platform/upstream/efl.git
2 years agoeina_value_util: fix time_t value for aarch64 13/279013/2
Hubert Stachowiak [Fri, 29 Jul 2022 09:29:35 +0000 (11:29 +0200)]
eina_value_util: fix time_t value for aarch64

On aarch64 mktime was returning -1 due to daylight saving time flag not being set in the struct tm. The struct is initialized with 0s and the flag is set to -1 (not available)

Change-Id: Id5958ad5769eaf53d3db571e34e07a7fd1bbd122

2 years agoelm_atspi: Return package name as application name 49/278949/2
Lukasz Oleksak [Thu, 28 Jul 2022 14:29:57 +0000 (16:29 +0200)]
elm_atspi: Return package name as application name

This patch unifies the behavior with DALi.

@tizen_only

Change-Id: I7a3eb090914e99efe90ffcf30e1c9e0a9fe5ed96

2 years agoecore_wl2: add a new window type - ECORE_WL2_WINDOW_TYPE_DESKTOP 19/278719/2 accepted/tizen/unified/20220728.044751 submit/tizen/20220725.072716 submit/tizen/20220727.010221
Doyoun Kang [Mon, 25 Jul 2022 09:17:26 +0000 (18:17 +0900)]
ecore_wl2: add a new window type - ECORE_WL2_WINDOW_TYPE_DESKTOP

@tizen_only

Change-Id: I1f0b62bc2a432cc95916523aafd401b205e516e5

2 years agoecore_wl2_window: send ECORE_WL2_EVENT_WINDOW_ROTATE event in _ecore_wl_window_cb_ang... 08/278508/1 accepted/tizen/unified/20220722.031016 submit/tizen/20220721.100327
Doyoun Kang [Thu, 21 Jul 2022 09:28:57 +0000 (18:28 +0900)]
ecore_wl2_window: send ECORE_WL2_EVENT_WINDOW_ROTATE event in _ecore_wl_window_cb_angle_change_with_resize

We add code to send a ECORE_WL2_EVENT_WINDOW_ROTATE event in the tizen_rotation's
angle_change_with_resize listener function (_ecore_wl_window_cb_angle_change_with_resize).

For this, a module can do what it wants when the window is rotated.

@tizen_only

Change-Id: I876e3bb877b65f55f5086c8b3fb73cc533c65581

2 years agoecore_wl2_window: fix condition for calling visibility change event 59/278259/1 accepted/tizen/unified/20220721.135054 submit/tizen/20220719.054534
Hosang Kim [Mon, 18 Jul 2022 07:21:50 +0000 (16:21 +0900)]
ecore_wl2_window: fix condition for calling visibility change event
callback.

Change-Id: Ia7fe752855c66455d4c19cfaa77d1a6aabe707dc

2 years agoefl_ui_win: set NULL when parent window is deleted 01/277401/2 accepted/tizen/unified/20220712.133056 submit/tizen/20220711.043600
Hosang Kim [Wed, 6 Jul 2022 08:15:42 +0000 (17:15 +0900)]
efl_ui_win: set NULL when parent window is deleted

Change-Id: If4822d5fb441628a98ffbf9cb04d3246845bf748

2 years agoevas: fix premultiply bug in webp after applying npatch 96/277296/1
tscholb [Tue, 5 Jul 2022 01:03:35 +0000 (10:03 +0900)]
evas: fix premultiply bug in webp after applying npatch

After applying npatch in webp, some of alpha value has a problem because of premultipy.
so i fix it.

Change-Id: I61425ade10659c3092f22b9854c932deb1fa468e

2 years agoecore_evas: set ecore_evas height 0 when frame is enabled. 32/274632/3
Hosang Kim [Mon, 2 May 2022 08:01:00 +0000 (17:01 +0900)]
ecore_evas: set ecore_evas height 0 when frame is enabled.

Change-Id: I3b2ec17c4f41bd81ed69b14a179de39a1eb60cbb

2 years agoelm_config: fix getting environment variable for win_no_border config. 97/274297/2 accepted/tizen/unified/20220624.143411 submit/tizen/20220623.015128 submit/tizen/20220623.083528
Hosang Kim [Tue, 26 Apr 2022 03:34:38 +0000 (12:34 +0900)]
elm_config: fix getting environment variable for win_no_border config.

It is always set to EINA_TRUE regardless of the environment variable
value.

Change-Id: Id93104027b5f199a837fc4ff855325e06d46404b

2 years agoecore_wl2_window: add visibility callback set 16/276616/2
Hosang Kim [Wed, 22 Jun 2022 02:43:27 +0000 (11:43 +0900)]
ecore_wl2_window: add visibility callback set

When the main thread is busy, WINDOW_VISIBILITY_CHANGE event is not
delivered on time. So add bypass callback for enhancing perpomance

Change-Id: I1ce27edf73f8753e8715e3e699ddf5f9b5f5cd07

2 years agoecore_wl2: prevent wrong use of wtz_exported_shell_set_destination 61/275461/3
Shinwoo Kim [Wed, 25 May 2022 06:51:26 +0000 (15:51 +0900)]
ecore_wl2: prevent wrong use of wtz_exported_shell_set_destination

It is an error if you are using wtz_exported_shell_set_destination
with negative or zero values.

Please refer to following description.

[wtz_foreign.xml]
  <interface name="wtz_exported_shell" version="1">
    <description summary="an exported shell handle">
    ...
    </description>
    <enum name="error">
      <entry name="bad_value" value="0"
         summary="negative or zero values in width or height"/>

Change-Id: Iea52527c9bd2c51b2f260464443af1437837ef17

2 years agospec: disable line coverage test 19/276219/1 accepted/tizen/unified/20220616.141909 submit/tizen/20220614.035254 submit/tizen/20220615.014424
Taehyub Kim [Mon, 13 Jun 2022 05:29:30 +0000 (14:29 +0900)]
spec: disable line coverage test

- except efl test because of maintain policy

Change-Id: If1b6233e2c188444d6a9c426f18fbb52c6dc40a2

2 years agoecore_wl2_display: remove checking TMP_WM_READY. 42/275742/1
Hosang Kim [Tue, 31 May 2022 07:56:07 +0000 (16:56 +0900)]
ecore_wl2_display: remove checking TMP_WM_READY.

