platform/upstream/efl.git
7 years agoefl readme - add libunwind to highly recommended for err bt
Carsten Haitzler (Rasterman) [Fri, 28 Apr 2017 10:01:26 +0000 (19:01 +0900)]
efl readme - add libunwind to highly recommended for err bt

7 years agoefl readme - remove liblz4 from recommendations
Carsten Haitzler (Rasterman) [Fri, 28 Apr 2017 09:59:44 +0000 (18:59 +0900)]
efl readme - remove liblz4 from recommendations

it's probably not a good idea as it adds some dirty pages for a tiny
lib that we already supply in our source. if your distro demands you
do this - fine. but i wouldnt recommend this highly.

7 years agoefl readme - vlc really shouldn't be recommended list if gst is def
Carsten Haitzler (Rasterman) [Fri, 28 Apr 2017 09:56:17 +0000 (18:56 +0900)]
efl readme - vlc really shouldn't be recommended list if gst is def

recommending vlc and having gts default is probably not good. choose
one or the other ... but not both. so remove from the recommended list
just ebcause it's causing excess deps by packagers apparently

7 years agoevas font code - build on older freetypes without new definitions
Carsten Haitzler (Rasterman) [Fri, 28 Apr 2017 09:37:40 +0000 (18:37 +0900)]
evas font code - build on older freetypes without new definitions

use the actual #35 as:

in freetype and it cant change without breaking abi...

@fix

7 years agothread fence code - move elsewhere in file to avoid conflicts
Carsten Haitzler (Rasterman) [Fri, 28 Apr 2017 02:24:48 +0000 (11:24 +0900)]
thread fence code - move elsewhere in file to avoid conflicts

7 years agoedje_cc: Fix warning caused by conflict between local & global variable
Jeeyong Um [Fri, 28 Apr 2017 02:54:11 +0000 (11:54 +0900)]
edje_cc: Fix warning caused by conflict between local & global variable

7 years agoecore_evas: fix test in the hide callback in Windows module
Vincent Torri [Wed, 26 Apr 2017 20:39:39 +0000 (22:39 +0200)]
ecore_evas: fix test in the hide callback in Windows module

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
7 years agoedje: Fix resource leak
Chris Michael [Thu, 27 Apr 2017 15:05:30 +0000 (11:05 -0400)]
edje: Fix resource leak

The function parse_str returns allocated memory which should be freed
before we exit this function.

Fix Coverity CID1374644

Signed-off-by: Chris Michael <cp.michael@samsung.com>
7 years agoedje: Fix resource leak
Chris Michael [Thu, 27 Apr 2017 15:02:51 +0000 (11:02 -0400)]
edje: Fix resource leak

The function parse_str returns allocated memory which should be freed
before we exit this function.

Fix Coverity CID1374647

Signed-off-by: Chris Michael <cp.michael@samsung.com>
7 years agowayland engines: Don't store evas in engine info
Derek Foreman [Wed, 26 Apr 2017 19:20:47 +0000 (14:20 -0500)]
wayland engines: Don't store evas in engine info

We don't need it anymore.

7 years agowayland_shm: Stop using the evas directly to block for async render
Derek Foreman [Wed, 26 Apr 2017 19:16:37 +0000 (14:16 -0500)]
wayland_shm: Stop using the evas directly to block for async render

Use the new fence function instead.

This makes Cedric less unhappy.

7 years agoevas: Add a function to block for render thread completion
Derek Foreman [Wed, 26 Apr 2017 18:44:33 +0000 (13:44 -0500)]
evas: Add a function to block for render thread completion

This is needed by dmabuf engine fallback when it realizes it locally
allocated a buffer, has been rendering to it, but the compositor can't use
it.

So the engine copies its buffer contents into a new wl_shm buffer and
continues from there - however we need to make sure the async renderer
has finished first, so we don't copy a partial buffer.

This allows us to block for all previously submit actions in the render
queue to complete.

7 years agogl_drm:Rename import_simple_dmabuf to gl_import_simple_dmabuf
Derek Foreman [Tue, 25 Apr 2017 20:57:43 +0000 (15:57 -0500)]
gl_drm:Rename import_simple_dmabuf to gl_import_simple_dmabuf

There's also an ecore_drm2 function to import dmabuf as fb, so it's a
little less confusing to be a little more explicit here.

