platform/upstream/efl.git
9 years agoautotools: more cross platform dlfuncs/shm_open checks
Daniel Kolesa [Wed, 5 Aug 2015 11:40:12 +0000 (12:40 +0100)]
autotools: more cross platform dlfuncs/shm_open checks

9 years agoevas_3d_utils: fix compiler warning
Daniel Kolesa [Wed, 5 Aug 2015 10:12:24 +0000 (11:12 +0100)]
evas_3d_utils: fix compiler warning

9 years agoEdje entry: Fix to delete selection when preedit status has not ended.
Subodh Kumar [Wed, 5 Aug 2015 10:02:53 +0000 (11:02 +0100)]
Edje entry: Fix to delete selection when preedit status has not ended.

Summary:
Fix to delete selection when preedit status has not ended

In preedit mode when selection is made and input any letter,
there is duplicate character appeared because the selection
is cleared when commit happens so delete the selection in
preedit last stage.

@fix

Test Plan: Test in Tizen devices.

Reviewers: woohyun, id213sin, shilpasingh, tasn

Subscribers: rajeshps, cedric, govi

Differential Revision: https://phab.enlightenment.org/D2837

9 years agoevas_check_engine: Remove tests for specific version of EGL
Chris Michael [Tue, 4 Aug 2015 15:38:21 +0000 (11:38 -0400)]
evas_check_engine: Remove tests for specific version of EGL

Summary: As per cedric, remove tests for egl specific version(s) when
building wayland engines.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas gl: add support for new partial update extension.
mythri.venugopal [Wed, 29 Jul 2015 21:12:35 +0000 (23:12 +0200)]
evas gl: add support for new partial update extension.

Summary:
If EGL_KHR_partial_update extension is implemented by the driver,
set the damage region. This is done before the draw calls.

@feature

Reviewers: wonsik, spacegrapher, jpeg

Reviewed By: spacegrapher

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2828

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoedje_embryo: add methods to handle event flags.
Jee-Yong Um [Wed, 29 Jul 2015 20:29:41 +0000 (22:29 +0200)]
edje_embryo: add methods to handle event flags.

Summary:
Add embryo methods that change the state
whether Edje_Part will handle Evas_Event or not.

Reviewers: Hermet, woohyun, cedric

Reviewed By: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2807

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas: add tangent space algorithm
perepelits.m [Wed, 29 Jul 2015 20:24:02 +0000 (22:24 +0200)]
evas: add tangent space algorithm

Summary: This is a logic of tangent data calculation, we need it in order to avoid bugs with normal mapping after building with opengl=yes.

Reviewers: Hermet, raster, cedric

Reviewed By: cedric

Subscribers: Oleksander, cedric, artem.popov

Differential Revision: https://phab.enlightenment.org/D2802

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas: twice faster NEON intrinsics scaling up implementation
Cedric BAIL [Wed, 29 Jul 2015 19:22:17 +0000 (21:22 +0200)]
evas: twice faster NEON intrinsics scaling up implementation

Summary: Already checked it for tizen 2.3/2.4. Scaling function works properly and much faster.

Reviewers: raster, jolfzverb, cedric

Reviewed By: cedric

Subscribers: cedric

Projects: #efl

Differential Revision: https://phab.enlightenment.org/D2881

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoEdje entry: Clean up conditions a bit.
Tom Hacohen [Tue, 4 Aug 2015 14:34:57 +0000 (15:34 +0100)]
Edje entry: Clean up conditions a bit.

Merge two conditions to clean up the code a bit.

9 years agoEdje entry: Make filter logic consistent across all input methods
Subodh Kumar [Tue, 4 Aug 2015 14:32:47 +0000 (15:32 +0100)]
Edje entry: Make filter logic consistent across all input methods

Summary:
When commit string comes from key_down text filter is applied before deleting
the range but when commit string comes from imf filter is applied after deleting
the range. Moreover when format is prepended range is deleted before applying
insert filter. Please check the _edje_key_down_cb and
_edje_entry_imf_event_commit_cb  routines.

