Jean-Philippe Andre [Thu, 7 Apr 2016 06:36:20 +0000 (15:36 +0900)]
Eo: Mark class_get() as weak APIs
While eolian-gen was generating EWAPI (weak) class_get()
symbol declarations, they were implemented as EAPI (strong).
Not sure if this mismatch had any significant effect.
This patch tries to address T3423 (windows build).
The mismatch between EAPI and EWAPI might be the problem.
Jean-Philippe Andre [Thu, 7 Apr 2016 06:25:42 +0000 (15:25 +0900)]
Ecore: Try to fix build for windows
Move efl.h above ecore.h to not mess with EAPI's dllimport
vs. dllexport definition. This addresses T3423.
Jean-Philippe Andre [Thu, 7 Apr 2016 05:40:10 +0000 (14:40 +0900)]
Efl: Add internal strong symbol to fix build on GCC < 5.3
This fixes a crash in ecore_init, calling a weak function from
libefl that was resolved to NULL.
So, here's a fun thing happening with GCC < 5.3. Since
a1a506e13e2
all EOAPI and EO class_get() functions are weak symbols. This means
that all APIs inside libefl.so are weak.
As a result, gcc linker with --as-needed skipped linking to libefl
since not a single strong symbol from libefl was required by
libecore. This is actually a bug in gcc linker since we do in fact
use symbols from libefl, just weak ones.
GCC 5.3 seems to be fixed, so people with GCC 5.3+ will not
experience any build/runtime issue. The current patch is
a workaround that bug, by artifically creating a strong symbol
required by ecore.
Other libraries than ecore might also need to call
__efl_internal_init, if they end up not being linked to libefl.
Cedric Bail [Wed, 6 Apr 2016 21:34:45 +0000 (14:34 -0700)]
ecore: attempt to fix windows build by putting header in a more logical place.
Felipe Magno de Almeida [Sat, 2 Apr 2016 01:50:28 +0000 (22:50 -0300)]
eolian: add Eolian support for Eina Promises
Add a promise object to allows Eolian interface to include promises
as a way to have asynchronous value return and composibility.
The usage is like this in a .eo file:
class Foo {
methods {
bar {
params {
@inout promise: Promise<int>;
}
}
}
}
Which will create the following API interface:
void foo_bar(Eo* obj, Eina_Promise** promise);
and a Eina_Promise_Owner for the implementation, like this:
void _foo_bar(Eo* obj, Private_Data* pdata, Eina_Promise_Owner* promise);
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Felipe Magno de Almeida [Fri, 11 Mar 2016 20:22:59 +0000 (17:22 -0300)]
eo: add before and after macro hooks for API generation functions
Add two parameters for macros that generate API functions in Eo so
that the generation can be customized with macros used by Eolian.
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Tom Hacohen [Wed, 6 Apr 2016 10:17:05 +0000 (11:17 +0100)]
Revert "Eo: Fix rare crash after call_resolve"
I'm reverting this because according to jpeg it was possibly fixed in
5284b62e930f0bef0ed3125b3a485e0599451ef8.
I reverted this patch after his fix and followed his reproduction cases
and it seems that his second patch does indeed fix this issue so this
patch is no longer needed.
This reverts commit
0862b9d08384bc1d862b90952130ec988f56b33b.
Tom Hacohen [Wed, 6 Apr 2016 09:16:24 +0000 (10:16 +0100)]
Eo tests: Also test function calls in reinit test.
Since we cache ops we also need to check function calls work
when we reinit eo, not just class_get functions.
This commit essentially verifies that
5284b62e930f0bef0ed3125b3a485e0599451ef8 was done correctly.
Carsten Haitzler (Rasterman) [Wed, 6 Apr 2016 09:18:58 +0000 (18:18 +0900)]
efl build - ecore - add efl link as a PUBLIC link as efl headers it in
I added Efl.h to Ecore.h - because we really want to use efl
interfaces ... like everywhere. so add it to the link too.
Jean-Philippe Andre [Wed, 6 Apr 2016 07:36:55 +0000 (16:36 +0900)]
elm_test: Fix warning with clang
Invalid enum type
Jean-Philippe Andre [Wed, 6 Apr 2016 07:34:53 +0000 (16:34 +0900)]
elm_glview: Fix warning with clang
I'm using the same gcc construct to initiliaze a complex
struct with {} instead of {0}.
