Tom Hacohen [Wed, 11 May 2016 12:00:57 +0000 (13:00 +0100)]
Efl network: Remove the no longer needed .Base hack.
Tom Hacohen [Wed, 11 May 2016 11:57:06 +0000 (12:57 +0100)]
Ector generic: Remove .Generic and .Base (hack no longer needed).
Tom Hacohen [Wed, 11 May 2016 11:38:27 +0000 (12:38 +0100)]
Ector renderer cairo: Remove the no longer needed .Base hack.
Tom Hacohen [Wed, 11 May 2016 11:33:30 +0000 (12:33 +0100)]
Ector renderer gl: Remove the no longer needed .Base hack.
Tom Hacohen [Wed, 11 May 2016 11:23:57 +0000 (12:23 +0100)]
Ector renderer software: Remove the no longer needed .Base hack.
Thiep Ha [Wed, 11 May 2016 08:54:45 +0000 (17:54 +0900)]
elm cnp: simplify data preparer
Use eina_strbuf to simplify the data preparer for uri.
This helps remove the complex for loop and also
reduce the memory allocated for uri array.
This also removes redundant checking.
Stefan Schmidt [Wed, 11 May 2016 08:17:11 +0000 (10:17 +0200)]
tests: eio: make sure we cleanup eio_test_xattr_set correctly
We missed tmpstr_del and shutting down eina here. This test segfaults from
time to time on Jenkins. Not sure this was the actual cause but better clean
it up anyway.
Thiep Ha [Wed, 11 May 2016 02:06:42 +0000 (11:06 +0900)]
elm win: change to use legacy api for cocoa engine
Some APIs are move from eo to legacy as in
ab3058d3,
but not all api calls are changed. This causes build break
in mac with cocoa backend.
This patch fixes it.
@fix
Jee-Yong Um [Tue, 10 May 2016 22:46:55 +0000 (15:46 -0700)]
edje_cc: cleanup queued jobs for removed program
Summary:
When a new edje program is created, some jobs like lookup part
(Part_Lookup) or compiling embryo script (Code_Program) are queued.
If program is removed, queued jobs should be removed also.
Reviewers: jpeg, cedric
Reviewed By: cedric
Differential Revision: https://phab.enlightenment.org/D3900
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Andrii Kroitor [Tue, 10 May 2016 22:44:30 +0000 (15:44 -0700)]
edje_edit: fix edje_edit_group_copy
Summary:
Writing copy directly to file.
Old way (adding directly to collection cache) leads to possible segfaults
on edje_collection_cache_flush.
Reviewers: reutskiy.v.v, cedric
Subscribers: jpeg
Differential Revision: https://phab.enlightenment.org/D3904
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Jee-Yong Um [Tue, 10 May 2016 22:18:33 +0000 (15:18 -0700)]
eet: add a helper to setup hash with generic value storage
Summary:
add a macro to setup eet_data_descriptor for hash with
generic value storage.
Test Plan: make check (test case is included in eet test suite)
Reviewers: cedric, raster
Subscribers: jpeg
Differential Revision: https://phab.enlightenment.org/D3925
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
Chris Michael [Tue, 10 May 2016 19:12:53 +0000 (15:12 -0400)]
elput: Add API function to set max allowed pointer position
This patch adds an API function that can be used to restrict mouse
movement from going outside a screen
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 10 May 2016 18:36:35 +0000 (14:36 -0400)]
elput: Fix issue of NULL seat named being passed during input init
When calling elput_input_init we need to check for a valid seat name
being passed in. If no seat name is provided, we will use the default
seat name.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 10 May 2016 18:26:01 +0000 (14:26 -0400)]
ecore-wl2: Remove need for including <sys/param.h>
As we already have MIN & MAX defined in Eina, we no longer need to
include sys/param.h header to get those defines
ref
d00643fcc4ebe2de3d9abf21edd0a7f55d99972c
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 10 May 2016 17:49:38 +0000 (13:49 -0400)]
elput: Add API function to return a list of Elput_Devices on a given
seat
This patch adds a new API function which can be called to retrieve a
list of existing Elput_Devices on a given Elput_Seat.
