platform/upstream/efl.git
9 years agoeina: add test case for eina_file_copy function.
vivek [Mon, 18 May 2015 08:52:54 +0000 (10:52 +0200)]
eina: add test case for eina_file_copy function.

Summary:
Added test case for eina_file_copy function to check copying of two files

Signed-off-by: vivek <vivek.ellur@samsung.com>
Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2449

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoeina: use less stack on Windows for eina_environment_home_get, and no snprintf()
Vincent Torri [Fri, 15 May 2015 07:03:48 +0000 (09:03 +0200)]
eina: use less stack on Windows for eina_environment_home_get, and no snprintf()

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas font draw - fix up some whitespace/indenting/formatting
Carsten Haitzler (Rasterman) [Mon, 18 May 2015 08:13:38 +0000 (17:13 +0900)]
evas font draw - fix up some whitespace/indenting/formatting

9 years agoedje edc examples - remobe trailing \ from edc list
Carsten Haitzler (Rasterman) [Mon, 18 May 2015 08:00:05 +0000 (17:00 +0900)]
edje edc examples - remobe trailing \ from edc list

this may be a build issue - i don't see it though.

9 years agoset_tween_state_anim API added.
kumar navneet [Mon, 18 May 2015 06:22:00 +0000 (15:22 +0900)]
set_tween_state_anim API added.

Summary:
Issue: Current embryo script do not provide any mechanism to set transition type in animation, by default supports only linear
e.g: set_tween_state API
Solution: To not break backward compatibility, a new API set_tween_state_anim is added in which we can even specify type of transition required.

Signed-Off by: Kumar Navneet <k.navneet@samsung.com>
Signed-Off by: Shilpa Singh <shilpa.singh@samsung.com>

Test Plan: A test edc (embryo_tween_anim.edc) is added to edje examples to test set_tween_state_anim API.

Reviewers: raster, cedric, shilpasingh, Hermet

Reviewed By: shilpasingh

Subscribers: poornima.srinivasan, SubodhKumar, rajeshps, cedric, govi

Differential Revision: https://phab.enlightenment.org/D2344

9 years agoevas canvas: fix insane mouse move events on proxy source.
ChunEon Park [Fri, 15 May 2015 15:31:52 +0000 (00:31 +0900)]
evas canvas: fix insane mouse move events on proxy source.

There is no need compare - transformed mouse pos and origin pos.
This compare just block the mouse move event trigger on source.

@fix

9 years agoeolian: use the new decl storage to retrieve REGULAR base types
Daniel Kolesa [Fri, 15 May 2015 14:16:48 +0000 (15:16 +0100)]
eolian: use the new decl storage to retrieve REGULAR base types

9 years agoeolian: much better and stricter redefinition checking
Daniel Kolesa [Fri, 15 May 2015 14:10:58 +0000 (15:10 +0100)]
eolian: much better and stricter redefinition checking

We can now check redefinitions between different types of declarations,
such as redefinition of struct as variable etc.

@feature

9 years agoeina_cpu - up max number of threads to 32
Carsten Haitzler (Rasterman) [Fri, 15 May 2015 10:07:42 +0000 (19:07 +0900)]
eina_cpu - up max number of threads to 32

this should be better for todays systems - max number of cores we can
query for now is 32.

@feature

9 years agoeina - cpu count - fic cpu count to count # of cores correctly
Carsten Haitzler (Rasterman) [Fri, 15 May 2015 10:05:33 +0000 (19:05 +0900)]
eina - cpu count - fic cpu count to count # of cores correctly

@fix

this fixes the cpu count to count the numebr of available cores
correctly. before if core 0 happened to be off, we'd get 0 cpu cores
available due to the break

9 years agoecore: add documentation for ecore_timer_dump
Vinícius dos Santos Oliveira [Thu, 14 May 2015 12:58:18 +0000 (09:58 -0300)]
ecore: add documentation for ecore_timer_dump

9 years ago[ecore_evas_extn] add function to block mouse event.
Shinwoo Kim [Fri, 15 May 2015 07:46:21 +0000 (16:46 +0900)]
[ecore_evas_extn] add function to block mouse event.