Change-Id: Ibb60d6a5e757c527abbc7c16c813dd813284b370

2 years agoeo_base_class: fixed possible null dereference 22/275622/2
Michal Szczecinski [Fri, 27 May 2022 08:30:00 +0000 (10:30 +0200)]
eo_base_class: fixed possible null dereference

Change-Id: I399b323d23610daf51becd878bac24ae0b38847d

2 years agoevas: proper image scaling 91/268691/3 accepted/tizen/unified/20220526.143700 submit/tizen/20220525.023233
mgrudzinska [Wed, 29 Dec 2021 23:59:30 +0000 (00:59 +0100)]
evas: proper image scaling

In the case an image is cached, the scaling should depend
on the SVG's preserveAspectRatio attribute. If the tvg loader
is used, this value is known only inside TVG, so the transformation
matrix established in EFL can not be applied and the scaling has to
be done in TVG.

Change-Id: I620fcb29d262bad278274815bd8a22dea1d587c8

2 years agogenlist: Fix for null check. 06/275206/1
jykeon [Wed, 18 May 2022 07:16:28 +0000 (16:16 +0900)]
genlist: Fix for null check.

Change-Id: I64e7afd3e3deead4a0d76880b565e17ddae9544d
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agoevas: Add for missing return value. 01/275201/2
jykeon [Wed, 18 May 2022 05:57:34 +0000 (14:57 +0900)]
evas: Add for missing return value.

Change-Id: I1f202f67d768e19da73ebfd0efe6776f3dc7b56f
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agogenlist: Fix for null check 59/275159/2
jykeon [Tue, 17 May 2022 10:37:51 +0000 (19:37 +0900)]
genlist: Fix for null check

Change-Id: Ic907d3a3c22f5fc9446b4fc0c65f395696ea95f4
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agoEnable AT-SPI by default for suppressed screen-reader 62/275162/1
Shinwoo Kim [Tue, 17 May 2022 11:06:16 +0000 (20:06 +0900)]
Enable AT-SPI by default for suppressed screen-reader

The ELM_SUPPRESSED_SCREEN_READER is enabled if the application wants to
speak directly without going through the screen-reader.

In an environment where the ELM_SUPPRESSED_SCREEN_READER is enabled,
the screen-reader does not use "IsEnabled" property but only
"ScreenReaderEnabled" property.

The "IsEnabled" is enabled by another AT-client such as the Aurum.

So this patch makes the AT-SPI bridge be enabled only if the "IsEnabled"
property is enabled, and ignore the changes of "ScreenReaderEnabled".

Change-Id: I9a006b559083ed19eb6985d7faa75ffac6edc3e5

2 years agoRevert "efl: remove unnecessary packed files" 37/275037/2
Shinwoo Kim [Fri, 13 May 2022 12:16:09 +0000 (21:16 +0900)]
Revert "efl: remove unnecessary packed files"

This reverts commit e0cdd9f65d99ce93c84395cb1be2a80b6efa1115.

Reason:

The rpm macro %{_libdir} is /usr/lib64 on the 64-bit build.
whereas %{_libdir} is /usr/lib on the 32-bit build.

But
the prefix of following files is /usr/lib on the 64-bit build.
No the prefix is not /usr/lib64 on the 64-bit build.

/usr/lib/debug/.build-id
/usr/lib/debug/.build-id/00
/usr/lib/debug/.build-id/00/6352bb9b60c759f3a3fbe71636aa1af486cbaf
/usr/lib/debug/.build-id/00/6352bb9b60c759f3a3fbe71636aa1af486cbaf.debug

For more details, I could get following information from 64-bit build.

rpmbuild/BUILDROOT$ ls -alt efl-1.25.1-0.aarch64/usr/lib/debug/.build-id/a1

df87ab6a793ab172fab8d72d500a1537f4170.debug -> ../../usr/lib64/libefl.so.1.25.1.debug
df87ab6a793ab172fab8d72d500a1537f4170 -> ../../../../lib64/libefl.so.1.25.1

I have no idea how the commit e0cdd9f removed the unexpected files above
We need to check if the same problem the commit e0cdd9f tried to solve happens again.

This patch is solving a rpm 64-bit build error on the TV product.
The build error is:

[  591s] Checking for unpackaged file(s): /usr/lib/rpm/check-files /home/abuild/rpmbuild/BUILDROOT/efl-1.25.1-0.aarch64
[  591s] error: Installed (but unpackaged) file(s) found:
[  591s]    /usr/lib/debug/.build-id/03/6e8ba7777cd243f78b8d0000b50f2479126c08
[  591s]    /usr/lib/debug/.build-id/03/6e8ba7777cd243f78b8d0000b50f2479126c08.debug
[  591s]    /usr/lib/debug/.build-id/1a/9a4c7f66ae4caea8a537f3af0ab7c77d65013b
[  591s]    /usr/lib/debug/.build-id/1a/9a4c7f66ae4caea8a537f3af0ab7c77d65013b.debug
[  591s]    /usr/lib/debug/.build-id/1e/a8bda77f406def3d779347c5e7c7ac619c75c2
[  591s]    /usr/lib/debug/.build-id/1e/a8bda77f406def3d779347c5e7c7ac619c75c2.debug

Yes, again, the prefix is /usr/lib, not /usr/lib64.

Change-Id: Ia8cd3f3b7543492bf579a4cfaf2c5543c08244ae

2 years agoevas: support npatch image in webp 85/274885/7 submit/tizen/20220513.082339
tscholb [Wed, 11 May 2022 10:48:50 +0000 (19:48 +0900)]
evas: support npatch image in webp

npatch image is only support in png
so i added some code to support it in webp

Change-Id: I2c553d72ae8d470219c5ec1b98c0d98d4ed26e05

2 years agoecore_wl2_display: waiting for window server is ready. 90/273190/4
Hosang Kim [Wed, 30 Mar 2022 10:27:16 +0000 (19:27 +0900)]
ecore_wl2_display: waiting for window server is ready.