@fix

Test Plan: Tested in Tizen device

Reviewers: herdsman, thiepha, shilpasingh, raster, tasn

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2904

9 years agoecore-wl: Support wl_keyboard_send_repeat_info for adjusting keyboard repeat rate...
Chris Michael [Wed, 15 Jul 2015 14:06:20 +0000 (10:06 -0400)]
ecore-wl: Support wl_keyboard_send_repeat_info for adjusting keyboard repeat rate and delay

Summary: This adds support for allowing a client/compositor to adjust
the keyboard repeat rate and delay via a call to
wl_keyboard_send_repeat_info.

@feature

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoEcore audio: Fix typo
Tom Hacohen [Tue, 4 Aug 2015 10:35:18 +0000 (11:35 +0100)]
Ecore audio: Fix typo

We really need to fix dependency tracking for .eo files. It didn't regenerate
dependencies...

9 years agoEina strbuf: Remove unused variable.
Tom Hacohen [Tue, 4 Aug 2015 10:29:34 +0000 (11:29 +0100)]
Eina strbuf: Remove unused variable.

It's never used because it's the base type and the others override the
magic anyway.

9 years agoEina quad: Remove unused function.
Tom Hacohen [Tue, 4 Aug 2015 10:19:50 +0000 (11:19 +0100)]
Eina quad: Remove unused function.

It may be useful for debugging, so you keeping it in source.
This silences a clang warning.

9 years agoEcore audio: Fix eo references.
Tom Hacohen [Tue, 4 Aug 2015 10:13:01 +0000 (11:13 +0100)]
Ecore audio: Fix eo references.

9 years agoMigrate more eo files to use Eina.Rectangle.
Tom Hacohen [Wed, 15 Jul 2015 16:34:07 +0000 (17:34 +0100)]
Migrate more eo files to use Eina.Rectangle.

9 years agoEvas out: migrate more eolian types.
Tom Hacohen [Wed, 15 Jul 2015 14:10:29 +0000 (15:10 +0100)]
Evas out: migrate more eolian types.

9 years agoEvas canvas3d: Move evas 3d types to evas_types.eot
Oleksandr Shcherbina [Tue, 14 Jul 2015 10:19:35 +0000 (11:19 +0100)]
Evas canvas3d: Move evas 3d types to evas_types.eot

Summary:
Move evas_canvas3d types from Evas_Eo.h to evas_types.eot
Change enum constatnt for vertex_arrtib, material_attrib, blend_func
Correct namespace for Evas.Real type

Reviewers: cedric, tasn

Reviewed By: tasn

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2734

9 years agoconfigure: Switch to dev mode again. Merge window for 1.16 is open now
Stefan Schmidt [Tue, 4 Aug 2015 09:47:22 +0000 (11:47 +0200)]
configure: Switch to dev mode again. Merge window for 1.16 is open now

9 years agorelease: Update NEWS and bump version for 1.15.0 release v1.15.0
Stefan Schmidt [Tue, 4 Aug 2015 08:52:34 +0000 (10:52 +0200)]
release: Update NEWS and bump version for 1.15.0 release

9 years agoecore_evas: Fix crash in async render
jiin.moon [Tue, 4 Aug 2015 02:49:04 +0000 (11:49 +0900)]
ecore_evas: Fix crash in async render

Summary:
The root cause of the crash is freed memory.
The evas_object_image_data_set api called with extn buffer in _ipc_server_data.
But there is no inform to render when free this buffer.
@fix

Reviewers: Hermet, jpeg, jypark

Subscribers: stefan_schmidt, cedric

Differential Revision: https://phab.enlightenment.org/D2902

Signed-off-by: Jean-Philippe Andre <jp.andre@samsung.com>
9 years agoevas image unload - fix free path to remove from pending list
Carsten Haitzler (Rasterman) [Tue, 4 Aug 2015 01:47:38 +0000 (10:47 +0900)]
evas image unload - fix free path to remove from pending list

this should fix T2580

9 years agoRevert "[HACK] configure.ac: Forces linking with -lEGL"
Nicolas Aguirre [Mon, 3 Aug 2015 12:35:00 +0000 (14:35 +0200)]
Revert "[HACK] configure.ac: Forces linking with -lEGL"

This reverts commit 9f8f64b697f974dc204acd340560e9547b21d743.

