platform/upstream/efl.git
10 years agoEvas: 3D: API documentation
Taekyun Kim [Wed, 8 Jan 2014 10:04:52 +0000 (19:04 +0900)]
Evas: 3D: API documentation

Conflicts:

src/lib/evas/Evas_3D.h

10 years agoEvas: 3D: Updated proxy texture example
Taekyun Kim [Tue, 7 Jan 2014 09:07:38 +0000 (18:07 +0900)]
Evas: 3D: Updated proxy texture example

10 years agoEvas: 3D: Refined proxy texture implementation
Taekyun Kim [Tue, 7 Jan 2014 08:39:23 +0000 (17:39 +0900)]
Evas: 3D: Refined proxy texture implementation

Added new APIs for controlling source object's visibility.
Fixed bugs related to updating proxy textures.

10 years agoEvas: 3D: Fix typo error in preprocessor for GLES
Taekyun Kim [Mon, 6 Jan 2014 06:12:02 +0000 (15:12 +0900)]
Evas: 3D: Fix typo error in preprocessor for GLES

10 years agoEvas: 3D: evas_3d_node_look_at_set() bug fix
Taekyun Kim [Mon, 6 Jan 2014 06:10:59 +0000 (15:10 +0900)]
Evas: 3D: evas_3d_node_look_at_set() bug fix

10 years agoEvas: 3D: Example: MD2 mesh file example
Taekyun Kim [Fri, 27 Dec 2013 12:14:59 +0000 (21:14 +0900)]
Evas: 3D: Example: MD2 mesh file example

MD2 model file is from http://mb.srb2.org/showthread.php?t=34800
Thanks to Jeck Jim.

10 years agoEvas: 3D: Example: Picking example
Taekyun Kim [Fri, 27 Dec 2013 11:31:39 +0000 (20:31 +0900)]
Evas: 3D: Example: Picking example

10 years agoEvas: 3D: Example: Proxy texture example
Taekyun Kim [Fri, 27 Dec 2013 11:29:07 +0000 (20:29 +0900)]
Evas: 3D: Example: Proxy texture example

10 years agoEvas: 3D: Example: Rotating Cube 2
Taekyun Kim [Fri, 27 Dec 2013 09:45:41 +0000 (18:45 +0900)]
Evas: 3D: Example: Rotating Cube 2

Demonstrating key frame material animation and normal mapping.

10 years agoEvas: 3D: Example: Rotating cube
Taekyun Kim [Fri, 27 Dec 2013 08:39:58 +0000 (17:39 +0900)]
Evas: 3D: Example: Rotating cube

10 years agoevas/evas3d: include evas3d header file properly.
ChunEon Park [Fri, 25 Apr 2014 07:17:43 +0000 (16:17 +0900)]
evas/evas3d: include evas3d header file properly.

10 years agoEvas: 3D: Introducing 3D scene rendering features
Taekyun Kim [Fri, 27 Dec 2013 07:56:30 +0000 (16:56 +0900)]
Evas: 3D: Introducing 3D scene rendering features

Enable 3D features using --enable-evas-3d=yes when configuring.
APIs are exposed through Evas_3D.h.
Currently, evas-3d is being supported only on gl_x11 engine.

Conflicts:

src/lib/evas/Evas_Eo.h

10 years agoEvas: gif decoding bug fix
Jaeun Choi [Fri, 25 Apr 2014 06:26:48 +0000 (15:26 +0900)]
Evas: gif decoding bug fix

Gif decoder decodes prior frames sequentially to decode a specific frame.
The last frame of sequential decoding, which is the frame we want to decode,
remains un-decoded until the while loop stops.
The frame count should be incremented after the comparison statement.

