platform/upstream/efl.git
4 years agoelementary_tizen/genlist: unfocus old content before it is deleted
Yeongjong Lee [Thu, 14 May 2020 04:19:00 +0000 (13:19 +0900)]
elementary_tizen/genlist: unfocus old content before it is deleted

This patch prevent unexpected focusing by focused_revert on the old content
deletion. because item content focus should be handled by genlist.

@tizen_fix

Change-Id: I5a8cf1e75a8cd3107c24a5822fcd22a9e6b326f4

4 years agoci: travis: switch exactness build to Fedora32 image
Stefan Schmidt [Tue, 12 May 2020 14:44:59 +0000 (16:44 +0200)]
ci: travis: switch exactness build to Fedora32 image

4 years agoelm: refactor legacy _icon_signal_emit
Mike Blumenkrantz [Wed, 13 May 2020 14:02:36 +0000 (10:02 -0400)]
elm: refactor legacy _icon_signal_emit

Summary:
this was duplicated for a number of widgets in slightly different ways,
so it can be unified as a private layout function to reduce maintenance
costs

Reviewers: Hermet, YOhoho, Jaehyun_Cho, jsuya

Reviewed By: jsuya

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11818

Change-Id: I5fdacb9b8244eebd585315a565c4291e97cb530a

4 years agoedje_calc: Add unload file for already loaded vector lottie file.
JunsuChoi [Wed, 13 May 2020 12:58:17 +0000 (21:58 +0900)]
edje_calc: Add unload file for already loaded vector lottie file.

Summary:
Users can use multiple jsons per description in one part in edc.
At that time, multiple files are loaded,
if a file is already loaded, it must be unloaded.

Test Plan:
edje_cc -beta -id ./ json_edc.edc
make
./test
Push Change button -> shutdown application

{F3882716}

Reviewers: Hermet, herb, kimcinoo

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11820

4 years agoRevert "evas_object_textblock: fix to free style properly by edje's style_user_pop"
Jongmin Lee [Wed, 13 May 2020 09:31:05 +0000 (18:31 +0900)]
Revert "evas_object_textblock: fix to free style properly by edje's style_user_pop"

This reverts commit 4937c2bb4a799e2a01840a06d51ea9dff315a907.

4 years agoRevert "canvas render: don't apply mask if the objects in the map."
Hermet Park [Wed, 13 May 2020 07:53:45 +0000 (16:53 +0900)]
Revert "canvas render: don't apply mask if the objects in the map."

This reverts commit eaafb9e4c53863cfd0ae4aaf24f04d333458ff2b.

Found some regression bugs. revert it.

4 years agoEfl.Ui.Radio_Legacy: Prevent unnecessary signal emit for contents
JunsuChoi [Tue, 12 May 2020 15:20:25 +0000 (11:20 -0400)]
Efl.Ui.Radio_Legacy: Prevent unnecessary signal emit for contents

Summary:
When call elm_radio_add, theme_apply of layout is called and
_efl_ui_radio_legacy_efl_ui_widget_theme_apply is called.
Then it calls the signal emit for the icon.
this call is unnecessary.

Test Plan:
Evas_Object *bt;
clock_t start, finish;
double sum= 0 ;
double avg = 0;
double cnt = 1000;

for(int i =0 ; i<(int)cnt; i++)
  {
     start = clock();
     bt = elm_radio_add(win);
     finish = clock();
     sum += (double)(finish-start)/CLOCKS_PER_SEC;
  }
avg = sum / cnt ;
printf("radio avg : %f\n",avg);

[before]
radio avg : 0.000232

[after]
radio avg : 0.000197

Reviewers: akanad, Jaehyun_Cho, Hermet, YOhoho, zmike

Reviewed By: zmike

Subscribers: bu5hm4n, zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11813

4 years agoEfl.Ui.Check_Legacy: Prevent unnecessary signal emit for contents
JunsuChoi [Tue, 12 May 2020 15:20:21 +0000 (11:20 -0400)]
Efl.Ui.Check_Legacy: Prevent unnecessary signal emit for contents

Summary:
When call elm_check_add, theme_apply of layout is called and
_efl_ui_check_legacy_efl_ui_widget_theme_apply is called.
Then it calls the signal emit for the icon.
this call is unnecessary.

Test Plan: N/A

Reviewers: akanad, Jaehyun_Cho, Hermet, YOhoho, zmike

Reviewed By: zmike

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11812

4 years agoEfl.Ui.Progressbar_Legacy: Prevent unnecessary signal emit for contents
JunsuChoi [Tue, 12 May 2020 15:20:18 +0000 (11:20 -0400)]
Efl.Ui.Progressbar_Legacy: Prevent unnecessary signal emit for contents

Summary:
When call elm_progressbar_add, theme_apply of layout is called and
_efl_ui_progressbar_legacy_efl_ui_widget_theme_apply is called.
Then it calls the signal emit for the icon.
this call is unnecessary.

Test Plan: N/A

Reviewers: Hermet, bowonryu, Jaehyun_Cho, YOhoho, zmike

Reviewed By: zmike

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11811

4 years agoEfl.Ui.Button_Legacy: Prevent duplicate signal emit of contents
JunsuChoi [Tue, 12 May 2020 15:20:15 +0000 (11:20 -0400)]
Efl.Ui.Button_Legacy: Prevent duplicate signal emit of contents

Summary:
When call elm_button_add, theme_apply of layout is called and
_efl_ui_button_legacy_efl_ui_widget_theme_apply is called.
Then it calls the signal emit for the icon.
In the button theme, the default of the icon is hidden,
so this call is unnecessary.

