Tom Hacohen [Mon, 8 Jul 2013 15:31:36 +0000 (16:31 +0100)]
eo2: Eo2 first commit.
Daniel Zaoui [Thu, 3 Apr 2014 13:20:58 +0000 (16:20 +0300)]
Eolian: add support for functions scope.
This patch adds support for protected functions.
In the .eo file, the scope (public by default) has to be added before
the function name e.g:
protected foo ...
To access the protected APIs, #define (CLASS)_PROTECTED is needed e.g:
#define ELM_BUTTON_PROTECTED
Daniel Zaoui [Sun, 30 Mar 2014 13:22:14 +0000 (16:22 +0300)]
Eolian: generation of Eo2.
The generation of Eo2 can be enabled by defining EO flag in
common_funcs.h.
Daniel Kolesa [Wed, 9 Apr 2014 14:48:41 +0000 (15:48 +0100)]
eina: a highly irritating error in eina_file documentation
Wonguk Jeong [Wed, 9 Apr 2014 10:45:30 +0000 (19:45 +0900)]
ecore_evas_extn: update plug image object after disconnection
Summary:
plug image object was not cleaned after server disconnection.
specifically, image data was cleaned (NULL), but it was not updated for real on screen
--> set dirty of plug image object
@fix
Test Plan: elementary_test -> open window socket -> open window plug -> close window socket -> plug should be cleaned up
Reviewers: raster, cedric
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D703
wonguk.jeong [Wed, 9 Apr 2014 10:21:30 +0000 (19:21 +0900)]
ecore_evas_extn: initialize server handle on server deletion
Summary:
Since plug sends message by using deleted server handle (dangling pointer),
I could see glorious error message as below:
*** ECORE ERROR: Ecore Magic Check Failed!!!
*** IN FUNCTION: ecore_con_server_send()
ERR<842>:ecore lib/ecore/ecore.c:729 _ecore_magic_fail() Input handle pointer is NULL!
ERR<842>:ecore lib/ecore/ecore.c:740 _ecore_magic_fail() *** NAUGHTY PROGRAMMER!!!
*** SPANK SPANK SPANK!!!
*** Now go fix your code. Tut tut tut!
Therefore, I initialized server handle on server deletion
@fix
Test Plan: create socket -> create plug -> destroy socket -> mouse move on plug area
Reviewers: raster, cedric
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D701
Felipe Magno de Almeida [Wed, 9 Apr 2014 10:12:46 +0000 (19:12 +0900)]
eina-cxx: Modified eina C++ log use syntax
Summary:
Modified syntax for eina C++ log to imitate the macro call as returning a stream. So, instead of:
EINA_CXX_DOM_LOG_CRIT(efl::eina::global_domain, "foo " << 5);
It is now used as:
EINA_CXX_DOM_LOG_CRIT(efl::eina::global_domain) << "foo " << 5;
Which more closely resambles using IOStreams in C++.
@feature
Reviewers: cedric, barbieri, smohanty
Reviewed By: barbieri
CC: cedric
Differential Revision: https://phab.enlightenment.org/D623
Felipe Magno de Almeida [Wed, 9 Apr 2014 10:11:46 +0000 (19:11 +0900)]
eet-cxx: Renamed a few local variables and avoided other warnings
Summary: Renamed a few local variables to avoid excessive warnings with -Wshadow warning. Also made construction of a few objects pass all arguments for explicit initialization to avoid other harmless warning.
Reviewers: cedric, tasn, smohanty, raster
Reviewed By: raster
CC: cedric
Differential Revision: https://phab.enlightenment.org/D698
Chris Michael [Wed, 9 Apr 2014 08:44:16 +0000 (09:44 +0100)]
ecore-evas-wl: Reduce useless compositor redraws during render update
and update input & opaque regions after resizing.
@bugfix: We do not need to call ecore_wl_window_damage & commit here.
The damages are already handled in the evas engine for both shm & egl.