9 years agoRevert "[HACK] ecore_evas_fb: Comments ecore_fb_ts calls that aren't resolved at...
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:59 +0000 (14:34 +0200)]
Revert "[HACK] ecore_evas_fb: Comments ecore_fb_ts calls that aren't resolved at runtime"

This reverts commit ff5f8a0412d02e245754e4a90122a7715036cb64.

9 years agoRevert "configure.ac: Don't check for Xprint extension. cf:"
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:58 +0000 (14:34 +0200)]
Revert "configure.ac: Don't check for Xprint extension. cf:"

This reverts commit 9e0cd04cdcd1e7b6b02a6b34ca5bea5f9a73f1ad.

9 years agoRevert "Include <Eet.h> where needed. cf:"
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:57 +0000 (14:34 +0200)]
Revert "Include <Eet.h> where needed. cf:"

This reverts commit 877338659b4176e1a77f7b6cf37f6225cca763bc.

9 years agoRevert "Adds support of the eglfs module in the autotools config"
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:57 +0000 (14:34 +0200)]
Revert "Adds support of the eglfs module in the autotools config"

This reverts commit 3b2074aa710a095c2379702334bfa125bcc1990a.

9 years agoRevert "evas: Adds an eglfs module"
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:56 +0000 (14:34 +0200)]
Revert "evas: Adds an eglfs module"

This reverts commit 270215889d5ba4e93f9ec05cfa9a9510a345c96c.

9 years agoRevert "ecore_evas: Adds an eglfs module"
Nicolas Aguirre [Mon, 3 Aug 2015 12:34:54 +0000 (14:34 +0200)]
Revert "ecore_evas: Adds an eglfs module"

This reverts commit a254725d6dedbcfec638f8151bd89e9db068df8f.

9 years ago[HACK] configure.ac: Forces linking with -lEGL
Florent Revest [Wed, 22 Jul 2015 16:56:25 +0000 (18:56 +0200)]
[HACK] configure.ac: Forces linking with -lEGL

9 years ago[HACK] ecore_evas_fb: Comments ecore_fb_ts calls that aren't resolved at runtime
Florent Revest [Wed, 22 Jul 2015 16:55:16 +0000 (18:55 +0200)]
[HACK] ecore_evas_fb: Comments ecore_fb_ts calls that aren't resolved at runtime

9 years agoconfigure.ac: Don't check for Xprint extension. cf:
Florent Revest [Wed, 22 Jul 2015 16:53:30 +0000 (18:53 +0200)]
configure.ac: Don't check for Xprint extension. cf:
https://github.com/openembedded/meta-oe/blob/master/meta-efl/recipes-efl/efl/efl/0001-configure.ac-Don-t-check-for-Xprint-extension.patch

9 years agoInclude <Eet.h> where needed. cf:
Florent Revest [Wed, 22 Jul 2015 16:50:38 +0000 (18:50 +0200)]
Include <Eet.h> where needed. cf:
https://github.com/openembedded/meta-oe/blob/master/meta-efl/recipes-efl/efl/efl/0002-evas_3d-Add-Eet.h-includes.patch

9 years agoAdds support of the eglfs module in the autotools config
Florent Revest [Wed, 22 Jul 2015 16:50:13 +0000 (18:50 +0200)]
Adds support of the eglfs module in the autotools config

9 years agoevas: Adds an eglfs module
Florent Revest [Wed, 22 Jul 2015 16:49:14 +0000 (18:49 +0200)]
evas: Adds an eglfs module

9 years agoecore_evas: Adds an eglfs module
Florent Revest [Wed, 22 Jul 2015 16:48:41 +0000 (18:48 +0200)]
ecore_evas: Adds an eglfs module

9 years agoecore_con: Correcting ece_svr to ece_cl in test case.
Srivardhan Hebbar [Mon, 3 Aug 2015 10:17:22 +0000 (12:17 +0200)]
ecore_con: Correcting ece_svr to ece_cl in test case.

Summary: Signed-off-by: Srivardhan Hebbar <sri.hebbar@samsung.com>

Reviewers: cedric

Subscribers: stefan_schmidt, cedric

Differential Revision: https://phab.enlightenment.org/D2903

9 years agoeio: grammar and typo fixes in docs.
Yomi [Mon, 3 Aug 2015 09:50:10 +0000 (11:50 +0200)]
eio: grammar and typo fixes in docs.