Jean-Philippe Andre [Wed, 6 Apr 2016 07:33:16 +0000 (16:33 +0900)]
elm_config: Fix typo and warning
A major typo (hard to find with the naked eye) was present in
elm_config's list of text & color classes. See D3487.
I'm pretty sure this feature has not been used at all.
Jean-Philippe Andre [Wed, 6 Apr 2016 06:16:37 +0000 (15:16 +0900)]
els_box: Fix warning with clang
warning: comparison of constant 100 with expression of type
'Evas_Aspect_Control' is always true
[-Wtautological-constant-out-of-range-compare]
Jean-Philippe Andre [Wed, 6 Apr 2016 06:22:20 +0000 (15:22 +0900)]
elm: Fix some warnings with clang
warning: missing field 'desc' initializer
[-Wmissing-field-initializers]
Solution: use gcc extension to init structs with {}.
This is a bit ugly, but having too many warnings leads to
ignoring them and not noticing valid ones.
The warning is triggered because the first member of Eo_Event
is not a primitive type (it's a struct _Eo_Opaque *).
Jean-Philippe Andre [Wed, 6 Apr 2016 07:24:33 +0000 (16:24 +0900)]
elm_web: Fix warning (missing initializer)
Jean-Philippe Andre [Wed, 6 Apr 2016 05:10:15 +0000 (14:10 +0900)]
elm_image: Fix warning (wrong enum type)
Jean-Philippe Andre [Wed, 6 Apr 2016 05:02:05 +0000 (14:02 +0900)]
Eo: Fix function cache after eo reinit
The function call resolve cache may be broken after an eo
shutdown + init cycle, leading to calls to invalid functions.
This adds an static uint for each and every single EO API
entry point, as well as an extra if() check.
Now I'm not sure if the previous commit
0862b9d08384bc1d8 is
still necessary.
Jean-Philippe Andre [Wed, 6 Apr 2016 04:27:09 +0000 (13:27 +0900)]
Efl vpath: Fix usage of @class function
Calling an @class function with a real object is not safe.
Missing check somewhere? I believe this should have failed safely.
Jean-Philippe Andre [Wed, 6 Apr 2016 02:25:29 +0000 (11:25 +0900)]
Eo: Fix rare crash after call_resolve
It seems that calling a @class function with an EO object
(that was not the required Eo_Class) lead to a situation
where func->func was NULL. And that meant a crash after
call_resolve.
The proper fix is to properly call a @class function with a
class object.
Yeshwanth Reddivari [Wed, 6 Apr 2016 01:46:04 +0000 (10:46 +0900)]
Interface: Flip and orientation interface
Summary:
Added flip and orientation interface and used them in evas_image.
Removed efl_image_orientation_set API and used efl_orientation_set and efl_flip_set API.
In implementation part, converted enums back and forth in order to keep current implementation as it is.
Test Plan: src/examples/evas/evas-images5.c
Reviewers: singh.amitesh, raster, tasn, herdsman, woohyun, cedric, felipealmeida, jpeg
Subscribers: cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3844
Jean-Philippe Andre [Mon, 4 Apr 2016 12:39:29 +0000 (21:39 +0900)]
elm_layout: Attach object to iterator
This is part of a new API, not a fix
This was missing in
3c40ebb99887ba4c74c40e4509c49cc675877886
Jean-Philippe Andre [Mon, 4 Apr 2016 09:57:26 +0000 (18:57 +0900)]
elm_box: Minor fixes
Stephen Houston [Wed, 6 Apr 2016 01:39:35 +0000 (20:39 -0500)]
Elementary: Don't send selected signal to a part with no selected state. Removes warnings.
Mike Blumenkrantz [Tue, 5 Apr 2016 19:10:57 +0000 (15:10 -0400)]
elm_win: update wayland opaque region during pre-flush...always
@fix
Felipe Magno de Almeida [Tue, 5 Apr 2016 18:38:49 +0000 (15:38 -0300)]
eo-cxx: Fix warning on weak symbols with inline functions
Removed the EOAPI which is not necessary as the function is already inlined
Felipe Magno de Almeida [Sat, 2 Apr 2016 01:24:19 +0000 (22:24 -0300)]
ecore: add tests for promises with ecore_thread_promise_run
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Felipe Magno de Almeida [Fri, 1 Apr 2016 18:37:42 +0000 (15:37 -0300)]
ecore: add promise for Ecore_Thread
Add ecore_thread_promise_run function that returns a Promise
and runs function in another thread which you can set the
value on a Eina_Promise_Owner.