Those damages are sent to the compositor Already from the evas engine,
so we don't need to send the same damages twice. This reduces more
useless compositor redraws as we are not constantly sending damages &
calling commit twice for every frame.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 9 Apr 2014 08:41:02 +0000 (09:41 +0100)]
ecore-wl: Optimize setting of window input region
@bugfix: Store the input region into the window structure, and don't
re-add a duplicate input region (done by comparison of saved region).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 9 Apr 2014 08:38:16 +0000 (09:38 +0100)]
ecore-wl: Store window input region into structure
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 9 Apr 2014 08:34:18 +0000 (09:34 +0100)]
ecore-wl: Optimize opaque region setting for surface and reduce
unnecessary calls to surface commit.
@bugfix: We don't need to call surface_commit if we are just setting
the opaque (or input) region(s). This reduces some overhead on the
compositor side (and thus reduces useless redraws). Also, optimize
opaque_region_set to not Reset to the same region; we do this by
storing the opaque region into the window structure, and on calls to
opaque_region_set we can compare those values and not re-apply the
same opaque region.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Yossi Kantor [Thu, 3 Apr 2014 10:58:34 +0000 (13:58 +0300)]
Eolian: Legacy generator generates const return value for legacy header if needed
Yossi Kantor [Wed, 2 Apr 2014 13:12:23 +0000 (16:12 +0300)]
Eolian: Legacy generator - no semicolon after EINA_WARN_UNUSED_RESULT
Stefan Schmidt [Tue, 8 Apr 2014 14:22:59 +0000 (16:22 +0200)]
rg_etc: Fix local shadow problems
Make it clear which local variable we really want to use by changing
the names. All of these seem to be fine but this can really bite us
so have better clarity here.
Stefan Schmidt [Tue, 8 Apr 2014 09:06:49 +0000 (11:06 +0200)]
build: Give correct path for test files
Another breakage coming in with the rg_etc changes. Image files
have been missing the correct location to pick them up for distcheck.
Stefan Schmidt [Tue, 8 Apr 2014 08:56:42 +0000 (10:56 +0200)]
build: Fix make distcheck for tgv saver
Classical copy and paste bug in Makefle.am. The saver src dir only has
the *save* code.
Jean-Philippe Andre [Tue, 8 Apr 2014 08:23:04 +0000 (17:23 +0900)]
Evas gl: Remove unused variable
Stefan Schmidt [Tue, 8 Apr 2014 08:45:44 +0000 (10:45 +0200)]
rg_etc: Really fix debug build.
Finally finishing what
2e8c7cad3a89dfce5b9bf95f42a91da625036ea5 tried to fix.
Jean-Philippe Andre [Tue, 8 Apr 2014 06:57:27 +0000 (15:57 +0900)]
ETC1 encoding: fix typo in TGV and EET savers
If quality is >30 && <=70 then choose medium params, not low.
Spotted by Snacker, thanks.
Yossi Kantor [Tue, 1 Apr 2014 14:35:04 +0000 (17:35 +0300)]
Eolian: Fix legacy generation.
Semicolon was misplaced when function flags are needed.
Daniel Zaoui [Sun, 6 Apr 2014 14:10:05 +0000 (17:10 +0300)]
Eolian: clean API.
defines used as keys for internal hash tables are now replaced by
functions giving access to the internal data.
Daniel Zaoui [Sun, 6 Apr 2014 13:41:45 +0000 (16:41 +0300)]
Eolian: rename enums.
Add the EOLIAN namespace to the function types.
zmike [Mon, 7 Apr 2014 17:15:40 +0000 (13:15 -0400)]
reformat edje_player
ffs these tabs
Mike Blumenkrantz [Mon, 7 Apr 2014 13:54:42 +0000 (09:54 -0400)]
+eeze_udev_get()
@feature
Ryuan Choi [Mon, 7 Apr 2014 01:21:19 +0000 (10:21 +0900)]
configure: Fix the possible build break after
83af91c8b
Added eet into internal dependencies for ecore_evas_connect, ethumb, ethumb_client
I got below errors for ecore_evas_connect, ethumb, ethumb_client
lib/evas/.libs/libevas.so: undefined reference to `eet_data_image_colorspace_get'
lib/evas/.libs/libevas.so: undefined reference to `eet_data_image_read_to_cspace_surface_cipher'
collect2: error: ld returned 1 exit status
Carsten Haitzler (Rasterman) [Sat, 5 Apr 2014 13:21:13 +0000 (22:21 +0900)]
fix swap buffers with damage to not detect if ext str is not there
@fix
Carsten Haitzler (Rasterman) [Fri, 4 Apr 2014 10:35:16 +0000 (19:35 +0900)]
ecore-con - deal with internal buffer growing over 2g in size
@fix this fixes a corner case where you may buffer 2g or more of data
in ecore-con buffers. this leads to a stall. @fix
Jean-Philippe Andre [Fri, 4 Apr 2014 09:35:22 +0000 (18:35 +0900)]
Evas filters: Remove useless include
Jean-Philippe Andre [Fri, 4 Apr 2014 09:06:26 +0000 (18:06 +0900)]
Evas: Fix harmless typo in eet image loader
Spotted by Snacker, thanks.