10 years agoedje: Remove build warnning by clang
bluezery [Fri, 25 Apr 2014 05:27:15 +0000 (14:27 +0900)]
edje: Remove build warnning by clang

Summary:
This can fix below warning:
bin/edje/edje_cc_handlers.c:9839:42: warning: equality comparison with
extraneous parentheses [-Wparentheses-equality]

Reviewers: raster, cedric, seoz

Reviewed By: seoz

CC: cedric
Differential Revision: https://phab.enlightenment.org/D792

10 years agoEvas TGV: Fix distortion of TGV images (ETC1)
Jean-Philippe Andre [Wed, 23 Apr 2014 09:08:36 +0000 (18:08 +0900)]
Evas TGV: Fix distortion of TGV images (ETC1)

Due to some invalid geometry considerations, there was a
1 pixel distortion in images, varying with the lz4 compressed
macro-block size.

Anyhow, I couldn't wrap my head around Cedric's code. So I rewrote
the whole thing instead, fixed it and improved the block size
selection (based on the image size, to optimize lz4 compression).

10 years agoEdje/Edje_edit: corrected type of input parameter for offset value
se.osadchy [Fri, 25 Apr 2014 02:09:35 +0000 (11:09 +0900)]
Edje/Edje_edit: corrected type of input parameter for offset value

Summary: Input type for offset value must be Integer in the "edje_edit_rel1(2)_offset_x(y)_set" functions

Reviewers: cedric, Hermet

CC: cedric, Hermet, raster
Differential Revision: https://phab.enlightenment.org/D562

10 years agoClear the values of pipe before calling handler.
Ryuan Choi [Fri, 25 Apr 2014 01:29:01 +0000 (10:29 +0900)]
Clear the values of pipe before calling handler.

Summary:
Applications are stuck when handler of pipe made nested loop.
In the nested loop, _ecore_pipe_read() tried to read new data based on previous information which is not cleared yet.

Spotted by gyuyoung.kim, sy302.park.

Related webkit bug is https://bugs.webkit.org/show_bug.cgi?id=129294

Reviewers: cedric, seoz, raster

CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D790

10 years agoEolian: Remove support of legacy into implement section.
Daniel Zaoui [Thu, 24 Apr 2014 13:51:04 +0000 (16:51 +0300)]
Eolian: Remove support of legacy into implement section.

It has been decided that this support is not needed.

10 years agoEolian: Removal of the support for custom destructors.
Daniel Zaoui [Thu, 24 Apr 2014 13:30:40 +0000 (16:30 +0300)]
Eolian: Removal of the support for custom destructors.

For no reason, I added it a long time ago.

10 years agoEolian/Tests: add test for ctors and dtors.
Daniel Zaoui [Thu, 24 Apr 2014 12:43:37 +0000 (15:43 +0300)]
Eolian/Tests: add test for ctors and dtors.

10 years agoTextBlock: fix porting to Eolian.
Daniel Zaoui [Thu, 24 Apr 2014 11:17:47 +0000 (14:17 +0300)]
TextBlock: fix porting to Eolian.

evas_object_text_clear was no more defined in the headers. Since the
function code is not generated by Eolian, the declaration in the headers
has to be done manually too.

10 years agoEolian/Tests: Clean makefile + add return comment tests.
Daniel Zaoui [Thu, 24 Apr 2014 11:07:18 +0000 (14:07 +0300)]
Eolian/Tests: Clean makefile + add return comment tests.

10 years agoemotion: play set before file open callback bug fix
Wonguk Jeong [Thu, 24 Apr 2014 09:10:10 +0000 (18:10 +0900)]
emotion: play set before file open callback bug fix

Summary:
Before "open_done" callback, play unset is ineffective.
is it intentional logic?

if we set play and unset play again before "open_done" callback, the video will be played.

@fix

Reviewers: raster, cedric

CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D786

10 years agoEolian: Remove useless functions.
Daniel Zaoui [Wed, 23 Apr 2014 13:26:15 +0000 (16:26 +0300)]
Eolian: Remove useless functions.