Summary: add ecore_evas_extn_socket_events_block_set/get

Test Plan: add mouse event callback, and check whether it could get event or not

Reviewers: raster, woohyun, jaehwan, Sergeant_Whitespace

Reviewed By: Sergeant_Whitespace

Subscribers: Sergeant_Whitespace, seoz, cedric

Differential Revision: https://phab.enlightenment.org/D2268

9 years agoedje entry: Improves anchors handling
Subodh Kumar [Fri, 15 May 2015 05:18:08 +0000 (14:18 +0900)]
edje entry: Improves anchors handling

Summary:
When edje entry is not inside the canvas view port, anchors update is aborted to improve the scrolling performance
in case of large number of anchors.

@feature

Test Plan:
1. Should have many entries inside scroller.
2. Each entry should contain large number of anchors.
3. Scroll it, scrolling is not smooth.

Reviewers: seoz, cedric, thiepha, woohyun, tasn, raster, shilpasingh, herdsman, JackDanielZ

Reviewed By: shilpasingh

Subscribers: poornima.srinivasan, rajeshps, cedric, govi

Differential Revision: https://phab.enlightenment.org/D2108

9 years agoeina - add comment on evlog api stability etc.
Carsten Haitzler (Rasterman) [Fri, 15 May 2015 02:32:04 +0000 (11:32 +0900)]
eina - add comment on evlog api stability etc.

9 years agoeina: handle fallback to HOMEDIR/HOMEPATH on WIN32.
Cedric BAIL [Thu, 14 May 2015 18:14:56 +0000 (20:14 +0200)]
eina: handle fallback to HOMEDIR/HOMEPATH on WIN32.

9 years agoeina: fix eina_inarray_search to do linear search
Vivek Ellur [Thu, 14 May 2015 17:37:06 +0000 (19:37 +0200)]
eina: fix eina_inarray_search to do linear search

Summary:
Currently eina_inarray_search was using binary search to search elements
which would not work on unsorted array so modified it to work as linear search.
There is already a function eina_inarray_search_sorted to work on sorted array.

Signed-off-by: Vivek Ellur <vivek.ellur@samsung.com>
Reviewers: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2501

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoeina: add test cases for various eina_inarray functions
Vivek Ellur [Thu, 14 May 2015 17:36:25 +0000 (19:36 +0200)]
eina: add test cases for various eina_inarray functions

Summary:
Added test cases for eina_inarray_search, eina_inarray_search_sorted,
eina_inarray_foreach_remove and other functions

Signed-off-by: Vivek Ellur <vivek.ellur@samsung.com>
Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2504

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoedje: fix back edje_watch.
Cedric BAIL [Thu, 14 May 2015 17:00:04 +0000 (19:00 +0200)]
edje: fix back edje_watch.

@fix

Thanks Thanatermesis for reporting the issue.

9 years agoecore_x: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:41:29 +0000 (18:41 +0200)]
ecore_x: use portable environment lookup.

9 years agoecore_file: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:41:17 +0000 (18:41 +0200)]
ecore_file: use portable environment lookup.

9 years agoecore_con: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:40:47 +0000 (18:40 +0200)]
ecore_con: use portable environment lookup.

9 years agoethumb: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:40:28 +0000 (18:40 +0200)]
ethumb: use portable environment lookup.

9 years agoefreet: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:40:13 +0000 (18:40 +0200)]
efreet: use portable environment lookup.

9 years agoevas: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:39:48 +0000 (18:39 +0200)]
evas: use portable environment lookup.

9 years agoeina: use portable environment lookup.
Cedric BAIL [Thu, 14 May 2015 16:38:59 +0000 (18:38 +0200)]
eina: use portable environment lookup.

9 years agoeina: fixup.
Cedric BAIL [Thu, 14 May 2015 16:38:36 +0000 (18:38 +0200)]
eina: fixup.

9 years agoecore_con: it doesn't make sense to get port on local protocol.
Cedric BAIL [Thu, 14 May 2015 16:19:57 +0000 (18:19 +0200)]
ecore_con: it doesn't make sense to get port on local protocol.

9 years agoeina: add crosss platforme API to retrieve tmp and home directories from environment.
Vincent Torri [Thu, 14 May 2015 05:40:33 +0000 (07:40 +0200)]
eina: add crosss platforme API to retrieve tmp and home directories from environment.