Test Plan:
Evas_Object *bt;
clock_t start, finish;
start = clock();
bt = elm_button_add(win);
finish = clock();
printf("duration : %f\n", (double)(finish-start)/CLOCKS_PER_SEC);

Evas_Object *bt;
clock_t start, finish;
double sum= 0 ;
double avg = 0;
double cnt = 1000;

for(int i =0 ; i<(int)cnt; i++)
  {
     start = clock();
     bt = elm_button_add(win);
     finish = clock();
     sum += (double)(finish-start)/CLOCKS_PER_SEC;
  }
avg = sum / cnt ;
printf("button avg : %f %f\n",avg, sum);

[before]
button avg : 0.000215

[after]
button avg : 0.000138

Reviewers: Hermet, YOhoho, Jaehyun_Cho, zmike

Reviewed By: zmike

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11810

4 years agobuild: enable -Wpointer-arith for dev builds
Mike Blumenkrantz [Tue, 12 May 2020 14:56:01 +0000 (16:56 +0200)]
build: enable -Wpointer-arith for dev builds

Summary:
this is a useful warning to avoid errors when calculating pointer offsets
Depends on D11816

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11817

4 years agobuild: fix void* use in pointer maths
Mike Blumenkrantz [Tue, 12 May 2020 14:55:59 +0000 (16:55 +0200)]
build: fix void* use in pointer maths

Summary:
void* is an invalid type for calculating pointer offsets, so ensure that
this is always cast to something else (e.g., char*) in the few cases
where it's been misused

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11816

4 years agobuild: enable -Wfloat-compare by default for dev builds
Mike Blumenkrantz [Tue, 12 May 2020 07:21:52 +0000 (09:21 +0200)]
build: enable -Wfloat-compare by default for dev builds

Summary: Depends on D11799

Subscribers: segfaultxavi, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11800

4 years agostatic/triangulator: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:21:35 +0000 (09:21 +0200)]
static/triangulator: resolve float compare warnings

Summary: Depends on D11798

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11799

4 years agoedje: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:21:26 +0000 (09:21 +0200)]
edje: resolve float compare warnings

Summary: Depends on D11797

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11798

4 years agoexactness: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:21:17 +0000 (09:21 +0200)]
exactness: resolve float compare warnings

Summary: Depends on D11796

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11797

4 years agoexamples: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:21:09 +0000 (09:21 +0200)]
examples: resolve float compare warnings

Summary: Depends on D11795

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11796

4 years agoevas/gl: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:21:00 +0000 (09:21 +0200)]
evas/gl: resolve float compare warnings

Summary: Depends on D11794

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11795

4 years agoevas/filter: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:20:50 +0000 (09:20 +0200)]
evas/filter: resolve float compare warnings

Summary: Depends on D11793

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11794

4 years agoevas: resolve float compare warnings
Mike Blumenkrantz [Tue, 12 May 2020 07:20:42 +0000 (09:20 +0200)]
evas: resolve float compare warnings

Summary: Depends on D11792

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11793

4 years agoEfl.Canvas.Object.Animation: Prevent crash when called .animation_pause_set.
JunsuChoi [Tue, 12 May 2020 14:08:30 +0000 (23:08 +0900)]
Efl.Canvas.Object.Animation: Prevent crash when called .animation_pause_set.

Summary:
When call efl_canvas_object_animation_animation_pause_set,
pd->in is freed by calling efl_canvas_object_animation_stop in _animator_cb.
So when set pause_state, crash occurred
This patch add null check to prevent null access.

Test Plan: N/A

Reviewers: Hermet, kimcinoo, herb, bu5hm4n

Reviewed By: Hermet, bu5hm4n

Subscribers: bu5hm4n, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11815

4 years agoevas_object_textblock: reduce layout calculations
Ali Alzyod [Tue, 12 May 2020 14:00:15 +0000 (10:00 -0400)]
evas_object_textblock: reduce layout calculations

Summary:
This patch reduces calculations for layouting textblock when it is not needed.
Exactly in **evas_object_textblock_render_pre**, layouting was done (if needed) regardless of object visibility.

evas_object_render_pre will async called if object status is changed, for example show->hide

In short words: **We do not layout textblock content if textblock t is hidden.**

```
// suppose textblock is shown
// and user want to hide it and set content in it
// to be visible later on if needed
evas_object_hide(textblock);
evas_object_textblock_text_markup_set(textblock, "Hello World");
//Layouting will be done on textblock regardless of its visiblity, becase render_pre
//will be called and will make it relayout
```

Reviewers: woohyun, zmike, tasn, raster, bu5hm4n

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11508

4 years agobuild: ensure we set ENABLE_LUA_OLD when using a lua interpreter
Stefan Schmidt [Mon, 11 May 2020 15:17:47 +0000 (17:17 +0200)]
build: ensure we set ENABLE_LUA_OLD when using a lua interpreter

We missed to actually set the ENABLE_LUA_OLD. Ross run into some
problems with Lua 5.2 builds and we hope this fixes the issue for him.

Found and fixed by Marcel Hollerbach (only put in the system by me).

Fixes T8705

4 years agoevas_object_textblock: fix to free style properly by edje's style_user_pop
WooHyun Jung [Tue, 12 May 2020 07:46:29 +0000 (16:46 +0900)]
evas_object_textblock: fix to free style properly by edje's style_user_pop

Summary:
When calling style_user_pop in edje, current style was not freed by
style_user_pop in evas properly (since old_ts was not cared at all).

Because of this, memory was leaked by none accessible textblock styles which
should have been freed.