Default ctor/dtor are considered as implemented functions and no more as
special functions.

10 years agorevert edje ellipsis doc changes
Mike Blumenkrantz [Wed, 23 Apr 2014 13:27:31 +0000 (09:27 -0400)]
revert edje ellipsis doc changes

if we're reverting functional changes, then it's probably a good idea to revert the accompanying documentation

10 years agoEolian: Legacy integration of Evas class
Yossi Kantor [Thu, 3 Apr 2014 15:15:14 +0000 (18:15 +0300)]
Eolian: Legacy integration of Evas class

10 years agoEolian: Legacy integration of Evas Box and Evas Out
Yossi Kantor [Thu, 3 Apr 2014 12:14:46 +0000 (15:14 +0300)]
Eolian: Legacy integration of Evas Box and Evas Out

10 years agoEolian: Legacy integration of Evas Image
Yossi Kantor [Thu, 3 Apr 2014 11:07:33 +0000 (14:07 +0300)]
Eolian: Legacy integration of Evas Image

10 years agoEolian: Legacy integration of Evas Grid
Yossi Kantor [Wed, 2 Apr 2014 14:53:53 +0000 (17:53 +0300)]
Eolian: Legacy integration of Evas Grid

10 years agoEolian: Legacy integration of Evas Object
Yossi Kantor [Wed, 2 Apr 2014 13:15:28 +0000 (16:15 +0300)]
Eolian: Legacy integration of Evas Object

10 years agoEolian: Legacy integration of Evas Common Interface
Yossi Kantor [Wed, 2 Apr 2014 10:46:36 +0000 (13:46 +0300)]
Eolian: Legacy integration of Evas Common Interface

10 years agoEolian: Legacy partial integration of Evas Smart
Yossi Kantor [Wed, 2 Apr 2014 07:27:36 +0000 (10:27 +0300)]
Eolian: Legacy partial integration of Evas Smart

10 years agoEolian: Fix distcheck
Daniel Zaoui [Wed, 23 Apr 2014 12:54:12 +0000 (15:54 +0300)]
Eolian: Fix distcheck

scope.eo was not included in EXTRA_DIST.

10 years agoEolian: Legacy integration of Ecore
Yossi Kantor [Mon, 7 Apr 2014 08:28:53 +0000 (11:28 +0300)]
Eolian: Legacy integration of Ecore

10 years agoEolian: Legacy integration of Edje
Yossi Kantor [Sun, 6 Apr 2014 10:42:36 +0000 (13:42 +0300)]
Eolian: Legacy integration of Edje

10 years agopartly revert mike's ellipsis changes as they break edje
Carsten Haitzler (Rasterman) [Wed, 23 Apr 2014 10:57:15 +0000 (19:57 +0900)]
partly revert mike's ellipsis changes as they break edje

partly revert adcc32329128b13fda47405a896319c7c96ea85f as the default
ellipsis value was 0 as per the document, and must stay, as changing
this breaks edc descirptions as now text is no longer ellipsised by
default. this ACTUALLY broke titlebars on the default theme - just
have a title that is too long and see how it no longer goes:

This is a title he...

it instead covers the screen for as long as the title is.

if you want -1 for ellipsis... then set it. :)

10 years agoEo: adapt benchmarks to Eo2.
Daniel Zaoui [Wed, 23 Apr 2014 08:23:39 +0000 (11:23 +0300)]
Eo: adapt benchmarks to Eo2.

10 years agoEolian: Space after const missing in function parameters of legacy header
Yossi Kantor [Thu, 10 Apr 2014 08:44:27 +0000 (11:44 +0300)]
Eolian: Space after const missing in function parameters of legacy header

10 years agoEolian/Generator: support --eo.
Daniel Zaoui [Wed, 23 Apr 2014 05:55:15 +0000 (08:55 +0300)]
Eolian/Generator: support --eo.

--eo1 is still supported but it will be removed soon.