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Tue, 10 May 2016 17:46:31 +0000 (13:46 -0400)]
elput: Add API function to return a list of seats
This patch adds a new API function which can be used to return a list
of existing Elput_Seat objects from a given manager.
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Mike Blumenkrantz [Tue, 10 May 2016 17:20:56 +0000 (13:20 -0400)]
Revert "elm entry: remove unused funtions"
This reverts commit
2e68debe7b6853dff2d58e1fa89ba7939708871f.
ref T3455
Mike Blumenkrantz [Tue, 10 May 2016 17:16:46 +0000 (13:16 -0400)]
Revert "elm cnp/Wl: Remove duplicated data receive callback"
This reverts commit
f5d10f1961fcfae5ca0204c1bc0c787c96fb4f2c.
this breaks pasting into efl applications on wayland
fix T3455
Daniel Kolesa [Tue, 10 May 2016 17:06:56 +0000 (18:06 +0100)]
eolian: fix C type serialization (handle undefined)
Mike Blumenkrantz [Tue, 10 May 2016 16:29:27 +0000 (12:29 -0400)]
theme: add clip for disabling focus effects on some windows
ref T3408
Tom Hacohen [Tue, 10 May 2016 14:07:44 +0000 (15:07 +0100)]
Eldbus: remove shadowing variable.
FFS people, turn on your compiler warnings.
Tom Hacohen [Tue, 10 May 2016 14:06:04 +0000 (15:06 +0100)]
Efl gfx gradient: Remove the no longer needed .Base hack.
Chris Michael [Tue, 10 May 2016 15:28:26 +0000 (11:28 -0400)]
ecore-drm: Fix potential segfault when setting output mode
If a NULL mode is passed into ecore_drm_output_mode_set, then we try
to disable the given output. If disabling the output fails, we were
trying to print an ERR message which tried to access the NULL mode.
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Amitesh Singh [Tue, 10 May 2016 13:32:33 +0000 (19:02 +0530)]
add work email id into AUTHORS
Tom Hacohen [Tue, 10 May 2016 11:25:44 +0000 (12:25 +0100)]
Efl gfx: Remove the no longer needed .Base hack.
Tom Hacohen [Tue, 10 May 2016 10:36:13 +0000 (11:36 +0100)]
Efl model: Remove the no longer needed .Base hack.
Tom Hacohen [Fri, 6 May 2016 12:24:16 +0000 (13:24 +0100)]
Ecore audio: Correctly namespace now that Eolian supports it better.
Eolian had a restriction due to the C++ generator that classes and
namespaces would be named differently. Now that the C++ generator is
fixed, eolian dropped the restriction and we can finally fix the wrong
namespaces in ecore audio.
Stefan Schmidt [Tue, 10 May 2016 09:10:24 +0000 (11:10 +0200)]
examples: eolian_cxx: fix some problems after changing the colourable namespace
Fix commit
3e8db298f70932ce7b2f5296d44b262a615ea87c the namepsace changed but
only a few files have been fixed for this. Many files did still not compile
with make examples. I fixed some more but stopped at the cxx files with the hope
that the original author would actually do this.
Stefan Schmidt [Tue, 10 May 2016 08:00:42 +0000 (10:00 +0200)]
examples: evas-3d: fix executable name for evas 3d fog example
This list contains the executable names and not the source file names.
Amitesh Singh [Tue, 10 May 2016 06:31:53 +0000 (12:01 +0530)]
elm: rename Elm.Nstate to Efl.Ui.Nstate
Thiep Ha [Tue, 10 May 2016 06:13:27 +0000 (15:13 +0900)]
elm test dnd: replace strcat with eina_strbuf
replace strcat with eina_strbuf.
Cedric BAIL [Mon, 9 May 2016 23:58:31 +0000 (16:58 -0700)]
efl: everyone should now rely on Eina MIN/MAX redefinition.
Cedric BAIL [Mon, 9 May 2016 23:56:38 +0000 (16:56 -0700)]
eina: add MIN/MAX redefine as we use it everywhere.
This is done on an attempt to permanently fix our Windows port. Windows
doesn't have MIN/MAX, so we should always do a ifndef/define in every
piece of code that use it. Of course we always forget and it take times
to notice and fix. We have over the year added it in many private
headers, but as the issue continue to raise again and again, I prefer
to get this fixed in our main header.