7 years agoecore_evas_wayland: Make local functions static
Derek Foreman [Tue, 25 Apr 2017 16:24:53 +0000 (11:24 -0500)]
ecore_evas_wayland: Make local functions static

7 years agoecore_evas_wayland: Remove unused functions
Derek Foreman [Tue, 25 Apr 2017 16:18:22 +0000 (11:18 -0500)]
ecore_evas_wayland: Remove unused functions

7 years agoecore_evas_wayland: Remove unnecessary function prototypes
Derek Foreman [Tue, 25 Apr 2017 16:00:34 +0000 (11:00 -0500)]
ecore_evas_wayland: Remove unnecessary function prototypes

7 years agoecore_evas_wayland_common: Reorder functions
Derek Foreman [Tue, 25 Apr 2017 15:59:33 +0000 (10:59 -0500)]
ecore_evas_wayland_common: Reorder functions

Try to put these in order so we don't need prototypes

No functional changes

7 years agoecore_evas: Add doxygen for ecore_evas_gl_x11_options_new()
Bryce Harrington [Wed, 26 Apr 2017 17:34:52 +0000 (13:34 -0400)]
ecore_evas: Add doxygen for ecore_evas_gl_x11_options_new()

Reviewers: devilhorns

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4828

7 years agoeo: move all the key_ properties to C
Daniel Kolesa [Wed, 26 Apr 2017 14:18:49 +0000 (16:18 +0200)]
eo: move all the key_ properties to C

7 years agoeo: move wref_add/wref_del to C
Daniel Kolesa [Wed, 26 Apr 2017 13:01:44 +0000 (15:01 +0200)]
eo: move wref_add/wref_del to C

7 years agoeo: move future_link to C
Daniel Kolesa [Wed, 26 Apr 2017 12:25:37 +0000 (14:25 +0200)]
eo: move future_link to C

7 years agoecore: efl_promise: remove superfluous NULL check
Stefan Schmidt [Wed, 26 Apr 2017 13:12:53 +0000 (15:12 +0200)]
ecore: efl_promise: remove superfluous NULL check

f is already checked against NULL in the eina array macro. No need to check
again.

CID: 1362730

7 years agoeldbus: check return of dbus_signature_iter_next()
Stefan Schmidt [Wed, 26 Apr 2017 12:49:37 +0000 (14:49 +0200)]
eldbus: check return of dbus_signature_iter_next()

CID: 1341856

7 years agoevas_filter: initialize was_empty to false to have a defined state
Stefan Schmidt [Wed, 26 Apr 2017 12:16:14 +0000 (14:16 +0200)]
evas_filter: initialize was_empty to false to have a defined state

It is used further below in an if clause but might not have been
initialized at that point.

CID: 1374276

7 years agoevas: gl_filters: fix leak of evas image
Stefan Schmidt [Wed, 26 Apr 2017 11:56:59 +0000 (13:56 +0200)]
evas: gl_filters: fix leak of evas image

Make sure we free the allocated use_map evas image before leaving scope.

CID: 1374275

7 years agoevas: fix logic in evas_events.c
Jaeun Choi [Wed, 26 Apr 2017 08:16:16 +0000 (17:16 +0900)]
evas: fix logic in evas_events.c

Summary:
when collecting the objects under a mouse pointer,
evas uses the geometry of an object to decide
whether the mouse pointer is inside the area of the object,
which is inappropriate for a mapped object.
so mapped objects don't receive mouse events when they should.
this patch fixes the issue.

Test Plan: A sample code will be added as comments

Reviewers: jpeg, raster

Subscribers: cedric, eunue

Differential Revision: https://phab.enlightenment.org/D4826

7 years agoelm_fileselector: made fileselector styling more flexible
Vyacheslav Reutskiy [Wed, 26 Apr 2017 08:16:01 +0000 (11:16 +0300)]
elm_fileselector: made fileselector styling more flexible

Previously the progressbar in fileselector use hardcoded style name
"wheel", that made unpossible to create different style for
fileselector. This commit made it possible.

@fix

7 years agoelm_cursor: add style name for sub cursor
Vyacheslav Reutskiy [Wed, 26 Apr 2017 05:57:22 +0000 (08:57 +0300)]
elm_cursor: add style name for sub cursor

Missing style name for sub cursor call the next error output:

ERR<24193>:elementary lib/elementary/els_cursor.c:734 elm_cursor_theme()
Could not apply the theme to the cursor style=(null)

Also this error call extra recalc for cursor hot spots.