10 years agoEo: adapt examples to Eo2.
Daniel Zaoui [Tue, 22 Apr 2014 15:39:38 +0000 (18:39 +0300)]
Eo: adapt examples to Eo2.

It includes Eo and Evas porting to Eo2.

10 years agoedje: the ellipsis fuckup preservation commit.
Mike Blumenkrantz [Wed, 23 Apr 2014 00:18:46 +0000 (20:18 -0400)]
edje: the ellipsis fuckup preservation commit.

what a huge, colossal cock-up of a clusterfuck. it's a good thing nobody ever uses ellipses or edje. otherwise we'd probably get complaints about this kind of thing.

10 years agoedje now correctly forces ellipsis setting on text parts
Mike Blumenkrantz [Tue, 22 Apr 2014 23:17:13 +0000 (19:17 -0400)]
edje now correctly forces ellipsis setting on text parts

this should always be applied, not just when min_x isn't set; failure to do so causes erroneous ellipsizing in some cases where it should be explicitly prohibited

@fix

10 years agoEolian: add methods and properties scope support tests
Jérémy Zurcher [Tue, 22 Apr 2014 18:31:11 +0000 (20:31 +0200)]
Eolian: add methods and properties scope support tests

10 years agoEolian: fix properties and methods scope support
Jérémy Zurcher [Tue, 22 Apr 2014 18:24:27 +0000 (20:24 +0200)]
Eolian: fix properties and methods scope support

define scope keywords 'protected' and 'public'
disallow scope keywords as property or method name
use move_ts action to move token start after scope modifier if found
use end_property_scope and end_method_scope actions to handle it

10 years agoimprove eo call stack
Carsten Haitzler (Rasterman) [Tue, 22 Apr 2014 11:22:35 +0000 (20:22 +0900)]
improve eo call stack

it now does up to 8192 entries and madvise unused upper pages when
dropping. also delay dropping to avoi too many syscalls

10 years agoEvas gl: Don't swap RGBA colors with ETC1
Jean-Philippe Andre [Tue, 22 Apr 2014 10:53:55 +0000 (19:53 +0900)]
Evas gl: Don't swap RGBA colors with ETC1

Since rg_etc1 now outputs proper BGRA data, the shaders should not
swizzle the colors around. Stick to the normal fragment shaders.

Note: This is not tested.

10 years agoeo - callstack. realloc is a bad idea. use mmap and keep addr fixed
Carsten Haitzler (Rasterman) [Tue, 22 Apr 2014 10:17:26 +0000 (19:17 +0900)]
eo - callstack. realloc is a bad idea. use mmap and keep addr fixed

this fixes and eo2 problem where when callstack grows (or shrinks)
and realloc nas to relocate memory, the frame ptrs like fptr become
invalid and all sorts of hell ensues.

this uses mmap so blowing the stack will segv, not scribble over
memory, also its separated from malloc heap, and now big enough to not
need to size ... ever (1024 entries).

10 years agorg_etc1: Fix RGBA vs BGRA mishandling of the ETC1 codec
Jean-Philippe Andre [Tue, 22 Apr 2014 02:52:53 +0000 (11:52 +0900)]
rg_etc1: Fix RGBA vs BGRA mishandling of the ETC1 codec

Evas uses BGRA data while rg_etc1 uses RGBA data, so there
were incompatibilities between the two.

Now, rg_etc1 will take BGRA data as input and output.

10 years agoEvas gl_x11: Fix abort() in evas_object_image_data_get()
Jean-Philippe Andre [Tue, 22 Apr 2014 08:12:02 +0000 (17:12 +0900)]
Evas gl_x11: Fix abort() in evas_object_image_data_get()

Since the introduction of new colorspaces for GL_X11
(GRY8, AGRY88 and ETC1), stride_get() would return
an invalid value and data_get() would just abort().