@fix

Test Plan:
1. call edje_object_part_text_style_user_pop
2. call edje_object_part_text_style_user_push with new style-text
3. repeat #1 and #2 several times and check the memory (ps -u)

Reviewers: ali.alzyod, bowonryu, bu5hm4n, zmike

Reviewed By: ali.alzyod

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11801

4 years agoevas_vg_node: deprecate evas_vg_node_geometry_set.
JunsuChoi [Tue, 12 May 2020 02:13:27 +0000 (11:13 +0900)]
evas_vg_node: deprecate evas_vg_node_geometry_set.

Summary:
The geometry info of evas_vg_node is based on Efl.Path_Bounds.
The size of vg_node does not change unless path data is changed.
Therefore, this API does not work normally.

Test Plan: N/A

Reviewers: Hermet, kimcinoo, herb

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11809

4 years agoefl_canvas_vg_node: fix the implemenation function of the efl_canvas_vg_node(evas_obj...
Taehyub Kim [Tue, 12 May 2020 00:12:26 +0000 (09:12 +0900)]
efl_canvas_vg_node: fix the implemenation function of the efl_canvas_vg_node(evas_object_above_get, evas_object_below_get)

Summary:
There are wrong implemenation codes in the above and below get objects of efl_canvas_vg_node functin.
and made it working correctly.

Reviewers: Hermet, jsuya

Reviewed By: jsuya

Subscribers: singh.amitesh, cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8700

Differential Revision: https://phab.enlightenment.org/D11781

4 years agoatspi: Listen to socket when widget window is shown
Prasoon Singh [Tue, 28 Apr 2020 05:48:41 +0000 (11:18 +0530)]
atspi: Listen to socket when widget window is shown

[Problem] in widget app socket is not being listened
[Solution] listen on show

Change-Id: Ia7b7000746ea10aacd22f6c11de936f86cd96c3f
Signed-off-by: Prasoon Singh <prasoon.16@samsung.com>
4 years agocanvas wayland_egl: fix evas gl partial rendering issue.
Hermet Park [Mon, 11 May 2020 05:43:59 +0000 (14:43 +0900)]
canvas wayland_egl: fix evas gl partial rendering issue.

swap mode should not be working if the partial rendering is disabled.
evas gl could toggle on/off this partial rendering,
it should consider to check the condition properly.

support_partial_rendering is a value for global system,
partial_rendering is a value for current context value.

Change-Id: I33ee4d6204897f49dd4808e7885864547c0ab3bf

4 years agotests: ecore_con another cert update to extend the time its valid to 100 years
Stefan Schmidt [Mon, 11 May 2020 15:05:44 +0000 (17:05 +0200)]
tests: ecore_con another cert update to extend the time its valid to 100 years

Instead of forcing someone to look at this again in a couple of years we
can simply extend it to 100 years and expect this code does no longer
exist. :-)

For more details on the steps I used to generate and the problem in
ecore_con please see the original commit 3b835b57ff70659c12a89158cd8d29e3187a42a9

4 years agoecore_evas(module): fix compatibility issue for adding dnd end event handler
Taehyub Kim [Fri, 8 May 2020 04:40:55 +0000 (13:40 +0900)]
ecore_evas(module): fix compatibility issue for adding dnd end event handler

if we don't use this event handler, we cannot know the moment of dnd end from wayland
@fix

Change-Id: I9392a62cde7e1558263f961a411df8b4645b92ee

4 years agocanvas render: don't apply mask if the objects in the map.
Hermet Park [Sun, 10 May 2020 07:03:10 +0000 (16:03 +0900)]
canvas render: don't apply mask if the objects in the map.

Current mask is shared across the objects hierarchy,
children would try to apply this mask on the rendering
when it's enabled by the parent.

but if the map is enabled from one of parents, children should not apply
the parents mask but leaves it to the parent's job.

@fix

4 years agoevas + gl - fix callback-to-app-changes-state issue with pixel data
Carsten Haitzler (Rasterman) [Sat, 9 May 2020 13:02:04 +0000 (14:02 +0100)]
evas + gl - fix callback-to-app-changes-state issue with pixel data

so if we call a callback that has e do something like set new pixel
data in a pixel data provider and we do it from the is pixel inside
call ebcause pixels were marked as dirty - pixels may no longer exist
after that, so re-get to make sure and handle properly.

this was found with tiling + e + shaped windows + moving them around
in tiling. it's a valid efl bug though.

@fix

4 years agoevas/events: fix conditional for emitting source mouse move events
Mike Blumenkrantz [Sat, 9 May 2020 08:25:27 +0000 (09:25 +0100)]
evas/events: fix conditional for emitting source mouse move events

Summary:
this should match all other similar cases when determining whether to
emit the event by checking whether at least one coord has changed, not both
Depends on D11791

Reviewers: devilhorns

Reviewed By: devilhorns

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11792

4 years agoelm: resolve float comparison warnings
Mike Blumenkrantz [Sat, 9 May 2020 08:25:21 +0000 (09:25 +0100)]
elm: resolve float comparison warnings

Summary: Depends on D11790

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11791

4 years agobin/elm: resolve float compare warnings
Mike Blumenkrantz [Sat, 9 May 2020 08:25:17 +0000 (09:25 +0100)]
bin/elm: resolve float compare warnings

Summary: Depends on D11789

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11790

4 years agotests: resolve float comparison warnings
Mike Blumenkrantz [Sat, 9 May 2020 08:25:14 +0000 (09:25 +0100)]
tests: resolve float comparison warnings

Summary: Depends on D11788

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11789

4 years agoedje_cc: consistently use eina macros for all float comparisons
Mike Blumenkrantz [Sat, 9 May 2020 08:25:11 +0000 (09:25 +0100)]
edje_cc: consistently use eina macros for all float comparisons