Fixes T5408
@fix

7 years agoEcore_Evas: use window states and event property to fix ecore_evas_fullscreen_set()
Vincent Torri [Wed, 26 Apr 2017 05:32:11 +0000 (07:32 +0200)]
Ecore_Evas: use window states and event property to fix ecore_evas_fullscreen_set()

This implies that elm_win_fullscreen_set() is also fixed

@fix

7 years agoEcore_Win32: add Property event API
Vincent Torri [Wed, 26 Apr 2017 05:17:04 +0000 (07:17 +0200)]
Ecore_Win32: add Property event API

@feature

7 years agoEcore_Win32: add ecore_win32_window_state_get() API to retrieve a window states
Vincent Torri [Wed, 26 Apr 2017 05:10:21 +0000 (07:10 +0200)]
Ecore_Win32: add ecore_win32_window_state_get() API to retrieve a window states

@feature

7 years agoEcore_Win32: update fullscreen state
Vincent Torri [Wed, 26 Apr 2017 04:56:42 +0000 (06:56 +0200)]
Ecore_Win32: update fullscreen state

7 years agoEcore_Win32: follow a bit more X code
Vincent Torri [Wed, 26 Apr 2017 04:53:57 +0000 (06:53 +0200)]
Ecore_Win32: follow a bit more X code

7 years agoEcore_Evas: clean up Windows module
Vincent Torri [Wed, 26 Apr 2017 04:47:13 +0000 (06:47 +0200)]
Ecore_Evas: clean up Windows module

7 years agoEcore_Win32: fix information message
Vincent Torri [Wed, 26 Apr 2017 04:19:20 +0000 (06:19 +0200)]
Ecore_Win32: fix information message

7 years agoEcore_Win32: fix spelling
Vincent Torri [Wed, 26 Apr 2017 04:17:08 +0000 (06:17 +0200)]
Ecore_Win32: fix spelling

7 years agoedje_cc: Convert anchors to beta feature (require -beta when compiled)
Jeeyong Um [Wed, 26 Apr 2017 02:33:47 +0000 (11:33 +0900)]
edje_cc: Convert anchors to beta feature (require -beta when compiled)

7 years agoedje_cc: Allow to set float position for anchors
Jeeyong Um [Tue, 25 Apr 2017 09:29:26 +0000 (18:29 +0900)]
edje_cc: Allow to set float position for anchors

7 years agoedje_cc: Allow part to be anchored to GROUP and the part named as "GROUP"
Jeeyong Um [Tue, 25 Apr 2017 09:27:52 +0000 (18:27 +0900)]
edje_cc: Allow part to be anchored to GROUP and the part named as "GROUP"

7 years agoedje_cc: Fix warning caused by missing const modifier
Jeeyong Um [Tue, 25 Apr 2017 09:03:04 +0000 (18:03 +0900)]
edje_cc: Fix warning caused by missing const modifier

7 years agoefl_ui_focus_manager: check for null return from node_get
Bryce Harrington [Tue, 25 Apr 2017 23:29:43 +0000 (16:29 -0700)]
efl_ui_focus_manager: check for null return from node_get

Summary:
node_get() can return NULL sometimes.  Add a missing check, and add dox
for the function to document the error return.

>>>     CID 1374434:  Null pointer dereferences  (NULL_RETURNS)
>>>     Assigning: "node" = null return value from "node_get".

@fix CID1374434

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4824

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
7 years agoexamples/evas: fix spelling
Bryce Harrington [Tue, 25 Apr 2017 23:28:15 +0000 (16:28 -0700)]
examples/evas: fix spelling

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4807

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
7 years agoedje_cc: introduce "Anchors" - easy way to set parts relationship
Jeeyong Um [Tue, 25 Apr 2017 23:25:11 +0000 (16:25 -0700)]
edje_cc: introduce "Anchors" - easy way to set parts relationship

Test Plan: compile src/examples/edc-anchors.c and run

Reviewers: zmike, raster, cedric, jpeg

Reviewed By: raster, jpeg

Subscribers: raster, barbieri, zmike, SanghyeonLee, taxi2se, Jaehyun_Cho

Differential Revision: https://phab.enlightenment.org/D4775

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
7 years agoevas: do not rely on Evas canvas for Evas Ector engine backend.
Cedric BAIL [Tue, 25 Apr 2017 23:15:37 +0000 (16:15 -0700)]
evas: do not rely on Evas canvas for Evas Ector engine backend.