Stefan Schmidt [Fri, 4 Apr 2014 07:39:28 +0000 (09:39 +0200)]
rg_etc: Fix debug build
In
4053911e we tried to fix the debug build and failed. The function API
actually asks for coords as first parameter. Our nightly builds expose
this debug build and have been failing due to this.
It also makes clear that the debug part of this code was never really
used upstream...
Chris Michael [Fri, 4 Apr 2014 06:51:06 +0000 (07:51 +0100)]
evas-drm: Remove all async_page_flip options/references
Async page flip can cause tearing, is not supported on all cards, and
apparently requires a specific libdrm patchlevel...in general, more
trouble than it's worth, so let's just remove it.
@bugfix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Fri, 4 Apr 2014 06:31:45 +0000 (07:31 +0100)]
evas-drm: Default async page flip to false as this can cause tearing
Async page flip can cause visual tearing, so disable by default for
all cards.
@bugfix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
wonguk.jeong [Fri, 4 Apr 2014 03:16:39 +0000 (12:16 +0900)]
ecore_evas_extn: fix doxygen sample
Summary:
fix doxygen sample according to D690
@fix
Reviewers: raster
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D697
Jean-Philippe Andre [Fri, 4 Apr 2014 01:28:45 +0000 (10:28 +0900)]
Evas cserve2: Fix crash in elm_test GLView
This whole cache/cache2 API duplication is a complete mess.
Jean-Philippe Andre [Thu, 3 Apr 2014 10:18:52 +0000 (19:18 +0900)]
Evas gl: Fix colors when using ETC1 textures
We need to use the RGBA shader to invert R and B otherwise
those will be inverted on the screen.
Jean-Philippe Andre [Thu, 3 Apr 2014 09:39:53 +0000 (18:39 +0900)]
Evas: Fix TGV loader to properly copy ETC1 data
There was some invalid arithmetics with the buffer offset.
Jean-Philippe Andre [Thu, 3 Apr 2014 08:28:57 +0000 (17:28 +0900)]
Evas gl: Fix dlsym usage for glCompressedTexImage2d
This symbol should be part of the loaded libraries, can be found
using dlsym, even if eglGetProcAddress() returns NULL.
Add etc1 flag in the debug output.
Jean-Philippe Andre [Thu, 3 Apr 2014 03:26:46 +0000 (12:26 +0900)]
Evas gl: Use implicit cast to DATA8 with image.data8
data8 is there precisely to avoid casting the iamge data when
handling it as DATA8.
This is purely a cosmetic change
Jean-Philippe Andre [Wed, 2 Apr 2014 10:03:03 +0000 (19:03 +0900)]
Evas: Add support for AGRY88 in image creation functions
- evas_common_rgba_image_from_data
- evas_common_rgba_image_colorspace_set
- evas_common_image_colorspace_normalize
Jean-Philippe Andre [Wed, 2 Apr 2014 10:01:05 +0000 (19:01 +0900)]
Evas: Add support for AGRY88 in evas_common_rgba_image_from_copied_data
This should fix a potential abort().
Carsten Haitzler (Rasterman) [Thu, 3 Apr 2014 11:57:04 +0000 (20:57 +0900)]
since event is for server, bdata->image is null anyway
wonguk.jeong [Thu, 3 Apr 2014 11:55:38 +0000 (20:55 +0900)]
ecore_evas_extn: give proper event data to callbacks
Summary:
extn send ecore event when plug is connected/disconnect
(ECORE_EVAS_EXTN_CLIENT_ADD/ECORE_EVAS_EXTN_CLIENT_ADD)
By the way, the event data is currently "Ecore_Evas_Engine_Buffer_Data" which should be hidden
Moreover, when I make multiple sockets,
I need a event data to distinguish which socket is connected by client.
