anaconda-patches April 2013

anaconda-patches@lists.fedorahosted.org

18 participants
101 discussions

[PATCH] Pass open file to execWithRedirect for vncpasswd (#948638)
by Brian C. Lane 08 Apr '13

08 Apr '13

From: "Brian C. Lane" <bcl(a)redhat.com> --- pyanaconda/vnc.py | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/pyanaconda/vnc.py b/pyanaconda/vnc.py index d86c36c..f016299 100644 --- a/pyanaconda/vnc.py +++ b/pyanaconda/vnc.py @@ -68,12 +68,13 @@ class VncServer: r, w = os.pipe() os.write(w, "%s\n" % self.password) - # the -f option makes sure vncpasswd does not ask for the password again - rc = iutil.execWithRedirect("vncpasswd", ["-f"], - stdin=r, stdout=self.pw_file) + with open(self.pw_file, "w") as pw_file: + # the -f option makes sure vncpasswd does not ask for the password again + rc = iutil.execWithRedirect("vncpasswd", ["-f"], + stdin=r, stdout=pw_file) - os.close(r) - os.close(w) + os.close(r) + os.close(w) return rc -- 1.8.1.4

2 1

[master] Something more for libreport
by Vratislav Podzimek 08 Apr '13

08 Apr '13

ABRT/libreport's bug deduplication should work in a better way if we provide it some more data. Patches 2/3 and 3/3 are needed for that together with the associated patch for python-meh. PATCH 1/3 just fixes the makeupdates script to place _isys.so where it is expected to be. Maybe we should change the place where it lives (from pyanaconda directory to pyanaconda/isys), but for now let's use this fix. Vratislav Podzimek (3): Fix _isys.so location in the updates.img Add release number to the result of getAnacondaVersion Tell python-meh architecture of the anaconda package pyanaconda/exception.py | 1 + pyanaconda/isys/Makefile.am | 2 +- scripts/makeupdates | 5 +---- 3 files changed, 3 insertions(+), 5 deletions(-) -- 1.7.11.7

3 9

rewrite of device factory classes
by David Lehman 05 Apr '13

05 Apr '13

This is a big cleanup and refactor that is needed to support more complex device stacks like lvm-on-md and lvm thin provisioning. Patch 1 is probably much easier to understand by applying it than by trying to read it. The same may or may not be true for patch 2. I've tested these a good bit. I've spent at least a week testing them and fixing bugs I find.

2 3

[master] Friday's documenation&refactorization
by Vratislav Podzimek 05 Apr '13

05 Apr '13

The patch was generated with the following command: find . -type f -name '*.py' -exec sed -ir 's?^$\s\+$@$.*:.*$$?\1:\2?' '{}' \; It migrates our epydoc-based docstrings to the new syntax used by the Sphinx tool which we use also for the new code. Since it affects only docstrings, there should cause no harm to the git history. Vratislav Podzimek (1): Use Sphinx syntax for docstrings pyanaconda/exception.py | 6 +-- pyanaconda/flags.py | 4 +- pyanaconda/iutil.py | 8 ++-- pyanaconda/keyboard.py | 66 +++++++++++++++--------------- pyanaconda/localization.py | 10 ++--- pyanaconda/network.py | 6 +-- pyanaconda/ntp.py | 24 +++++------ pyanaconda/safe_dbus.py | 58 +++++++++++++------------- pyanaconda/timezone.py | 18 ++++---- pyanaconda/ui/gui/__init__.py | 4 +- pyanaconda/ui/gui/spokes/datetime_spoke.py | 10 ++--- pyanaconda/ui/gui/spokes/welcome.py | 2 +- 12 files changed, 108 insertions(+), 108 deletions(-) -- 1.7.11.7

2 2

various blivet patches
by David Lehman 05 Apr '13

05 Apr '13

These are all pretty self-explanatory.

2 5

anaconda custom spoke patches
by David Lehman 05 Apr '13

05 Apr '13

This is stuff I turned up testing the new device factory classes.