7 years agoelm_box/table: set the order directly when a container is registered
Marcel Hollerbach [Tue, 25 Apr 2017 15:04:50 +0000 (17:04 +0200)]
elm_box/table: set the order directly when a container is registered

The following situation:
 - A Box in a naviframe, with n children.
 - All added children register to the focus graph with the box as parent,
   order gets set correctly.
 - Naviframe hides this item, so box property tree unfocusable gets set
   to true, it gets unregistered from the focus graph, even every single
   child gets unregistered.
 - The item gets shown - every child and the table are getting
   registered again.
 - Order is not set again, since the box does not get changed
 - Order of the children is mixed up.

This should fix this case since the order is flushed every time the box
gets registered.

7 years agoefl_ui_focus_manager: safe fields from the node
Marcel Hollerbach [Tue, 25 Apr 2017 14:51:30 +0000 (16:51 +0200)]
efl_ui_focus_manager: safe fields from the node

There could be the case that the item gets freed due to some handling in
a event handler of the event EFL_UI_FOCUS_MANAGER_EVENT_FOCUSED.
So the code now sets the node to NULL after the event is called and
saves the fields that are rfom use later.

7 years agoefl.net: rename Ip.Address to Ip_Address.
Gustavo Sverzut Barbieri [Tue, 25 Apr 2017 14:16:40 +0000 (11:16 -0300)]
efl.net: rename Ip.Address to Ip_Address.

It's unlikely that we'll have other stuff under Ip namespace, also not
that likely to have other than Ip Addresses (to invert it to
Address.Ip), thus make a toplevel entry Ip_Address as suggested by
DaveMDS.

7 years agoelm_entry: reset 'icon' and 'end' swallow when style changed
Vyacheslav Reutskiy [Tue, 25 Apr 2017 13:34:03 +0000 (16:34 +0300)]
elm_entry: reset 'icon' and 'end' swallow when style changed

Fixes T5397

7 years agoefl/gfx: Remove unused function
Jean-Philippe Andre [Tue, 25 Apr 2017 12:21:54 +0000 (21:21 +0900)]
efl/gfx: Remove unused function

This fixes a warning with -Wunused-function

7 years agoevas: Fix crash in idle_flush
Jean-Philippe Andre [Tue, 25 Apr 2017 12:18:22 +0000 (21:18 +0900)]
evas: Fix crash in idle_flush

As reported by raster on the ML. The pointer passed to the
engines is not an engine output but an Efl_Canvas_Output.

7 years agoitests: elm_focus: rename local variable to avoid clash with main()
Stefan Schmidt [Tue, 25 Apr 2017 09:54:25 +0000 (11:54 +0200)]
itests: elm_focus: rename local variable to avoid clash with main()

tests/elementary/elm_test_focus_sub.c:88:25: warning: â€˜main’ is usually a function [-Wmain]

7 years agodocs: efl_ui_focus_manager: document last missing items in this new file
Stefan Schmidt [Tue, 25 Apr 2017 09:32:36 +0000 (11:32 +0200)]
docs: efl_ui_focus_manager: document last missing items in this new file

7 years agodocs: elm_widget: document new foc us_register method
Stefan Schmidt [Tue, 25 Apr 2017 09:32:03 +0000 (11:32 +0200)]
docs: elm_widget: document new foc us_register method

7 years agodocs: efl_ui_focus_manager: document focus relations struct members
Stefan Schmidt [Tue, 25 Apr 2017 09:13:12 +0000 (11:13 +0200)]
docs: efl_ui_focus_manager: document focus relations struct members

7 years agoevas render: Fix issue with map render
Jean-Philippe Andre [Tue, 25 Apr 2017 05:34:03 +0000 (14:34 +0900)]
evas render: Fix issue with map render

Test case:
  elementary_test -to "Gesture Layer"

Just move the pictures around (eg. to the bottom). They could
disappear entirely.

This is because the geometry used was based on the smart
object "bounding box" rather than the mapped output.

@fix

7 years agoedje_cc: Fix external param smart parser
Jeeyong Um [Tue, 25 Apr 2017 00:41:46 +0000 (09:41 +0900)]
edje_cc: Fix external param smart parser

7 years agoevas: remove engine.data.output.
Cedric BAIL [Mon, 24 Apr 2017 22:09:25 +0000 (15:09 -0700)]
evas: remove engine.data.output.