Summary:
Fix a sentence.

Changes:

     src/examples/eio/eio_file_ls.c

Reviewers: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2894

9 years agoeina: add intermediate inline function for eina_crc.
Cedric BAIL [Sun, 2 Aug 2015 19:56:08 +0000 (21:56 +0200)]
eina: add intermediate inline function for eina_crc.

This is necessary to make my life easier when adding assembling implementation
of eina_crc.

9 years agoecore_pipe: Fix potential double close
Jean-Philippe Andre [Fri, 31 Jul 2015 07:58:19 +0000 (16:58 +0900)]
ecore_pipe: Fix potential double close

p->fd_read can be closed in _ecore_pipe_read (from a previous loop)

CID 1267459

9 years agoeina/btlog: Fix some Coverity issues
Jean-Philippe Andre [Fri, 31 Jul 2015 06:46:08 +0000 (15:46 +0900)]
eina/btlog: Fix some Coverity issues

CID 1297406
CID 1297408
CID 1297409

9 years agoEeze: Fix build break with --disable-libmount
Ryuan Choi [Fri, 31 Jul 2015 15:08:00 +0000 (00:08 +0900)]
Eeze: Fix build break with --disable-libmount

While compiling eeze_disk_dummy.c, eeze_mount_tabs_watch does not have visibility
because EAPI is cleared by Ecore.h in Eeze_Disk.h.

9 years agoEvas text: Migrate docs to new Eolian format.
Tom Hacohen [Fri, 31 Jul 2015 15:00:08 +0000 (16:00 +0100)]
Evas text: Migrate docs to new Eolian format.

9 years agoeolian: different error for validating @since
Daniel Kolesa [Fri, 31 Jul 2015 14:55:01 +0000 (15:55 +0100)]
eolian: different error for validating @since

9 years agoEvas table: Migrate docs to new Eolian format.
Tom Hacohen [Fri, 31 Jul 2015 14:19:33 +0000 (15:19 +0100)]
Evas table: Migrate docs to new Eolian format.

9 years agoevas-wayland-shm: Use wl_display_dispatch_pending to wait for buffers
Chris Michael [Fri, 31 Jul 2015 13:58:08 +0000 (09:58 -0400)]
evas-wayland-shm: Use wl_display_dispatch_pending to wait for buffers

Summary: This fixes T2625 where launching filemanager would cause a
deadlock. The issue here is that wl_display_dispatch actually uses
poll() which will block and wait, while using
wl_display_dispatch_pending does not case any blocking.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agospecify ector licenses.
ChunEon Park [Fri, 31 Jul 2015 07:23:33 +0000 (16:23 +0900)]
specify ector licenses.

9 years agoEvas filters: Fix invalid sizeof
Jean-Philippe Andre [Fri, 31 Jul 2015 03:05:44 +0000 (12:05 +0900)]
Evas filters: Fix invalid sizeof

CID 1308609

9 years agowayland-shm: When we run out of shm buffers kick the server to get some back
Derek Foreman [Thu, 30 Jul 2015 19:37:14 +0000 (15:37 -0400)]
wayland-shm: When we run out of shm buffers kick the server to get some back

Summary:
This fixes a rendering issue with wayland apps. We consume all buffers then
drop any attempted renders that take place while we have no buffers.
This fix now waits for the server to give us some buffers back.
This is the cause of T2612

@fix

 #Kansas

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2897

9 years agoecore-evas-wayland: Remove unused variable
Chris Michael [Thu, 30 Jul 2015 19:30:36 +0000 (15:30 -0400)]
ecore-evas-wayland: Remove unused variable

Summary: As we no longer need the wdata here (see previous commit), we
can remove the usage of this variable

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-evas-wayland: Remove call to ecore_wl_window_update_location
Chris Michael [Thu, 30 Jul 2015 19:17:57 +0000 (15:17 -0400)]
ecore-evas-wayland: Remove call to ecore_wl_window_update_location