@feature

No tests added as it is highly dependent on the system and it would make little sens.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoeina: fix warning on Windows.
Vincent Torri [Thu, 14 May 2015 07:19:53 +0000 (09:19 +0200)]
eina: fix warning on Windows.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoecore_con: add missing declarations on Windows.
Vincent Torri [Thu, 14 May 2015 07:15:41 +0000 (09:15 +0200)]
ecore_con: add missing declarations on Windows.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoecore_avahi: fix definition of EAPI on Windows.
Vincent Torri [Thu, 14 May 2015 07:14:59 +0000 (09:14 +0200)]
ecore_avahi: fix definition of EAPI on Windows.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoeina, efl: add copyright notice.
Vincent Torri [Thu, 14 May 2015 06:41:11 +0000 (08:41 +0200)]
eina, efl: add copyright notice.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoautotools: cleanup Makefile_Eina.am.
Vincent Torri [Thu, 14 May 2015 06:33:46 +0000 (08:33 +0200)]
autotools: cleanup Makefile_Eina.am.

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoefreet: lower error message to warning.
Cedric BAIL [Thu, 14 May 2015 09:55:30 +0000 (11:55 +0200)]
efreet: lower error message to warning.

9 years agoedje: support translation on static strings in edc.
katpavalli [Wed, 13 May 2015 16:08:04 +0000 (18:08 +0200)]
edje: support translation on static strings in edc.

Summary:
   Internationalisation of the static text specified as part of the edc is implemented.
   Problem: Static text when specified in the edc, remains unchanged when the system language is changed.
   Solution: Language support is provided even for the static strings in the edc.

Test Plan:
   Test code to test this implementation is done as part of efl/src/examples/edje/edje-text.c and efl/src/examples/edje/text.edc
   Compile the code with the below  command
   edje_cc -md <dir path>/efl/src/examples/edje/ text.edc && gcc -o edje-text edje-text.c `pkg-config --libs --cflags ecore-evas edje evas ecore`
   ./edje-text

   1) change the language of the system using the command
      export LANGUAGE=hi
      ./edje.text

      Not the text Loading gets displayed in hindi language

   2) change the language of the system using the command
      export LANGUAGE=ta
      ./edje.text

      Not the text Loading gets displayed in tamil language

   3) change the language of the system using the command
      export LANGUAGE=en
      ./edje.text

      Not the text Loading gets displayed in english language
As the number of .mo files in the /edje folder can be increased, those many languages can be supported

Reviewers: cedric, shilpasingh

Reviewed By: shilpasingh

Subscribers: cedric, rajeshps, govi, poornima.srinivasan

Differential Revision: https://phab.enlightenment.org/D2336

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoedje: use efreet for cache directory get.
Cedric BAIL [Wed, 13 May 2015 15:14:09 +0000 (17:14 +0200)]
edje: use efreet for cache directory get.

9 years agoevas-drm: Mark framebuffer as dirty before sending
Chris Michael [Thu, 14 May 2015 15:49:51 +0000 (11:49 -0400)]
evas-drm: Mark framebuffer as dirty before sending

Summary: We should be marking the framebuffer dirty regions Before we
send the pageflip to the kernel so that the kernel can properly
optimize drawing dirty regions before scheduling a pageflip.

NB: Slightly optimizes evas drm rendering a bit more. Thanks to Daniel
Stone for the help :)

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoeolian: use fprintf(stderr) rather than eina_log for compile errors
Daniel Kolesa [Thu, 14 May 2015 15:33:09 +0000 (16:33 +0100)]
eolian: use fprintf(stderr) rather than eina_log for compile errors

9 years agoecore-drm: Ignore output enabled flag when setting output mode
Chris Michael [Thu, 14 May 2015 13:17:02 +0000 (09:17 -0400)]
ecore-drm: Ignore output enabled flag when setting output mode

Summary: We really do not need to check the enable flag here because
if we are setting a valid mode, then we will be enabled anyway

NB: This makes it actually possible to Enable/Disable outputs in the
RandR config dialog of E-Wl ;)

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix segfault if sending output event and output has no current mode
Chris Michael [Thu, 14 May 2015 12:36:57 +0000 (08:36 -0400)]
ecore-drm: Fix segfault if sending output event and output has no current mode

Summary: This fixes a segfault which could happen if we enable an
output before setting the output current mode. Now we test for an
output having a current_mode, and if not we use the crtc size when
sending the output event.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore_evas_wayland: prevent duplicated shutdown of ecore_wl
MinJeong Kim [Thu, 14 May 2015 11:54:16 +0000 (07:54 -0400)]
ecore_evas_wayland: prevent duplicated shutdown of ecore_wl