Eina_Promise* promise;
Ecore_Thread* thread = ecore_thread_promise_run
( &function_heavy, &cancellation_function, private_data,
sizeof(ValueType), &promise);
This calls function_heavy on another thread and returns
the Ecore_Thread and a Eina_Promise as an out-parameter.
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Felipe Magno de Almeida [Fri, 1 Apr 2016 18:30:37 +0000 (15:30 -0300)]
eina: add tests for promises
Felipe Magno de Almeida [Fri, 1 Apr 2016 17:49:58 +0000 (14:49 -0300)]
eina: add promise
Add a promise object that will allows Eolian interface to include promises
as a way to have asynchronous value return and composibility.
To understand better, let see the coming usage in a .eo file:
class Foo {
methods {
bar {
params {
@inout promise: Promise<int>;
}
}
}
}
Which will create the following API interface:
void foo_bar(Eo* obj, Eina_Promise** promise);
and the equivalent declaration for implementation.
However, the API function will instantiate the Promise for the user
and the implementer of the class automatically. So the user of this
function will treat it as a @out parameter, while the developer of the
function will treat it like a @inout parameter.
So, the user will use this function like this:
Eina_Promise* promise; // No need to instantiate
foo_bar(obj, &promise);
eina_promise_then(promise, callback);
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 15:43:14 +0000 (11:43 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing
ecore_x_randr_crtc_panning_area_set function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 14:14:19 +0000 (10:14 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing
ecore_x_randr_edid_display_interface_type_get function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:58:11 +0000 (09:58 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing function
ecore_x_randr_screen_backlight_level_set
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:46:57 +0000 (09:46 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing API function
ecore_x_randr_edid_dpms_off_available_get
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:44:29 +0000 (09:44 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing
ecore_x_randr_edid_dpms_standby_available_get function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:41:40 +0000 (09:41 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing
ecore_x_randr_edid_display_serial_get function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:39:33 +0000 (09:39 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing
ecore_x_randr_edid_display_ascii_get function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:24:42 +0000 (09:24 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing API function
ecore_x_randr_edid_dpms_standby_available_get
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:23:22 +0000 (09:23 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing function
ecore_x_randr_edid_dpms_available_get
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:20:53 +0000 (09:20 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for missing API function
ecore_x_randr_edid_manufacturer_serial_number_get
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:18:10 +0000 (09:18 -0400)]
ecore-xcb: Add implementation for missing Ecore_X API
This patch adds an xcb implementation for ecore_x_randr_edid_model_get
function which was missing from ecore-xcb.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:16:49 +0000 (09:16 -0400)]
ecore-xcb: Add implementation for missing API function
This patch adds an xcb implementation for
ecore_x_randr_edid_manufacturer_model_get function
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:11:41 +0000 (09:11 -0400)]
ecore-xcb: Add missing API function
ecore_x_randr_edid_manufacturer_name_get
This patch adds an implementation inside ecore-xcb for missing API
function ecore_x_randr_manufacturer_name_get
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 13:06:06 +0000 (09:06 -0400)]
ecore-xcb: Add missing API function ecore_x_randr_edid_info_has_valid_checksum
This patch adds an implementation for
ecore_x_randr_edid_info_has_valid_checksum which was missing in the
ecore-xcb codebase.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 12:59:24 +0000 (08:59 -0400)]
ecore-xcb: Add missing ecore_x_randr_output_crtc_set API function
This patch adds definition of missing API function
ecore_x_randr_output_crtc_set.
NB: This function also has no code in the xlib implementation
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 5 Apr 2016 11:42:56 +0000 (07:42 -0400)]
evas-wayland-shm: Update copyright notice
As portions of this code have been derived from existing code in
Weston, we should also be including their copyright/license text to
give credit.
Fixes T3421
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Jean-Philippe Andre [Tue, 5 Apr 2016 11:44:57 +0000 (20:44 +0900)]
eolian: Fix test cases
So sorry I missed this in my weak API patch! Bad JP didn't run
make check.
Jean-Philippe Andre [Tue, 5 Apr 2016 11:10:17 +0000 (20:10 +0900)]
Makefile: Add efl as internal dependency for ecore
This commit broke the build:
-
7c8b2da2863bb734236f5758288a00eb0d8c047a
Each and every single library including ecore now must
also include efl as well (for Efl.h).