Summary: This function should really not be called here as it triggers
an xdg_surface_set_window_geometry call which (in turn) should only be
getting called when the window geometry (meaning visible region)
itself has changed.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Set wl_display field inside the Shm_Surface structure
Chris Michael [Thu, 30 Jul 2015 19:11:09 +0000 (15:11 -0400)]
evas-wayland-shm: Set wl_display field inside the Shm_Surface structure

Summary: As we need access to the wl_display in order to dispatch
events on the queue while we wait for free buffers, pass it in to the
surface creation function and set it in the Shm_Surface structure

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Pass along the wl_display parameter to the surface creation function
Chris Michael [Thu, 30 Jul 2015 19:10:49 +0000 (15:10 -0400)]
evas-wayland-shm: Pass along the wl_display parameter to the surface creation function

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Modify existing engine functions to take a wl_display parameter
Chris Michael [Thu, 30 Jul 2015 19:09:55 +0000 (15:09 -0400)]
evas-wayland-shm: Modify existing engine functions to take a wl_display parameter

Summary: These functions need to have an additional wl_display
parameter so that we can set the field inside the Surface structure

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Provide wl_display to the surface structure
Chris Michael [Thu, 30 Jul 2015 19:09:02 +0000 (15:09 -0400)]
evas-wayland-shm: Provide wl_display to the surface structure

Summary: In order to fix an existing rendering issue, we need access
to the wl_display so that we can dispatch events on the que while we
wait for a free buffer

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-evas-wayland: Set wl_display in the evas engine structure
Chris Michael [Thu, 30 Jul 2015 18:37:58 +0000 (14:37 -0400)]
ecore-evas-wayland: Set wl_display in the evas engine structure

Summary: This sets the wl_display field of the Evas engine info
structure so that the wl_display can later be used for dispatching
queue while we await a free buffer

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Provide wl_display field in engine info structure
Chris Michael [Thu, 30 Jul 2015 18:37:10 +0000 (14:37 -0400)]
evas-wayland-shm: Provide wl_display field in engine info structure

Summary: This just adds a field so we have some place to store the
wl_display for use in the evas wayland engine

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: If no depth is specified during reconfigure, use existing depth
Chris Michael [Thu, 30 Jul 2015 17:53:10 +0000 (13:53 -0400)]
evas-wayland-shm: If no depth is specified during reconfigure, use existing depth

Summary: This fixes a crash which Could occur if no outbuf depth was
given during a call to reconfigure.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowayland-shm: Introduce buffer ages
Derek Foreman [Thu, 30 Jul 2015 15:45:38 +0000 (11:45 -0400)]
wayland-shm: Introduce buffer ages

Summary:
We now track each shm buffer's time since last draw so evas can tell
what it needs to re-render.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2893

9 years agowayland-shm: Assign buffers in the swap_mode callback
Derek Foreman [Thu, 30 Jul 2015 15:45:32 +0000 (11:45 -0400)]
wayland-shm: Assign buffers in the swap_mode callback

Summary:
This is when we're about to start a new render, so it's the best
time to set up the next buffer to draw to.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2892

9 years agowayland-shm: rename _evas_outbuf_swapmode_get() to _evas_outbuf_swap_mode_get()
Derek Foreman [Thu, 30 Jul 2015 15:45:26 +0000 (11:45 -0400)]
wayland-shm: rename _evas_outbuf_swapmode_get() to _evas_outbuf_swap_mode_get()

Summary: Everyone else calls it swap_mode, let's be trendy.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2891

9 years agowayland-shm: Remove curr_buff from Shm_Surface
Derek Foreman [Thu, 30 Jul 2015 15:45:17 +0000 (11:45 -0400)]
wayland-shm: Remove curr_buff from Shm_Surface

Summary:
We already have *current, but we're not using it properly.  Let's throw
out curr_buff and just use *current.

