Benedikt Morbach [Thu, 19 Jan 2012 18:10:38 +0000 (19:10 +0100)]
Install a systemd service file if systemd is used
This makes dbus defer starting udisks to systemd instead of doing it
itself and thus puts the daemon under proper systemd supervision.
Can be disabled by passing --without-systemdsystemunitdir to configure
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sat, 21 Jan 2012 21:45:50 +0000 (16:45 -0500)]
Fix use of memset(3)
Pointed out here https://bugs.freedesktop.org/show_bug.cgi?id=45039
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sat, 21 Jan 2012 21:26:46 +0000 (16:26 -0500)]
Ensure PATH is set
Apparently PATH may be empty if activated by recent D-Bus daemons.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 20 Jan 2012 21:44:50 +0000 (16:44 -0500)]
Post-release version bump to version 1.91.0
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 20 Jan 2012 21:38:56 +0000 (16:38 -0500)]
Update NEWS for release
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 20 Jan 2012 18:08:11 +0000 (13:08 -0500)]
Fix typos in udisks(8) man page and reword a couple of things
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 20 Jan 2012 17:39:02 +0000 (12:39 -0500)]
Clarify access control rules in udisks(8) man page
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 20 Jan 2012 16:12:19 +0000 (11:12 -0500)]
Fix hints for PC floppy drives
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 19 Jan 2012 18:36:22 +0000 (13:36 -0500)]
Don't accept a blank options string when adding/updating fstab entries
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 19 Jan 2012 15:22:31 +0000 (10:22 -0500)]
Change how devices referenced in /etc/fstab and /etc/crypttab are handled
Hitherto we've required extra authorization (typically requiring the
user to authenticate as root) if a device is referenced in fstab or
crypttab and this actually limits the utility of these files -
e.g. why should the user need to authenticate as root just because
there are options in the fstab entry for the device? In fact, this
behavior is so annoying that people use e.g. the 'users' option for a
device which opens up extra security hole insofar that remote users
can now access the device.
Therefore, change the behavior so we treat fstab and crypttab devices
just like any other device, e.g. require the filesystem and
filesystem-system authorizations.
We also introduce options comment=udisks-auth [1] and x-udisks-auth
that can be used in fstab and crypttab respectively to force admin
authentication (the polkit actions in question are filesystem-fstab
encrypted-unlock-crypttab).
This has also been hooked up in GNOME's Disks
application
http://people.freedesktop.org/~david/udisks-fstab-option-udisks-auth.png
http://people.freedesktop.org/~david/udisks-crypttab-option-x-udisks-auth.png
Also mention this behavior in the udisks(8) man page.
Note that if a device is mounted at start-up, a user just can't
willy-nilly unmount the device - by default, he will still need to
authenticate as root in order to unmount the device (the polkit
action in question is 'filesystem-unmount-others')
[1] : this will be changed to x-udisks-auth when a future util-linux
release supporting arbitrary x-* options lands
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 18 Jan 2012 20:59:31 +0000 (15:59 -0500)]
Just call it "fstab device" instead of "non-user fstab device"
Signed-off-by: David Zeuthen <davidz@redhat.com>
Martin Pitt [Wed, 18 Jan 2012 06:26:13 +0000 (07:26 +0100)]
Add some safe and useful ntfs-3g allowed mount options
David Zeuthen [Tue, 17 Jan 2012 21:59:58 +0000 (16:59 -0500)]
Remove invalid assertion that all managed mounts are in /media
This is not true as we support mounting items referenced in /etc/fstab.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 10 Jan 2012 21:31:17 +0000 (16:31 -0500)]
Don't set Drive:Ejectable to TRUE for floppy drives
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 10 Jan 2012 21:15:48 +0000 (16:15 -0500)]
Adjust sort-key calculation
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 10 Jan 2012 21:01:35 +0000 (16:01 -0500)]
Fix up how Drive:SortKey and Drive:TimeMediaDetected is calculated
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 6 Jan 2012 20:08:56 +0000 (15:08 -0500)]
Add work-around for kernel devices not going on media removal if mounted
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 5 Jan 2012 19:56:28 +0000 (14:56 -0500)]
Use reading/writing instead of backup/restore in authentication prompts
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 5 Jan 2012 19:09:07 +0000 (14:09 -0500)]
Add convenience routines to get block devices by label and uuid
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 5 Jan 2012 18:30:56 +0000 (13:30 -0500)]
Fix escaping issues when handling LABEL= and UUID= entries in fstab and crypttab
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 12 Dec 2011 15:14:12 +0000 (10:14 -0500)]
Nuke debug spew
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 9 Dec 2011 19:30:50 +0000 (14:30 -0500)]
Add some more thumb drive matches
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 9 Dec 2011 19:22:07 +0000 (14:22 -0500)]
Support new ID_DRIVE_THUMB udev property and 'thumb' media type
This will also set Drive:Removable to FALSE to override what the
hardware (wrongly) reports. So also add new Drive:Ejectable property
since we still want to display an Eject icon for thumb drives.