Jean-Philippe Andre [Tue, 5 Apr 2016 08:19:55 +0000 (17:19 +0900)]
Evas smart: Mark all smart clippers as no_render
This allows doing ugly things such as setting a clipper/mask
on any smart object, without having to worry about seeing a
white rectangle cover the whole screen.
While the above scenario is not a great idea (bypass elementary in
its handling of the clipper), having a big white rect covering the
entire UI seems even worse.
I wonder if elm objects shouldn't simply allow user clippers, and
somehow manage them along with the smart object internal clipper.
Jean-Philippe Andre [Tue, 5 Apr 2016 05:54:29 +0000 (14:54 +0900)]
Evas: Fix GL shader selection for afill
AFILL should be used only for certain images, which are
actually native external or dynamic content images.
All normal image don't need the AFILL flag, since they should
have proper argb and alpha flag.
Jinyong Park [Tue, 5 Apr 2016 08:29:12 +0000 (13:59 +0530)]
popup : fix create scroller bug when list item add.
Summary:
data-> scr = scroller object
data-> scroll = scrollable flag.
in _list_add function, when scroller object is not created, do _create_scroller.
if (!sd->scroll) _create_scroller(obj);
must change to
if (!sd->scr) _create_scroller(obj);
@fix
Reviewers: jaehwan, id213sin, cedric, raster, singh.amitesh
Reviewed By: singh.amitesh
Subscribers: herb, jpeg
Differential Revision: https://phab.enlightenment.org/D3866
Carsten Haitzler (Rasterman) [Mon, 4 Apr 2016 12:50:30 +0000 (21:50 +0900)]
evas: image obj - use vath to find images (sync not async)
Carsten Haitzler (Rasterman) [Mon, 4 Apr 2016 08:24:29 +0000 (17:24 +0900)]
elm - add vpath paths for application resources
this now allows vpath files to be looked uop using app directory
resources like @app.data/file.jpg ... if the object uses vpath to do
its hunting.
@feature
Carsten Haitzler (Rasterman) [Mon, 4 Apr 2016 08:12:38 +0000 (17:12 +0900)]
efl ecore - init vpath on ecore init so we have a base object working
this inits a new vpath object and adds it at priority 0 to the vpath
manager so you can use the vpath manager to create vpath file objects
and look things up.
@feature
Carsten Haitzler (Rasterman) [Fri, 1 Apr 2016 01:45:07 +0000 (10:45 +0900)]
efl: vpath subsystem
this adds a core vpath subsystem to efl that allows paths like:
~/file.jpg
~user/file.jpg
(:tmp/file.jpg
(:config/file.jpg
(:videos/file.mp4
(:pictures/file.jpg
(:app.config/mycfg.cfg
etc. to be translated/looked up. it is desitgned to be async and call
event callbacks when ready. the reason for this complexity is fo in
future also handle:
file:///whatever/file.jpg
http://blah.com/file.jpg
https://blah.com/file.jpg
ssh://blah.com:~/file.jpg
etc.
@feature
Daniel Zaoui [Tue, 5 Apr 2016 05:39:26 +0000 (08:39 +0300)]
Popup: don't set wrap type if ELM_WRAP_NONE is given
If ELM_WRAP_NONE is forbidden, the internal variable should not be set
if this value is given as parameter.
Taehyub Kim [Tue, 5 Apr 2016 04:46:38 +0000 (10:16 +0530)]
elc_popup: support siganl emit for the main layout of popup
Summary:
Popup has the main layout to show popup view, but it is not resize object of popup.
As the result, we cannot emit the signal to the main layout when we want to change the state of the popup.
So, popup inherited a layout signal emit and redirect the signal to the main layout.
Test Plan:
1. add the custom signal in elm/popup/base/default layout
2. send the custom siganl in application code
Reviewers: woohyun, raster, cedric, Hermet, singh.amitesh
Reviewed By: singh.amitesh
Subscribers: jpeg
Differential Revision: https://phab.enlightenment.org/D3852
WooHyun Jung [Tue, 5 Apr 2016 02:23:04 +0000 (11:23 +0900)]
elm_win: newest is always unfocused when window just gets focus
Derek Foreman [Mon, 4 Apr 2016 21:57:30 +0000 (16:57 -0500)]
wayland_shm: Fix resize optimization
We're supposed to allocate a large pool at startup and use it for
resizing to save pool allocations. However, this was broken and
we ended up allocating both a large pool and a proper sized pool
every resize.