Felipe Magno de Almeida [Mon, 9 May 2016 22:10:26 +0000 (19:10 -0300)]
eina: Add eina_promise_race composition function
Added eina_promise_race function that composes multiple
promise objects into a new promise which is fulfilled
when one of the promises are fulfilled, or fails
when one of the promises have failed.
Chris Michael [Mon, 9 May 2016 16:00:35 +0000 (12:00 -0400)]
elput: Add code to handle keybindings for vt switching
This patch makes elput handle receiving the key combinations for vt
switching and perform the actual switch itself. This is done so that
ecore_drm2, ecore_fb, etc, etc do not each need their own code to
handle vt switching.
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Mon, 9 May 2016 15:35:48 +0000 (11:35 -0400)]
elput: Add API function to switch to a given vt
This patch adds a new API function to Elput that can be used to switch
to a given VT. This allows drm2 or enlightenment to switch to a given
virtual terminal in response to keybindings.
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Mon, 9 May 2016 15:05:21 +0000 (11:05 -0400)]
elput: Remove erroneous debug print out
NB: No functional changes, just cleanup of extra debug message and
commented out codeblock
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Mon, 9 May 2016 15:01:30 +0000 (11:01 -0400)]
elput: Support sending of session activate/deactivate event
This patch adds support for notifying when a session gets activated or
deactivated. This will be used in drm2 to notify compositor when to
pause/restart rendering during a vt switch
@feature
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Chris Michael [Mon, 9 May 2016 14:40:23 +0000 (10:40 -0400)]
elput: Remove useless dbus argument fetching
As we do not need minor or fd from the dbus message during input
device resume, we can remove these variables
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Carsten Haitzler (Rasterman) [Mon, 9 May 2016 10:37:48 +0000 (19:37 +0900)]
edje - file set - use vpath for edje files now too like evas images
Thiep Ha [Mon, 9 May 2016 08:28:02 +0000 (17:28 +0900)]
elm entry: add more control for dnd
Adding dnd's enter, leave, position callbacks to let users know
when drag item is entered, leaved; move cursor to follow
drag item's position.
Thiep Ha [Mon, 9 May 2016 08:10:01 +0000 (17:10 +0900)]
elm test dnd: fix multiple items dnd
When we do dnd with multiple items in genlist, gengrid in dnd test,
the items are correctly transfered.
It is caused by incorrect sent data.
This patch fixes by adding new line separation between items' data.
Jean-Philippe Andre [Mon, 9 May 2016 05:37:22 +0000 (14:37 +0900)]
Edje: Fix test case to show proper use of part APIs
The one and only valid use case is calling a function directly on
the part handle, like:
func(part_get(obj, "part"), args).
This does not change the actual proxy object lifecycle or resolve
any potential issues, but this should clear up things in terms
of scope for proper (supported) use.
efl_content_get may become efl_part() and return a temporary object,
valid for a single function call. Those are "details" still to be
refined.
Jean-Philippe Andre [Mon, 9 May 2016 02:46:11 +0000 (11:46 +0900)]
Evas: Remove invalid code
Since Efl.Canvas.Image and Efl.Canvas.Scene3d inherit from
Evas.Image, the two CHECK(isa) can not be reached, and also the
cast from eo_data_scope_get() would be invalid (it should return
NULL as those classes have no private data).
Also, I believe the CHECK() for texture3d can not be reached
either since texture3d objects are not evas objects.
Ping @zmike
Dave Andreoli [Sun, 8 May 2016 20:34:26 +0000 (22:34 +0200)]
Improve elm standard icons test
* use a genlist to speedup and make use of group ability
* manually list the icons (otherwise old and deprecated icons show up)
NOTE: as always this expose a new genlist bug: just change the size
of the icons (with the radio button) and scroll the list...new realized
items get a wrong size.
