" functions to retrieve the results.
Also, make nm_client_activate_connection_finish(),
nm_client_add_and_activate_finish(), and
nm_remote_settings_add_connection_finish() be (transfer full) rather
than (transfer none), because the refcounting semantics become
slightly confusing in some edge cases otherwise.
Merge nm_remote_settings_add_connection() and
nm_remote_settings_add_connection_unsaved(), and likewise
nm_remote_connection_commit_changes() and
nm_remote_connection_commit_changes_unsaved(), by adding a boolean
flag to each saying whether to save to disk.
The entry can't be empty. Otherwise removing the value and saving connection
would result in the primary option containing only the first character of the
original value.
