platform/upstream/efl.git
6 years agoeolian_gen: remove reference to elm_entry because it is not installed
Sungtaek Hong [Wed, 24 Jan 2018 09:51:25 +0000 (18:51 +0900)]
eolian_gen: remove reference to elm_entry because it is not installed

Summary:
elm_entry.eo is not installed to system directory like
         "/usr/local/share/eolian/include/elementary-1", and
         eolian_gen tries to refer elm_entry.eo which results in failure.

Test Plan:
Create any eo class file which inherits Efl.Ui.Layout.
           eolian_gen eo_file.eo
           observe eolian_gen finishes

Reviewers: jpeg, herdsman, woohyun, Jaehyun_Cho, cedric

Subscribers: conr2d, id213sin, JongminLee

Differential Revision: https://phab.enlightenment.org/D5760

6 years agoefl_ui_focus_layer: avoid a call to NULL
YeongJong Lee [Wed, 24 Jan 2018 09:49:04 +0000 (18:49 +0900)]
efl_ui_focus_layer: avoid a call to NULL

Summary:
efl_ui_focus_layer_enable_set(obj, EINA_FALSE) can be called before
registered_manager assigned

Test Plan:
1. EINA_LOG_LEVELS=eo:2 elementary_test -to menu
2. terminate the elemetary_test
3. check that there is no focus_manager warning about a call to NULL

Reviewers: bu5hm4n

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5759

6 years agoefl_ui_legacy: add new interface to indicate legacy widget
Sungtaek Hong [Wed, 24 Jan 2018 06:50:58 +0000 (15:50 +0900)]
efl_ui_legacy: add new interface to indicate legacy widget

Summary:
For now, how to check whether a widget is legacy or not
is to check flags in private data or static flag, which is set
during elm_legacy_add.
If Efl.Ui.Legacy interface is added, it can be easilly checked
by efl_isa(obj, EFL_UI_LEGACY_INTERFACE)

Reviewers: woohyun, jpeg, cedric, Jaehyun_Cho

Subscribers: conr2d, cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5748

6 years agoefl_ui_win: Avoid a call to NULL
YeongJong Lee [Wed, 24 Jan 2018 06:36:08 +0000 (15:36 +0900)]
efl_ui_win: Avoid a call to NULL

Summary:
parent_get and smart_parent_get are called in parent_widget_get
this also remove the duplicated code

Reviewers: jpeg

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5757

6 years agoeina: Make return doxygen consistent
Bryce Harrington [Wed, 24 Jan 2018 02:35:01 +0000 (11:35 +0900)]
eina: Make return doxygen consistent

Summary:
Ensure @return defines error returns consistently.  In several cases the
errors were explained in the body but not mentioned in the @return docs.

Drop redundant return value documentation for clarity.  Some routines
were defining the return values both in @return and in the doxygen body.

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5756

6 years agopyolian test: just rename the main eolian state
Dave Andreoli [Tue, 23 Jan 2018 20:31:31 +0000 (21:31 +0100)]
pyolian test: just rename the main eolian state

6 years agoefl_ui_focus_manager: listen to deletion for regular and logical
Marcel Hollerbach [Tue, 23 Jan 2018 19:58:59 +0000 (20:58 +0100)]
efl_ui_focus_manager: listen to deletion for regular and logical

otherwise we might exit with a load of errors

6 years agoelm_genlist: prepare the item when its set
Marcel Hollerbach [Tue, 23 Jan 2018 19:58:59 +0000 (20:58 +0100)]
elm_genlist: prepare the item when its set

so the border element is set accordingly on the parent manager which
then can be picked up for focus relations.

ref T6453

6 years agoefl_ui_focus_composition: only rely on manager
Marcel Hollerbach [Tue, 23 Jan 2018 19:56:49 +0000 (20:56 +0100)]
efl_ui_focus_composition: only rely on manager

for the case of a custom manager this resulted in wrong behaviour.

6 years agoelput: Fix null pointer dereference
Chris Michael [Tue, 23 Jan 2018 12:00:19 +0000 (07:00 -0500)]
elput: Fix null pointer dereference

Coverity reports this as a null pointer dereference, so check for
'seat' existence before comparing with strcmp

Fixes CID1385369

Signed-off-by: Chris Michael <cp.michael@samsung.com>
6 years agoelementary config: Fix to use ELEMENTARY_BASE_DIR for configure path
Youngbok Shin [Mon, 22 Jan 2018 12:01:27 +0000 (21:01 +0900)]
elementary config: Fix to use ELEMENTARY_BASE_DIR for configure path

Summary:
"--with-elementary-base-dir" option was ignored by recent patches on elm_config.
The macro is being used in elm_theme. It should be syncronized.
The default value of the macro is ".elementary".