Use the icon 'media-removable' for drives of this type.
http://people.freedesktop.org/~david/gdu2-ID_DRIVE_THUMB-support.png
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 9 Dec 2011 17:08:35 +0000 (12:08 -0500)]
Force ReadOnly to TRUE for optical drives
At least do this until the kernel sets the 'ro' property properly.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 7 Dec 2011 21:19:13 +0000 (16:19 -0500)]
Fix loop device cleanup if backing device is not known
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 7 Dec 2011 21:09:03 +0000 (16:09 -0500)]
Don't fail LoopSetup() because fstat(2) fails
With this change, this method now works with fds from a GVfs FUSE
mount (where not even uid 0 may open files on).
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 7 Dec 2011 16:59:35 +0000 (11:59 -0500)]
Add Encrypted.ChangePassphrase() method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 7 Dec 2011 16:59:07 +0000 (11:59 -0500)]
Fix up docs
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 5 Dec 2011 18:16:54 +0000 (13:16 -0500)]
Add Block.OpenForRestore() method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 1 Dec 2011 17:19:04 +0000 (12:19 -0500)]
Add Block.OpenForBackup() method used for creating disk images
See
http://people.freedesktop.org/~david/gdu2-create-disk-image-1.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-2.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-3.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-4.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-5.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-6.png
http://people.freedesktop.org/~david/gdu2-create-disk-image-7.png
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 29 Nov 2011 21:59:33 +0000 (16:59 -0500)]
Don't cleanup loop devices if backing device has major 0
E.g. NFS mounts.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 29 Nov 2011 21:59:18 +0000 (16:59 -0500)]
Fix compile warning
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 29 Nov 2011 14:43:44 +0000 (09:43 -0500)]
UDisksCleanup: Nuke error handling from user-level API
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 29 Nov 2011 14:19:29 +0000 (09:19 -0500)]
Greatly simplify cleanup logic and routines
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 28 Nov 2011 15:32:51 +0000 (10:32 -0500)]
Use /etc/bash_completion.d, not /etc/profile.d for bash completion script
Also use _DATA, not _SCRIPTS, to ensure that the installed file is not
executable.
Signed-off-by: David Zeuthen <davidz@redhat.com>
Martin Pitt [Mon, 7 Nov 2011 06:08:11 +0000 (07:08 +0100)]
integration tests: small simplification
Introduce "no_options" constant for an empty options variant.
David Zeuthen [Wed, 23 Nov 2011 14:55:00 +0000 (09:55 -0500)]
Use non-deprecated mutex and thread primitives in GLib 2.31 and later
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 22 Nov 2011 18:48:50 +0000 (13:48 -0500)]
Add new Job:Objects property and set it
This can be used to render spinners, e.g.