Add proper support for these functions.
ETC1 data will NOT be returned from data_get() and
stride_get() will return 0. This is to avoid people from
messing up badly with encoded color spaces.

10 years agoEo: Add tests dir to .gitignore
Daniel Zaoui [Mon, 21 Apr 2014 14:02:36 +0000 (17:02 +0300)]
Eo: Add tests dir to .gitignore

10 years agoEolian: infras for tests + first test
Daniel Zaoui [Mon, 21 Apr 2014 06:36:44 +0000 (09:36 +0300)]
Eolian: infras for tests + first test

10 years agoEcore: Fix compilation warnings
Jean-Philippe Andre [Tue, 22 Apr 2014 03:41:07 +0000 (12:41 +0900)]
Ecore: Fix compilation warnings

When HAVVE_LIBTS is not defined, we have 3 unused variable warnings.
Fix that.

10 years agorg_etc1: Disable code producing artifacts
Jean-Philippe Andre [Mon, 21 Apr 2014 02:41:28 +0000 (11:41 +0900)]
rg_etc1: Disable code producing artifacts

So I must have been a bit tired last Friday when "fixing" some
code producing artifacts, as I was just basically disabling part
of the code without realizing it :)
Let's just disable it then.

10 years agoedje now alerts the user to its failure when running recursive embryo programs
Mike Blumenkrantz [Mon, 21 Apr 2014 14:04:31 +0000 (10:04 -0400)]
edje now alerts the user to its failure when running recursive embryo programs

10 years agotypo from edje program info commit
Mike Blumenkrantz [Mon, 21 Apr 2014 03:32:08 +0000 (23:32 -0400)]
typo from edje program info commit

10 years agoedje programs now provide slightly more verbose errors when disabling due to recursio...
Mike Blumenkrantz [Mon, 21 Apr 2014 03:30:13 +0000 (23:30 -0400)]
edje programs now provide slightly more verbose errors when disabling due to recursion limits

10 years agoemotion: fix triggering condition of fallback module loading
Wonguk Jeong [Sun, 20 Apr 2014 13:58:47 +0000 (09:58 -0400)]
emotion: fix triggering condition of fallback module loading

@fix

Summary:
In emotion_engine_instance_new(), if given module name is wrong or if loding is failed, fallback module loading should be triggered.a (previously, it check just whether name is NULL or not)

In case of T1025, "vlc" was given name. Therefore, fallback was not triggered.

Fixes T1025

Test Plan: enlightenment -> efm -> mouse over multimedia files -> check whether preview is shown properly or not

Reviewers: raster, cedric

CC: seoz, cedric, zmike
Maniphest Tasks: T1025

Differential Revision: https://phab.enlightenment.org/D773

10 years agoecore_evas_convert: Fix default quality setting
Jean-Philippe Andre [Fri, 18 Apr 2014 10:02:28 +0000 (19:02 +0900)]
ecore_evas_convert: Fix default quality setting

Quality should not default to 100 unless specified in the
command line. In particular, we don't want to save ETC1 at
high quality by default since it can take hours (literally).

10 years agorg_etc1: Fix rare artifacts + add comments
Jean-Philippe Andre [Fri, 18 Apr 2014 09:17:57 +0000 (18:17 +0900)]
rg_etc1: Fix rare artifacts + add comments

Add some comments as I'm reading and understanding the code.
Fix some rare artifacts happening mostly with medium quality
encoding, where a few pixels (2x2, 2x4 or 4x2) will have a
horrible contrast with their surroundings (eg. pink over black).

10 years agoEvas: Add comments to the TGV saver (FIXME)
Jean-Philippe Andre [Fri, 18 Apr 2014 09:32:20 +0000 (18:32 +0900)]
Evas: Add comments to the TGV saver (FIXME)

The block size has a visual effect. This is obviously a bug.