Summary: Depends on D11787

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11788

4 years agoedje_cc: fix comparisons for possible non-float values
Mike Blumenkrantz [Sat, 9 May 2020 08:25:08 +0000 (09:25 +0100)]
edje_cc: fix comparisons for possible non-float values

Summary:
when defined, BUILD_EDJE_FP causes some of these struct members to not
be floating types and additionally defines special macros to be used
for comparisons to handle this case
Depends on D11786

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11787

4 years agoeina: resolve float comparison warnings
Mike Blumenkrantz [Sat, 9 May 2020 08:25:04 +0000 (09:25 +0100)]
eina: resolve float comparison warnings

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11786

4 years agoatspi: Establish connection and add monitors again after fork reset
Prasoon Singh [Sat, 2 May 2020 08:50:45 +0000 (14:20 +0530)]
atspi: Establish connection and add monitors again after fork reset

[Problem] when fork reset is done dbus handlers are deleted and
connection is unreffed and freed
[Solution] establish connection again

Change-Id: Icba95eb148d2367ac84388265b721a6f84959545
Signed-off-by: Prasoon Singh <prasoon.16@samsung.com>
4 years agodoxygen docs: Fix problems in Evas_Canvas
Xavi Artigas [Fri, 8 May 2020 10:30:13 +0000 (12:30 +0200)]
doxygen docs: Fix problems in Evas_Canvas

- Put some methods in a group so they are accessible and can be linked to.
- Remove references to Efl_ methods, unavailable in Legacy.
- Incorrect param names and typos

4 years agodoxygen docs: Put Eina XAttrs in their own group
Xavi Artigas [Fri, 8 May 2020 10:05:38 +0000 (12:05 +0200)]
doxygen docs: Put Eina XAttrs in their own group

To stop polluting the Eina Tools group.

4 years agodoxygen docs: Complete docs for Eina Thread
Xavi Artigas [Fri, 8 May 2020 09:54:49 +0000 (11:54 +0200)]
doxygen docs: Complete docs for Eina Thread

4 years agodoxygen docs: Fix grouping mess in Eina Simple XML
Xavi Artigas [Fri, 8 May 2020 09:48:32 +0000 (11:48 +0200)]
doxygen docs: Fix grouping mess in Eina Simple XML

Seriously, who came up with such convoluted doxygen code?

4 years agodoxygen docs: Complete docs for Eina Module
Xavi Artigas [Fri, 8 May 2020 09:35:56 +0000 (11:35 +0200)]
doxygen docs: Complete docs for Eina Module

And fix yet another doxygen grouping mess

4 years agodoxygen docs: Complete docs for Eina Log
Xavi Artigas [Fri, 8 May 2020 09:05:50 +0000 (11:05 +0200)]
doxygen docs: Complete docs for Eina Log

4 years agodoxygen docs: Complete docs for Eina CPU functions
Xavi Artigas [Fri, 8 May 2020 08:47:26 +0000 (10:47 +0200)]
doxygen docs: Complete docs for Eina CPU functions

4 years agodoxygen docs: Complete docs for Eina_Rectangle
Xavi Artigas [Fri, 8 May 2020 08:19:40 +0000 (10:19 +0200)]
doxygen docs: Complete docs for Eina_Rectangle

4 years agoevas: use SEQUENTIAL for decoding png file
Shinwoo Kim [Thu, 7 May 2020 07:56:40 +0000 (08:56 +0100)]
evas: use SEQUENTIAL for decoding png file

Summary:
There was a mistake caused by "2a0eeba evas: fix png regression issue".

The flag is_for_data was close_file before this commit.
_evas_image_load_file_internal_head_png closes file if it is called only for
header. So the close_file does not mean is_for_data.

This patch is changing is_for_data to is_for_head, and make
_evas_image_load_file_internal_head_png use EINA_FILE_SEQUENTIAL
for data decoding.

Reviewers: Hermet, jsuya, herb, raster

Reviewed By: raster

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11783

4 years agotests/elm: add callback testing to image memfile test
Mike Blumenkrantz [Fri, 8 May 2020 01:45:27 +0000 (10:45 +0900)]
tests/elm: add callback testing to image memfile test

Summary:
verify that ready is called (preload done) and error is not called

ref D11758

Reviewers: Hermet, jsuya, kimcinoo

Reviewed By: Hermet

Subscribers: stefan_schmidt, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11760

4 years agoci: travis: update Debian to 10.3 and Ubuntu to 20.04
Stefan Schmidt [Thu, 13 Jun 2019 13:08:04 +0000 (15:08 +0200)]
ci: travis: update Debian to 10.3 and Ubuntu to 20.04

Now that the ecore_con tes suite failure are fixed with a renewed ssl
certificate we can update these two distros as well.

4 years agotests: ecore_con: renew test ssl certificate which expired in 2016
Stefan Schmidt [Thu, 7 May 2020 12:41:48 +0000 (14:41 +0200)]
tests: ecore_con: renew test ssl certificate which expired in 2016

For our unit testing of ecore_con we have a sself-signed certificate for
localhost to do some of the tests. It expired in 2016 and this was not
really a problem so far. Starting from Debian 10 and Ubuntu 20.04 this
resulted in failures thought. Either the newer openssl version is more
strict or the config for Debian and Ubuntu is.