Change-Id: I1c7a99615ff50869e78107a34fe5e0f71d6c86fa

2 years agogenlist: Fix build error. 16/274916/1
jykeon [Thu, 12 May 2022 04:44:43 +0000 (13:44 +0900)]
genlist: Fix build error.

Change-Id: I9f3588b3b7e683728a9812155d21043c93f495f8
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agogenlist: unset FOCUSABLE, HIGHLIGHTABLE for DISPLAY_ONLY item 73/274873/1
Shinwoo Kim [Wed, 11 May 2022 07:40:27 +0000 (16:40 +0900)]
genlist: unset FOCUSABLE, HIGHLIGHTABLE for DISPLAY_ONLY item

Application uses the DISPLAY_ONLY item as a seperator.
This item does not have to get focus or highlight.

Change-Id: Ic49a81ea96f194cf8de44a231cd5c7df683c1249

2 years agoatspi,genlist: Fix parent-children incosistencies in atspi tree 10/274610/1
Shinwoo Kim [Tue, 3 May 2022 09:32:15 +0000 (18:32 +0900)]
atspi,genlist: Fix parent-children incosistencies in atspi tree

This patch is lost part of following commit.

commit ab83efe24c1c212cb5b67af0444350e65dd876be
Author: Radoslaw Cybulski <r.cybulski@partner.samsung.com>
Date:   Mon Oct 22 10:57:07 2018 +0200

    Fix parent-children incosistencies in atspi tree

    Fixes incosistencies (where obj A was child of object B, but object B's
    parent was different, than A) in atspi tree. Now atspi parent is
    retrieved by calling efl_access_object_access_parent_set.

    Change-Id: I3e20af88bd7f066f3f9faa5185e4028d0e5e05db

Change-Id: I38b55c5c6bf48a490a667af24e88949ae6f6ab5d

2 years agoecore_wl2_subsurf: make subsurface locate below parent 30/274330/1 accepted/tizen/unified/20220428.162702 submit/tizen/20220427.042343
Shinwoo Kim [Tue, 26 Apr 2022 10:04:52 +0000 (19:04 +0900)]
ecore_wl2_subsurf: make subsurface locate below parent

Application is not able to call.

  - tizen_policy_place_subsurface_below_parent

So ecore_wl2_subsurf is calling this when a subsurface is exported.

Change-Id: I80738fa8ee15a781ca69daeaaf8bccd5301d6335

2 years agoecore_wl2_window: modify configure event's size 27/274127/3
Doyoun Kang [Thu, 21 Apr 2022 23:49:46 +0000 (08:49 +0900)]
ecore_wl2_window: modify configure event's size

There was an ambiguous configure event's size. So we modified the code to determine
the size of the configure event for each case.

@tizen_only

Change-Id: I13001bedf9a1035b5f8023d97a76a39b6d05370e

2 years agoefl_ui_widget: add null check for smart data 72/273672/2 accepted/tizen/unified/20220422.021246 submit/tizen/20220421.004950
Bowon Ryu [Tue, 12 Apr 2022 05:13:41 +0000 (14:13 +0900)]
efl_ui_widget: add null check for smart data

Change-Id: Ic10b5170ea4da28c307367e6e58d06baf16f50e7
Signed-off-by: Bowon Ryu <bowon.ryu@samsung.com>
2 years ago[UI Thread] Fix duplicated initialization of ecore signal handler 28/273928/2 submit/tizen/20220419.070532
Hwankyu Jhun [Mon, 18 Apr 2022 06:49:57 +0000 (15:49 +0900)]
[UI Thread] Fix duplicated initialization of ecore signal handler

To prevent duplicated initialization of the ecore signal handler, this patch add
a new flag to check whether the ecore signal is initialized or not.
Before creating a new child process, the ecore signal handler blocks signals to
stop inheriting signal handlers. After creating a new child process,
the ecore signal handler unblocks signals.

Change-Id: I5f8e2efffec98e9cd3e4387420ceb7126aa01d84
Signed-off-by: Hwankyu Jhun <h.jhun@samsung.com>
2 years agoecore_wl2_display: modify code for ecore_wl2_display_screen_size_get 73/273573/1 accepted/tizen/unified/20220418.141309 submit/tizen/20220414.061035
Doyoun Kang [Fri, 8 Apr 2022 01:20:52 +0000 (10:20 +0900)]
ecore_wl2_display: modify code for ecore_wl2_display_screen_size_get

We modified code that this API returns only current screen size not all screens size.
If an user wants to get whole screen size, then use below API.
- ecore_wl2_display_all_screens_size_get

@tizen_only

Change-Id: Iea71ea890d5141e4d5e303db5023311c93ec2113

2 years agoecore_wl2_display: add API for getting all screens size 31/273531/2
Doyoun Kang [Thu, 7 Apr 2022 05:40:46 +0000 (14:40 +0900)]
ecore_wl2_display: add API for getting all screens size

we add a new API for getting a size of all connected screens
- ecore_wl2_display_all_screens_size_get

@tizen_only

Change-Id: I11489c1be17c748445ce424ce9fdff0574d4dce1

2 years ago[UI Thread] Fix ecore signal handler 21/272621/7
Hwankyu Jhun [Tue, 22 Mar 2022 02:10:23 +0000 (11:10 +0900)]
[UI Thread] Fix ecore signal handler

To forward the signal to the previous signal handler, the ecore signal
saves previous actions.

Change-Id: I5b5a6088e98df3ae8928014f5d149d9bcff9394f
Signed-off-by: Hwankyu Jhun <h.jhun@samsung.com>
2 years ago[UI Thread] Use tizen glib context 75/273175/2
Hwankyu Jhun [Thu, 31 Mar 2022 06:59:16 +0000 (15:59 +0900)]
[UI Thread] Use tizen glib context

After this patch is applied, ecore main loop dispaches tizen glib context.
The application core sets the tizen glib context using setenv() with
"TIZEN_GLIB_CONTEXT".

Change-Id: Ib1405fe6fb8a646a946b8c666b7d0ef74e0bfcc4
Signed-off-by: Hwankyu Jhun <h.jhun@samsung.com>
2 years agoevas_gl: add logs in evgl_tls 60/273160/4
Daekwang Ryu [Thu, 31 Mar 2022 05:45:09 +0000 (14:45 +0900)]
evas_gl: add logs in evgl_tls

To check the reason that _evgl_tls_resource_get() failed, logs are
added.
The issue causes that doesn't destroy eglContext in
evgl_destroy_context().