2 5

Let systemd/udev activate storage during installer boot
by David Lehman 05 Apr '13

05 Apr '13

Stop fighting all the things that want to automatically activate storage devices. This allows us to remove our udev rules, which is a good thing. It also moves activation of multipath and fwraid into the domains of the appropriate tools and out of our hands, which is also good. One of the problems with our current/old approach is that our udev rules run lvm/md information gathering commands for partitions on every path of a multipath, which is insane for those with large numbers of paths. This way we only run those commands as needed on devices that we know are directly usable.

2 5

[PATCH] Network spoke: fix refresh of device IP configuration
by Radek Vykydal 05 Apr '13

05 Apr '13

Register signal instead of silly polling which even doesn't work in some cases (eg change ipv6 congig of a device to manual and switch it off and on -> only ipv6 config is showed) --- pyanaconda/ui/gui/spokes/network.py | 49 ++++++++++++++++--------------------- 1 file changed, 21 insertions(+), 28 deletions(-) diff --git a/pyanaconda/ui/gui/spokes/network.py b/pyanaconda/ui/gui/spokes/network.py index 04f01ce..7ee3618 100644 --- a/pyanaconda/ui/gui/spokes/network.py +++ b/pyanaconda/ui/gui/spokes/network.py @@ -336,6 +336,8 @@ class NetworkControlBox(object): def initialize(self): for device in self.client.get_devices(): + device.connect("notify::ip4-config", self.on_device_config_changed) + device.connect("notify::ip6-config", self.on_device_config_changed) self.add_device_to_list(device) treeview = self.builder.get_object("treeview_devices") @@ -389,9 +391,12 @@ class NetworkControlBox(object): if new_state == NetworkManager.DeviceState.SECONDARIES: return self._refresh_carrier_info() - read_config_values = (new_state == NetworkManager.DeviceState.ACTIVATED) if device == self.selected_device(): - self.refresh_ui(device, read_config_values, new_state) + self.refresh_ui(device, new_state) + + def on_device_config_changed(self, device, *args): + if device == self.selected_device(): + self.refresh_ui(device) def on_wireless_ap_changed_cb(self, combobox, *args): if self._updating_device: @@ -673,7 +678,7 @@ class NetworkControlBox(object): for row in rows_to_remove: del(row) - def refresh_ui(self, device, read_config_values=True, state=None): + def refresh_ui(self, device, state=None): if not device: notebook = self.builder.get_object("notebook_types") @@ -686,27 +691,12 @@ class NetworkControlBox(object): self._refresh_parent_vlanid(device) self._refresh_speed_hwaddr(device, state) self._refresh_ap(device, state) - if read_config_values: - num_of_tries = 3 - else: - num_of_tries = 0 - self._refresh_device_cfg((device, num_of_tries), state) + self._refresh_device_cfg(device, state) - def _refresh_device_cfg(self, dev_tries, state): - device, num_of_tries = dev_tries - ipv4cfg = None - ipv6cfg = None + def _refresh_device_cfg(self, device, state): - # We might need to wait for config objects to become available - if num_of_tries > 0: - ipv4cfg = device.get_ip4_config() - ipv6cfg = device.get_ip6_config() - if not ipv4cfg and not ipv6cfg: - GLib.timeout_add(300, - self._refresh_device_cfg, - (device, num_of_tries-1), - state) - return False + if state is None: + state = device.get_state() dev_type = device.get_device_type() if dev_type == NetworkManager.DeviceType.ETHERNET: @@ -718,10 +708,14 @@ class NetworkControlBox(object): elif dev_type == NetworkManager.DeviceType.VLAN: dt = "wired" - if state is None: - state = device.get_state() - if (ipv4cfg - and state == NetworkManager.DeviceState.ACTIVATED): + if state == NetworkManager.DeviceState.ACTIVATED: + ipv4cfg = device.get_ip4_config() + ipv6cfg = device.get_ip6_config() + else: + ipv4cfg = None + ipv6cfg = None + + if ipv4cfg: addr = socket.inet_ntoa(struct.pack('=L', ipv4cfg.get_addresses()[0].get_address())) self._set_device_info_value(dt, "ipv4", addr) @@ -747,8 +741,7 @@ class NetworkControlBox(object): # TODO NM_GI_BUGS - segfaults on get_addres(), get_prefix() ipv6_addr = None - if (ipv6cfg - and state == NetworkManager.DeviceState.ACTIVATED): + if ipv6cfg: config = dbus.SystemBus().get_object(NM_SERVICE, ipv6cfg.get_path()) addr, prefix, gw = getNMObjProperty(config, ".IP6Config", "Addresses")[0] -- 1.7.11.7