This temporarily makes swapmode_get do full renders, this will be
fixed in a future commit that adds proper buffer ages.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2890

9 years agowayland-shm: refactor _evas_shm_surface_data_get()
Derek Foreman [Thu, 30 Jul 2015 15:45:05 +0000 (11:45 -0400)]
wayland-shm: refactor _evas_shm_surface_data_get()

Summary:
Split out the buffer picking stuff because we're going to want to call
it from somewhere else.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2889

9 years agowayland-shm: Use MAX_BUFFERS macro instead of magic 4
Derek Foreman [Thu, 30 Jul 2015 15:44:57 +0000 (11:44 -0400)]
wayland-shm: Use MAX_BUFFERS macro instead of magic 4

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2888

9 years agowayland-shm: rename _evas_shm_surface_redraw() to _evas_shm_surface_post()
Derek Foreman [Thu, 30 Jul 2015 15:44:49 +0000 (11:44 -0400)]
wayland-shm: rename _evas_shm_surface_redraw() to _evas_shm_surface_post()

Summary:
It doesn't redraw anything, it posts a buffer full of data to the
compositor.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2887

9 years agowayland-shm: remove _evas_shm_surface_swap()
Derek Foreman [Thu, 30 Jul 2015 15:44:39 +0000 (11:44 -0400)]
wayland-shm: remove _evas_shm_surface_swap()

Summary: It no longer does anything useful

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2886

9 years agoeolian/generator: actually overrun-safe buffer fill
Daniel Kolesa [Thu, 30 Jul 2015 12:59:16 +0000 (13:59 +0100)]
eolian/generator: actually overrun-safe buffer fill

9 years agoeolian/generator: no point in creating tmp_env, it's not used
Daniel Kolesa [Thu, 30 Jul 2015 12:50:46 +0000 (13:50 +0100)]
eolian/generator: no point in creating tmp_env, it's not used

9 years agolicenses: fix and generalize copyright information
Daniel Kolesa [Thu, 30 Jul 2015 12:38:14 +0000 (13:38 +0100)]
licenses: fix and generalize copyright information

9 years agoeolian/common_funcs: Play safe and use strncpy
Stefan Schmidt [Thu, 30 Jul 2015 12:21:55 +0000 (14:21 +0200)]
eolian/common_funcs: Play safe and use strncpy

Not very likely but better play safe and make sure we don not overrun
our destination buffer here.

CID: 1311352

9 years agoevas textblock: convert docs
Daniel Kolesa [Thu, 30 Jul 2015 12:16:05 +0000 (13:16 +0100)]
evas textblock: convert docs

9 years agoevas_common_interface: remove bad doc comment
Daniel Kolesa [Thu, 30 Jul 2015 12:01:29 +0000 (13:01 +0100)]
evas_common_interface: remove bad doc comment

9 years agoevas canvas3d: doc conversion of mesh
Daniel Kolesa [Thu, 30 Jul 2015 12:00:25 +0000 (13:00 +0100)]
evas canvas3d: doc conversion of mesh

9 years agoevas canvas3d: doc conversion of scene
Daniel Kolesa [Thu, 30 Jul 2015 11:38:51 +0000 (12:38 +0100)]
evas canvas3d: doc conversion of scene

9 years agoevas canvas3d: doc conversion of texture
Daniel Kolesa [Thu, 30 Jul 2015 11:30:40 +0000 (12:30 +0100)]
evas canvas3d: doc conversion of texture

9 years agogl_common/evas_gl_3d: Assign attenuation values correctly.
Stefan Schmidt [Thu, 30 Jul 2015 10:47:47 +0000 (12:47 +0200)]
gl_common/evas_gl_3d: Assign attenuation values correctly.

This really looks like a copy and paste error to me. The assigned const and
linear value would always be overridden by quad.

CID: 1267462, 1267460

9 years agocopying: add license information
Daniel Kolesa [Thu, 30 Jul 2015 09:55:29 +0000 (10:55 +0100)]
copying: add license information

9 years agoevas/gl_drm: Fix compilation with mesa 10.6.3 on Debian
Jean-Philippe Andre [Thu, 30 Jul 2015 05:41:48 +0000 (14:41 +0900)]
evas/gl_drm: Fix compilation with mesa 10.6.3 on Debian

Thanks @aerodynamik for the report.

Fixes T2616

9 years agoevas engine: add null point exception.
Jaehwan Kim [Thu, 30 Jul 2015 02:08:39 +0000 (11:08 +0900)]
evas engine: add null point exception.