Summary:
When the ecore_evas of wayland_egl/shm is freed with ecore_evas_free(),
ecore_wl_shutdown() is called by _ecore_evas_wl_common_free().
so ecore_wl_shutdown() after ecore_evas_free() has to be skipped.
@fix

Reviewers: devilhorns, raster, zmike, gwanglim

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2503

9 years agoecore-evas: Added support for getting window auxiliary hint ID and value
Doyoun Kang [Thu, 14 May 2015 11:54:32 +0000 (20:54 +0900)]
ecore-evas: Added support for getting window auxiliary hint ID and value

Summary:
There are no APIs for getting window auxiliary hint ID and value which was set by a user.

Below API can get the ID of the window auxiliary hint.
- ecore_evas_aux_hint_id_get

Below API can get the value of the window auxiliary hint id.
- ecore_evas_aux_hint_val_get

Test Plan: N/A

Reviewers: seoz, bryceharrington, ManMower, devilhorns, cedric, raster, Hermet

Reviewed By: Hermet

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2493

9 years agoRevert "ecore-evas: Added support for setting/getting window auxiliary hints"
ChunEon Park [Thu, 14 May 2015 11:40:17 +0000 (20:40 +0900)]
Revert "ecore-evas: Added support for setting/getting window auxiliary hints"

This reverts commit 365e390a370d6464ec6b606123272e44fd018158.

Eeeek. Sorry I didn't mean this submitting.

9 years agoevas canvas: add exceptional handling in invalid input case.
ChunEon Park [Thu, 14 May 2015 10:50:50 +0000 (19:50 +0900)]
evas canvas: add exceptional handling in invalid input case.

@fix

9 years agoecore-evas: Added support for setting/getting window auxiliary hints
Doyoun Kang [Tue, 12 May 2015 08:58:09 +0000 (17:58 +0900)]
ecore-evas: Added support for setting/getting window auxiliary hints

Summary:
There are no APIs for getting window auxiliary hint value which was set by a user.
Below APIs can get the window auxiliary hint value.
- ecore_evas_aux_hint_val_get
- ecore_evas_aux_hint_string_val_get

And below API can set the window auxiliary hint value by using string not id.
- ecore_evas_aux_hint_string_val_set

Test Plan: N/A

Reviewers: raster, cedric, seoz, Hermet

Reviewed By: Hermet

Subscribers: c, cedric

Differential Revision: https://phab.enlightenment.org/D2493

9 years agoefl_model_base: fix enum/struct syntax in comment
Daniel Kolesa [Thu, 14 May 2015 10:30:20 +0000 (11:30 +0100)]
efl_model_base: fix enum/struct syntax in comment

9 years agoeolian: force specification of inner type for all complex types
Daniel Kolesa [Thu, 14 May 2015 10:27:00 +0000 (11:27 +0100)]
eolian: force specification of inner type for all complex types

9 years agoefl - add more evlog points
Carsten Haitzler (Rasterman) [Thu, 14 May 2015 08:35:22 +0000 (17:35 +0900)]
efl - add more evlog points

@feature

9 years agoecore-drm: Remove debug output
Chris Michael [Wed, 13 May 2015 18:57:39 +0000 (14:57 -0400)]
ecore-drm: Remove debug output

Summary: This should not have been included in the previous push as it
was just debug noise which was added during testing

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Add API function to set a new mode on an output
Chris Michael [Wed, 13 May 2015 18:41:12 +0000 (14:41 -0400)]
ecore-drm: Add API function to set a new mode on an output

Summary: This adds a new API function (ecore_drm_output_mode_set) that
we can use from within RandR code to set the resolution of an output
(or disable an output if NULL is passed in).