Therefore, I recommend to use Ecore_Evas as event data.
@fix
Reviewers: raster, Hermet, woohyun, cedric, seoz
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D690
Carsten Haitzler (Rasterman) [Thu, 3 Apr 2014 10:56:04 +0000 (19:56 +0900)]
it seems the if 0'd code was wrong and ecore win32 doesnt support nocursor
Tom Hacohen [Thu, 3 Apr 2014 10:52:00 +0000 (11:52 +0100)]
Revert "tests: Add a case for ellipsis in complex markup text."
Opened a task (1151) for it and should revert this when ready to fix.
This reverts commit
9128ac82a0c0b6081d0065891d746768be21ccd5.
Daniel Zaoui [Thu, 3 Apr 2014 07:24:44 +0000 (10:24 +0300)]
Eolian: use legacy_prefix: null for interfaces.
legacy set to null was used for every function to not generate a legacy
API. legacy_prefix set to null indicates no legacy has to be generated
for all the class.
Carsten Haitzler (Rasterman) [Wed, 2 Apr 2014 23:06:04 +0000 (08:06 +0900)]
unbreak edje multisense after eolian changes to ecore_audio
fixes break in
70b39368e61e6740e741f8ff34bb0c1bddbaaacd
Tom Hacohen [Wed, 2 Apr 2014 13:54:16 +0000 (14:54 +0100)]
Ecore X: Correctly handle autogen generated ecore_x_version.h
This is by no means a built_source, and it should not be cleaned.
Fixes T1148.
Tom Hacohen [Wed, 2 Apr 2014 13:34:47 +0000 (14:34 +0100)]
Ecore audio: Fix casting of int to void *.
When doing that, intptr_t shoud be used as well.
Yossi Kantor [Thu, 27 Mar 2014 12:05:41 +0000 (14:05 +0200)]
Eolian: Integration of Ecore Audio In Tone
Yossi Kantor [Thu, 27 Mar 2014 11:34:01 +0000 (13:34 +0200)]
Eolian: Integration of Ecore Audio Out Pulse
Yossi Kantor [Thu, 27 Mar 2014 09:53:26 +0000 (11:53 +0200)]
Eolian: Integration of Ecore Audio Out Sndfile
Yossi Kantor [Wed, 26 Mar 2014 16:30:35 +0000 (18:30 +0200)]
Eolian: Integration of Ecore Audio In Sndfile
Yossi Kantor [Wed, 26 Mar 2014 16:01:36 +0000 (18:01 +0200)]
Eolian: Integration of Ecore Audio Out
Yossi Kantor [Wed, 26 Mar 2014 15:01:08 +0000 (17:01 +0200)]
Eolian: Integration of Ecore Audio In
Yossi Kantor [Wed, 26 Mar 2014 08:17:38 +0000 (10:17 +0200)]
Eolian: Integration of Ecore Audio
Carsten Haitzler (Rasterman) [Wed, 2 Apr 2014 11:29:03 +0000 (20:29 +0900)]
ecore-evas - fix object cursor to not delete the same cursor when set
this fixes a misbehavior with ecore evas object cursors when you set
one, the old one is deleted, but if the old is the same, the new one
you set gets deleted, rather than just updated.
@fix
Daniel Zaoui [Wed, 2 Apr 2014 11:23:26 +0000 (14:23 +0300)]
Eolian: remove Eo API defined in evas_private.h
We don't need it more because Eolian generates it already.
Daniel Zaoui [Wed, 2 Apr 2014 11:04:12 +0000 (14:04 +0300)]
Eolian/Lexer: support data type for events.
Tom Hacohen [Wed, 2 Apr 2014 09:06:57 +0000 (10:06 +0100)]
Evas textblock: Minor code cleanup.
Moved a global variable into a local scope, as it's no longer needed globally.
Cedric BAIL [Wed, 2 Apr 2014 07:40:59 +0000 (16:40 +0900)]
evas: always allocate an image.data pixels buffers even YUV content.
fix T1146.
Daniel Zaoui [Wed, 2 Apr 2014 07:08:27 +0000 (10:08 +0300)]
Eolian: fix warnings.