Sometimes the data parameter can be null.
It makes a segment fault.
test: evas_new > add method_output as buffer > evas_free

@fix

9 years agoecore_evas_x: make gl_sync_draw_done an explicit signed char type
Dongyeon Kim [Thu, 30 Jul 2015 01:14:14 +0000 (10:14 +0900)]
ecore_evas_x: make gl_sync_draw_done an explicit signed char type

Summary:
char type is interpreted as unsigned char in some implementations,
so setting -1 will be read as 255.
So we should explicitly make this a signed char type.
@fix

Test Plan: Local tests

Reviewers: cedric, jpeg

Reviewed By: jpeg

Subscribers: wonsik, cedric

Differential Revision: https://phab.enlightenment.org/D2895

9 years agoecore-evas-x: resize window when it is rotated by window manager.
Seunghun Lee [Wed, 29 Jul 2015 15:14:11 +0000 (11:14 -0400)]
ecore-evas-x: resize window when it is rotated by window manager.

Summary:
window manager can send arguments and its meaning as follows.

1) resize:0
it means client window doesn't need to resize its window by rotation.
this case is a ELM_WIN_BASIC window in mobile profile.

2) resize:1, ee->w != w, ee->h != h (deprecated)
it means client window should be resized by rotation, and wm already resize its window.
so, client don't need to resize its window.
it's just for backward compatibility.

3) resize:1, ee->w == w, ee->h == h (addition)
it means client window should be resized by rotation, and wm don't resize it.
so, client should resize its window.

Test Plan: N/A

Reviewers: gwanglim, raster, jypark, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2831

9 years agoNEWS: Small manual tweaks
Stefan Schmidt [Wed, 29 Jul 2015 15:00:54 +0000 (17:00 +0200)]
NEWS: Small manual tweaks

Most notably removed a merge commit and a commit stating to take EO out of
beta which is not happening for 1.15.

9 years agoecore_evas_input: add value to enable mouse button cancel feature
Ji-Youn Park [Wed, 29 Jul 2015 12:56:15 +0000 (21:56 +0900)]
ecore_evas_input: add value to enable mouse button cancel feature

mike reported this feature break the behavior in E19, but I have trouble to reproduce it.
so I first add value to enable this feature and after that,

9 years agoeina str: remove useless and wrong ifdef for FreeBSD
Daniel Kolesa [Wed, 29 Jul 2015 11:34:14 +0000 (12:34 +0100)]
eina str: remove useless and wrong ifdef for FreeBSD

9 years agoEvas filters: Fix make doc, check when EFL is not installed
Jean-Philippe Andre [Wed, 29 Jul 2015 11:00:35 +0000 (20:00 +0900)]
Evas filters: Fix make doc, check when EFL is not installed

Thanks @aerodynamik for even giving the proper environment
variable to set :)

Fixes T2614

9 years agoevas: fix size of borders in case need generate texture without atlases
Oleksandr Shcherbina [Tue, 28 Jul 2015 20:33:49 +0000 (22:33 +0200)]
evas: fix size of borders in case need generate texture without atlases

Summary:
Set size of texture unit without 2 pixels for borders in case use it without
atlses. Just one case if texture for 3D use repeat mode and non-normalized
tuxture coordinates

Reviewers: Hermet, cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2805

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoecore-evas-wayland-shm: Remove duplicated code on configure event
Chris Michael [Tue, 28 Jul 2015 19:12:17 +0000 (15:12 -0400)]
ecore-evas-wayland-shm: Remove duplicated code on configure event

Summary: As we end up calling ecore_evas_wl_common_resize anyway,
there is no need for all these size calculations in the configure
callback. The resize function will deal with this anyway.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoefl: Bump required wayland library versions
Chris Michael [Tue, 28 Jul 2015 16:53:20 +0000 (12:53 -0400)]
efl: Bump required wayland library versions

Summary: As a lot of functionality has changed in the relevant efl/e
wayland codebase, we should bump the required wayland library versions
as we are now dependant on updated wayland libraries.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoevas-wayland-shm: Fix issue of surface_swap spamming invalid warning
Chris Michael [Tue, 28 Jul 2015 13:43:16 +0000 (09:43 -0400)]
evas-wayland-shm: Fix issue of surface_swap spamming invalid warning

Summary: Since the surface_swap functionality is changed now, we no
longer need to bother checking if the leaf is busy as we don't use the
leaf itself inside this function anymore.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowayland-shm: Remove "mapped"
Derek Foreman [Tue, 28 Jul 2015 13:16:04 +0000 (09:16 -0400)]
wayland-shm: Remove "mapped"

Summary:
Looks like the recent changes to swap/redraw have fixed a problem with
damage tracking and we no longer need this.