@feature

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix issue with output_current_resolution_get function
Chris Michael [Wed, 13 May 2015 18:39:57 +0000 (14:39 -0400)]
ecore-drm: Fix issue with output_current_resolution_get function

Summary: This fixes ecore_drm_output_current_resolution_get function
to properly operate if an output does not have a current mode assigned.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Only send output events when we enable/disable an output
Chris Michael [Wed, 13 May 2015 18:37:35 +0000 (14:37 -0400)]
ecore-drm: Only send output events when we enable/disable an output

Summary: This reduces the number of output events that wl_drm module
will receive from ecore-drm. We now only send those events when an
output gets enabled/disabled where previously we were also sending
them during output creation.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Greatly improve drm rendering speed
Chris Michael [Wed, 13 May 2015 18:33:08 +0000 (14:33 -0400)]
ecore-drm: Greatly improve drm rendering speed

Summary: This greatly improves rendering speed in evas drm engine.
Previously we would always call drmModeSetCrtc regardless if it was
needed or not. These changes greatly improve rendering speed in drm as
we now only call drmModeSetCrtc if it is needed.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix purpose of ecore_drm_output_enable/disable functions
Chris Michael [Wed, 13 May 2015 14:19:32 +0000 (10:19 -0400)]
ecore-drm: Fix purpose of ecore_drm_output_enable/disable functions

Summary: These API functions should be used for enable/disable of a
given output. They were previously being misused to stop/start
rendering on an output when we VT switch away so now we add an
internal function we can call to disable/enable rendering.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoeolian: precompute all enum field values (faster runtime, constness)
Daniel Kolesa [Wed, 13 May 2015 17:10:02 +0000 (18:10 +0100)]
eolian: precompute all enum field values (faster runtime, constness)

9 years agoeolian: allow forced retrieval of enum field values
Daniel Kolesa [Wed, 13 May 2015 16:56:18 +0000 (17:56 +0100)]
eolian: allow forced retrieval of enum field values

9 years agoeolian: use the actual enum type in tests
Daniel Kolesa [Wed, 13 May 2015 16:22:25 +0000 (17:22 +0100)]
eolian: use the actual enum type in tests

9 years agoEeze tests: Remove an unreliable test.
Tom Hacohen [Wed, 13 May 2015 16:18:18 +0000 (17:18 +0100)]
Eeze tests: Remove an unreliable test.

I hope this will be enough to make the suite less broken.
This fails often on jenkins and cedric's box. This should
either be made reliable, or removed, but the current state
is definitely not good if we would like to increase the trust
in Jenkins.

9 years agoeolian: fix evaluation of "undefined" enum fields
Daniel Kolesa [Wed, 13 May 2015 16:15:20 +0000 (17:15 +0100)]
eolian: fix evaluation of "undefined" enum fields

9 years agoedje: fix parsing issue in edje_cc with bezier program transition.
kumar navneet [Wed, 13 May 2015 14:18:47 +0000 (16:18 +0200)]
edje: fix parsing issue in edje_cc with bezier program transition.

Summary:
Issue: "CURRENT" param not considered while parsing cubic bezier transition in edje cc
Soln: parse 5,6 parameter if Curr is set else parse 4,5 parameter

Reviewers: cedric, shilpasingh

Reviewed By: shilpasingh

Subscribers: poornima.srinivasan, rajeshps, cedric, govi

Differential Revision: https://phab.enlightenment.org/D2488

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas: correct normal map shader.
Oleksandr Shcherbina [Wed, 13 May 2015 14:12:59 +0000 (16:12 +0200)]
evas: correct normal map shader.

Summary:
Due to using Evas_GL_Image for generation texture unit for Evas_3D_Texture it is
need adjusting texture coordinates in shader
It has been missed here
https://git.enlightenment.org/core/efl.git/commit/?id=d88ccf06a5f6ebcfc68dcc21e55a241f64ff9aa9

Reviewers: Hermet, cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2474

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas: add support for BUILD_NEON_INTRINSICS to evas_convert_rgb_32.c
Yury Usishchev [Wed, 13 May 2015 13:33:15 +0000 (15:33 +0200)]
evas: add support for BUILD_NEON_INTRINSICS to evas_convert_rgb_32.c

Summary: This fixes build for aarch64 when TILE_ROTATE is disabled and BUILD_NEON is enabled(it is enabled by default for aarch64 since https://phab.enlightenment.org/D2309).

Reviewers: cedric, raster

Subscribers: cedric

Projects: #efl

Differential Revision: https://phab.enlightenment.org/D2498

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas: fix coverity CID 1297159 and CID 1297158.
Oleksandr Shcherbina [Wed, 13 May 2015 13:32:10 +0000 (15:32 +0200)]
evas: fix coverity CID 1297159 and CID 1297158.