@fix

Test Plan: N/A

Reviewers: raster, cedric, jpeg

Reviewed By: jpeg

Differential Revision: https://phab.enlightenment.org/D5755

6 years agoecore-x: add more null checks for functions
Mike Blumenkrantz [Mon, 22 Jan 2018 22:34:20 +0000 (17:34 -0500)]
ecore-x: add more null checks for functions

@fix

6 years agoecore-x: perform internal shutdown on io error if callback is set
Mike Blumenkrantz [Mon, 22 Jan 2018 21:09:00 +0000 (16:09 -0500)]
ecore-x: perform internal shutdown on io error if callback is set

avoid crashing in callback when referencing dead display or handling events

@fix

6 years agoecore-wl2: correctly translate spacebar keyname into key events
Mike Blumenkrantz [Mon, 22 Jan 2018 18:02:44 +0000 (13:02 -0500)]
ecore-wl2: correctly translate spacebar keyname into key events

fix T6620

@fix

6 years agoecore_drm2: fix seat matching when checking for a device with a null seat
Chris Michael [Mon, 22 Jan 2018 16:26:13 +0000 (11:26 -0500)]
ecore_drm2: fix seat matching when checking for a device with a null seat

Summary: Check to make sure that the seat from the matching device is still null.

Test Plan: Make sure that devices are not being rejected while on seat1. This depends on another patch

Reviewers: ManMower, devilhorns

Reviewed By: devilhorns

Subscribers: ManMower, devilhorns, cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5711

6 years agoecore_drm2: Only use a virtual terminal on seat0
Chris Michael [Mon, 22 Jan 2018 16:25:15 +0000 (11:25 -0500)]
ecore_drm2: Only use a virtual terminal on seat0

Summary: Only seat0 supports virtual terminals.

Test Plan: Make sure enlightenment does not error out on non-seat0 due to a virtual terminal error. Enlightenment will not start yet on seat1 with this alone

Reviewers: ManMower, devilhorns

Reviewed By: devilhorns

Subscribers: ManMower, devilhorns, cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5710

6 years agogenlist: Avoid a call to NULL
Jean-Philippe Andre [Mon, 22 Jan 2018 10:10:48 +0000 (19:10 +0900)]
genlist: Avoid a call to NULL

elementary_test -to genlist
-> LOTS of warnings, when creating the genlist, as the items aren't
ready yet. This just avoids the WRN as recalc will happen later anyway.

6 years agoefl_selection_manager: fix compilation on osx after rename
Stefan Schmidt [Mon, 22 Jan 2018 08:31:35 +0000 (09:31 +0100)]
efl_selection_manager: fix compilation on osx after rename

In commit c91ae98fd3ecf532036b16499ffd521a2e91e2a3 data was renamed but
the ifdefed osx part forgotten to update.

6 years agoecore_evas: Check for null ptr deref
Myoungwoon Roy, Kim [Mon, 22 Jan 2018 09:28:24 +0000 (18:28 +0900)]
ecore_evas: Check for null ptr deref

Summary:
This patch fixes a tentative crash owing to ptr dereference.
Unlike ecore_evas_object_cursor_set and ecore_evas_object_cursor_device_set,
ecore_evas_cursor_set uses Ecore_Evas *ee before calling internal function which
internally checks null ptr dereference of Ecore_Evas *ee.

Test Plan: Executes test suite

Reviewers: cedric, raster, stefan, Jaehyun_Cho

Reviewed By: Jaehyun_Cho

Subscribers: jpeg

Differential Revision: https://phab.enlightenment.org/D5753

Change-Id: Id24aa10e044dc63c09b29aa0d29ddf52cd3f3e80

6 years agofix break of hoversels
Carsten Haitzler (Rasterman) [Sun, 21 Jan 2018 10:26:38 +0000 (19:26 +0900)]
fix break of hoversels

this fixes T6627

6 years agoelementary: fix compilation of efl_selection_manager on windows
Vincent Torri [Fri, 19 Jan 2018 21:36:08 +0000 (13:36 -0800)]
elementary: fix compilation of efl_selection_manager on windows

Test Plan: compilation

Reviewers: cedric

Subscribers: jpeg

Differential Revision: https://phab.enlightenment.org/D5751

Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
6 years agowayland: Reduce number of unneccessary display flushes
Derek Foreman [Thu, 18 Jan 2018 23:21:27 +0000 (17:21 -0600)]
wayland: Reduce number of unneccessary display flushes

b48781aa6c85c72f620c57cebff96580ddd67899 fixed multiple bugs where the
display wasn't flushed correctly, however it was a little overzealous.