This restores correct behaviour.
Mike Blumenkrantz [Mon, 4 Apr 2016 21:20:49 +0000 (17:20 -0400)]
elm_win: conditionally emit focus and maximize signals in frame creation
@fix
Mike Blumenkrantz [Mon, 4 Apr 2016 20:30:59 +0000 (16:30 -0400)]
elm_win: update opaque region while fullscreened
@fix
Mike Blumenkrantz [Mon, 4 Apr 2016 20:30:33 +0000 (16:30 -0400)]
elm_win: unset opaque region when alpha is set
@fix
Derek Foreman [Mon, 4 Apr 2016 18:08:55 +0000 (13:08 -0500)]
wayland-egl: Fix use after free
eng_window_use() uses outbuf->redirect, so the clever code in
eng_outbuf_reconfigure that avoided setting it to NULL to check
if we needed to recreate the redirect was not very clever at all.
Derek Foreman [Mon, 4 Apr 2016 16:12:06 +0000 (11:12 -0500)]
wayland_egl: Fix redirect to texture
Previous redirect to texture approach broke multiple window applications
by having only a single redirect per context. This approach allows
multiple redirections.
Chris Michael [Mon, 4 Apr 2016 12:54:45 +0000 (08:54 -0400)]
ecore-xcb: Implement missing keyrouter and keygrab functions for xcb
This patch implements missing API functions (in ecore xcb) for
ecore_x_window_keygrab_set, ecore_x_window_keygrab_unset,
ecore_x_e_keyrouter_set, ecore_x_e_keyrouter_get. These
were missing from the initial commit of keygrab/keyrouter code.
Fixes T3377
ref
5c3a08433aa5797582f6cda12b3bb51b89f367d0
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Mon, 4 Apr 2016 12:41:27 +0000 (08:41 -0400)]
ecore-xcb: Fix wrong initialization
As Ecore_X_Window is an 'int' type, we should not be initializing it
to NULL. This causes compiler warnings about initialization makes
integer from pointer without a cast.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Jean-Philippe Andre [Mon, 4 Apr 2016 05:26:20 +0000 (14:26 +0900)]
Evas x11: Fix warnings and invalid calls in engine
This follows commit
f10672dd7429dd98dd3b7d88d9c3c63aac392a40
which apparently forgot some changes in xlib and eglfs.
I hope there are no remains of the old native bind/unbind
functions. Also... I can't build everything, so I'm just hoping
this works. Note: GL and SW native images are not using the
same internal api prototypes.
Ji-Youn Park [Mon, 4 Apr 2016 01:44:49 +0000 (10:14 +0830)]
Elm_image: remove resizable_up and resizable_down Eo api
risizable_up and resizable_down api will be deprecated
after creating the apis to deal with image content using scale type
Carsten Haitzler (Rasterman) [Mon, 4 Apr 2016 00:51:12 +0000 (09:51 +0900)]
evas - gif loader - handle missing colormap
follow on from
dd90b6afadf706aafec9e53a6b1efa8f899ab277 - this handled
if a gif file has no colormap (it'll decode as blank now as a colormap
of all 0's is used instead).
@fix
Jean Guyomarc'h [Sat, 2 Apr 2016 20:19:29 +0000 (22:19 +0200)]
evas_gl_cocoa: fix segmentation fault at program exit
Programs crashed on a segmentation fault when the last window was
closed. The eng_output_idle_flush() function was removed... but
since gl_cocoa does not properly use *_generic modules, the
output_idle_flush() function called by the render engine was
garbage (hence the segfault).
Now nothing is done... but at least we don't crash anymore.
Jean Guyomarc'h [Sat, 2 Apr 2016 20:17:53 +0000 (22:17 +0200)]
autotools: on OSX warn about prefix being /usr
With Apple's SIP, nobody is allowed to modify the system
(excepted /usr/local). If SIP is enabled, the installation
will just fail.
Mike Frysinger [Sat, 2 Apr 2016 20:19:52 +0000 (16:19 -0400)]
minor spelling fixes
Dave Andreoli [Sat, 2 Apr 2016 11:18:29 +0000 (13:18 +0200)]
tab--
Carsten Haitzler (Rasterman) [Sat, 2 Apr 2016 07:02:07 +0000 (16:02 +0900)]
elm win - null out freed data on smart del
this fixes double-frees of strings in oddball cases where a deleted
window that is not done deleting is accessed.