Dave Andreoli [Sun, 8 May 2016 17:06:03 +0000 (19:06 +0200)]
FDO icons: more work on Places icons
Felipe Magno de Almeida [Sun, 8 May 2016 15:20:27 +0000 (12:20 -0300)]
elementary: Fix example compilation when compiling in another directory
Felipe Magno de Almeida [Sun, 8 May 2016 15:19:48 +0000 (12:19 -0300)]
eolian-cxx: Disable example and fix new naming standard
Fixed naming standard in some tests and disabled the callback test
while Ecore Eo API is still rapidly changing
Felipe Magno de Almeida [Sun, 8 May 2016 14:58:56 +0000 (11:58 -0300)]
evas: Disabled C++ example while Eo API is still rapidly changing
Dave Andreoli [Sun, 8 May 2016 12:23:30 +0000 (14:23 +0200)]
FDO icons: redesigned all the actions icons
Felipe Magno de Almeida [Fri, 6 May 2016 19:32:10 +0000 (16:32 -0300)]
eina: Fix memory leaks in promise
Dave Andreoli [Sat, 7 May 2016 14:55:40 +0000 (16:55 +0200)]
FDO icons: reenable all the new icons, this time in png format
#SecondTry
Dave Andreoli [Sat, 7 May 2016 13:31:54 +0000 (15:31 +0200)]
FDO icons: convert the Status ctx to png
Dave Andreoli [Sat, 7 May 2016 13:00:16 +0000 (15:00 +0200)]
FDO icons: convert the Actions ctx to png
Dave Andreoli [Sat, 7 May 2016 12:36:04 +0000 (14:36 +0200)]
FDO icons: convert the Categories ctx to png
Dave Andreoli [Sat, 7 May 2016 12:16:01 +0000 (14:16 +0200)]
FDO icons: convert the Emblems ctx to png
Dave Andreoli [Sat, 7 May 2016 12:00:48 +0000 (14:00 +0200)]
FDO icons: convert the Places ctx to png
Cedric BAIL [Sat, 7 May 2016 00:01:10 +0000 (17:01 -0700)]
evas: let's reuse what we know when possible to avoid more useless eo_data_scope_get.
Cedric BAIL [Fri, 6 May 2016 23:47:56 +0000 (16:47 -0700)]
edje: do not check against eet if strings where never allocated.
Cedric BAIL [Fri, 6 May 2016 23:47:07 +0000 (16:47 -0700)]
evas: cache result of eo_data_scope_get that are heavily used.
This alone save a good 3% time to our first frame being displayed.
Cedric BAIL [Fri, 6 May 2016 23:18:51 +0000 (16:18 -0700)]
edje: reduce memory usage by 57Kb.
If we do a copy of a string living in the object we are storing in the
same eina_hash, it is obviously better to use the direct pointer version.
Cedric BAIL [Fri, 6 May 2016 22:34:47 +0000 (15:34 -0700)]
ecore: add tests case for Efl.Loop.Fd.
Cedric BAIL [Fri, 6 May 2016 22:33:36 +0000 (15:33 -0700)]
ecore: add Efl.Loop.Fd.
This allow you to monitor fd and get notification using Eo events. I
have not implemented the buffered read as used by X. I think that if
this is useful, we should just do another class to handle bufferred fd.
Cedric BAIL [Fri, 6 May 2016 22:32:32 +0000 (15:32 -0700)]
ecore: update documentation of Efl.Timer.
I am wondering how this is going to affect our legacy API documentation
which has indeed a different lifecycle.
Derek Foreman [Fri, 6 May 2016 20:08:33 +0000 (15:08 -0500)]
ecore_drm: Fix off by 1000 in output rate calculation
This results in wayland wl_output modes having their refresh rates
1000x too high, which has the magical effect of making some native
Qt apps fail with a GTK error when Xwayland crashes on 32-bit systems.
@fix
Fixes T3587
Jean Guyomarc'h [Fri, 6 May 2016 20:12:39 +0000 (22:12 +0200)]
ecore_file: fix memory leak
@fix
Jean Guyomarc'h [Fri, 6 May 2016 20:11:51 +0000 (22:11 +0200)]
eina: fix typos in tmpstr doc
Jean Guyomarc'h [Wed, 4 May 2016 19:59:52 +0000 (21:59 +0200)]
ecore: don't use an unspecified amount of arguments
Jean Guyomarc'h [Wed, 13 Apr 2016 19:34:26 +0000 (21:34 +0200)]
ecore/getopt: slightly improve documentation
Jean Guyomarc'h [Mon, 11 Apr 2016 20:41:03 +0000 (22:41 +0200)]
ecore: fix typo in documentation of Ecore_Exe_Event_Del
Chris Michael [Fri, 6 May 2016 19:48:14 +0000 (15:48 -0400)]
efl: Make building Ecore_Wayland optional
This patch makes building Ecore_Wayland library disabled by default
but provides an option to enable it (if needed).