Some of the flushes were added after calls that only updated internal
state, some in internal functions in which the caller was already going
to flush, and some were after wayland protocol calls that are double
buffered anyway and won't do anything until a following commit.

Also, I've removes at least one long standing flush where the recently
added flush is in a better location than the original.

6 years agowayland_shm: Clean up duplicate prototypes
Derek Foreman [Thu, 4 Jan 2018 22:49:56 +0000 (16:49 -0600)]
wayland_shm: Clean up duplicate prototypes

These were supposed to be removed when I moved the associated functions
into ecore_wl2.

6 years agowidget: Ooops! Fix focus break in previous patch.
Jean-Philippe Andre [Fri, 19 Jan 2018 08:48:15 +0000 (17:48 +0900)]
widget: Ooops! Fix focus break in previous patch.

Sorry... Weird that elementary_test worked fine. I might have badly
tested.

6 years agowidget: Move on_focus_update to Focus.Object
Jean-Philippe Andre [Fri, 19 Jan 2018 08:25:47 +0000 (17:25 +0900)]
widget: Move on_focus_update to Focus.Object

Following @taxi2se's recommendation. This is indeed a focus method, and
Widget already inherits from Focus.Object.

Ping @bu5hm4n who probably wants to adapt this further.

Ref T5363

6 years agowidget: Remove item from on_focus_update
Jean-Philippe Andre [Fri, 19 Jan 2018 08:13:07 +0000 (17:13 +0900)]
widget: Remove item from on_focus_update

It's legacy stuff, and always NULL.

6 years agoRevert "efl_ui_win : resolve ABI break"
Wonki Kim [Tue, 3 Apr 2018 05:24:01 +0000 (14:24 +0900)]
Revert "efl_ui_win : resolve ABI break"

This reverts commit 6319f09a18feced8c5df429df5cf9b66e77ff7fb.

6 years agowidget: Use enum for orientation_mode
Jean-Philippe Andre [Fri, 19 Jan 2018 06:55:19 +0000 (15:55 +0900)]
widget: Use enum for orientation_mode

This was proposed by Dave on the ML, I think it makes sense. Right now
the enum is just like the boolean, feature-wise, but it makes more sense
semantically (mode is not a bool) and allows for future extension (eg.
only apply orientation update for landscape vs. portrait modes).

There is absolutely zero testing for this in our existing codebase. Yay.

6 years agocxx: Fix a trivial warning
Jean-Philippe Andre [Fri, 19 Jan 2018 06:58:52 +0000 (15:58 +0900)]
cxx: Fix a trivial warning

6 years agoelementary: simplify Efl.Ui.Text.Factory test.
Cedric BAIL [Thu, 18 Jan 2018 18:48:43 +0000 (10:48 -0800)]
elementary: simplify Efl.Ui.Text.Factory test.

6 years agoelementary: simplify Efl.Ui.Text.Factory.Emoticons.
Cedric BAIL [Thu, 18 Jan 2018 18:48:21 +0000 (10:48 -0800)]
elementary: simplify Efl.Ui.Text.Factory.Emoticons.

6 years agoCanvas image: add Efl.Canvas.Text.Factory + use in Ui.Text
Daniel Hirt [Thu, 9 Nov 2017 15:53:20 +0000 (17:53 +0200)]
Canvas image: add Efl.Canvas.Text.Factory + use in Ui.Text

This interface has a simple 'create' method to create Efl.Canvas.Object
given a key.
This is used higher-up in Ui.Text in the next commit.

Ui text: add ability to set item factories

Added API to set an item factory object.
This is similar to the previous item providers (that worked with
callbacks).

You instantiate a factory object and set it on the Ui.Text object.
Each factory implements the "create" method from
Efl.Canvas.Text.Item_Factory.

This also includes 3 public factories (Image, Emoticon and Fallback):
  - Image factory: creates images from added entries (key strings)
  - Emoticon factory: creates emoticons by querying the theme
  - Fallback: creates image, then falls back to emoticon

If no factory is set, then the fallback (internal) factory is used.

See the added "Ui.text Item Factory" test in elementary_test for an
example of usage.

