This branch fixes several issues with ecore_drm:
In particular, it address the issue of not properly restoring the tty
on shutdown, saves the current tty keyboard mode and properly restores
that on shutdown, we now get the VT from systemd (if available) and
properly setup the VT.
@fix
The functions to take or release a session no longer accept session
name as a parameter, so fix calls to those functions
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
The session take/release functions no longer need the session name
paramater, so remove that.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
The session take/release functions no longer require the name of the
session, so remove the function parameter
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: As the launcher code now uses some fields from the
Ecore_Drm_Device, we need to free the device After the launcher has
been disconnected.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: These functions have been moved to the private header file
now, and thus this file is no longer needed.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: Since we removed the need for the logind header file, let's
remove that from the build order.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Ecore_Drm_Device
Summary: This moves storing the session name and VT fields inside the
device structure. This also adds a field so we can store the keyboard
mode of the tty, and restore it on close.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: This moves the logind function prototypes to the private
header and thus removes the need for a separate logind header file.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: This moves the VT switch handling code to the tty file, and
sets the proper flags on the VT when opened.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: This fixes a few issues with the logind code such as: Getting
the VT from systemd (if available), Properly setup the VT using proper
'open' flags, adding a 'restore' function to reset the tty properly,
and handle take/release session.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: This cleans up the tty code a bit, and makes it the
responsibility of the tty code to handle VT switching signals
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary: This sets the default session to NULL when a device is
created. Also a formatting fix
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
take/release session control inside dbus functions
Summary: These changes move the storage of session name to inside the
Ecore_Drm_Device structure and remove calls to take/release session
control to outside of the dbus code
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Some complex examples of masking with mapped smart objects
would fail miserably, rendering the object without any mask,
and/or showing the mask itself somewhere in white color...
Summary:
If the file size of RLE compressed image is bigger than original image,
BMP loader doesn't work as well.
@fix
Reviewers: Hermet, cedric
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1892
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
This @fix the parsing of dirs from the xdg env vars. Now always remove
the leading / char from the paths.
This was causing lots of trouble on my system, where XDG_DATA_DIRS is:
/usr/local/share/enlightenment:/usr/local/share:/usr/local/share/:/usr/share/
At first /usr/local/share was added 2 times in the list, one with the / and one
witout, causing a double lookup for each file.
Secondly the icon cache was totally unusable as the cached paths ended up
as: /usr/share//icons/Mint-X/places/32/folder.svg. The double / in there
was making the cache lookup to fail and anways return the biggest icon
available. Causing a big system slowdown whe searching for icons.
As a bonus the function now use eina_str_split instead of the custom splitting
code that require a bad special handling for the last item.
actually have a grabbed button
Summary: If we don't have an actual grabbed mouse button, then don't
send a mouse up event. Caught this while adding e_grabinput support to
Enlightenment.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
This adjusts the clip to match the mask image data.
To be fair, I'm not sure in which situation those two
don't match well. This patch fixes a rare crash.
The memory usage graph was going up and to the right!
I was told this is always a good thing!
... maybe not this time :)
Hopefully I didn't forget a case. An intense session of
genlist scrolling with masks all over the place and masks
of masks didn't show any glitch, crash or memory leak.
The main difference between 1.12 and 1.13 memory foot print is actually
related to this two pointer to mask. I am wondering if there is not an
issue here also has we do have a duplicated pointer. We have prev_mask
and mask in both cur and prev state of an Evas_Object, but only mask
and prev_mask from the cur state seems to be accessed.
If we can remove two pointers from those state, we should have a decent
win in expedite benchmark. Hopefully 300KB to win there (Close to half
the additional cost in memory).