| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Mount operation for encrypted volumes consists from two udisks2 calls
(unlock and mount). Volume monitor is able to mount already unlocked
volumes. However password prompt doesn't disappear if the volume is
unlocked externally and any reaction causes mount operation failure.
The volume may be unlocked externally by e.g. project Tang, see
https://github.com/latchset/tang for more details. Abort mount
operation and continue with mounting in such situations.
https://bugzilla.gnome.org/show_bug.cgi?id=763890
|
|
|
|
|
|
|
|
|
|
| |
The eject/stop operation fails currently with "Cannot eject drive in
use: Encrypted device /dev/sdb1 is unlocked" if any volume is unlocked
and isn't mounted. The volume monitor is able to mount already unlocked
volumes, so it should be able to eject unlocked volumes also. Lock the
unlocked volumes if there are any before eject/stop operation.
https://bugzilla.gnome.org/show_bug.cgi?id=763890
|
| |
|
|
|
|
|
|
|
|
|
| |
Rereading of trash dirs is pretty expensive and unfortunately it is
performed more often with recent changes. Check modification time of
each trash dir in order to avoid needless rereading, because simple
query_info call is much faster than enumeration.
https://bugzilla.gnome.org/show_bug.cgi?id=711459
|
|
|
|
|
|
|
|
|
|
|
|
| |
Patches to use G_FILE_ATTRIBUTE_TRASH_ITEM_COUNT were recently proposed
for Nautilus and GTK+. We have to be sure that the item count is valid
and therefore we have to rescan the trash dirs before. Otherwise the
count is 0 if enumeration isn't called before, which is obviously wrong.
Similarly, trash dirs have to be rescanned before other operations
(i.e. read, delete, pull), otherwise the operations may fail with
G_IO_ERROR_NOT_FOUND if enumeration wasn't called before.
https://bugzilla.gnome.org/show_bug.cgi?id=711459
|
|
|
|
| |
Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
As we can only start HouseArrest for applications that have
UIFileSharingEnabled turned on, there's absolutely no need to list the
other applications.
See https://github.com/libimobiledevice/libimobiledevice/issues/193
https://bugzilla.gnome.org/show_bug.cgi?id=763606
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Only sideloaded applications (development versions for example) can use
the VendContainer method of accessing application documents.
Always use the VendDocuments method instead.
Note that we'll also re-root the Documents folder as the root of the app
container is inaccessible ("EPERM") to avoid users fiddling with it.
See https://github.com/libimobiledevice/libimobiledevice/issues/193#issuecomment-95280639
https://bugzilla.gnome.org/show_bug.cgi?id=763606
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=749639
|
|
|
|
|
|
| |
As can happen with some iOS and libplist combinations.
https://bugzilla.gnome.org/show_bug.cgi?id=749639
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There's no interesting data for users in afc:/// so it's best to show
iDevices through MTP instead. The only reason we'd want to use this
mount is if we had support for accessing the iTunes database.
Unfortunately, this database needs to be encrypted, and we don't know
how to encrypt it.
Even then, we would probably want the default AFC service to be hidden
so only applications access it, and not users.
https://bugzilla.gnome.org/show_bug.cgi?id=749639
|
|
|
|
|
|
|
| |
There's no interesting data for users in afc:/// so it's best to show
iDevices through MTP instead. Stop hiding iDevices through the MTP.
https://bugzilla.gnome.org/show_bug.cgi?id=749639
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
iDevices had a list of thumbnails available through afc:/// for photos
stored in the DCIM/ sub-directory, but the code to handle this was
always pretty fragile, relying on the iOS version to know which of a
number of different ways the thumbnails could be called or where they
would be stored.
It's easier to create the thumbnails ourselves from the JPG files, or
even access photos through PTP so that thumbnails are readily
accessible and in a way that doesn't change with each iOS release.
https://bugzilla.gnome.org/show_bug.cgi?id=749639
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Fuse removes its signals handlers in the middle of exit procedure,
before vfs_destroy() call. Consequenlty fuse daemon may be killed
instead of proper termination if signal is recieved. Reimplement
fuse_main helper using lowlevel API in order to ignore new signals
during the whole exit procedure.
https://bugzilla.gnome.org/show_bug.cgi?id=762250
|
|
|
|
|
|
|
|
|
| |
There may be a race when exiting between a fuse main loop and
a subthread's main loop. Therefore call g_main_loop_quit always
from vfs_destroy on the main thread and wait there for a termination
of the subthread to avoid the race.
https://bugzilla.gnome.org/show_bug.cgi?id=762250
|
| |
|
| |
|
|
|
|
|
|
| |
Result from g_vfs_get_default shouldn't be freed as per documentation.
https://bugzilla.gnome.org/show_bug.cgi?id=762250
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Unmount notification is sent for each volume when stopping drive,
which is wrong. It may lead to data loss, because user may unplug
the device after the first notification, before the device is really
stopped. It should be handled similary as an eject operation and
send the notification only once for the drive.
https://bugzilla.gnome.org/show_bug.cgi?id=760523
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
Commit c014b64 was pushed to prevent race between unmount reply and
retry timer. Result of mount operation reply should be stored if
unmount operation is in progress, however it isn't, because the
conditional statement is always true. Fix the condition accordingly.
https://bugzilla.gnome.org/show_bug.cgi?id=678555
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Wrong file size is cached if query_info is executed during write
operation, because dir cache is invalidated on open_for_write. Wrong
file size might cause that e.g. POSIX editors fails to read file
content over fuse daemon. Invalidate the cache also on close_write
to fix this issue.
https://bugzilla.gnome.org/show_bug.cgi?id=760159
|
|
|
|
|
|
|
|
|
| |
Unwanted attributes are sent from some backends (e.g. ftp), because
g_file_info_copy_info is used and thus source attribute matcher is replaced.
Apply attribute mask before sending as it is done in gvfsjobeneumerate
to fix this issue.
https://bugzilla.gnome.org/show_bug.cgi?id=745212
|
| |
|
| |
|