http://people.freedesktop.org/~david/gdu2-spinners.png
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 22 Nov 2011 15:28:29 +0000 (10:28 -0500)]
Don't distribute generated sources in tarballs
... instead, insist that these sources are regenerated at build time.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 22 Nov 2011 15:01:47 +0000 (10:01 -0500)]
Update misc files
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 22 Nov 2011 14:55:39 +0000 (09:55 -0500)]
Move udisks man page to section 8
... from section 7. This is to avoid conflicts with the udisks(7) man
page from udisks 1.x.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 21 Nov 2011 18:36:22 +0000 (13:36 -0500)]
Add shell-snippet showing how to update the ATA smart data
gdbus call --system --dest org.freedesktop.UDisks2 \
--object-path /org/freedesktop/UDisks2/drives/WDC_WD1002FAEX_00Y9A0_WD_WCAW30039835 \
--method org.freedesktop.UDisks2.Drive.Ata.SmartUpdate \
"{'atasmart_blob': <'$ATA_SMART_BLOB'>}"
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 16 Nov 2011 21:27:22 +0000 (16:27 -0500)]
Add new Block:HintPartitionable property
This can be used in the UI to determine if a "Format Disk" (to
e.g. create a partition table) button should be offered.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 16 Nov 2011 16:10:38 +0000 (11:10 -0500)]
Add ReadOnly property to the Block interface
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 15 Nov 2011 20:13:31 +0000 (15:13 -0500)]
Wipe device before formatting it
It's better to not trust various mkfs tools etc. to get it right.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 15 Nov 2011 20:04:42 +0000 (15:04 -0500)]
Actually wait for the right object when formatting a block device
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 14 Nov 2011 20:35:11 +0000 (15:35 -0500)]
Allow user to manipulate partitions of a loop device if he set it up
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 14 Nov 2011 19:28:22 +0000 (14:28 -0500)]
Use LO_FLAGS_PARTSCAN when setting up a loop device
Actually, use its value of 8 until 3.2 has been out long enough for us
to require it to be available and installed when compiling the code.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 14 Nov 2011 17:03:46 +0000 (12:03 -0500)]
Pass UDisksObject so $(udisks2.device) is properly expanded
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 14 Nov 2011 16:44:58 +0000 (11:44 -0500)]
Allow creating partition tables with Block.Format()
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 10 Nov 2011 19:36:36 +0000 (14:36 -0500)]
Support creating dos partitions
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 10 Nov 2011 18:08:54 +0000 (13:08 -0500)]
UDisksClient: add get_partitions() convenience method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 18:35:46 +0000 (13:35 -0500)]
Don't support dos partitions of type 0x00
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 18:35:13 +0000 (13:35 -0500)]
For dos partitions of type 0, output 0x00, not 0x0
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 18:17:19 +0000 (13:17 -0500)]
Refuse to change partition type to an extended partition
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 16:52:44 +0000 (11:52 -0500)]
Introduce UDisksPartitionTypeInfoFlags
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 16:40:22 +0000 (11:40 -0500)]
When creating a partition, ensure size is a multiple of 512
Also account for parted(8) expecting the @end parameter to be
inclusive...
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 16:18:46 +0000 (11:18 -0500)]
When creating a GPT partition, don't insist that the size is divisible by 1 MiB
... only insist that the partition start is. Also take the last 33
LBAs into account - this is not space we can use.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 9 Nov 2011 16:01:17 +0000 (11:01 -0500)]
Fix ref count bug
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 8 Nov 2011 18:13:56 +0000 (13:13 -0500)]
Add the concept of partition table subtypes
This makes it possible to clean up the UI like this
http://people.freedesktop.org/~david/gdu2-partition-table-subtypes.png
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 7 Nov 2011 20:19:11 +0000 (15:19 -0500)]
Add PartitionTable.CreatePartition() method
Right now it only works for 'gpt' (because parted(8) expects different
formats depending on the partition table type (!)) and it doesn't yet
set the partition type. Apart from that it works great.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 7 Nov 2011 20:16:39 +0000 (15:16 -0500)]
Synthesize 'change' uevent when deleting a partition
Otherwise the udev database is not updated...
Signed-off-by: David Zeuthen <davidz@redhat.com>
Martin Pitt [Mon, 7 Nov 2011 06:06:13 +0000 (07:06 +0100)]
integration tests: Test loop device creation/teardown
Test in read/write and readonly modes.
Martin Pitt [Sat, 5 Nov 2011 23:22:12 +0000 (19:22 -0400)]
integration tests: Test LUKS
Martin Pitt [Sat, 5 Nov 2011 19:57:26 +0000 (15:57 -0400)]
Add integration test suite
This is the beginning of porting udisks1's "tests/run" integration test suite.
When running the suite from the source tree, this tests the daemon and library
from source tree; falls back on the installed system ones otherwise.
Currently covered are:
- Creation and label handling on all supported file systems (including swap
and empty)
- cleanup after drive removal without unmount
- UDisksDrive properties
- picks up changes from command line tools like mkfs
- correct file/dir/mountpoint permissions
- SMART
David Zeuthen [Thu, 3 Nov 2011 20:52:03 +0000 (16:52 -0400)]
Call the right completion method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 3 Nov 2011 18:04:36 +0000 (14:04 -0400)]
Add a Partition.Delete() method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Thu, 3 Nov 2011 12:44:47 +0000 (08:44 -0400)]
Export a Drive.Removable property
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 21:41:29 +0000 (17:41 -0400)]
Add Btrfs to list of known filesystems
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 21:16:20 +0000 (17:16 -0400)]
Handle the odd case when there is no partition table for a partition
This can happen when a device is ejected and it changes size to
zero. In that situation, old stale partition objects may be lying
around.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 20:23:25 +0000 (16:23 -0400)]
Add support for creating encrypted devices in Block.Format()
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 20:22:09 +0000 (16:22 -0400)]
Use luks-${UUID} instead of LUKS-udisks2-${UUID} to match the rest of the OS
In particular, systemd uses luks-${UUID}
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 18:02:57 +0000 (14:02 -0400)]
Support option 'take-ownership' in the Block.Format() method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 2 Nov 2011 15:32:34 +0000 (11:32 -0400)]
Rename Block.CreateFilesystem to Block.Format
... and other misc changes.