Reviewers: Hermet, b.devichev, cedric

Reviewed By: cedric

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2494

Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
9 years agoevas grid smart - fix coord overflows when grid size and vsize are large
Carsten Haitzler (Rasterman) [Wed, 13 May 2015 13:22:32 +0000 (22:22 +0900)]
evas grid smart - fix coord overflows when grid size and vsize are large

if your virtual size is fairly big AND your actual object size is also
big, you easily overflow a signed int for intermediate coordinate
calculations, resulting in seeing only a small fractin of your objects
correctly. this fixes that by expanding up to long longs internally to
allow for the added space needed for the multiplications

@fix

9 years agoeina evlog - fix overflow tracking
Carsten Haitzler (Rasterman) [Mon, 11 May 2015 07:55:23 +0000 (16:55 +0900)]
eina evlog - fix overflow tracking

9 years agoeolian: check inherit validity before using it
Daniel Kolesa [Wed, 13 May 2015 12:44:46 +0000 (13:44 +0100)]
eolian: check inherit validity before using it

9 years agoecore_x: Add key router feature related with client side.
Ji-Youn Park [Wed, 13 May 2015 11:47:36 +0000 (20:47 +0900)]
ecore_x: Add key router feature related with client side.
Currently app only grab key using Xgrabkey.
Keyrouter will support several keygrab mode and apps can use it.

9 years agoeolian: remove unneeded cruft (struct/enum cannot be a base for alias)
Daniel Kolesa [Wed, 13 May 2015 10:25:49 +0000 (11:25 +0100)]
eolian: remove unneeded cruft (struct/enum cannot be a base for alias)

9 years agoeolian: generate underscored structs as well as correct opaque struct typedefs
Daniel Kolesa [Wed, 13 May 2015 10:14:49 +0000 (11:14 +0100)]
eolian: generate underscored structs as well as correct opaque struct typedefs

9 years agoEvas filters: Fix glReadPixels usage for EGL
Jean-Philippe Andre [Wed, 13 May 2015 01:16:17 +0000 (10:16 +0900)]
Evas filters: Fix glReadPixels usage for EGL

EGL might very well not support RGBA read mode, so we
need to check for it first.

Also remove some error logs (see previous commit), and useless
initialization of the Evas GL engine.

@fix

9 years agoEvas GL common: Add very basic GL calls tracing method
Jean-Philippe Andre [Wed, 13 May 2015 01:17:12 +0000 (10:17 +0900)]
Evas GL common: Add very basic GL calls tracing method

Rewrap the GL calls to print them out. The arguments are not
interpreted. See the GL_ERRORS #define.

9 years agoEvas filters: Disable excessive debug
Jean-Philippe Andre [Wed, 13 May 2015 01:12:36 +0000 (10:12 +0900)]
Evas filters: Disable excessive debug

9 years agoeolian: do not check alias/struct/enum base for builtin types
Daniel Kolesa [Tue, 12 May 2015 16:23:42 +0000 (17:23 +0100)]
eolian: do not check alias/struct/enum base for builtin types

9 years agoecore-drm: Skip not connected outputs when calculating geometry
Chris Michael [Tue, 12 May 2015 16:05:22 +0000 (12:05 -0400)]
ecore-drm: Skip not connected outputs when calculating geometry

Summary: When we make a call to get the geometry of all outputs, we
should be skipping ones which are not connected.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix finding possible crtcs
Chris Michael [Tue, 12 May 2015 16:03:24 +0000 (12:03 -0400)]
ecore-drm: Fix finding possible crtcs

Summary: This fixes an issue when searching for possible crtcs that an
output can work on. Previously, we would end up not returning any
possible crtcs due to not looping the crtcs of the resource.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Add a 'name' field to Ecore_Drm_Event_Output
Chris Michael [Tue, 12 May 2015 15:59:40 +0000 (11:59 -0400)]
ecore-drm: Add a 'name' field to Ecore_Drm_Event_Output

Summary: This adds a new 'name' field to the Ecore_Drm_Event_Output
structure so that when we catch drm output events in E, we can compare
this name to find an e_randr screen and update compositor's outputs.