This is the first step toward handling multi output. This patch
remove engine.data.output from Evas structure and use an Eina_List
for it instead. It also start moving code around to fetch an output
or an engine context (which are the same at the moment, but will be
split in a later patch).

7 years agoevas: start setting up output independently of the engine.
Cedric BAIL [Mon, 24 Apr 2017 21:53:52 +0000 (14:53 -0700)]
evas: start setting up output independently of the engine.

7 years agoevas: evas_render actually need to access that structure.
Cedric BAIL [Wed, 19 Apr 2017 22:48:22 +0000 (15:48 -0700)]
evas: evas_render actually need to access that structure.

7 years agoevas: Fix grammar of which vs. that
Bryce Harrington [Mon, 24 Apr 2017 12:28:31 +0000 (08:28 -0400)]
evas: Fix grammar of which vs. that

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4809

7 years agoevas: Fix grammar for which vs. that
Bryce Harrington [Mon, 24 Apr 2017 12:27:49 +0000 (08:27 -0400)]
evas: Fix grammar for which vs. that

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4808

7 years agoecore_evas: Fix grammar for at/to/for/on/which/etc.
Bryce Harrington [Mon, 24 Apr 2017 12:26:18 +0000 (08:26 -0400)]
ecore_evas: Fix grammar for at/to/for/on/which/etc.

Summary:
Fixes some grammar confusion for in that/this, that/which, to/at,
to/for, at/by, etc.

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4806

7 years agoFix spelling of 'alpha channel'
Bryce Harrington [Mon, 24 Apr 2017 12:25:22 +0000 (08:25 -0400)]
Fix spelling of 'alpha channel'

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4817

7 years agoevas: round glyph's advance before adding it to pen_x
Youngbok Shin [Sun, 23 Apr 2017 06:22:42 +0000 (09:22 +0300)]
evas: round glyph's advance before adding it to pen_x

Summary:
Rounding the sum of glyph's advance could cause inconsistency of
each glyph's positions. When Evas enables Harfbuzz library,
Each glyph's position has to be handled by only nearby glyphs.
But, currently, totally unrelated glyph's advacne could change
other glyphs positions.

ex) 1. "connect."
    2. "Tap here to connect."

You can see different gap between "c" and "o" of word "connect".
It should be same even if there was a different text before the word "connect".
@fix

Test Plan: N/A

Reviewers: raster, herdsman, jpeg

Reviewed By: raster

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D4782

7 years agotests: Fix make check with build dir
Jean-Philippe Andre [Mon, 24 Apr 2017 02:56:15 +0000 (11:56 +0900)]
tests: Fix make check with build dir

This changes how the eo files are compiled for the focus test
case, and adds the include folder with -I.

7 years agoefreetd - make ability to not connect to efreetd non-critical
Carsten Haitzler (Rasterman) [Sun, 23 Apr 2017 01:58:07 +0000 (10:58 +0900)]
efreetd - make ability to not connect to efreetd non-critical

it's an error we can survive, so make it that level

7 years agoefreetd - make efreetd use eio and not ecore_file_monitor for monitoring
Carsten Haitzler (Rasterman) [Sun, 23 Apr 2017 01:33:41 +0000 (10:33 +0900)]
efreetd - make efreetd use eio and not ecore_file_monitor for monitoring

this should fix T5130 - freebsd's efreetd polling regularly

@fix

7 years agoelementary: fix typo on comment
Shinwoo Kim [Sun, 23 Apr 2017 08:02:14 +0000 (17:02 +0900)]
elementary: fix typo on comment

change "an widget" to "a widget".

7 years agoefl_ui_focus_manager: change the new order that gets set
Marcel Hollerbach [Sat, 22 Apr 2017 08:41:07 +0000 (10:41 +0200)]
efl_ui_focus_manager: change the new order that gets set

Lets say there is a box with the following ordered children:
|Button|Box|Button|Box|Button| the two boxes do not have any children
at the time of the setup. The logic of the order_update will only order
the children like that:
|Button|Button|Button| Which is correct by that time, the two boxes dont
have any children.
Now the two boxes are also getting children, the order will not
selfupdate or anything so the logical chain would be:
|Button|Button|Button|Box|Box|. Which is wrong. To solve that the
manager keeps the order that got set last, and reapplies the order again
if something gets added to the parent where the order was set.