Signed-off-by: David Zeuthen <davidz@redhat.com>
Martin Pitt [Tue, 25 Oct 2011 06:37:33 +0000 (08:37 +0200)]
Bug 41931 — Add file system creation support
For now this only supports the "label" option. Further features (LUKS, taking
ownership, etc.) will be delivered in future patches.
We still have a rather weird bug here: When udisksd calls mkfs.ext4, there is
no change uevent on the block device when it finishes, thus udev and udisks are
missing the newly created file system. This does not happen when calling
mkfs.ext4 from a shell. Work around this by triggering a change uevent after
mkfs.ext4 ourselves.
https://bugs.freedesktop.org/show_bug.cgi?id=41931
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 31 Oct 2011 20:28:54 +0000 (16:28 -0400)]
Return "Unknown" for unknown partition types
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 31 Oct 2011 19:57:14 +0000 (15:57 -0400)]
Partition: Add Set{Name,Type,Flags} methods and implement for dos and gpt
Using the parted(8), sfdisk(8) and sgdisk(8) commands.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 31 Oct 2011 16:31:43 +0000 (12:31 -0400)]
UDisksClient: rename some methods and parameters
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sun, 30 Oct 2011 16:25:55 +0000 (12:25 -0400)]
Export the dev_t instead of major/minor for block devices
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sun, 30 Oct 2011 16:00:20 +0000 (12:00 -0400)]
UDisksClient: Include partition flags in get_partition_info()
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sat, 29 Oct 2011 15:28:52 +0000 (11:28 -0400)]
Ensure 'dos' partition types always have two digits, e.g. 0x0c, not 0xc
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 28 Oct 2011 18:40:53 +0000 (14:40 -0400)]
Use a uint64 for partition flags instead of a string
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 28 Oct 2011 16:36:08 +0000 (12:36 -0400)]
UDisksClient: add get_partition_info() method
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 28 Oct 2011 15:29:32 +0000 (11:29 -0400)]
For serial numbers, only use ID_SERIAL_SHORT, never ID_SERIAL
In the case that no real serial number exists, ID_SERIAL is made up
from vendor/model e.g. ST360021_A-0:0 or MediaTek_DVD-ROM_1309. We
don't want to convey such lies.
On the other hand, we should be guaranteed if ID_SERIAL_SHORT is set,
then it is something we can actually trust to be somewhat unique.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 20:19:02 +0000 (16:19 -0400)]
udisksctl: When setting up a loop device, don't free fd-list too soon
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 20:13:35 +0000 (16:13 -0400)]
Initialize variable to NULL
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 18:59:07 +0000 (14:59 -0400)]
UDisksClient: make settle() method emit any outstanding changes
... and also document this.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 18:27:32 +0000 (14:27 -0400)]
Don't leak job objects
In particular, since each job has a GCancellable, this meant that we
were leaking an eventfd per job.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 17:32:31 +0000 (13:32 -0400)]
Use new /dev/loop-control device to create loop devices
See
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=
770fe30a46a12b6fb6b63fbe1737654d28e84844
for the commit that introduced this feature. This is available as of
Linux 3.1.