Anyway, time to generate a new one. Its valied until Sep 23 12:33:13
2024 GMT and I used the follwoing commands to generate the cert/key and
the second command to rmeove the key passphrase.

openssl req -x509 -newkey rsa:4096 -keyout server.key -out server.pem -days 1600 -subj '/CN=127.0.0.1'
openssl rsa -in server.key -out server2.key

Someone has to do this again in 4 years. if we stillhave this code
around.

4 years agoUse __func__ C99 identifier instead of __FUNCTION__ compiler extension
Vincent Torri [Thu, 7 May 2020 13:27:07 +0000 (09:27 -0400)]
Use __func__ C99 identifier instead of __FUNCTION__ compiler extension

Summary: see http://www.open-std.org/JTC1/SC22/wg14/www/docs/n1124.pdf section 6.4.2.2 page 52

Test Plan: compilation

Reviewers: raster, devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11785

4 years agocxx: Initialize _list and _node with nullptr
João Paulo Taylor Ienczak Zanette [Mon, 4 May 2020 14:48:38 +0000 (14:48 +0000)]
cxx: Initialize _list and _node with nullptr

Fixes CID 1409576 pointing that _node was unitialized.

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Reviewed-by: Felipe Magno de Almeida <felipe@expertisesolutions.com.br>
Differential Revision: https://phab.enlightenment.org/D11770

4 years agoevas: fix png regression issue
Shinwoo Kim [Thu, 7 May 2020 03:15:02 +0000 (12:15 +0900)]
evas: fix png regression issue

Summary:
Accidentally commit "382c580 evas: add support for .9.png file to PNG loader."
adding the 9 patch feature with small code refactoring made use of setjmp
incorrectly.

[Problem]
evas_image_load_file_data_png calls _evas_image_load_file_internal_head_png,
and _evas_image_load_file_internal_head_png calls setjmp and returns without
problem. And png_read_row calls longjmp. This causes jumping into a function
which was exited. Problematic png file will be attached.

[Solution]
Save calling environment i.e. call setjmp, after returning from
_evas_image_load_file_internal_head_png.

Test Plan:
Problematic png file
{F3876983}

And example code.
{F3876986}

Reviewers: Hermet, jsuya, herb

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11782

4 years agocanvas map: draw direct image as possible.
Hermet Park [Thu, 7 May 2020 02:25:13 +0000 (11:25 +0900)]
canvas map: draw direct image as possible.

Since we moved to floating calculation for msaa, this comparsion
is not working for direct image drawing though it's doable.

To resolve this, we can compare floatings regardless of msaa enable.

4 years agoecore_evas: add checkpoints for profiling
Wonki Kim [Tue, 28 Apr 2020 02:01:59 +0000 (11:01 +0900)]
ecore_evas: add checkpoints for profiling

Change-Id: Ibcf73f837ca65af05822c02d6677a94f4f8338f2

4 years agoRevert "ecore_anim: fix animator ticking logic to gain accuarte"
Wonki Kim [Tue, 28 Apr 2020 02:04:34 +0000 (11:04 +0900)]
Revert "ecore_anim: fix animator ticking logic to gain accuarte"

commit 9159047a is a workaround patch which has a chance to make it worse.

This reverts commit 9159047a7b09658419d88021850e0a798fe40c32.

Change-Id: I196c017538641641d7b16b5b07a8090d6b89c5f4

4 years agodocs: Makes Lock(eina_lock) internal API groups in Tizen
Myoungwoon Roy, Kim [Wed, 6 May 2020 07:04:22 +0000 (16:04 +0900)]
docs: Makes Lock(eina_lock) internal API groups in Tizen

Change-Id: Icbfb42f57fb31856789a31ab548cbf10a88280c9

4 years agoelm_genlist: initialize local variables to prevent unexpected results
Jongmin Lee [Wed, 6 May 2020 00:21:07 +0000 (09:21 +0900)]
elm_genlist: initialize local variables to prevent unexpected results

Change-Id: I27a0dd417aab5b48e635e4362c075cfdd1c2ede7

4 years agoecore-evas-drm: Remove useless check for visibility
Christopher Michael [Wed, 6 May 2020 14:06:53 +0000 (10:06 -0400)]
ecore-evas-drm: Remove useless check for visibility

This is already checked above so having a second check here is just
silliness

4 years agoremove arguments of LOGFN in ecore_wayland modules as the are always the same
Vincent Torri [Wed, 6 May 2020 13:18:31 +0000 (09:18 -0400)]
remove arguments of LOGFN in  ecore_wayland modules as the are always the same

Reviewers: raster, devilhorns

Reviewed By: devilhorns

Subscribers: devilhorns, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11777

4 years agoci: travis: switch coverity build back to Fedora 31 image
Stefan Schmidt [Wed, 6 May 2020 12:51:32 +0000 (14:51 +0200)]
ci: travis: switch coverity build back to Fedora 31 image

While it builds there is a problem to pick up the analyzing units for
the scanner. My best guess is that gcc10 is to new and we need to wait
for the next cov-tools release before we can bump this.

4 years agotravis: update asan build to Fedora 32
Stefan Schmidt [Wed, 6 May 2020 12:32:06 +0000 (14:32 +0200)]
travis: update asan build to Fedora 32

With the updated gcc to version 10 the AddressSanitizer plugin seems to
have switched the OneDefionitionRule to 2, which includes reports for
definitions with the same size.

https://github.com/google/sanitizers/wiki/AddressSanitizerOneDefinitionRuleViolation

In our case that breaks heavily due to our usage in eina_module:

==147526==ERROR: AddressSanitizer: odr-violation (0x7fca71b29300):
  [1] size=8 '__eina_module_shutdown' ../src/modules/ecore_imf/xim/ecore_imf_xim.c:821:1
  [2] size=8 '__eina_module_shutdown' ../src/modules/evas/engines/buffer/evas_engine.c:185:1

Disabeling it for our runs.