@feature

6 years agoEdje: remove item provider from eo api
Daniel Hirt [Tue, 7 Nov 2017 15:33:56 +0000 (17:33 +0200)]
Edje: remove item provider from eo api

6 years agoelm_win: move wayland display flush to after method calls
Mike Blumenkrantz [Thu, 18 Jan 2018 15:47:01 +0000 (10:47 -0500)]
elm_win: move wayland display flush to after method calls

ref b48781aa6c85c72f620c57cebff96580ddd67899

6 years agoeina: Clarify that Eina_Matrix_Type can describe any matrix
Bryce Harrington [Thu, 18 Jan 2018 13:19:08 +0000 (08:19 -0500)]
eina: Clarify that Eina_Matrix_Type can describe any matrix

Summary:
The Eina_Matrix_Type enum is returned by eina_matrix4_type_get and
eina_matrix2_type_get; it is not Matrix3-specific.  Update doxygen
accordingly.

Reviewers: devilhorns

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5744

6 years agoeina: Cleanup spelling/grammar/punctuation for Eina List
Bryce Harrington [Thu, 18 Jan 2018 13:16:52 +0000 (08:16 -0500)]
eina: Cleanup spelling/grammar/punctuation for Eina List

Summary:
This fixes some typos and misspellings, massages grammar in a few
places, tidys up a little whitespace, and fixes incorrect docs in a spot
or two.

Signed-off-by: Bryce Harrington <bryce@osg.samsung.com>
Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5745

6 years agoeolian: fix redefined function checks
Daniel Kolesa [Thu, 18 Jan 2018 11:49:05 +0000 (12:49 +0100)]
eolian: fix redefined function checks

For one, the hash has to be populated once per inheritance tree
and the skipping on already-validated funcs was interfering with
that as the function might have been validated in another
inheritance tree already.

Also, if a class appeared multiple times in an inheritance tree,
as is common with e.g. Efl.Object, it would get added into the
hash the first time and then checked against the second time,
which would result in a strange error message about the
function being redefined in its own class.

So now we prevent both cases from happening.

6 years agocnp: Rename selection data to content
Jean-Philippe Andre [Thu, 18 Jan 2018 09:39:47 +0000 (18:39 +0900)]
cnp: Rename selection data to content

This solves an issue with C# bindings where a field of a struct can not
have the same name as the struct itself. Go figure why.

6 years agoall: Simplify definition of EAPI
Vincent Torri [Thu, 18 Jan 2018 09:04:03 +0000 (18:04 +0900)]
all: Simplify definition of EAPI

This will help in the transition from Autotools to Meson. This has been
tested on Windows for which EFL_XXX_BUILD were first introduced.

6 years agowidget: Remove comment
Jean-Philippe Andre [Thu, 18 Jan 2018 08:40:21 +0000 (17:40 +0900)]
widget: Remove comment

6 years agoefl: Merge Flipable and Orientation
Jean-Philippe Andre [Thu, 18 Jan 2018 08:18:48 +0000 (17:18 +0900)]
efl: Merge Flipable and Orientation

No need to split: those two are used in all the same classes, since the
split between Orientation and Ui.Dir.

Note that the enum types remain in the main namespace.

6 years agowidget: Rename orientation_mode_disabled
Jean-Philippe Andre [Thu, 18 Jan 2018 08:13:59 +0000 (17:13 +0900)]
widget: Rename orientation_mode_disabled

Name is stupid and long, why not just invert this bool?

6 years agovg: Fix warning (unused function)
Jean-Philippe Andre [Thu, 18 Jan 2018 08:24:22 +0000 (17:24 +0900)]
vg: Fix warning (unused function)

6 years agoa11y: Remove import elm_general
Jean-Philippe Andre [Thu, 18 Jan 2018 07:31:20 +0000 (16:31 +0900)]
a11y: Remove import elm_general

Ref T5329

6 years agoui.text: Use proper EO type for cnp mode
Jean-Philippe Andre [Thu, 18 Jan 2018 07:19:01 +0000 (16:19 +0900)]
ui.text: Use proper EO type for cnp mode

It's not working. Just "fixing" the API for consistency.
Also, we're lacking a proper hint for "markup without images".

So I think EFL_SELECTION_FORMAT_MARKUP should be without images, while
EFL_SELECTION_FORMAT_MARKUP | EFL_SELECTION_FORMAT_IMAGE would allow
markup with images.