This should fix strange next / prev operations in ephoto.

7 years agoecore - fix epoll ifdefs and case handling
Carsten Haitzler (Rasterman) [Sat, 22 Apr 2017 03:20:06 +0000 (12:20 +0900)]
ecore - fix epoll ifdefs and case handling

7 years agoecore_drm2: Fix ticking when atomic mode set is enabled
Derek Foreman [Fri, 21 Apr 2017 21:37:18 +0000 (16:37 -0500)]
ecore_drm2: Fix ticking when atomic mode set is enabled

Putting the PAGE_FLIP_EVENT flag on the set rotation request resulted
in an extra event on the drm device fd that screwed up page flipping
badly from that point on.

@fix

7 years agotests: add test for large eina tiler rect
Mike Blumenkrantz [Fri, 21 Apr 2017 20:14:57 +0000 (16:14 -0400)]
tests: add test for large eina tiler rect

7 years agoeina_tiler: use int size matching Eina_Rectangle internally to avoid overflow
Mike Blumenkrantz [Fri, 21 Apr 2017 20:14:57 +0000 (16:14 -0400)]
eina_tiler: use int size matching Eina_Rectangle internally to avoid overflow

since this code's creation it seems that the internal int size was set to use
short in order to micro-optimize memory usage, while the api function parameters
used Eina_Rectangle which had a larger int size. when initializing the internal
rect struct, this would lead to overflows which resulted in broken tilers which
returned iterators with no valid rects after having valid rects added

test case: run weston-subsurfaces

@fix

7 years agoelm_win: do not require mouse buttons to be pressed to activate csd move/resize
Mike Blumenkrantz [Fri, 21 Apr 2017 20:14:57 +0000 (16:14 -0400)]
elm_win: do not require mouse buttons to be pressed to activate csd move/resize

the api function requires this, but the unified handler for api+edje handler does
not, since edje singals are deferred and the button which triggered the move
may be released before the signal is processed

ref ea7bbfe47dc1df5f51725be05eadf7ebce1d8361
@fix

7 years agoecore-wl2: move ecore_wl2_display_inputs_get() into ecore_wl2_display.c
Mike Blumenkrantz [Fri, 21 Apr 2017 20:14:57 +0000 (16:14 -0400)]
ecore-wl2: move ecore_wl2_display_inputs_get() into ecore_wl2_display.c

7 years agoeina: some architecture don't like prefetch on NULL apparently.
Cedric Bail [Fri, 21 Apr 2017 18:29:38 +0000 (11:29 -0700)]
eina: some architecture don't like prefetch on NULL apparently.

7 years agoeo: hide dbg_info from eo files
Daniel Kolesa [Fri, 21 Apr 2017 15:58:38 +0000 (17:58 +0200)]
eo: hide dbg_info from eo files

Continues b780cf2af299ce17be260486bb7ab64badddbaec.

7 years agoeo: move event related structures and callback to C
Daniel Kolesa [Fri, 21 Apr 2017 13:47:12 +0000 (15:47 +0200)]
eo: move event related structures and callback to C

We don't need to keep this in eo files anymore because the APIs
using them are now fully in C. This also allows removal of the
event callback builtin from Eolian.

7 years agoelm test - unused vars - fix warning and remove them
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:49:29 +0000 (16:49 +0900)]
elm test - unused vars - fix warning and remove them

7 years agoevas gl filter blur - fix uninitialized vare warning for sum
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:40:36 +0000 (16:40 +0900)]
evas gl filter blur - fix uninitialized vare warning for sum

7 years agogl filter blur - fix warning about possible uninit vars by initting them
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:39:20 +0000 (16:39 +0900)]
gl filter blur - fix warning about possible uninit vars by initting them

7 years agoeio monitor - remove unused vars and fix warning
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:39:06 +0000 (16:39 +0900)]
eio monitor - remove unused vars and fix warning

7 years agoeina mmap - fix warnings thanks to missing include
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:38:40 +0000 (16:38 +0900)]
eina mmap - fix warnings thanks to missing include

add include for eina_file.h

7 years agoeina debug - fix include of eina_file.h that was missing
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:38:18 +0000 (16:38 +0900)]
eina debug - fix include of eina_file.h that was missing

fixes new warnings

7 years agoefl_io_closer - fix warning about error print
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:23:16 +0000 (16:23 +0900)]
efl_io_closer - fix warning about error print

lazy removal of fcntl without fixing up error print... fixed.