@feature

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix ecore_drm_output_edid_get to return a hex string
Chris Michael [Tue, 12 May 2015 15:56:52 +0000 (11:56 -0400)]
ecore-drm: Fix ecore_drm_output_edid_get to return a hex string

Summary: As we will use the edid string inside RandR code to store
unique information about an output, we should be returning this edid
in a "readable" form.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-drm: Fix issue of edid parsing not ignoring string
Chris Michael [Tue, 12 May 2015 15:51:01 +0000 (11:51 -0400)]
ecore-drm: Fix issue of edid parsing not ignoring string

Summary: When we are parsing the edid string, if the string is random
junk, then we need to ignore it. Prior to this commit, we were not
setting the returned text properly.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoeolian: remove unused variables
Daniel Kolesa [Tue, 12 May 2015 16:07:26 +0000 (17:07 +0100)]
eolian: remove unused variables

9 years agoeolian: new API eolian_type_enum_field_c_name_get
Daniel Kolesa [Tue, 12 May 2015 16:04:54 +0000 (17:04 +0100)]
eolian: new API eolian_type_enum_field_c_name_get

This API allows you to retrieve the C name of an arbitrary enum field,
respecting the legacy prefix and properly uppercasing the name.

9 years agoeolian: butts
Daniel Kolesa [Tue, 12 May 2015 15:15:29 +0000 (16:15 +0100)]
eolian: butts

9 years agoeolian_cxx: Fix @beta on generated bindings for C++
Vitor Sousa [Tue, 12 May 2015 14:45:16 +0000 (11:45 -0300)]
eolian_cxx: Fix @beta on generated bindings for C++

9 years agoeolian: fix wrong enum tests
Daniel Kolesa [Tue, 12 May 2015 13:37:16 +0000 (14:37 +0100)]
eolian: fix wrong enum tests

9 years agoeolian: update eolian_type_base_type_get for REGULAR types
Daniel Kolesa [Tue, 12 May 2015 13:27:02 +0000 (14:27 +0100)]
eolian: update eolian_type_base_type_get for REGULAR types

9 years agoecore-wayland: Input grab_count is unsigned int, no need for comparison < 0
Chris Michael [Tue, 12 May 2015 13:14:26 +0000 (09:14 -0400)]
ecore-wayland: Input grab_count is unsigned int, no need for comparison < 0

Summary: As input->grab_count is an unsigned int there is no need for
the < 0 comparison as that will always return false

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoecore-wayland: Add a grab_count variable for synchronization wl_input_grab/ungrab...
jhyuni.kang [Tue, 12 May 2015 12:57:03 +0000 (08:57 -0400)]
ecore-wayland: Add a grab_count variable for synchronization wl_input_grab/ungrab with pointer and touch event

Summary:
The touch screen device generates touch events.
         But in some special enviroments, a first finger will be matched to a pointer event(not touch event).
         And other fingers (second, third, ...) will be matched touch events.
         In that case ecore_wl_input_ungrab() is called abnormally.
         A first finger pressed, _ecore_wl_input_cb_pointer_button() call ecore_wl_input_grab().
         A second finger pressed, _ecore_wl_input_cb_touch_down() is called but not grab.
         But when a second finger is released, _ecore_wl_input_cb_touch_up() call ecore_wl_input_ungrab()
         So ungrab function generate two mouse up events and a first finger is released.
         In other case, first finger pressed -> second finger pressed -> first finger release.
         That case when a first finger released a second finger release event is generated.
         So after that application doesn't get a release event about a second finger
         when a second finger is really released.

         I think in a multitouch case, ungrab function will be called when a all finger are released.
         So I add a grab_count variable for count currently touched fingers.
         And only called a ungrab funtion all fingers are released.

Test Plan:
In a touch screen supported multitouch, press two or more fingers and release.
           And watch events generation.

Reviewers: raster, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2481

9 years agoecore-drm: Add API function to test if an output can go on a given crtc
Chris Michael [Mon, 11 May 2015 16:37:24 +0000 (12:37 -0400)]
ecore-drm: Add API function to test if an output can go on a given crtc

Summary: This adds a new API function to test if a given
Ecore_Drm_Output can be used on a given crtc. This is needed for DRM
RandR support

@feature

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoEina log: Fix previous commit (ERR & CRI -> bt)
Jean-Philippe Andre [Tue, 12 May 2015 10:34:55 +0000 (19:34 +0900)]
Eina log: Fix previous commit (ERR & CRI -> bt)

This was a < comparison, not <=
I don't totally agree on taking the bts for each error... just try
to cat a binary file in terminology and have fun.