Ping @thiep.ha

Ref T5329

6 years agocnp: Use EO types to define legacy types
Jean-Philippe Andre [Thu, 18 Jan 2018 07:18:23 +0000 (16:18 +0900)]
cnp: Use EO types to define legacy types

They are the same.
Ref T5329

6 years agowin: Avoid a warning
Jean-Philippe Andre [Thu, 18 Jan 2018 07:20:21 +0000 (16:20 +0900)]
win: Avoid a warning

I still hate those types.
warning: case value not in enumerated type 'Efl_Ui_Win_Type' [-Wswitch]

6 years agoexamples: evas vg: fix the compilation of vg batman
Amitesh Singh [Thu, 18 Jan 2018 06:19:53 +0000 (15:19 +0900)]
examples: evas vg: fix the compilation of vg batman

6 years agowin: Remove unusable type "dnd"
Jean-Philippe Andre [Thu, 18 Jan 2018 05:20:49 +0000 (14:20 +0900)]
win: Remove unusable type "dnd"

Override mode is not available in the unified api. That's because you
can't do it in Wayland.

I still think those types are stupid and should all die: Win subclasses
handle them for you. Odd types can't be used anyway.

Ref T5322

6 years agoui.list: Fix shadow variable warnings
Jean-Philippe Andre [Thu, 18 Jan 2018 05:13:05 +0000 (14:13 +0900)]
ui.list: Fix shadow variable warnings

node2 didn't conflict, just making it consistent with nodedata2

6 years agowin: Remove "ioerr" event
Jean-Philippe Andre [Thu, 18 Jan 2018 05:06:45 +0000 (14:06 +0900)]
win: Remove "ioerr" event

Unless it's implemented for Wayland as well, AND provides more
information than a NULL event_info, I see no point in this being an EO
event. Keep legacy as-is: a smart callback only.

Also, minor cleanups to the EO file.

Ref T5322

6 years agoedje: Return "" as state for invalid parts
Jean-Philippe Andre [Thu, 18 Jan 2018 04:54:56 +0000 (13:54 +0900)]
edje: Return "" as state for invalid parts

Instead of "default". This is consistent with legacy, and allows for
some kind of error checking. Not returning NULL helps avoiding crashes.

6 years agoRevert "efl wl - fix segv where garbage ptr to comp surface is still there"
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
Revert "efl wl - fix segv where garbage ptr to comp surface is still there"

This reverts commit 574ef9a5e53ac2ecabc14fd02be7e55cd90a9b25.

this should be resolved by 863a6e159e99c5f69ecc7461106aba2102f1da11 and is
only a possible case when an unmapped surface was activated, something which
is prevented by spec and can only occur if bugs exist elsewhere

6 years agowayland: always flush client display when protocol sends are pending
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
wayland: always flush client display when protocol sends are pending

in the case where a connection was not actively rendering, there was nothing
which would trigger a display flush, leading to applications potentially
deadlocking

@fix

6 years agoefl-wl: only perform mouse-out operations for a seat if the mouse was "in"
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: only perform mouse-out operations for a seat if the mouse was "in"

this could lead to cases where the original application's cursor was permanently
lost

@fix

 #TheDisappointer

6 years agoefl-wl: ensure parent surface is mapped before attempting to revert activation
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: ensure parent surface is mapped before attempting to revert activation

6 years agoefl-wl: immediately unset a destroyed surface's cursor
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: immediately unset a destroyed surface's cursor

@fix

6 years agoefl-wl: enforce xdg-shell spec requiring a commit to an unmapped surface
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: enforce xdg-shell spec requiring a commit to an unmapped surface

a separate commit is required for a newly-unmapped surface in order for
the surface to be able to attach a buffer and be mapped

6 years agoefl-wl: only send keyboard enter and activation for mapped surfaces
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: only send keyboard enter and activation for mapped surfaces

setting active for unmapped surfaces bypasses the expected path for
unsetting activate during surface unmap if the surface is destroyed
while active and unmapped

@fix

6 years agoefl-wl: remove some broken logic for activating toplevel parents
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: remove some broken logic for activating toplevel parents

parents of parents should not be marked as activated...probably

@fix

6 years agoefl-wl: set event ON_HOLD flag when they are sent to a surface
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: set event ON_HOLD flag when they are sent to a surface

use a flag to indicate when processed events have been internally processed

@fix

6 years agoefl-wl: propagate surface activation back to parent if child is hidden
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: propagate surface activation back to parent if child is hidden

a toplevel child will always be stacked right above its parent

@fix

6 years agoefl-wl: send more mouse buttons to clients
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: send more mouse buttons to clients

@fix

6 years agoefl-wl: split case for toplevel vs popup during surface hide activate propagation
Mike Blumenkrantz [Wed, 17 Jan 2018 22:33:57 +0000 (17:33 -0500)]
efl-wl: split case for toplevel vs popup during surface hide activate propagation

this should have no functional differences due to spec, but make sure it reads
more clearly