Change-Id: I97692933144a1902dd80c7d432b34d36b9620a2f

2 years agoecore_wl2_window: added APIs for set min / max size of window 12/272812/4
Junseok Kim [Fri, 25 Mar 2022 02:43:26 +0000 (11:43 +0900)]
ecore_wl2_window: added APIs for set min / max size of window

@tizen_only

Change-Id: Iebd4193320fb55d33a489e46f20dde37d2e64eec
Signed-off-by: Junseok Kim <juns.kim@samsung.com>
2 years agoevas: fix memleak during a picture duplication 10/268910/3
Mira Grudzinska [Tue, 4 Jan 2022 23:55:12 +0000 (00:55 +0100)]
evas: fix memleak during a picture duplication

A tvg picture is created twice - by efl and by tvg.
The object created by efl has to be deleted in order
to prevent memory the leaks.

Change-Id: I74629c8cb8ea47e476e102cefcb45ee571837856

2 years agoelm_entry: do not send clicked signal if disabled 21/272821/1
Prateek Thakur [Fri, 25 Mar 2022 04:54:24 +0000 (10:24 +0530)]
elm_entry: do not send clicked signal if disabled

Clicked signal was sent even when entry was disabled.
Due to this application would receive clicked signal on disabled entry and
can cause issue if they are taking any decision on this signal.

Change-Id: I72ddfbfa1c9cade57cf9936440392b077e4caffb
Signed-off-by: Prateek Thakur <prateek.th@samsung.com>
2 years agoecore_wl2, elm: add resource id to window attribute 15/272715/4 accepted/tizen/unified/20220325.133343 submit/tizen/20220325.025506
Shinwoo Kim [Wed, 23 Mar 2022 05:02:12 +0000 (14:02 +0900)]
ecore_wl2, elm: add resource id to window attribute

The window resource id is unique value in the system.
This value can be used in screen-reader to support multi-window.
The screen-reader will use this to find a window at xy position.

Change-Id: Ibf19a2aaf4b698cb518b677220a100e14cb1be6f

2 years ago[UI Thread] Use thread default context 74/270474/11
Hwankyu Jhun [Thu, 3 Feb 2022 06:43:34 +0000 (15:43 +0900)]
[UI Thread] Use thread default context

If the process sets the thread default context, the ecore main loop uses
it instead of the global default context.

Change-Id: Icf01d6098ad3c3c9e82d5fbf8045c45e85877928
Signed-off-by: Hwankyu Jhun <h.jhun@samsung.com>
2 years agoecore_evas_tbm: fix a 'NO_RETURN_VALUE' issue 25/272525/3
duna.oh [Fri, 18 Mar 2022 08:18:45 +0000 (17:18 +0900)]
ecore_evas_tbm: fix a 'NO_RETURN_VALUE' issue

Change-Id: Ied351b15f24a18a813659c11ff52efe9a0f13d7b

2 years agoevas: Fixed Null defect issue. 01/272501/2
jykeon [Fri, 18 Mar 2022 02:58:15 +0000 (11:58 +0900)]
evas: Fixed Null defect issue.

Change-Id: I941a6a3e87b4297f5939028c7590b1f9d1f9d777
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agoecore_evas_tbm: use a longer string's length in strncmp() 89/272489/1
duna.oh [Fri, 18 Mar 2022 00:38:05 +0000 (09:38 +0900)]
ecore_evas_tbm: use a longer string's length in strncmp()

Change-Id: I0c8b323ec4518f79905699cb3e68243a98100ff2

2 years agoecore_evas_tbm: fix a wrong seat device as parent of a evas_device 84/272484/1
duna.oh [Thu, 17 Mar 2022 10:38:30 +0000 (19:38 +0900)]
ecore_evas_tbm: fix a wrong seat device as parent of a evas_device

It was a bug that all of evas_devices in tbm backend have a default seat,
though a device sits in different seat, other than 'Enlightenment seat'.
This is fixed by using seatname in Ecore_Event_Device_Info struct.
A device now has a right seat device as a parent of it.

Change-Id: Ica419b3c56eb18a376fe462346da70b234790466

2 years agoevas: ++safety 13/271913/6
Michal Szczecinski [Thu, 3 Mar 2022 06:23:24 +0000 (07:23 +0100)]
evas: ++safety

Stability tests detected crash in evas_object_pointer_grab_del function.
This commit checks NULL dereference in this API.

Change-Id: I2109755e2c38e44ebbf048b32988dd41f7dc0bb0

2 years agoecore_wl2: do not destroy tizen_position in ecore_wl2_window_hide 72/272372/1
Doyoun Kang [Tue, 15 Mar 2022 09:43:32 +0000 (18:43 +0900)]
ecore_wl2: do not destroy tizen_position in ecore_wl2_window_hide

There was a bug that the window's position was wrong if a window called
ecore_wl2_window_position_set after calling ecore_wl2_window_hide.
In this case, the display server didn't get an information about new position
because the tizen_position was destroyed. When the window was shown again, the
display server sent configure event using old position not new position.
By this, window moved to wrong(old) position.

This patch resolve this problem.

@tizen_only

Change-Id: I28b377ca5fd77106a69f442e4300feb891236d84

2 years agoevas vg json: hide parent layer if invisible mask applied 09/269909/3 accepted/tizen/unified/20220316.182839 submit/tizen/20220307.054027 submit/tizen/20220315.013931
Michal Maciola [Fri, 21 Jan 2022 10:09:36 +0000 (11:09 +0100)]
evas vg json: hide parent layer if invisible mask applied

If a layer has an invisible mask applied, parent layer will be set
invisible too. It will help displaying some json files correctly.

Change-Id: Ibaacb3f9617eace384eba0e5fa38032ed9ba9ebf

2 years agoelm_atspi: use timer to recover bridge init failure 45/271645/3
Shinwoo Kim [Thu, 24 Feb 2022 05:17:59 +0000 (14:17 +0900)]
elm_atspi: use timer to recover bridge init failure

An idler has been used to recover bridge init failure.
The idler callback can be called within 1 ms. By the result,
an application was not able to initialize the ATSPI bridge.