9 years agoEina log: Enable backtrace only for CRI and ERR by default
Jean-Philippe Andre [Tue, 12 May 2015 10:00:57 +0000 (19:00 +0900)]
Eina log: Enable backtrace only for CRI and ERR by default

Using EINA_LOG_LEVEL=4 for standard debugging has now become
absolutely horrible (and slow!). Backtraces may make sense in
case of ERR and CRI messages, but are just pollution for other
levels.

WRN could be argued over but the old env variable is still there
so just use it if you want backtraces:

$ export EINA_LOG_BACKTRACE=2

9 years agoEvas GL common: Skip shaders generation if there is no change
Jean-Philippe Andre [Tue, 12 May 2015 02:30:57 +0000 (11:30 +0900)]
Evas GL common: Skip shaders generation if there is no change

Based on a quick git diff we check that the glsl code has not changed.
This should fix out-of-tree builds and avoid all source modification
unless required.

When compiling from a tarball there should be no git tree (err 129),
or if there's one the files should not be checked in (ie. no diff).
If you changed the glsl files in a tarball... too bad for you.

If this is still not enough to fix the build, then go ahead and disable
the script from Makefile_Evas.am

I would like to note that the auto-generation during make is extremely
useful when working on the shaders, which is why I'd rather keep it enabled.

@fix

9 years agodata: Silent another mkdir when usign automake silent rules
Stefan Schmidt [Mon, 11 May 2015 16:36:48 +0000 (18:36 +0200)]
data: Silent another mkdir when usign automake silent rules

Visible when running as verbose.

9 years agoeolian: fix up remaining doc comments in the tree
Daniel Kolesa [Mon, 11 May 2015 14:37:19 +0000 (15:37 +0100)]
eolian: fix up remaining doc comments in the tree

9 years agoeolian: fix doc comments across the tree
Daniel Kolesa [Mon, 11 May 2015 14:16:17 +0000 (15:16 +0100)]
eolian: fix doc comments across the tree

9 years agoecore_wl: get dpi after all pending events are handled
Boram Park [Mon, 11 May 2015 13:27:13 +0000 (09:27 -0400)]
ecore_wl: get dpi after all pending events are handled

Summary: ecore_wl_dpi_get will return the correct value after wl_output's events are handled

Reviewers: zmike, devilhorns, bryceharrington

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2479

9 years agorg_etc: Correct ifdef to keep function available for debug build
Stefan Schmidt [Mon, 11 May 2015 12:44:57 +0000 (14:44 +0200)]
rg_etc: Correct ifdef to keep function available for debug build

In the nightly builds we have debug enabled and this spotted the case where
rg_etc1_solution_coordinates_block_colors_get is actually still used:

lib/eet/.libs/libeet.so: undefined reference to `rg_etc1_solution_coordinates_block_colors_get'

Showed only after we switched back from release to dev mode.

@fix

9 years agoefreet: fix signed warning
Sebastian Dransfeld [Mon, 11 May 2015 11:31:21 +0000 (13:31 +0200)]
efreet: fix signed warning

Update to 6a0d23. Casting to int isn't a real solution, since we could
have values which overflows.

Since we want the absolute value, just make sure we subtract the larger
value from the smaller.

9 years agoedje: rename variable
Sebastian Dransfeld [Mon, 11 May 2015 10:21:54 +0000 (12:21 +0200)]
edje: rename variable

Give variable a name which better describes its purpose.

9 years agoevas: fix bug in evas gl texture.
Ji-Youn Park [Mon, 11 May 2015 11:07:12 +0000 (20:07 +0900)]
evas: fix bug in evas gl texture.
Summary: bytecount always 1,2,4. so changed for simple code.

9 years agobuild: Fix benchmark and exmaples dep on all target as well.
Stefan Schmidt [Mon, 11 May 2015 10:45:44 +0000 (12:45 +0200)]
build: Fix benchmark and exmaples dep on all target as well.

Same as the check-build target I fixed in my previous commit these two need
the all dependency to compile everything needed before being run.

9 years agobuild: Add correct dependency for check-build target.
Stefan Schmidt [Mon, 11 May 2015 10:34:54 +0000 (12:34 +0200)]
build: Add correct dependency for check-build target.

We need to build everythign else before. Without this dep running check-build
as first target from a fresh build will fail due to wrong dependency handling
(like no eolian run over the eo files, etc)

Inspired by D2489 from Kabeer Khan.