@fix
Signed-off-by: Chris Michael <cpmichael@osg.samsung.com>
Mike Blumenkrantz [Fri, 6 May 2016 19:43:01 +0000 (15:43 -0400)]
ecore-wl2: only print socket ERR messages when not attempting to reconnect
reduces spamming in logs
Mike Blumenkrantz [Fri, 6 May 2016 19:32:58 +0000 (15:32 -0400)]
eldbus: send exit/quit event upon disconnecting the session bus
if an application is connected to the session bus, being disconnected
should indicate that the session has been terminated and all applications
in that session should also be terminated
this ensures that no zombie processes will remain after the session bus dies
Felipe Magno de Almeida [Sun, 17 Apr 2016 18:47:33 +0000 (15:47 -0300)]
eina: Add progress notify callback feature for Promise Owners
Add a way for users of the promise owner to get notified when a
promise progress is registered. Also added a convenience composition
function that creates a promise which is fulfilled when another
promise has a progress notification.
Daniel Kolesa [Fri, 6 May 2016 11:56:16 +0000 (12:56 +0100)]
eolian: lift restriction on namespace/class name conflicts in library
This will allow us to name classes the same as namespaces.
Carsten Haitzler (Rasterman) [Fri, 6 May 2016 07:09:07 +0000 (16:09 +0900)]
Ecore.h wants efl_timer.eo.legacy.h so lets make sure it is there
so ensure it is generated AND installed. otherwise good luck building
anything against efl.
Yeshwanth Reddivari [Fri, 6 May 2016 05:53:17 +0000 (11:23 +0530)]
Toolbar: Use orientation interface APIs instead of horizontal set/get APIs
Test Plan: elementary_test
Reviewers: raster, jpeg, cedric, singh.amitesh
Reviewed By: singh.amitesh
Subscribers: cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3890
Cedric Bail [Fri, 6 May 2016 04:55:43 +0000 (21:55 -0700)]
ecore: fix legacy use of double free scenario.
Cedric Bail [Fri, 6 May 2016 04:32:37 +0000 (21:32 -0700)]
Revert "ecore timer - fuix up segv storm that has crept in with frees"
This reverts commit
a13570c17c97bb9407c24bcf78ab28eaa4541ad9.
This doesn't really fix the problem which is hidden by eo capability to not
crash on bad unref. With legacy API you are allowed to do a ecore_timer_del
and also return EINA_FALSE. In that case you have a double eo_del (which is
luckily protected) and a double free (that is not). It does crash on the
double free, but the issue is a lifecycle issue. Will bring a better patch
for this.
Carsten Haitzler (Rasterman) [Fri, 6 May 2016 02:06:56 +0000 (11:06 +0900)]
ecore timer - fuix up segv storm that has crept in with frees
freeing already freed data, not clearing up pointers to freed data
once freed etc. etc. etc. - segv land. fix that up so we can use efl
again. :(
Cedric BAIL [Thu, 5 May 2016 23:18:39 +0000 (16:18 -0700)]
ecore: make efl.timer depend on efl.loop_user.
Cedric BAIL [Thu, 5 May 2016 23:18:18 +0000 (16:18 -0700)]
ecore: add an efl.loop_user class.
Cedric BAIL [Thu, 5 May 2016 21:18:27 +0000 (14:18 -0700)]
elementary: make clean, clean edj correctly.
Cedric BAIL [Thu, 5 May 2016 21:11:38 +0000 (14:11 -0700)]
elementary: add resulting edj to allow make dist to work on a clean directory.
Due to dependencies from elementary codegen on the edj files triggered by the
generation of .h and .c, we actually need to provide the edj as it would
otherwise to build edje_cc prior to any make dist.