So this patch is using a timer and retry for 1 min.

Change-Id: I7d42ec6985b8fe932eb8020b5ea4cf1785365a0e

2 years agoecore_con: add missing goto for error case. 21/271721/1 submit/tizen/20220304.033428
Hosang Kim [Fri, 25 Feb 2022 09:51:15 +0000 (18:51 +0900)]
ecore_con: add missing goto for error case.

Change-Id: I8dc140ec95c43897db5cfabb4742c9608134a1c4
Signed-off-by: Hosang Kim <hosang12.kim@samsung.com>
2 years agoevas_object_smart: Set child_has_map flag to parent when parent is set 22/271622/1
Jaehyun Cho [Wed, 23 Feb 2022 12:16:35 +0000 (21:16 +0900)]
evas_object_smart: Set child_has_map flag to parent when parent is set

Previously, parent's child_has_map flag was set only when
evas_object_map_enable_set was called for its child.

This causes an issue that mouse touch may not be caught by the parent.
Because the parent's child_has_map is supposed to be true but it is
actually false.
So the bounding_box of the parent may not be updated properly but mouse
touch is checked by the intersection of the bounding_box since
child_has_map flag is false.

Now, parent's child_has_map flag is set when parent is set.
So child_has_map flag is set to parent and mouse touch is caught
correctly.

Change-Id: I2ed8e57e3a929a6d37f65c319e4306d58f4d88e0

2 years agoecore_evas:fix string comparison of device's identifier 39/271339/5
duna.oh [Fri, 18 Feb 2022 04:30:06 +0000 (13:30 +0900)]
ecore_evas:fix string comparison of device's identifier

Change-Id: I0ecfd04e021af8801dc5ae9c77bdc85a79cf0f0e

2 years agoFix:meson build error. 23/271423/2 accepted/tizen/unified/20220221.080255 submit/tizen/20220221.054736
jykeon [Mon, 21 Feb 2022 05:08:01 +0000 (14:08 +0900)]
Fix:meson build error.

Change-Id: I3d69cc97563e226416163563cb7f4fbea28816df
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agoedje_calc: fix check return of _edje_fetch 58/270858/2
Jaehyun Cho [Thu, 10 Feb 2022 07:26:25 +0000 (16:26 +0900)]
edje_calc: fix check return of _edje_fetch

Summary: checks if _edje_fetch returns null before calling _edje_recalc_do.

Reviewers: Hermet, raster, kimcinoo, jsuya

Reviewed By: jsuya

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12322

Change-Id: Id1c680e79abd348fa5cc888f0e50334fc93571c5

2 years agoUpdate gcove output patch 91/270691/1 accepted/tizen/unified/20220209.131748 submit/tizen/20220208.041216
jykeon [Tue, 8 Feb 2022 02:18:31 +0000 (11:18 +0900)]
Update gcove output patch

Change-Id: I3bc99b139538df6df48d82aa75d9aba401c2d853
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agogen : remove unnecessary focus_manager calls. 73/270173/3
EverLEEst(SangHyeon Lee) [Wed, 26 Jan 2022 07:39:26 +0000 (16:39 +0900)]
gen : remove unnecessary focus_manager calls.

Change-Id: Ic992497fd990cc0e6c3bd37ab5aeb8d41323baf6

2 years agoevas_device: remove pointer data for touch device. 43/269843/1
Hosang Kim [Tue, 23 Feb 2021 12:37:42 +0000 (21:37 +0900)]
evas_device: remove pointer data for touch device.

Pointer data of touch device is always created.
But it can't be removed when device is removed.
Because there is condition for checking mouse or not.

Change-Id: I9132cef203135a605f3aad9684baf79b1766b03e

2 years agoecore_wl2: do not destroy tizen_rotation in ecore_wl2_window_hide 02/269602/1 accepted/tizen/unified/20220125.121829 submit/tizen/20220120.091110
Doyoun Kang [Wed, 19 Jan 2022 04:12:53 +0000 (13:12 +0900)]
ecore_wl2: do not destroy tizen_rotation in ecore_wl2_window_hide

There was a bug that the display server (enlightenment) didn't get rotation done event if
a window called ecore_wl2_window_rotation_change_done_send after calling ecore_wl2_window_hide.

This patch resolve this problem.

@tizen_only

Change-Id: I20cd592e4744f033978d65977d72a1bb9968441b
Signed-off-by: Doyoun Kang <doyoun.kang@samsung.com>
2 years agoevas render: Fixed invalid sanity check 01/267101/2
Michal Szczecinski [Thu, 25 Nov 2021 10:58:31 +0000 (11:58 +0100)]
evas render: Fixed invalid sanity check

Change-Id: I1db4c81ca84a5c81205eeba946cce687a2c9721d

2 years agoevas examples: improved evas-vg-json-multiple.c example 65/267865/2
Michal Maciola [Fri, 10 Dec 2021 15:47:28 +0000 (16:47 +0100)]
evas examples: improved evas-vg-json-multiple.c example

Change-Id: I6ebf64ca1b8609869db97d3a6053cc44ba70da18

2 years agovg json : Don't create empty shape 11/268611/3
jykeon [Tue, 28 Dec 2021 05:20:53 +0000 (14:20 +0900)]
vg json : Don't create empty shape

Change-Id: I0d9dc77f0061b199e71491624ee6df9f080316ec
Signed-off-by: jykeon <jykeon@samsung.com>
2 years agovg_common: Invert LOTNode transformation matrix before translate. 21/268321/2
Michal Szczecinski [Mon, 20 Dec 2021 09:22:32 +0000 (10:22 +0100)]
vg_common: Invert LOTNode transformation matrix before translate.

Columns and rows in the LOTNode->image->data transformation matrix are
inverted. This commit uses valid layout.

Change-Id: Ia7647f13902db9de3045d78a8a32b18b76a48135

2 years agoevas vg json: refactor masks construct 67/267867/3
Michal Maciola [Fri, 10 Dec 2021 15:56:35 +0000 (16:56 +0100)]
evas vg json: refactor masks construct

This patch changes how nodes that have masks are handled.
TVG doesn't implement masking as mask Substract/Intersect/Difference.
If such masks was used, for most cases, the whole node was masked and
incorrectly invisible. After this patch it should be working correctly
(or just better) for most basic cases with a single mask.