Daniel Zaoui [Wed, 2 Apr 2014 06:24:30 +0000 (09:24 +0300)]
Eolian/Generator: don't generate legacy if legacy_prefix is "null".
For new classes that don't need legacy, instead of setting legacy null
for all the functions, legacy_prefix can be set to "null" to not
generate legacy.
However, if, for example, only one function among 50 need legacy, you
can specify it by setting for this function the legacy token.
Yossi Kantor [Mon, 24 Mar 2014 11:56:18 +0000 (13:56 +0200)]
Eolian: Integration of Ecore Parent
Yossi Kantor [Mon, 24 Mar 2014 11:23:14 +0000 (13:23 +0200)]
Eolian: Integration of Ecore Animator
Yossi Kantor [Mon, 24 Mar 2014 08:29:39 +0000 (10:29 +0200)]
Eolian: Integration of Ecore Idle Exiter
Yossi Kantor [Mon, 24 Mar 2014 08:09:18 +0000 (10:09 +0200)]
Eolian: Integration of Ecore Idle Enterer
Yossi Kantor [Mon, 24 Mar 2014 11:21:28 +0000 (13:21 +0200)]
Eolian: Explicit generation of either method or property in implemented functions
Youngbok Shin [Wed, 2 Apr 2014 06:57:16 +0000 (15:57 +0900)]
tests: Add a case for ellipsis in complex markup text.
Summary: Evas textblock can't cut off text properly when it has separated items.
Reviewers: tasn, woohyun, raster
Reviewed By: raster
CC: cedric, herdsman
Differential Revision: https://phab.enlightenment.org/D667
Cedric BAIL [Wed, 2 Apr 2014 06:18:46 +0000 (15:18 +0900)]
emotion: do not segv when no module are found.
Jean-Philippe Andre [Wed, 2 Apr 2014 03:43:58 +0000 (12:43 +0900)]
Eina: Fix mistake in a previous patch
Suffix length was not computed properly.
Also, the change on eina_mkstemps was a @feature:
add file extension support when creating temporary files
Jean-Philippe Andre [Wed, 2 Apr 2014 03:17:27 +0000 (12:17 +0900)]
Evas benchmark: Add .tgv extension to saved files
This will make the saver actually do something, instead of
just skipping the save as no extension = unknown format :)
Jean-Philippe Andre [Tue, 25 Mar 2014 07:54:57 +0000 (16:54 +0900)]
Evas filters: Add template optimizable files for box blur
The new files (i386, sse3 and neon) are basically empty and fallback
to the C version. This is just to pave the way for full low-level
optimization... if someone has the time and skills to do it :)
Add both Alpha and RGBA template files.
Cedric BAIL [Wed, 2 Apr 2014 05:21:59 +0000 (14:21 +0900)]
evas: fix use of unitialized data in eet loader.
should fix T1144.
Jean-Philippe Andre [Tue, 1 Apr 2014 08:55:09 +0000 (17:55 +0900)]
Evil: Add support for mkstemps
This adds filename suffix support for temporary files on Windows.
Jean-Philippe Andre [Tue, 1 Apr 2014 07:50:46 +0000 (16:50 +0900)]
Eina: Add support for file extensions in eina_mkstemp
Using mkstemps
Cedric BAIL [Wed, 2 Apr 2014 01:31:34 +0000 (10:31 +0900)]
autotools: better fix for rewriting content of src/Makefile
Cedric BAIL [Wed, 2 Apr 2014 01:31:06 +0000 (10:31 +0900)]
rg_etc: fix typo in debug code.
Daniel Zaoui [Tue, 1 Apr 2014 13:12:38 +0000 (16:12 +0300)]
Eolian: Proper use of legacy prefix in legacy function definition.
Yossi Kantor [Mon, 31 Mar 2014 06:28:13 +0000 (09:28 +0300)]
Eolian: Legacy generator - dont generate @param[out] ret if return value is specified
Yossi Kantor [Sun, 30 Mar 2014 13:13:23 +0000 (16:13 +0300)]
Eolian: No need for --eo version when generating legacy header
Tom Hacohen [Tue, 1 Apr 2014 13:21:48 +0000 (14:21 +0100)]
Eo: all classes are allowed in extension list.