4 years agoedje_textblock: content_fit size_range update
Ali Alzyod [Wed, 6 May 2020 11:33:42 +0000 (20:33 +0900)]
edje_textblock: content_fit size_range update

Summary: update size range will skip empty font_sizes

Reviewers: woohyun, bowonryu

Reviewed By: woohyun

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11507

4 years agoedje_textblock: optimize priority handling for content_fit options
Ali [Wed, 6 May 2020 11:30:14 +0000 (20:30 +0900)]
edje_textblock: optimize priority handling for content_fit options

Summary: this patch will reduce amount of unnessesary calculations, if user set both font_size_array and size_range for content fit.

Reviewers: woohyun, bowonryu

Reviewed By: woohyun

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11498

4 years agoevas_textblock: clear paragraphs during markup_set
Ali Alzyod [Wed, 6 May 2020 09:40:37 +0000 (10:40 +0100)]
evas_textblock: clear paragraphs during markup_set

Summary: Clear paragraph during markup set, instead of waiting until next text layout to happen, this will release unused memory right away

Reviewers: woohyun, zmike, bu5hm4n, raster

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11773

4 years agospec: resolve migration build error
Jongmin Lee [Wed, 6 May 2020 01:37:11 +0000 (10:37 +0900)]
spec: resolve migration build error

Change-Id: I4085fb2b0d868cb39b10d81a50fd099bd03a4d76

4 years agoci: travis: update more jobs to Fedora32 image
Stefan Schmidt [Tue, 5 May 2020 13:02:29 +0000 (15:02 +0200)]
ci: travis: update more jobs to Fedora32 image

With the edje test suite problems resolved almost all jobs can be moved.
Mingw, exactness and asan need more testing and debugging.

4 years agotests: edje: Fedora32 system fail resolve
Ali Alzyod [Tue, 5 May 2020 12:15:55 +0000 (12:15 +0000)]
tests: edje: Fedora32 system fail resolve

This test seems to fail on Fedora32, we increase font size margin to
prevent it from happening

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D11776

4 years agoci: travis: update first few jobs to Fedora32
Stefan Schmidt [Tue, 5 May 2020 11:50:53 +0000 (13:50 +0200)]
ci: travis: update first few jobs to Fedora32

The rest stays on 31 right now until we worked out some remaining
problems in test suites, asan, etc.

4 years agoecore_drm: ensure we have the symbol available at the right place.
Stefan Schmidt [Tue, 5 May 2020 10:28:39 +0000 (11:28 +0100)]
ecore_drm: ensure we have the symbol available at the right place.

Summary:
Linking under Fedora32 toolchain got more picky. Ecore_drm_device.c
needs the symbol so make sure we have it there.

Reviewers: devilhorns, raster

Reviewed By: raster

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11772

4 years agoexactness_player: do not test for delay_ms
Marcel Hollerbach [Tue, 5 May 2020 08:53:15 +0000 (10:53 +0200)]
exactness_player: do not test for delay_ms

that is a unsigned int, if its 0 the timer is called as fast as
possible. Not doing that breaks exactness.

@fix

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D11774

4 years agoevas_textblock: optimize calculate main format once in layout setup stage
Ali Alzyod [Mon, 4 May 2020 19:25:07 +0000 (15:25 -0400)]
evas_textblock: optimize calculate main format once in layout setup stage

Summary:
This change based on discussion on D9533 , where @smohanty example shows that unnecessary calculation happened on textblock related to lay-outing

Where now _layout_setup will not calculate main format using (_format_fill) unless style has been changed.

expedite commit:
https://git.enlightenment.org/tools/expedite.git/commit/?id=dc6c931dc2e6c240d3e240f24578980c689ab7fc
src/bin/textblock_text_fill_format.c

Test Plan:
```
#define EFL_EO_API_SUPPORT 1
#define EFL_BETA_API_SUPPORT 1
#include <Eina.h>
#include <Efl.h>
#include <Elementary.h>

EAPI_MAIN int
elm_main(int argc, char **argv)
{
   Evas_Object *win, *textblock;

   elm_policy_set(ELM_POLICY_QUIT, ELM_POLICY_QUIT_LAST_WINDOW_CLOSED);

   win = elm_win_util_standard_add("Main", "App");
   elm_win_autodel_set(win, EINA_TRUE);

   int l, r, t, b;
   textblock = evas_object_textblock_add(evas_object_evas_get(win));

   Evas_Textblock_Style *st = evas_textblock_style_new();
   evas_textblock_style_set (st, "DEFAULT='font=Sans font_size=10 color=#000000 wrap=word align=left outline_color=#000 shadow_color=#fff8 shadow_color=#0002 glow2_color=#fe87 glow_color=#f214 underline_color=#00f linesize=40'");
   evas_object_textblock_style_set(textblock, st);
   int sizes[] = {600, 700};
   evas_object_textblock_text_markup_set(textblock, "This test resize text block and keep style (style parsed only once)");
   clock_t start, end;
   start = clock();
   for (int i = 0; i < 10000; i++)
   {
      evas_object_resize(textblock, sizes[i % 2], sizes[i % 2]);
      evas_object_textblock_style_insets_get(textblock, &l, &r, &t, &b);
   }
   end = clock();
   double total_Time1 = ((double)(end - start)) / CLOCKS_PER_SEC;
   printf("total time = %f\n", total_Time1);

   evas_object_size_hint_weight_set(textblock, EVAS_HINT_EXPAND, EVAS_HINT_EXPAND);
   evas_object_size_hint_align_set(textblock, EVAS_HINT_FILL, EVAS_HINT_FILL);
   evas_object_show(textblock);

   evas_object_move(textblock, 0, 0);
   evas_object_resize(textblock, 640, 800);
   evas_object_resize(win, 640, 800);

   evas_object_show(win);
   elm_run();

   return 0;
}
ELM_MAIN()