6 years agoRevert "efl loop - provide efl namespace versions of begin/end locks on mainloop"
Carsten Haitzler (Rasterman) [Wed, 17 Jan 2018 12:30:49 +0000 (21:30 +0900)]
Revert "efl loop - provide efl namespace versions of begin/end locks on mainloop"

This reverts commit 76b837002eaea56b5ecb174bffe284012084dc74.

seems no one wants efl api's for this

6 years agoelm_test: Fix to call efl_add with parent
Jaehyun Cho [Wed, 17 Jan 2018 09:43:35 +0000 (18:43 +0900)]
elm_test: Fix to call efl_add with parent

To trace back to the main loop, call efl_add with parent.
(See https://phab.enlightenment.org/T6573)

6 years agoefl theme - fix bug in e init splash that would do hide anim 2x
Carsten Haitzler (Rasterman) [Wed, 17 Jan 2018 09:12:33 +0000 (18:12 +0900)]
efl theme - fix bug in e init splash that would do hide anim 2x

this should fix a "go black and stay black for a bit" bug...

@fix T6619

6 years agoelput - fix crash on shutdown where input.lib was NULL
Carsten Haitzler (Rasterman) [Wed, 17 Jan 2018 05:27:45 +0000 (14:27 +0900)]
elput - fix crash on shutdown where input.lib was NULL

this was null checked everywhere except in 1 place. fix. now more
crash on shutdown now

6 years agoexamples: Simplify & fix thread example
Vincent Torri [Wed, 17 Jan 2018 08:31:51 +0000 (17:31 +0900)]
examples: Simplify & fix thread example

This fixes the example for Windows, where apparently it ran into a lot
of errors with spinlock. Note that on Linux we have such errors too
(both before and after the patch):
EINA ERROR: 'Invalid argument' on lock 0x7ffd218af410
EINA ERROR: 'Invalid argument' on unlock 0x7ffd218af410

Patch by @vtorri

6 years agoglview: Remove trailing newline in ERR
Jean-Philippe Andre [Wed, 17 Jan 2018 06:23:52 +0000 (15:23 +0900)]
glview: Remove trailing newline in ERR

It's not needed.

6 years agoevas gl: Properly verify that Evas GL works
Jean-Philippe Andre [Wed, 17 Jan 2018 06:17:59 +0000 (15:17 +0900)]
evas gl: Properly verify that Evas GL works

For SW engine we need to verify that OSMesa is present. The patch
fb048e73120b39095cc87aa4f8d7f3d397a93c56 broke the logic.

Tested by temporarily removing OSMesa from my system.

Fixes T6617 (again)

6 years agoefl/vg: Refactor the svg path handling and interpolation.
subhransu mohanty [Wed, 17 Jan 2018 05:20:15 +0000 (14:20 +0900)]
efl/vg: Refactor the svg path handling and interpolation.

Summary:
This patch is taken from the tizen branch

1. Take care when the locale is different than POSIX.
See: strtod, strtol

2. Interpolate  all the property of a shape.

Reviewers: jpeg

Subscribers: cedric, jpeg

Differential Revision: https://phab.enlightenment.org/D5740

6 years agowin: Fix standard window bg with custom themes
Jean-Philippe Andre [Wed, 17 Jan 2018 05:12:34 +0000 (14:12 +0900)]
win: Fix standard window bg with custom themes

Ref T6579
See also D5740

6 years agoRevert "elm: disable interface theme loading"
Jean-Philippe Andre [Wed, 17 Jan 2018 04:59:22 +0000 (13:59 +0900)]
Revert "elm: disable interface theme loading"

This reverts commit 3d07b90461818804ed51b6dcaed9f05c0d0155bb.

Some commits should address the issues mentioned in T6579
  959571b9955babc40367cf3a634dddba7a7cdcc2
  fe346d2ee23eb0d9fca0544f6732c709ec3a4b8b

But to be honest I am still not sure exactly what the issue was?

Ref T6579

6 years agoRevert "warning remove - comment out unused vars"
Jean-Philippe Andre [Wed, 17 Jan 2018 05:03:00 +0000 (14:03 +0900)]
Revert "warning remove - comment out unused vars"

This reverts commit bbac253f11bd7a7a2320c52ab295c1ad9919bb6a.

6 years agoefl-wl: add lots of null checks to egl init
Mike Blumenkrantz [Tue, 16 Jan 2018 23:37:15 +0000 (18:37 -0500)]
efl-wl: add lots of null checks to egl init

ref T6617

6 years agoefl wl - fix segv where garbage ptr to comp surface is still there
Carsten Haitzler (Rasterman) [Tue, 16 Jan 2018 05:50:31 +0000 (14:50 +0900)]
efl wl - fix segv where garbage ptr to comp surface is still there

comp surface destruction doesnt clear the pointer to one in the comp
struct that then is garbage later. fix.