Second, since the kernel now (virtually) supports an unlimited number
of loop devices, allow local active users to do this without having to
authenticate.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Wed, 26 Oct 2011 17:13:55 +0000 (13:13 -0400)]
Nuke debug spew
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 25 Oct 2011 20:42:09 +0000 (16:42 -0400)]
Handle the case where blkid(8) detects both a filesystem and a partition table
The F15 x86_64 live cd is both an ISO filesystem and a partition
table, presumably on purpose:
# blkid -o udev -p ~davidz/Downloads/Fedora-15-x86_64-Live-Desktop.iso
ID_FS_LABEL=Fedora-15-x86_64-Live-Desktop.is
ID_FS_LABEL_ENC=Fedora-15-x86_64-Live-Desktop.is
ID_FS_TYPE=iso9660
ID_FS_USAGE=filesystem
ID_PART_TABLE_TYPE=dos
If just used on an optical disc this presents no problem per se
because the kernel does not create partitions. However, if you dd the
image to a USB stick, the kernel will in fact recognize the partition
table and create a partition for it:
# blkid -o udev -p /dev/sdb
ID_FS_LABEL=Fedora-15-x86_64-Live-Desktop.is
ID_FS_LABEL_ENC=Fedora-15-x86_64-Live-Desktop.is
ID_FS_TYPE=iso9660
ID_FS_USAGE=filesystem
ID_PART_TABLE_TYPE=dos
# blkid -o udev -p /dev/sdb1
ID_FS_LABEL=Fedora-15-x86_64-Live-Desktop.is
ID_FS_LABEL_ENC=Fedora-15-x86_64-Live-Desktop.is
ID_FS_TYPE=iso9660
ID_FS_USAGE=filesystem
ID_PART_TABLE_TYPE=dos
ID_PART_ENTRY_SCHEME=dos
ID_PART_ENTRY_TYPE=0x17
ID_PART_ENTRY_FLAGS=0x80
ID_PART_ENTRY_NUMBER=1
ID_PART_ENTRY_OFFSET=0
ID_PART_ENTRY_SIZE=1161216
ID_PART_ENTRY_DISK=8:16
There is nothing, per se, wrong with this ... except that applications
(such as udisks and GNOME sitting on top) need to be careful and not
e.g. offer *two icons* for the iso9660 filesystem (one for sdb and one
for sdb1).
The fix is fortunately simple - we only allow a 'disk' block device to
be a PartitionTable or a Filesystem and not both at the same time. The
way we do this is to look at whether the kernel has already
partitioned the device or not (and this works only because the kernel
partiitoner guarantees that partition kobjects are created prior to
announcing the 'disk' block device).
Also stress in the docs that applications shouldn't be looking at
IdUsage or IdType (a direct copy of whatever blkid(8) puts in the udev
database) and that they instead should be looking at what D-Bus
interfaces are implemented.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 24 Oct 2011 16:58:10 +0000 (12:58 -0400)]
Add properties to identify extended and logical dos partitions
This allows to remove special-casing of dos and 0x05, 0x0f and 0x85
partition types in each client.
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Mon, 24 Oct 2011 14:47:13 +0000 (10:47 -0400)]
Use glib top-level includes
glib HEAD complains about this now
/usr/include/glib-2.0/glib/gtypes.h:28:2: error: #error "Only <glib.h> can be included directly."
Also nuke some trailing whitespace.
Signed-off-by: David Zeuthen <davidz@redhat.com>
Martin Pitt [Mon, 17 Oct 2011 09:53:39 +0000 (11:53 +0200)]
Bug 41862 — Add label changing for previously supported file systems
Add support for label changing on vfat, ntfs, xfs, reiserfs, and nilfs2 file
systems.
Instead of having another huge if-then-else ladder as in udisks1, use a lookup
table _fs_info. For now this contains the commands for label changing and
erasing, but will be extended in the future to contain the commands for
creating and checking file systems.
https://bugs.freedesktop.org/show_bug.cgi?id=41862
David Zeuthen [Sun, 23 Oct 2011 16:26:59 +0000 (12:26 -0400)]
Update internal man page links
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sun, 23 Oct 2011 16:26:45 +0000 (12:26 -0400)]
Add links to udisks(7) man page) for the Hint* properties
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sun, 23 Oct 2011 16:06:26 +0000 (12:06 -0400)]
Move all utility functions to the UDisksClient type
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Sat, 22 Oct 2011 14:31:04 +0000 (10:31 -0400)]
Remove unused functions for doc sections file
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Fri, 21 Oct 2011 19:12:23 +0000 (15:12 -0400)]
Add Partition{,Table} interfaces and rely on blkid(8) partition properties
Signed-off-by: David Zeuthen <davidz@redhat.com>
David Zeuthen [Tue, 18 Oct 2011 17:51:47 +0000 (13:51 -0400)]
udisks: Also accept a --no-debug option and use it for activation
Otherwise syslog gets spammed via D-Bus' system bus activation helper
via systemd (got it? got it!).
Signed-off-by: David Zeuthen <davidz@redhat.com>