```

**Old  Code output : total time = 0.096900
New Code output : total time = 0.045580**

Reviewers: smohanty, woohyun, Hermet, bowonryu, cedric, bu5hm4n, zmike

Reviewed By: zmike

Subscribers: zmike, segfaultxavi, bu5hm4n, smohanty, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D9536

4 years agoremove arguments of LOGFN in ecre_x and ecore_wayland as the are always the same
Vincent Torri [Mon, 4 May 2020 13:21:03 +0000 (14:21 +0100)]
remove arguments of LOGFN in ecre_x and ecore_wayland as the are always the same

Reviewers: raster

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11769

4 years agoexamples: cxx: remove forced warning on unfinished APIs
Stefan Schmidt [Tue, 28 Apr 2020 12:04:18 +0000 (14:04 +0200)]
examples: cxx: remove forced warning on unfinished APIs

Personally I find it quite annoying to have these in every build.
Especially if they have not been attempted to get fixed even after we
started to stabilize some part of the unified API.

Reviewed-by: Felipe Magno de Almeida <felipe@expertisesolutions.com.br>
Differential Revision: https://phab.enlightenment.org/D11763

4 years agoecore_evas_x - fix maximize/unmaximize request corner case
Carsten Haitzler (Rasterman) [Sun, 3 May 2020 14:34:42 +0000 (15:34 +0100)]
ecore_evas_x - fix maximize/unmaximize request corner case

state was not managed well.. this will fix that. also a test in
elementary for it to you can test it out.

@fix

4 years agoecore-buffer: Fix formatting
Christopher Michael [Sat, 2 May 2020 18:44:59 +0000 (14:44 -0400)]
ecore-buffer: Fix formatting

NB: No functional changes

4 years agoecore-drm2: Add support for using page_flip_handler2
Christopher Michael [Fri, 1 May 2020 14:26:54 +0000 (10:26 -0400)]
ecore-drm2: Add support for using page_flip_handler2

In our current Ecore_Drm2_Context structure, we are missing support
for using page_flip_handler2. This patch adds that ability. This could
prove useful when working with ecore_drm2 as it allows
page_flip_handler callbacks to setup a handler for page_flip_handler2,
which when the callback gets received, passes back the crtc_id which
the pageflip occured on.

@feature

4 years agoedje - don't set font sizes less than 1
Carsten Haitzler (Rasterman) [Fri, 1 May 2020 11:50:53 +0000 (12:50 +0100)]
edje - don't set font sizes less than 1

this makes evas unhappy and results in noise.

@fix

4 years agobuild - remove useless header checks for uv.h
Carsten Haitzler (Rasterman) [Thu, 30 Apr 2020 23:07:12 +0000 (00:07 +0100)]
build - remove useless header checks for uv.h

we check for uv.h - but never do anything with it. this is configure
noise that just confuses people into thinking we might do something.
we still have dead libuv support but it's enabled with HAVE_LIBUV and
we'd need to link to libuv too... which we don't do and we don't enable
this. removing libuv supporrt entirely is a totally different question
to raise. this here is just removing noise.

4 years agoexactness: remove no longer needed _evas_new()
Stefan Schmidt [Thu, 30 Apr 2020 13:58:31 +0000 (15:58 +0200)]
exactness: remove no longer needed _evas_new()

This has been unused since the move to a preloaded lib and now makes
trouble durign compilation on Fedora32.

@fix

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11767

4 years agoecore_evas: only define these eina errors once
Marcel Hollerbach [Thu, 30 Apr 2020 13:22:01 +0000 (09:22 -0400)]
ecore_evas: only define these eina errors once

Summary:
they are not needed multiple times, infact this is creating multiple
definition errors.

@fix

Reviewers: stefan_schmidt, zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11766

4 years agoelm_fileselector: support expandable items
Marcel Hollerbach [Wed, 29 Apr 2020 15:47:05 +0000 (11:47 -0400)]
elm_fileselector: support expandable items

Summary:
the fileselector can store exactly 1 model and monitor that for changes,
which works normally just well. However, when we expand a items, another
model comes into play, which is not monitored, and no additional
reference is taken. Which means, _process_last will simply delete it,
and elements that did not come into play at the _populate call are never
displayed.

With this commit, we check that we do not have a parent item, and
explictly wait for the model to publish all its items, before we call
_process_last.

Additionally, that showed that the tracking of the list/grid item is
ultimatly buggy, as there is no assertion that the parent model is
really the parent directoy. Hence we need to hand pass the parent item
here, as a fallback option to set the correct parent item.

fixes T8689

Reviewers: zmike

Reviewed By: zmike

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8689

Differential Revision: https://phab.enlightenment.org/D11765

4 years agodoxygen docs: Add missing group braces
Xavi Artigas [Wed, 29 Apr 2020 15:32:41 +0000 (17:32 +0200)]
doxygen docs: Add missing group braces

Without them, these Evil docs had no group and therefore didn't appear anywhere.

4 years agodoxygen docs: Add tests folder to EXAMPLE_PATH
Xavi Artigas [Wed, 29 Apr 2020 15:23:09 +0000 (17:23 +0200)]
doxygen docs: Add tests folder to EXAMPLE_PATH

One Emotion example is located in src/tests instead of src/examples
so look for examples in that folder too.