6 years agoefl-wl: multiply axis events by 10 to send normalized values
Mike Blumenkrantz [Fri, 12 Jan 2018 21:02:40 +0000 (16:02 -0500)]
efl-wl: multiply axis events by 10 to send normalized values

ref T6443

6 years agoefl-wl: unset cursor role on existing cursor surface if new cursor is set
Mike Blumenkrantz [Fri, 12 Jan 2018 20:49:06 +0000 (15:49 -0500)]
efl-wl: unset cursor role on existing cursor surface if new cursor is set

this avoids killing clients which replace cursors

@fix

6 years agoefl-wl: return NULL without error when non-extracted parent is found
Mike Blumenkrantz [Thu, 11 Jan 2018 21:24:38 +0000 (16:24 -0500)]
efl-wl: return NULL without error when non-extracted parent is found

in  efl_wl_extracted_surface_extracted_parent_get()

6 years agoefl-wl: do not modify extracted surface visibility during commit
Mike Blumenkrantz [Wed, 10 Jan 2018 19:20:19 +0000 (14:20 -0500)]
efl-wl: do not modify extracted surface visibility during commit

extracted surfaces are wholly managed externally, so this should not be modified

ref T6443

@fix

6 years agowayland: add xdg-shell vfinal support for client-side
Mike Blumenkrantz [Mon, 8 Jan 2018 18:42:11 +0000 (13:42 -0500)]
wayland: add xdg-shell vfinal support for client-side

@feature

6 years agoefl-wl: move to xdg-shell vfinal
Mike Blumenkrantz [Mon, 8 Jan 2018 17:53:57 +0000 (12:53 -0500)]
efl-wl: move to xdg-shell vfinal

@feature

6 years agoefl_ui_focus_manager: NULL out fields correctly 48/174448/1
Marcel Hollerbach [Tue, 27 Mar 2018 09:33:51 +0000 (11:33 +0200)]
efl_ui_focus_manager: NULL out fields correctly

Change-Id: I97906e1a0cc30e2aca6ed8fdcb3cb3814676b2f2
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
6 years agoevas textblock: Fix wrong bidi decision caused by empty text item 03/174403/1
Youngbok Shin [Sat, 31 Mar 2018 08:01:01 +0000 (17:01 +0900)]
evas textblock: Fix wrong bidi decision caused by empty text item

An empty text item can be generated in the following squences.

ex) [LATIN]<preedit>[ARAB]</preedit>
1. Create a text item for LATIN script.
2. Create a text item for applying new format stage(<preedit>) to the next text.
3. Create a text item for ARAB script.
=> The text item which is created from #2 will be remained empty.
   And it will have 0 length, LATIN's bidi property and
   same text_pos value with the item which is created from #3.

These empty text item can ruin bidi informations that is working based on text_pos(= cursor's position).

@tizen_fix

Change-Id: Iae219918e3acb1f3c4c6a8eec1f3e12bf77ae00b

6 years agoevas_render: Don't render from norender 31/174331/1
Derek Foreman [Wed, 14 Mar 2018 18:12:17 +0000 (13:12 -0500)]
evas_render: Don't render from norender

commit 4dee873a appears to have accidentally flipped a flag in
_evas_canvas_norender causing it to actually try to render.

Change-Id: I01a4a050207c97816a0b20741f64412e8fb90a1f

6 years agoadd Requires libwayland-egl-tizen 00/174200/1
Joogab Yun [Thu, 29 Mar 2018 05:04:29 +0000 (14:04 +0900)]
add Requires libwayland-egl-tizen

Change-Id: If95d5ce9ededfa010740ab0808bf69fda0eb8e4b

6 years agoecore_evas: don't set in_async_render flag if not in use async render. 80/174080/1
Hosang Kim [Wed, 28 Mar 2018 02:23:13 +0000 (11:23 +0900)]
ecore_evas: don't set in_async_render flag if not in use async render.