As the edje file was added after in the git repository, you may have some
timestamp issue to fix before doing a make dist. Try touch on all codegen_example
files.
T3588
Dave Andreoli [Thu, 5 May 2016 21:17:46 +0000 (23:17 +0200)]
Disable the new svg icons for the moment :(
As explained in the ml we are hitting an issue with the svg loaders not included in efl
Dave Andreoli [Thu, 5 May 2016 20:41:09 +0000 (22:41 +0200)]
Fix distcheck wrt new fdo icons
distcheck is still not working for the lacks of the svg loader in efl
Dave Andreoli [Thu, 5 May 2016 18:44:07 +0000 (20:44 +0200)]
fdo icons: added Categories icons
this icons are blindly copyed from the RaveX/Faenza themes,
they are a bit alien in looks, I consider them as a
placeholder for some more eflish icons
Cedric BAIL [Thu, 5 May 2016 18:40:28 +0000 (11:40 -0700)]
po: update position in files.
Cedric BAIL [Thu, 5 May 2016 18:40:14 +0000 (11:40 -0700)]
elementary: unbreak make distcheck.
Cedric BAIL [Thu, 5 May 2016 18:39:59 +0000 (11:39 -0700)]
edje: unbreak EXTRA_DIST.
Dave Andreoli [Tue, 3 May 2016 18:07:04 +0000 (20:07 +0200)]
Elm icons: keep all old icons for compat
Dave Andreoli [Mon, 2 May 2016 21:41:17 +0000 (23:41 +0200)]
Add (quite) all the status icons
Dave Andreoli [Mon, 2 May 2016 18:13:33 +0000 (20:13 +0200)]
Add all the Action icons, and TODO placeholder for ALL FDO icon names.
I really don't like the look of those new icons, will improve them in a second step
Dave Andreoli [Sun, 1 May 2016 14:52:45 +0000 (16:52 +0200)]
Initial structure for a proper FDO compliant icon theme.
The idea here is to provide a "real" fdo icon theme along the icons in the elm edje theme.
To minimize file redundancy the icons are profided directly in a fdo structure and picked
in icon_fdo.edc. Also I'm using all the icons in svg format so that we don't need to provide
all the creazy sizes as per fdo spec.
More icons will comes in the next commits
Tom Hacohen [Thu, 5 May 2016 15:08:08 +0000 (16:08 +0100)]
Eo composite: change composite objects to not be tied to parent
This commit breaks behaviour!
Re-parenting no longer detaches composite objects, so watch out.
Now you can have an object be a composite object of an object although
it's not its child. This allows widgets to do things like having an
object as the child of a child object while still making it a composite
object to the main object.
With this change, composite objects don't keep a reference to the child,
but instead composite "bonds" are implicitly removed when either the
parent or the child are destructed.
Mike Blumenkrantz [Thu, 5 May 2016 14:45:59 +0000 (10:45 -0400)]
wayland: implement session recovery
add support for reconnecting wayland applications if the compositor dies
disconnect -> destroy gl ctx + image textures -> block rendering ->
reconnect -> create gl ctx -> create image textures -> unblock rendering ->
sprinkle special seasoning on top -> just like ma used to make
#SamsungFeatures
@feature
Mike Blumenkrantz [Thu, 5 May 2016 14:45:04 +0000 (10:45 -0400)]
evas: add internal functions for unsetting+regenerating image data
in the event that the global gl context changes, all engine image data
must be destroyed and then re-created in order to continue rendering
successfully
Mike Blumenkrantz [Thu, 5 May 2016 14:42:26 +0000 (10:42 -0400)]
wayland_egl: return immediately in eng_window_free() when null is passed
Mike Blumenkrantz [Thu, 5 May 2016 14:41:27 +0000 (10:41 -0400)]
wayland_egl: unset global egl context if used wl_display changes
as long as evas uses a global gl context any change to the wl_display
must trigger a full re-create on that context
Mike Blumenkrantz [Thu, 5 May 2016 14:38:31 +0000 (10:38 -0400)]
wayland_egl: ensure gl context destruction when unsetting wl_display
extra refcounting during setup in this case will prevent the context
destruction, leading to a dead gl context which is unable to render