7 years agoecore_pipe - fix warning of sign vs unsigned comparison
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 07:19:43 +0000 (16:19 +0900)]
ecore_pipe - fix warning of sign vs unsigned comparison

7 years agoelm widget - fix weak reference messup that causes segv's
Carsten Haitzler (Rasterman) [Fri, 21 Apr 2017 06:58:04 +0000 (15:58 +0900)]
elm widget - fix weak reference messup that causes segv's

welm widget weak refered the logical parent slot but never unreffed
the weak ref - ever. this should fix that. in fact it does. one crash
less with:

elementary_test -to "icon standard"

@fix

7 years agoevas_async_events: Integrated Ecore_Pipe in evas_async_events
Ivan Furs [Fri, 21 Apr 2017 02:30:05 +0000 (11:30 +0900)]
evas_async_events: Integrated Ecore_Pipe in evas_async_events

Summary: Integrated Ecore_Pipe in evas_async_events

Reviewers: vtorri, artem.popov, cedric, jpeg, raster, #eflete

Reviewed By: cedric

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D4622

7 years agofocus: Review EO file and adjust doc a bit
Jean-Philippe Andre [Fri, 21 Apr 2017 05:24:45 +0000 (14:24 +0900)]
focus: Review EO file and adjust doc a bit

I was reading it to understand this new focus system. So I also
made some cosmetic changes here and there in the file: wrap, fix
a few typos, add missing docs. I'm not pretending that the doc
is perfect now. This really was just a personal review.

7 years agofocus: Fix shadow variable warning
Jean-Philippe Andre [Fri, 21 Apr 2017 05:52:30 +0000 (14:52 +0900)]
focus: Fix shadow variable warning

7 years agoecore wl2: Fix compilation
Jean-Philippe Andre [Fri, 21 Apr 2017 01:08:52 +0000 (10:08 +0900)]
ecore wl2: Fix compilation

7 years agoelm_image: Improve test case from previous patch
Jean-Philippe Andre [Thu, 20 Apr 2017 09:59:02 +0000 (18:59 +0900)]
elm_image: Improve test case from previous patch

This refers to 331bf9f74683aad72415e007af14c3d4713f112b
See also, for 1.19: 4eda12b5e1d6e968fb0b057f11c280a87916e93f

7 years agoecore_wl2: Don't crash when drag and drop is unsupported by compositor
Derek Foreman [Thu, 20 Apr 2017 19:40:29 +0000 (14:40 -0500)]
ecore_wl2: Don't crash when drag and drop is unsupported by compositor

Check for presence of data device manager so we don't crash if it's not
there.

7 years agoeo: move unbindable event APIs to C
Daniel Kolesa [Thu, 20 Apr 2017 16:24:38 +0000 (18:24 +0200)]
eo: move unbindable event APIs to C

In a few classes, this requires some manual expansion. This should
not break anything but it's also fairly ugly; a better solution
would be appreciated, for now we do this.

Similar changes will be done to a few other Efl.Object APIs as
well at later point.

7 years agoeo: move Efl_Event definition back to C
Daniel Kolesa [Thu, 20 Apr 2017 14:14:18 +0000 (16:14 +0200)]
eo: move Efl_Event definition back to C

It's not used anywhere in eo files and has a void pointer.

7 years agoefl_ui_focus_manager: select the widget with the smallest distance
Marcel Hollerbach [Thu, 20 Apr 2017 08:15:53 +0000 (10:15 +0200)]
efl_ui_focus_manager: select the widget with the smallest distance

this improves the correctness-feeling when moving focus through a
elementary application.

7 years agoelm_interface_scrollable: add focus override
Marcel Hollerbach [Wed, 19 Apr 2017 12:11:02 +0000 (14:11 +0200)]
elm_interface_scrollable: add focus override

it seems this override got lost somewhere

7 years agoelementary: fix focus tests
Marcel Hollerbach [Wed, 19 Apr 2017 12:10:43 +0000 (14:10 +0200)]
elementary: fix focus tests

7 years agoelm_list: migrate to new focus system
Marcel Hollerbach [Wed, 11 Jan 2017 15:17:05 +0000 (16:17 +0100)]
elm_list: migrate to new focus system

7 years agofix .eo file breaks
Marcel Hollerbach [Mon, 2 Jan 2017 10:42:35 +0000 (11:42 +0100)]
fix .eo file breaks