| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| |
|
|
|
|
|
|
|
|
|
| |
I hit a bug triggered by repeatedly toggling encryption for a vg
while also doing some size changes that revealed a problem in the
current apply/propagate_actions approach. There's just too much
that can happen to try to duplicate/cover it all.
Just replacing the main Storage instance isn't an option since each
of the ui spokes holds a reference to it. However, all of the data
is really in just a few data attributes of Storage and DeviceTree.
|
| |
|
|
|
| |
Also make it a bit easier to see what's going on by performing removals
from the container's disk set separately from toggling member encryption.
|
| | |
|
| |
|
|
|
|
|
| |
In this case what's really being done is we're creating a new device
with the old parameters since we failed to apply the new settings.
Either way, we shouldn't be passing the new/failed encryption, raid
level, or disk set when trying to recreate the previous device setup.
|
| |
|
|
|
| |
If the user selects btrfs as device type and there is a new btrfs volume
defined, populate the raid options based on that volume's raid settings.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
old:
Wireless (wlan0) connected to anaconda_wifi_test4
new:
Wireless connected to anaconda_wifi_test4
old:
Connected devices: eth0, wlan0 (anaconda_wifi_test4)
new:
Connected: eth0, anaconda_wifi_test4
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
Depending on the filesystem choices for stage2 grub2's core.img may be
too large to fit into the area between the mbr and the first partition.
This adds a check that will show an error if the stage2 fs isn't extX
and the first partition starts lower than 512K.
|
| |
|
|
|
|
| |
We use PANGO_ELLIPSIZE_MIDDLE to show the status and one cannot
get the whole status. Also screen readers could probably work with
this.
|
| |
|
|
|
|
|
|
|
|
| |
Without these lines, the IO blocks everything in python until complete, which
means setting up a bad NFS mount in the source spoke gives you an unresponsive
hub until the NFS timeout hits.
This is only a patch. The better fix would be to use subprocess for mounting
instead of our own C blob. The only reason this C blob exists is because
loader also required it, and no one's found the time to convert this old code.
|
| |
|
|
| |
I guess I'm the only person to touch isys in a very long time, huh?
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
- default auto connections are run by NM
- device can be configured also using dracut options
- in rescue, nmcli can be used, it is going to be enhanced to
offer also configuration
- rescue is using obsolete text mode network configuration which
is incomplete, not maintained and doesn't work currently
|
| |
|
|
|
|
|
|
|
|
| |
Add a very basic U-Boot bootloader class for ARM platforms.
This addresses the following needs.
- avoids using GRUB2 as a placeholder for the bootloader.
- installs the required uboot-tools package by default.
- ensures the boot/uboot partition is bootable.
Signed-off-by: Brian C. Lane <bcl@redhat.com>
|
| | |
|
| |
|
|
|
|
|
|
|
| |
In the resize/reclaim dialog you can trigger creation of a disklabel
on a disk containing whole-disk formatting by selecting "Delete".
In custom, you can set mountpoint, reformat, encrypt. If you want to
remove the filesystem and set the disk up with a disklabel, hit the
minus/remove button while the whole-disk filesystem is selected.
|
| |
|
|
| |
Related: rhbz#870476
|
| |
|
|
|
|
|
|
|
|
|
| |
Stop using the AUTOPART_TYPE constants since there aren't enough of
them to cover all the device types we'll want to handle.
This also removes some conversion between DEVICE_TYPE constants
(previously used only in the custom spoke) and AUTOPART_TYPE constants
(previously used only in storage).
Related: rhbz#870476
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
Additional fix.
If the info refresh is caused by device state change, use the state
that triggered the callback (do not read the state in the callback)
to prevent race with SECONDARIES state.
|
| |
|
|
|
|
|
|
|
| |
With 1a1e7ee8405f44585801d368b139d9e0d3e3abfd we now can use this
trick, because the script (and encoding) is parsed from sr_RS@latin
(i.e. before mangleLocale call). The result is that english_name,
display_name and territory are taken from the babel.Locale object that
understands "sr_Latn_RS" and value for $LANG is valid locale
"sr_RS.UTF-8@latin" as returned by LocaleInfo.__repr__.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
We have to potentially unmangle a lang value from say es_ES.UTF-8 down
to just "es" in order to find the group within comps.
|
| |
|
|
| |
Do not check initially_connected in refresh() but in __init__()
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Only include the swap request size in the case where we're about to
tell them they can run autopart in the existing free space, since
that's the main place it can be a surprise to find that you do not
actually have space for the selected software.
For the bug referenced below we still need to redo the space check
each time a new software selection is applied.
Related: rhbz#853636
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Don't even allow toggling of encryption for existing devices that have
their encryption applied at some lower level (eg: what our autopart
produces).
If the encryption is applied to container members (lvm, btrfs), let
Storage.setContainerMembers handle the changes. If toggling encryption
for an existing device or for any device for which we apply encryption
to the leaves (md, partition), handle the change of state directly in
_save_right_side.
|
| |
|
|
|
| |
At some point we'll want the same support for managing encrytion of
leaf devices, but for now that's handled by the custom spoke.
|
| |
|
|
| |
Related: rhbz#873445
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
The error message isn't the best one ever, and unfortunately the real error
is buried so deep that there's not really much hope of getting it to
display. At least this way, the user might think to look at the log file.
|
| |
|
|
|
|
| |
This will require lorax to actually put the plugin files in the image,
but once they are there we get all the benefit of somebody else's code
managing conditional package installs.
|
| |
|
|
|
| |
ksdata only gets set at the beginning and ending of a spoke, not when things are
changed on the spoke without leaving. Thus, ksdata is invalid.
|
| |
|
|
|
|
| |
It makes sense to set timezone to UTC or GMT-XX. Please mind that
"Etc" category doesn't come from my mind, that's how these timezones
are defined in the /usr/share/zoneinfo tree.
|
| |
|
|
|
|
| |
This is needed for displaying UTC and GMT-X timezones. Setting
timezone to "" graphically sets "Europe/London", but doesn't add
the pin.
|
| | |
|
| |
|
|
|
| |
rsync tries to remove xattrs even though -x has been specified. They
don't need to be copied anyway so exclude /dev/ /proc/ and /sys/
|
| | |
|