@fix
Carsten Haitzler (Rasterman) [Sat, 2 Apr 2016 04:22:11 +0000 (13:22 +0900)]
evas - gif loader - be a little more optimal in pixel lookups on decode
Carsten Haitzler (Rasterman) [Sat, 2 Apr 2016 03:25:52 +0000 (12:25 +0900)]
evas: gif loader - fix out of bounds access on cmap of invalid pixels
if gif has example 4 colors in colormap, pixels provided still can
hold values higher than 3 (4, 8, 255 etc.) ass a pixel is still a
byte. it should not, but it could. technically it'd be nice for gitlib
to pad its palette out to 256 entires to ensure this cant be a
problem, but it doesn't have to , so make a local copy of the cmap
when decoding pixels and pad out to 256 entires (using color 0 as any
value > pallette ize is invalid anyway so any color will do).
this fixes a possible security attack vector in reading memory out of
bounds of an allocated array. not very far out of bounds - but enough
to cause a crash - ie a dos attack, (not to inject code though).
@fix
Cedric BAIL [Fri, 1 Apr 2016 20:54:40 +0000 (13:54 -0700)]
elementary: install elementary cxx pkgconfig file.
T3397
Cedric BAIL [Fri, 1 Apr 2016 18:50:20 +0000 (11:50 -0700)]
elementary: install desktop files in the correct directory.
Fix for T3397.
Mike Blumenkrantz [Fri, 1 Apr 2016 20:22:45 +0000 (16:22 -0400)]
theme: fix a bunch of the fixed: X Y edje errors
Mike Blumenkrantz [Fri, 1 Apr 2016 19:54:37 +0000 (15:54 -0400)]
ecore-drm: do not attempt to destroy sprite outputs which have never repainted
silences a surprisingly large number of errors on shutdown
@fix
Chris Michael [Fri, 1 Apr 2016 16:22:02 +0000 (12:22 -0400)]
ecore-wl2: Enable releasing keyboard, pointer, and touch if supported
This removes a fixme where we were not using wl_(pointer, keyboard,
touch)_release functions as they rely on newer versions of the wayland
protocol. As we now rely on 1.10 of wayland, these can be enabled now.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Stefan Schmidt [Fri, 1 Apr 2016 13:40:01 +0000 (15:40 +0200)]
ecore_timer: remove redundancy in property docs
Streamline this a bit to document the property itself and only extra information
the the methods if needed.
Stefan Schmidt [Fri, 1 Apr 2016 13:01:48 +0000 (15:01 +0200)]
ecore_timer: document properties
Make sure that we actually document the property itself and not only the
property methods.
Chris Michael [Fri, 1 Apr 2016 11:56:54 +0000 (07:56 -0400)]
evas-software-generic: remove native.func.data variable and data argument of native calblacks.
Evas Image should be independent of render engine.
So remove native.func.data member of RGBA_Image, Evas_GL_Image
struct. And remove data argument,too.
ref
f10672dd7429dd98dd3b7d88d9c3c63aac392a40
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Fri, 1 Apr 2016 11:34:43 +0000 (07:34 -0400)]
ecore-wl2: Move structures above functions
NB: This is just a cosmetic issue, no functional changes
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Derek Foreman [Mon, 28 Mar 2016 17:56:23 +0000 (12:56 -0500)]
evas-wayland-egl: Add www protocol handling to wayland-egl engine
Use the new post-processing API and www extension to implement
CSS effects for wayland-egl applications.
Signed-off-by: Derek Foreman <derekf@osg.samsung.com>
Signed-off-by: Mike Blumenkrantz <zmike@osg.samsung.com>
Derek Foreman [Mon, 28 Mar 2016 17:06:46 +0000 (12:06 -0500)]
gl_common: Add API for redirecting render to texture
New API allows a context to be redirected to texture. When rendering is
complete the texture can be unbound from the frame buffer and used for
post-processing effects.
Signed-off-by: Derek Foreman <derekf@osg.samsung.com>
Signed-off-by: Mike Blumenkrantz <zmike@osg.samsung.com>
Mike Blumenkrantz [Thu, 24 Mar 2016 17:46:42 +0000 (12:46 -0500)]
ecore_evas wayland: store relevant www data in engine info
this accumulates all data relevant to client-side www into
the Evas_Engine_Info_Wayland_Egl struct so the client can react to it.