Damage history seems ok without it now.

@fix

Reviewers: zmike, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2876

9 years agowayland-shm: re-organize some of the swap/redraw logic
Derek Foreman [Tue, 28 Jul 2015 13:14:31 +0000 (09:14 -0400)]
wayland-shm: re-organize some of the swap/redraw logic

Summary:
The way things were, we'd be committing a buffer we hadn't rendered to
yet.

Now redraw() contains all the attach/damage/commit logic, and swap() is
called afterwards to set up a new target buffer for the next render.

@fix

Reviewers: zmike, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2875

9 years agowayland_shm: Don't store frame_cb
Derek Foreman [Tue, 28 Jul 2015 13:13:16 +0000 (09:13 -0400)]
wayland_shm: Don't store frame_cb

Summary:
We don't need to - all the frame callback does for us is ensure a queue
push.

We should destroy any callback we receive when we receive it.

@fix

Reviewers: zmike, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2874

9 years agoEcore: Fix invalid lock count if DEBUG is enabled
Jean-Philippe Andre [Tue, 28 Jul 2015 06:49:47 +0000 (15:49 +0900)]
Ecore: Fix invalid lock count if DEBUG is enabled

With -DDEBUG, we can see an error message, like:
ERR: You are calling _ecore_lock() from outside of the main loop
     threads in lib/ecore/ecore_private at line 306

Looking at the code shows that ecore_lock fails immediately if
thread debugging is enabled, but ecore_unlock does not, so the
value _ecore_main_lock_count could go below 0.

This is not very important as the value is never used.

9 years agoEvas filters: Fix order of header file includes
Jean-Philippe Andre [Tue, 28 Jul 2015 06:13:59 +0000 (15:13 +0900)]
Evas filters: Fix order of header file includes

This fixes filters on Ubuntu 32 bits.

This was one hell of a weird bug to track down. Everything worked
like a charm on my 64 bit machines, but filters would simply fail
for no good reason: a safety check sees a NULL pointer when clearly
it was properly allocated.

Just after entering a function, the content of an RGBA_Image would
change, even though there was no memory write there. This made the
image data pointer NULL, and filters would fail miserably.

So I printed out the contents of the RGBA_Image, they changed. But
the memory itself had not changed. The size of the struct itself
had changed when jumping from one file to another! But its definition
had not! Non-sense!

Unless of course a system header file was included before config.h
and ino_t or off_t would switch between 32 and 64 bits...

@fix

9 years agoEvas filters: Fix COW usage in filter mixin
Jean-Philippe Andre [Tue, 28 Jul 2015 05:17:53 +0000 (14:17 +0900)]
Evas filters: Fix COW usage in filter mixin

9 years agoEvas filters: Disable debug code path
Jean-Philippe Andre [Tue, 28 Jul 2015 02:27:08 +0000 (11:27 +0900)]
Evas filters: Disable debug code path

9 years agoEvas filters: Fix invalid rendering in case of async render
Jean-Philippe Andre [Mon, 27 Jul 2015 08:20:23 +0000 (17:20 +0900)]
Evas filters: Fix invalid rendering in case of async render

This simplifies the code and fixes async buffer management.

@fix

9 years agoexamples: Update .gitignore for binaries
Stefan Schmidt [Mon, 27 Jul 2015 12:17:58 +0000 (14:17 +0200)]
examples: Update .gitignore for binaries

9 years agoecore_con_eet: fix dereferencing of object during callback that crashed some tests...
Cedric BAIL [Tue, 28 Jul 2015 01:04:38 +0000 (03:04 +0200)]
ecore_con_eet: fix dereferencing of object during callback that crashed some tests case.

9 years agoecore_con: fix test to actually send a NULL terminated string.
Cedric BAIL [Tue, 28 Jul 2015 00:55:32 +0000 (02:55 +0200)]
ecore_con: fix test to actually send a NULL terminated string.