Reviewers: jypark, raster, cedric
Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5844
Change-Id: I7d8ce1356bcb9f64d4d93a6854d69ebaed6fea78

6 years agoecore_imf: check validation before accessing context class 28/172728/3
Jihoon Kim [Thu, 15 Mar 2018 11:26:40 +0000 (20:26 +0900)]
ecore_imf: check validation before accessing context class

Change-Id: Id329b87167caa71aadd8ca454c50e9aaca275415
Signed-off-by: Jihoon Kim <jihoon48.kim@samsung.com>
6 years ago ecore_evas: fix ecore evas engine load fail related with gl_tbm and software_tbm 65/173465/1 accepted/tizen/unified/20180328.010251 submit/tizen/20180319.053649
Jiyoun Park [Thu, 22 Mar 2018 00:37:10 +0000 (09:37 +0900)]
   ecore_evas: fix ecore evas engine load fail related with gl_tbm and software_tbm

Change-Id: I88c18d33f77318f58e87a481f23015f753f50ab1

6 years ago elm_icon: support elm_icon_order_lookup_get/set until find alternative api 85/173185/1
Jiyoun Park [Tue, 20 Mar 2018 02:59:24 +0000 (11:59 +0900)]
elm_icon: support elm_icon_order_lookup_get/set until find alternative api

Change-Id: I6e9d35868736c19bce37a5c1f49e6b4e0b066962

6 years agoupdate debug library path correctly 71/173171/1
Jongmin Lee [Tue, 20 Mar 2018 02:06:37 +0000 (11:06 +0900)]
update debug library path correctly

Change-Id: Id2e010bf821fb825099cd161e53ac80dae6760bc

6 years ago ecore_wl2: add function declaration in Ecore_Wl2.h 18/173118/1
Jiyoun Park [Mon, 19 Mar 2018 23:50:15 +0000 (08:50 +0900)]
   ecore_wl2: add function declaration in Ecore_Wl2.h

Change-Id: If7f9846ec9aa15ff80066a144bc3bd900d86aed5

6 years agoecore_wl2: add keyboard repeat set API 96/172996/2
Hosang Kim [Mon, 19 Mar 2018 06:59:56 +0000 (15:59 +0900)]
ecore_wl2: add keyboard repeat set API

Application want to change keyboard repeat rate and delay time.

@tizen_only

Change-Id: Idb0c490f5d49849123c53b2828d2a50b74627b4a

6 years agoecore-evas-tbm : add tbm_alloc_func, pixel_acquire and release symbol 99/173099/1
Wonsik, Jung [Mon, 19 Mar 2018 12:44:14 +0000 (21:44 +0900)]
ecore-evas-tbm : add tbm_alloc_func, pixel_acquire and release symbol

Change-Id: I7139ff378b61978e9e4d6eead385299626ce0abd

6 years agoevas : rename evas_object_image_pixels_noti_callback_set() 79/173079/1
Joogab Yun [Mon, 19 Mar 2018 11:11:44 +0000 (20:11 +0900)]
evas : rename evas_object_image_pixels_noti_callback_set()

Change-Id: Id87c0bc8c4eb300593655998b026dcea10b2749c

6 years agoMerge branch 'tizen' into efl120 submit/sandbox/upgrade/efl120/20180319.053334
Wonki Kim [Mon, 19 Mar 2018 05:29:38 +0000 (14:29 +0900)]
Merge branch 'tizen' into efl120

This commit merge branch 'tizen' and 'efl120.
However, old changes on 'tizen' branch could be removed by applying this commit
because of the merge strategy(-s ours, not -X ours)

Change-Id: I6390c045ba9957a6343b89169c35247f69564f14

6 years agoconformant: add patchs for supporting backward compatibility. 23/172623/4
Hosang Kim [Thu, 15 Mar 2018 01:50:46 +0000 (10:50 +0900)]
conformant: add patchs for supporting backward compatibility.

1. conformant: Added "virtualkeypad,size,changed" smart callback.
2. elm_win/elm_conformant: add processing properties of window.

Change-Id: I67296f5413ac58c7fc47af9ef903f5c10dc08241

6 years agoelm: do not set input region 37/171537/3
Shinwoo Kim [Mon, 5 Mar 2018 10:29:50 +0000 (19:29 +0900)]
elm: do not set input region

If a window is not a full screen, then the window has used
geometry of "elm.spacer.opaque" part for its input region.

This geometry value is not correct at first launching time,
if window does not have its resize object.

For example the keyboard shows with "0, 0, 63, 0" input region,
and the keyboard could not get touch events from the window manager.

*tizen_only

Change-Id: I1824649f67674befe59514fa119fb5c000013580

6 years agoecore_evas: get evas device from ecore device 40/171840/1
JengHyun Kang [Wed, 7 Mar 2018 13:52:04 +0000 (22:52 +0900)]
ecore_evas: get evas device from ecore device

Change-Id: I9285fbf0ebdb0479f2444e8cd3eb08da60bd915d