Signed-off-by: Derek Foreman <derekf@osg.samsung.com>
Mike Blumenkrantz [Thu, 24 Mar 2016 17:44:21 +0000 (12:44 -0500)]
ecore-evas wayland: use www protocol when available
handle www protocol events and store the provided data into engine data
Signed-off-by: Derek Foreman <derekf@osg.samsung.com>
Mike Blumenkrantz [Thu, 24 Mar 2016 17:19:42 +0000 (12:19 -0500)]
ecore_wl2: implement www extension for client-side use
handling for global binding and signal prop
Signed-off-by: Derek Foreman <derekf@osg.samsung.com>
Carsten Haitzler (Rasterman) [Fri, 1 Apr 2016 08:37:37 +0000 (17:37 +0900)]
evas - fix leak because cutouts_fre .. doesnt free - it just resets to 0
this works with
7166e6b85994b19a29f05c9e2b6d75a314a3cb91 and fixes a
leak added because ... free does not free!
evas_common_draw_context_cutouts_real_free(0 now actually frees the
rects, but evas_common_draw_context_cutouts_free() before did not.
@fix (follow on from
7166e6b85994b19a29f05c9e2b6d75a314a3cb91)
Ji-Youn Park [Fri, 1 Apr 2016 08:49:14 +0000 (17:19 +0830)]
Elm_image: remove object_get from eo and object_size_get to view.size.get
remove elm_image_object_get frome eo.
change elm_image_object_size_get to efl_gfx_view_size_get
SangHyeon Lee [Fri, 1 Apr 2016 07:53:57 +0000 (16:53 +0900)]
support elm_object_item_style_set/get on genlist items
SangHyeon Lee [Fri, 1 Apr 2016 07:40:16 +0000 (16:40 +0900)]
fix item_class refcount bug when class updated
SangHyeon Lee [Fri, 1 Apr 2016 07:26:57 +0000 (16:26 +0900)]
fix genlist content leak in reausable case
Ji-Youn Park [Fri, 1 Apr 2016 06:12:01 +0000 (14:42 +0830)]
Elm_image: remove Elm_Image_Orient and orient property from EO.
Ji-Youn Park [Fri, 1 Apr 2016 06:06:55 +0000 (14:36 +0830)]
Elm_image: move preload_disabled API from eo to legacy.
Ji-Youn Park [Fri, 1 Apr 2016 05:44:21 +0000 (14:14 +0830)]
Elm_image: remove elm_image_fill_outside_get
This api will be removed.
elm_image_fill_outside_get is same as !elm_image_fill_inside_get.
elm_image_fill_outside_set function is too.
Jean-Philippe Andre [Thu, 31 Mar 2016 05:19:08 +0000 (14:19 +0900)]
Eolian: Mark all EO class_get() as weak
This follows the previous commit for the same reasons.
Jean-Philippe Andre [Thu, 31 Mar 2016 05:02:49 +0000 (14:02 +0900)]
Eo: Mark all Eo APIs as weak
This marks all EOAPI functions with GCC weak attribute.
This allows two things:
- replace functions
- link at runtime and check if functions exist
The purpose of this patch is to exploit these two features
of weak symbols. The first goal is to allow applications
to build binaries against later versions and run on earlier
versions of EFL without any run-time link error. Those errors
simply prevent applications to even start if they were using
any function that's not present in the old version of EFL.
Now all that needs to be done on the application side is to
do either of:
- if (efl_version > xx) { call_weak_symbol() }
- if (call_weak_symbol) { call_weak_symbol() }
In the future, we can also imagine providing a compatibility
library that would replace EFL's internal APIs with a newer
version. This would let apps use new EFLs on platforms that
don't update fast enough.
I am now pushing this patch as an experiment to see what breaks,
but I expect no problem.
@feature
Minkyoung Kim [Thu, 31 Mar 2016 06:55:15 +0000 (15:55 +0900)]
evas : remove native.func.data variable and data argument of native calblacks.
Summary:
Evas Image should be independent of render engine.
So remove native.func.data member of RGBA_Image, Evas_GL_Image struct.
And remove data argument,too.
Test Plan: Local test, Tizen3.0 mobile, Desktop englitenment
Reviewers: jpeg, spacegrapher, wonsik
Subscribers: cedric, dkdk
Differential Revision: https://phab.enlightenment.org/D3850