4 years agodoxygen docs: add missing example references
Xavi Artigas [Wed, 29 Apr 2020 15:14:35 +0000 (17:14 +0200)]
doxygen docs: add missing example references

4 years agodoxygen docs: Fix reference to Elm_Start
Xavi Artigas [Wed, 29 Apr 2020 14:51:23 +0000 (16:51 +0200)]
doxygen docs: Fix reference to Elm_Start

4 years agodoxygen docs: Replace broken URL with doxygen ref
Xavi Artigas [Wed, 29 Apr 2020 14:26:47 +0000 (16:26 +0200)]
doxygen docs: Replace broken URL with doxygen ref

4 years agodoxygen docs: Fix missing links in Ecore_Drm
Xavi Artigas [Wed, 29 Apr 2020 14:17:53 +0000 (16:17 +0200)]
doxygen docs: Fix missing links in Ecore_Drm

Ecore_Drm_Input had SOME docs, so a group for them is created.
Ecore_Drm_Sprite had NO DOCS at all, so the link is just removed.
This is a deprecated library anyway.

4 years agodoxygen docs: Fix broken links in main page
Xavi Artigas [Wed, 29 Apr 2020 13:39:52 +0000 (15:39 +0200)]
doxygen docs: Fix broken links in main page

4 years agoedje: efl: add new EFL_VERSION_1_25 define for new cycle
Stefan Schmidt [Wed, 29 Apr 2020 13:47:36 +0000 (15:47 +0200)]
edje: efl: add new EFL_VERSION_1_25 define for new cycle

Allow things to be marked for the upcoming 1.25 release version while we
develop it.

4 years agobuild: release freeeze over
Stefan Schmidt [Wed, 29 Apr 2020 10:14:20 +0000 (12:14 +0200)]
build: release freeeze over

Going back to normal development mode. Have fun!

4 years agoefl_net_dialer_unix: use compositor socket instead of socket4 for connection
Yeongjong Lee [Fri, 13 Mar 2020 04:28:29 +0000 (13:28 +0900)]
efl_net_dialer_unix: use compositor socket instead of socket4 for connection

since 80c7e8093f8c097b2999e07aa9934dd67509c8a1, fd is -1 when
efl_net_connect_async_with_fd_new is called. this patch change it to use
socket from E compositor because of tizen policy.
(See commit ebfaab3e05f755428aaee021253c8e5255fd388f)

Also, This reverts commit 80c7e8093f8c097b2999e07aa9934dd67509c8a1.

@tizen_fix

Change-Id: Ie824d15ace426f91baa2a9ac1547481f000c5804

4 years agogenlist: need to custom fish_eye effect
HyukSoon Choi [Fri, 10 Apr 2020 05:57:26 +0000 (14:57 +0900)]
genlist: need to custom fish_eye effect

Customisation of fish eye effect is required for spinner and more option
to achieve the UX

"vi_effect" in genlist: enable/disable all items opearation
"item_gap": padding values to be applied between items on fish eye effect in spinner
            parameter 1: position adjustment for the n-1 and n+1 item to center item n
            parameter 2: position adjustment for the n-2 and n+2 item to center item n
"proxy_gap": Padding for proxy for top and bottom item in more option
"proxy_top_offset": Offset adjustment for top item in more option
"proxy_color": To custom color transition

Change-Id: I1a7cfbb035daefca1173f83000c56c14ef9d7288
Signed-off-by: HyukSoon Choi <hs619.choi@samsung.com>
4 years ago[Scrollable Interface] Fix for genlist item selection issue
Godly T.Alias [Wed, 15 Apr 2020 05:09:54 +0000 (10:39 +0530)]
[Scrollable Interface] Fix for genlist item selection issue

Change-Id: I1261427ce251c676bcffc1090ecc6107f9be4466
Signed-off-by: Yukti Pandey <yukti.pandey@samsung.com>
4 years ago[efl]Support for rotary event on page change.
Godly T.Alias [Tue, 14 Apr 2020 04:37:59 +0000 (10:07 +0530)]
[efl]Support for rotary event on page change.
Signed-off-by: Yukti Pandey <yukti.pandey@samsung.com>
Change-Id: I0a10ddf685522d5685ba1beab1bfa469b2adc96c

4 years ago[efl] Fix for scroller not maintaining pan position on resize.
Godly T.Alias [Mon, 13 Apr 2020 08:08:07 +0000 (13:38 +0530)]
[efl] Fix for scroller not maintaining pan position on resize.
Signed-off-by: Yukti Pandey <yukti.pandey@samsung.com>
Change-Id: Iddbdb89c156ed14e0d3dfbb9787edb8715b1b81b

4 years ago[evas_textblock] Fix for [DEREF_OF_NULL.RET.ALLOC]
Prince Kumar Dubey [Tue, 21 Apr 2020 19:32:52 +0000 (01:02 +0530)]
[evas_textblock] Fix for [DEREF_OF_NULL.RET.ALLOC]

Change-Id: I46f5e585f0131273b6b3742b2175dd748d6214b5
Signed-off-by: Prince Kumar Dubey <prince.dubey@samsung.com>
4 years ago[elm_genlist] Fix for [DEREF_OF_NULL.RET.STAT]
Prince Kumar Dubey [Tue, 21 Apr 2020 17:34:08 +0000 (23:04 +0530)]
[elm_genlist] Fix for [DEREF_OF_NULL.RET.STAT]

Change-Id: I1bcf01456f66a169edc10f0a5b6f17acc7d679fd
Signed-off-by: Prince Kumar Dubey <prince.dubey@samsung.com>