All classes are allowed, because all classes can be used as interfaces in
order to override behaviour. This is especially needed for mixins and broke
the eo2 tests.
Cedric BAIL [Tue, 1 Apr 2014 12:58:36 +0000 (21:58 +0900)]
autotools: fix to many argument on shell script during make check
The autogenerated check macro by autotools lead to a huge number of file
being added on the command line of a script, when the only things it does
is to test for the definition of those strings.
Hoping this is going to fix the problem. If that doesn't we will need to write
our own CHECK macro here.
Cedric BAIL [Fri, 28 Mar 2014 08:51:44 +0000 (17:51 +0900)]
evas: make eet negociate it's colorspace with evas.
Cedric BAIL [Fri, 28 Mar 2014 07:05:45 +0000 (16:05 +0900)]
eet: add negociation colorspace support to Eet.
Cedric BAIL [Thu, 27 Mar 2014 04:12:12 +0000 (13:12 +0900)]
evas: add ETC1 texture format support to Evas.
Cedric BAIL [Mon, 24 Mar 2014 10:56:17 +0000 (19:56 +0900)]
Cedric BAIL [Mon, 24 Mar 2014 10:55:45 +0000 (19:55 +0900)]
evas: change TGV internal encoding to account for GPU needs of duplicated border.
With OpenGL, the border of a texture are not "well" defined. So interpolation at
the border can result in weird/bad looking texture border. To avoid that we do
duplicate the border in all direction at the time of the texture upload. But with
ETC1 it is not possible as the border are grouped with 15 others pixels. It needs
to be done at saving time. So internally we do have an image that would be of
size width + 2 pixels and height + 2 pixels.
Cedric BAIL [Fri, 21 Mar 2014 02:54:27 +0000 (11:54 +0900)]
evas: support uploading AGRY88 and GRY8 directly to GPU.
Cedric BAIL [Fri, 21 Mar 2014 02:53:57 +0000 (11:53 +0900)]
evas: handle premultiplying of AGRY88.
Cedric BAIL [Fri, 21 Mar 2014 01:48:04 +0000 (10:48 +0900)]
evas: don't forget that we can allocate AGRY8 now.
Cedric BAIL [Fri, 21 Mar 2014 01:47:03 +0000 (10:47 +0900)]
evas: fix png loader to actually produce lower resolution content when asked.
Cedric BAIL [Thu, 20 Mar 2014 10:29:57 +0000 (19:29 +0900)]
evas: add support for GL_LUMINANCE_ALPHA and GL_LUMINANCE when loading image from disk.
Cedric BAIL [Wed, 19 Mar 2014 03:20:07 +0000 (12:20 +0900)]
evas: add png support for other color space output.
Cedric BAIL [Wed, 19 Mar 2014 00:41:28 +0000 (09:41 +0900)]
autotools: let's be precise on how we do support ABI stability.
NOTE: We do use enum in structure. There size can be changed by changing
the parameter of the compiler, this obviously will break ABI. As long as
you use the same compiler (and normally any future version of that compiler)
with the same set of CFLAGS/CXXFLAGS, you will have ABI stability.
Cedric BAIL [Tue, 18 Mar 2014 09:11:49 +0000 (18:11 +0900)]
eet: add internal encoding to ETC1 as an alternate solution to Jpeg.
Cedric BAIL [Mon, 17 Mar 2014 10:28:22 +0000 (19:28 +0900)]
evas: let TGV loader choose which encoding it want.
If region is specified we will not allow ETC1 colorspace as it would
basically break at the frontier as we would be unable to generate a
duplicate of the border as GPU require if you want nice and correct
rendering. So no region and ETC1 output at the same time.
Cedric BAIL [Mon, 17 Mar 2014 10:31:36 +0000 (19:31 +0900)]
evas: handle allocation of ETC1 color space image surface.
Cedric BAIL [Mon, 17 Mar 2014 07:45:50 +0000 (16:45 +0900)]
evas: remove dead code.
Cedric BAIL [Mon, 17 Mar 2014 06:19:47 +0000 (15:19 +0900)]
evas: remove RGBA_IMAGE_ALPHA_ONLY flags and use EVAS_COLORSPACE_GRY8 instead.