Change-Id: I9204255b79bc6adf551a844de156a8d3e3e41665

2 years agoRevert "vg_common_json: No path information, do not create a shape" 63/268563/1 accepted/tizen/unified/20211228.140402 submit/tizen/20211227.054708
Chun [Mon, 27 Dec 2021 02:47:16 +0000 (02:47 +0000)]
Revert "vg_common_json: No path information, do not create a shape"

This reverts commit b688e1d312aa33ac3c71c7aeac783bdbefc0fb78.

Change-Id: I82a45e07b36057bfce1238fa56c80a423ff3f293

2 years agovg_common_json: No path information, do not create a shape 77/268377/1 submit/tizen/20211227.021527
JunsuChoi [Wed, 22 Dec 2021 04:18:26 +0000 (20:18 -0800)]
vg_common_json: No path information, do not create a shape

Change-Id: Ic016bf3e41d8fd706c31973a78b3c6825d0941b4

2 years agoecore_wl2_input: fix calibration logic for repeating key event 52/268252/1
Hosang Kim [Fri, 17 Dec 2021 09:47:10 +0000 (18:47 +0900)]
ecore_wl2_input: fix calibration logic for repeating key event

Change-Id: I6f444b2a71e6e4d2c0379b0ffad8e755d9375ee0

2 years agoevas canvas: remove tizen only code. 06/268206/2
Hermet Park [Fri, 17 Dec 2021 04:48:19 +0000 (13:48 +0900)]
evas canvas: remove tizen only code.

remove the temporary api usage,
correct the right api call.

Change-Id: I49077d27845a16c1b8722accc295513510eafdc7

2 years agoecore_wl2_input: fix order of calcuation 92/268092/1
Hosang Kim [Thu, 16 Dec 2021 05:51:36 +0000 (14:51 +0900)]
ecore_wl2_input: fix order of calcuation

Change-Id: I5e77796721ebfacc37bfc6ed9ca68067d7398b1a

2 years agoefl: remove unnecessary packed files 65/268065/1
Shinwoo Kim [Wed, 15 Dec 2021 11:31:11 +0000 (20:31 +0900)]
efl: remove unnecessary packed files

This patch will remove unexpected installed files such as

/usr/lib/debug/.build-id
/usr/lib/debug/.build-id/00
/usr/lib/debug/.build-id/00/6352bb9b60c759f3a3fbe71636aa1af486cbaf
/usr/lib/debug/.build-id/00/6352bb9b60c759f3a3fbe71636aa1af486cbaf.debug
...

Change-Id: Ife39da7694e887ac95edde9157e6a6e7baf16268

2 years agoefl: remove touch position log for improving security. 12/267912/1
Hosang Kim [Mon, 13 Dec 2021 07:03:13 +0000 (16:03 +0900)]
efl: remove touch position log for improving security.

Change-Id: Icb99aba71e486a71d6f7f87258ca246a005b7b9e

2 years agodnd: keep the order of enter and leave event for dnd 71/267671/2 accepted/tizen/unified/20211210.115209 submit/tizen/20211209.063123
Taehyub Kim [Wed, 8 Dec 2021 11:34:34 +0000 (20:34 +0900)]
dnd: keep the order of enter and leave event for dnd

Change-Id: I97ce6ee2a26d7c24d25a403cd4e44db383a4cc53

2 years agoelm_cnp: keep the copy and paste selection backword compatibility 74/267674/1
Taehyub Kim [Wed, 8 Dec 2021 11:39:17 +0000 (20:39 +0900)]
elm_cnp: keep the copy and paste selection backword compatibility

Change-Id: Icf64b185929bc2b04911adcb08ebdd98b63807d9

2 years agoevas common: remove lru optimization. 18/267518/2
Hermet Park [Mon, 6 Dec 2021 12:01:53 +0000 (21:01 +0900)]
evas common: remove lru optimization.

This lru optimization is useless because we have already reference count,
Actually, we can't remove the data even though the cache capacity is full
because those are completed referenced.

This mis-implementation occurs the memory leak.

Change-Id: If25e843bcbb6759df81e87c076a002bf771ea1b1

2 years agoevas: fixed memory leak on lottie images container destruction 86/267386/2
Michal Maciola [Thu, 2 Dec 2021 13:04:55 +0000 (14:04 +0100)]
evas: fixed memory leak on lottie images container destruction

Change-Id: I43798742ce06b69fe0691c1c98a0c6b1548b5e3e

2 years agoecore evas extn: Check queue pointer before use. 48/267248/3
Michal Szczecinski [Tue, 30 Nov 2021 08:42:30 +0000 (09:42 +0100)]
ecore evas extn: Check queue pointer before use.

This commit increates safety of _compositor_socket_get() API.

Change-Id: I44ca3a2a095cba9f1aa073bd7749ec008de358f0

2 years ago[AT-SPI] Fixed providing number of items for objects of AT_SPI role DIALOG and POPUP_MENU 69/267169/5
Maria Bialota [Fri, 26 Nov 2021 20:29:42 +0000 (21:29 +0100)]
[AT-SPI] Fixed providing number of items for objects of AT_SPI role DIALOG and POPUP_MENU

Change-Id: I074172a98064da6e1bc7fb957feb1c7ed4a8fe1a

2 years agoedje lua2: check string ptr before dereference 17/267217/1
Taehyub Kim [Tue, 30 Nov 2021 01:48:37 +0000 (10:48 +0900)]
edje lua2: check string ptr before dereference

Summary: fix null pointer dereference

Reviewers: Hermet, kimcinoo, jsuya, raster

Reviewed By: kimcinoo

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12309

Change-Id: I32e0edbdf7709b516acaa6d91c7b734b36adab9d

2 years agoevas: Fixed invalid sanity check 90/265490/4 accepted/tizen/unified/20211130.125551 submit/tizen/20211125.062540 submit/tizen/20211129.000112
Michal Szczecinski [Wed, 20 Oct 2021 11:53:06 +0000 (13:53 +0200)]
evas: Fixed invalid sanity check