2 2

[python-meh] [PATCH] Reword the 'Debug' button waring
by Vratislav Podzimek 05 Apr '13

05 Apr '13

Getting to tty1 happens only in Anaconda running a non-live installation, so we should just warn user it may happen. Also the GUI is now on tty7, but in any case there is no point in getting back to tty7 manually, because until the debugging session is finished the GUI is frozen. --- ui/exception-dialog.glade | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/ui/exception-dialog.glade b/ui/exception-dialog.glade index 9ae497b..0bed126 100644 --- a/ui/exception-dialog.glade +++ b/ui/exception-dialog.glade @@ -87,11 +87,9 @@ report the bug below or quit the program.</property> <child> <object class="GtkButton" id="reportButton"> <property name="label" translatable="yes">_Report Bug</property> - <property name="use_action_appearance">False</property> <property name="visible">True</property> <property name="can_focus">True</property> <property name="receives_default">True</property> - <property name="use_action_appearance">False</property> <property name="use_underline">True</property> <signal name="clicked" handler="on_report_clicked" swapped="no"/> </object> @@ -104,11 +102,9 @@ report the bug below or quit the program.</property> <child> <object class="GtkButton" id="quitButton"> <property name="label" translatable="yes">_Quit</property> - <property name="use_action_appearance">False</property> <property name="visible">True</property> <property name="can_focus">True</property> <property name="receives_default">True</property> - <property name="use_action_appearance">False</property> <property name="use_underline">True</property> <signal name="clicked" handler="on_quit_clicked" swapped="no"/> </object> @@ -186,13 +182,13 @@ report the bug below or quit the program.</property> <object class="GtkBox" id="bottomBox"> <property name="visible">True</property> <property name="can_focus">False</property> - <property name="spacing">4</property> + <property name="spacing">6</property> <child> <object class="GtkLabel" id="warningLabel"> <property name="visible">True</property> <property name="can_focus">False</property> - <property name="xalign">0.89999997615814209</property> - <property name="label" translatable="yes">'Debug' will take you to tty1. Press Ctrl + Alt + F6 to return.</property> + <property name="xalign">1</property> + <property name="label" translatable="yes">'Debug' may take you to tty1.</property> <attributes> <attribute name="style" value="italic"/> <attribute name="scale" value="0.80000000000000004"/> @@ -207,11 +203,9 @@ report the bug below or quit the program.</property> <child> <object class="GtkButton" id="debugButton"> <property name="label" translatable="yes">_Debug</property> - <property name="use_action_appearance">False</property> <property name="visible">True</property> <property name="can_focus">True</property> <property name="receives_default">True</property> - <property name="use_action_appearance">False</property> <property name="use_underline">True</property> <signal name="clicked" handler="on_debug_clicked" swapped="no"/> </object> -- 1.7.11.7

2 1

[PATCH 1/6] memInstalled has moved (#947261).
by Chris Lumens 05 Apr '13

05 Apr '13

--- anaconda | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/anaconda b/anaconda index 4260dd6..77b61ee 100755 --- a/anaconda +++ b/anaconda @@ -469,7 +469,7 @@ def setupDisplay(anaconda, opts, addon_paths=None): # disable VNC over text question when not enough memory is available if blivet.util.total_memory() < isys.MIN_GUI_RAM: - stdoutLog.warning("Not asking for VNC because current memory (%d) < MIN_GUI_RAM (%d)" % (iutil.memInstalled(), isys.MIN_GUI_RAM)) + stdoutLog.warning("Not asking for VNC because current memory (%d) < MIN_GUI_RAM (%d)" % (blivet.util.total_memory(), isys.MIN_GUI_RAM)) flags.vncquestion = False # disable VNC question if text mode is requested and this is a ks install -- 1.8.1.2

2 6

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

anaconda-patches April 2013