Change-Id: I35f2674d5a29c52cba5540439ec92fb59829da7c

2 years agoecore_wl2: add events for interactive move/resize done 78/266878/1
Doyoun Kang [Mon, 22 Nov 2021 04:22:58 +0000 (13:22 +0900)]
ecore_wl2: add events for interactive move/resize done

Event Type
- ECORE_WL2_EVENT_WINDOW_INTERACTIVE_MOVE_DONE
- ECORE_WL2_EVENT_WINDOW_INTERACTIVE_RESIZE_DONE

Event Structure
- Ecore_Wl2_Event_Window_Interactive_Move_Done
- Ecore_Wl2_Event_Window_Interactive_Resize_Done

Change-Id: I4752611802a3dc9268f7941d769dc20ae399f992

2 years agoevas vg shape: gradient taken from the proper source 88/266588/4
Mira Grudzinska [Tue, 16 Nov 2021 01:13:49 +0000 (02:13 +0100)]
evas vg shape: gradient taken from the proper source

When a shape with the gradient is duplicated, the gradient
adr was taken from the source obj instead of from the target
of the duplication.
In TVG, when a shape is duplicated all its properties, like grad,
are duplicated as well. EFL created independently a new gradient,
which has to be relased first and replaced by the one created
in TVG.

Change-Id: I0848b0591671b6e69148b0d248a0ba0c8bd66726

2 years agoecore_wl2_input: unref device before device is deleted. 99/266599/2
Hosang Kim [Tue, 16 Nov 2021 05:28:47 +0000 (14:28 +0900)]
ecore_wl2_input: unref device before device is deleted.

@fix

Change-Id: I96a09ae32b768ebb9c54bf8ade2a95f14721432a

2 years agovg: update the tvg apis 62/265762/3
Hermet Park [Thu, 28 Oct 2021 06:28:44 +0000 (15:28 +0900)]
vg: update the tvg apis

These apis has been modified under the beta.
now we correct it with the latest ones.

Change-Id: I52fdbe2d28da29057db8b61d3965b8351f5f069d

3 years agoevas_font_query: fix ARABIC HAMZA ABOVE render issue 78/266378/2
Bowon Ryu [Thu, 11 Nov 2021 07:22:35 +0000 (16:22 +0900)]
evas_font_query: fix ARABIC HAMZA ABOVE render issue

INHERITED skip prevents ARABIC_HAMZA_ABOVE from rendering standalone.
Do not skip when ARABIC_HAMZA_ABOVE is at the begginning of an arabic word.

Change-Id: I75d5206194543031fcfe81f5ed208a8b404fc26a
Signed-off-by: Bowon Ryu <bowon.ryu@samsung.com>
3 years agoefl_ui_win: delete ecore_evas when inlined image window is deleted 73/266073/3 accepted/tizen/unified/20211109.042919 submit/tizen/20211108.084528
Hosang Kim [Fri, 5 Nov 2021 07:02:35 +0000 (16:02 +0900)]
efl_ui_win: delete ecore_evas when inlined image window is deleted

When inlined image window is deleted, associated image object and ecore
evas are not deleted. So there is memory leak.

@tizen_only

Change-Id: Iab248b0cf0bfe3f3a9f81a8eb8604a68218be232

3 years agoefl_canvas_image : remove the animated image limitation 08/266108/1
Taehyub Kim [Thu, 4 Nov 2021 07:38:48 +0000 (16:38 +0900)]
efl_canvas_image : remove the animated image limitation

Summary: remove FRAME_MAX limitation to play the huge animated image which has more than 1024 frames

Test Plan: load gif image which has more than 1024 frames

Reviewers: Hermet, kimcinoo

Reviewed By: Hermet

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12303

Change-Id: I59bd00d0e2fb647805c70230df72f4e811a6e58c

3 years agowayland_egl: fix egl attribute type 06/265806/2
Changyeon Lee [Fri, 29 Oct 2021 04:58:59 +0000 (13:58 +0900)]
wayland_egl: fix egl attribute type

attribute of evas_gl_common_eglCreateImage is EGLAttrib

Change-Id: I8afbfc98c4aa8b363dd75d16a6ea86744eb3dee7

3 years agowayland_egl: fix checking egl extension for native image 79/265779/3
Changyeon Lee [Thu, 28 Oct 2021 06:05:49 +0000 (15:05 +0900)]
wayland_egl: fix checking egl extension for native image

initial value of has_tizen_native_surface and has_dma_buf_import
is EINA_TRUE.
it is changed to EINA_FALSE after checking extension in gl_extn_veto.

Change-Id: I74224e3ff696774d46063dc3e2cb3d4a5343d40c

3 years agoevas vg json: fix invisible layer 87/265787/2 accepted/tizen/unified/20211102.024819 submit/tizen/20211101.025718
Michal Maciola [Thu, 28 Oct 2021 11:35:01 +0000 (13:35 +0200)]
evas vg json: fix invisible layer

This patch fixes the reported rendering bug for particular json files.
If the layer was invisible, it was skipped and not rendered, but if it had
some masking layer, it was incorrectly applied for previous layer as a parent.
This fixes incorrect behaviour by skipping these masking layer.

Change-Id: I9015e47b83d479d6ac3e644fcd52551df8836042

3 years agotests: enhance evas_map test
Shinwoo Kim [Mon, 18 Oct 2021 08:10:17 +0000 (09:10 +0100)]
tests: enhance evas_map test

Reviewers: raster, Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12296

3 years agotests: add evas_map test
Shinwoo Kim [Tue, 12 Oct 2021 07:36:23 +0000 (08:36 +0100)]
tests: add evas_map test

Reviewers: raster, Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12295

3 years agoevas_render: fix a mistake of ++safety 89/265489/1
Shinwoo Kim [Wed, 20 Oct 2021 11:13:26 +0000 (20:13 +0900)]
evas_render: fix a mistake of ++safety

There was a mistake of following commit.

   928c008 evas: ++safety

And it caused performance issue by not adding obscuring objects.

*tizen_fix

Change-Id: Ie0b376b12e74a4d32925e6ea2500084ad1399f97

3 years agoEcore_Wl2: move ecore_wl2_offer functions for the internal usage 91/265391/3
Taehyub Kim [Mon, 18 Oct 2021 05:36:33 +0000 (14:36 +0900)]
Ecore_Wl2: move ecore_wl2_offer functions for the internal usage

To implement the drag and drop feature in dali,
ecore_wl2_offer functions should be added to Ecore_Wl2.h

Change-Id: I78f872fe6493854a094901c03346b569a83928ee

3 years agoecore_wayland2: include eina 48/265348/1
Shinwoo Kim [Fri, 15 Oct 2021 08:22:43 +0000 (17:22 +0900)]
ecore_wayland2: include eina

If application includes Ecore_Wayland2.h without Eina.h,
then application will have build break as below.

   "error: unknown type name 'Eina_Bool'"

So this patch makes Ecore_Wayland2.h include Eina.h
to fix the application side build error.

Change-Id: Ibd8a65e4cee111c00305ac70b1a930bbb8e4e08d

3 years agoevas map: check source image size for direct render 84/264784/1 accepted/tizen/6.5/unified/20211028.225531 accepted/tizen/unified/20211007.154801 submit/tizen/20211007.055733 submit/tizen_6.5/20211028.163301 tizen_6.5.m2_release
Shinwoo Kim [Wed, 29 Sep 2021 12:54:45 +0000 (21:54 +0900)]
evas map: check source image size for direct render

The direct render is working correctly only if source image size
is equal to the size of smart object.

So this patch is cheking the size of source image.

Change-Id: I2e6c70360a477283903e4c494410c6683dfaa0c0

3 years agoevas: reset change after map direct render 56/264656/3 accepted/tizen/unified/20210929.022336 submit/tizen/20210928.143216
Shinwoo Kim [Tue, 28 Sep 2021 02:30:13 +0000 (11:30 +0900)]
evas: reset change after map direct render

There is a flickering issue which is caused by absence of reset.

If the source image does not reset, then it remains in pending objects
array at the end of  evas_render_updates_internal, and it keeps the
'obj->changed' as TRUE for next frame. And it makes its parent 'changed'
value TRUE as well.

But after handling pending objects array, we are handling render objects
array. The parent 'changed' value is set to FALSE.

As a result, the parent object does not draw on the next frame, and its
children including the source image neither.
(no push to restack objects array, render_pre, output_redraws_rect_add)

Even though there is a change on the source image for the next frame,
(for example, evas_object_color_set)
it cannot makes the parent object 'changed' value TRUE,
(see the logic of evas_object_change)
because the source image 'changed' is already TRUE by the logic handling
pending objects array on the previous frame.

We are calling evas_object_change_reset after evas_render_mapped.
So, same logic should exist for the map direct render.

*tizen_fix

Change-Id: If8b57a8f423e25f0c1f91a90dc30ac1f2c8b4972

3 years agoevas vg: fix for invisible masking layer 72/264572/3 accepted/tizen/unified/20211001.001525 submit/tizen/20210928.074404
Michal Maciola [Fri, 24 Sep 2021 12:02:33 +0000 (14:02 +0200)]
evas vg: fix for invisible masking layer

This patch solves the problem of masking on layers that visibility
is set to false.

Change-Id: I0d96362709e551d8dadf151d7fd79ddb39e8ce7d

3 years agoeo_base_class: Avoid useless memory alloc 90/264590/1
Woochanlee [Thu, 24 Dec 2020 02:47:03 +0000 (11:47 +0900)]
eo_base_class: Avoid useless memory alloc

Summary:
The pd->ext will be NULL when the target object on invalidate state or edje object or isolated object etc..

In above case, If data is NULL and ext is not needed, it seems there is no need to create a node.

Reviewers: raster, cedric, Hermet

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12206

Change-Id: I7000fd4cbf48c0841865483092e5b778602886d1

3 years agoevas vg: json/lottie memory usage optimization 72/264272/1
Michal Maciola [Wed, 15 Sep 2021 08:41:02 +0000 (10:41 +0200)]
evas vg: json/lottie memory usage optimization

This patch fixes potential memory leaks and unnecessary allocations.

Change-Id: I515c5a809dc5b23e433514aad177dcbda47beb75

3 years agocommon json: fix for an empty layer used as a mask 35/264235/2 accepted/tizen/unified/20210916.123547 submit/tizen/20210915.070054
Michal Maciola [Tue, 14 Sep 2021 14:51:32 +0000 (16:51 +0200)]
common json: fix for an empty layer used as a mask

If a layer is empty, do not use it as a mask, so not to incorrectly completely
hide a parent layer.

Change-Id: Ia48b0b150faa3a1fd44cc79c16facab14ab00564

3 years agoevas vg: Added localization support in append path API. 46/264146/6
Michal Szczecinski [Mon, 13 Sep 2021 17:13:04 +0000 (19:13 +0200)]
evas vg: Added localization support in append path API.

In some languages strtod function don't parse '.' and ',' properly. Due
to this issue set locale part was added. Now path is parsed in the same
way as it was in efl_gfx API.

Change-Id: Ibcd0e460e0dad0e09a0ff08365418aab670194ca

3 years agoupdated edje license. 81/264181/2
Hermet Park [Tue, 14 Sep 2021 06:00:16 +0000 (15:00 +0900)]
updated edje license.

edje uses BSD license and its tools(epp) uses GPL.

Change-Id: Idc18fa7a0db0865d106f129eab94350837aa5f48

3 years agoevas_vg refactoring: thorvg svg loader: edje 49/261149/5 accepted/tizen/unified/20210910.004743 submit/tizen/20210909.071105
Michal Maciola [Mon, 12 Jul 2021 10:39:38 +0000 (12:39 +0200)]
evas_vg refactoring: thorvg svg loader: edje

This patch changes evas svg loader to using thorvg svg loader.
The patch extands .svg only; .svgz and .svg.gz will be loaded using old loader.
For edje_cc support (edj/edc files), svg file content is written into edj.

Patch is divided into two parts- edje & evas. This is edje part.

Change-Id: I716fa0a8b2e5bc23d2b262bb4b134a7972a545ce