10 years agoEdje: Use LOSSY_ETC1 instead of LOSSY
Jean-Philippe Andre [Fri, 18 Apr 2014 07:39:37 +0000 (16:39 +0900)]
Edje: Use LOSSY_ETC1 instead of LOSSY

Add a new flag in EDC files to specify ETC1 compression
should be enabled. It follows the same rules as the
current LOSSY flag for JPEG compression.

@feature

10 years agorg_etc1: Fix encoding of solid color blocks
Jean-Philippe Andre [Fri, 18 Apr 2014 06:31:43 +0000 (15:31 +0900)]
rg_etc1: Fix encoding of solid color blocks

The ETC1 encoder is expected to write all 8 bytes of the
output data. But in case of a solid color block, it was writing
only 1 of the first 3 bytes (R, G, B). So lots of solid blocks
were containing invalid data (for instance: R + dR < 0 or > 255).

10 years agoEdje: Fix ETC1 encoding with EET and Edje
Jean-Philippe Andre [Fri, 18 Apr 2014 04:12:03 +0000 (13:12 +0900)]
Edje: Fix ETC1 encoding with EET and Edje

There were a few critical issues:
- Invalid pointer arithmetics on the input data (char vs. int)
- Invalid logic in the pixel duplication code

All of these due to bad copy and paste :(

Also, use LZ4HC instead of LZ4 when compression is enabled.
ETC1 encoding is so damn slow you won't see the difference between
LZ4 and LZ4HC compression times.

10 years agoEdje: Add support for ETC1 encoding
Jean-Philippe Andre [Thu, 17 Apr 2014 10:04:31 +0000 (19:04 +0900)]
Edje: Add support for ETC1 encoding

Edje will compress images as ETC1 provided that they don't
contain any alpha AND the command line includes -etc1.

10 years agoEet: Use medium quality for ETC1 by default
Jean-Philippe Andre [Thu, 17 Apr 2014 10:03:19 +0000 (19:03 +0900)]
Eet: Use medium quality for ETC1 by default

And remove  dithering.
This is the same as  07700a799ce759

10 years agoEolian/Lexer: Fix coverity issues.
Daniel Zaoui [Fri, 18 Apr 2014 06:21:35 +0000 (09:21 +0300)]
Eolian/Lexer: Fix coverity issues.

Fix CID 1203411
Fix CID 1203412

10 years agoEolian/Generator: Fix coverity issues.
Daniel Zaoui [Fri, 18 Apr 2014 06:01:37 +0000 (09:01 +0300)]
Eolian/Generator: Fix coverity issues.

Fix CID 1203409
Fix CID 1203410

10 years agoEolian/Lexer: fix when function name is composed of only one letter.
Daniel Zaoui [Thu, 17 Apr 2014 14:07:23 +0000 (17:07 +0300)]
Eolian/Lexer: fix when function name is composed of only one letter.

10 years agoedje: let's call this function before returning.
Cedric Bail [Thu, 17 Apr 2014 19:29:31 +0000 (21:29 +0200)]
edje: let's call this function before returning.

CID 1194718.

10 years agorg_etc: check against meaningful size (pointer size is not)
Cedric Bail [Thu, 17 Apr 2014 19:26:57 +0000 (21:26 +0200)]
rg_etc: check against meaningful size (pointer size is not)

CID 1195440.

10 years agoevas: let's check fwrite work as expected.
Cedric Bail [Thu, 17 Apr 2014 19:21:54 +0000 (21:21 +0200)]
evas: let's check fwrite work as expected.

CID 1195441.

10 years agoevas: fix TGV loader to not leak ressource.
Cedric Bail [Thu, 17 Apr 2014 19:18:32 +0000 (21:18 +0200)]
evas: fix TGV loader to not leak ressource.

CID 1195443.

10 years agoecore_evas: avoid using non initialized value for cursor position in X backend.
Cedric Bail [Thu, 17 Apr 2014 19:16:12 +0000 (21:16 +0200)]
ecore_evas: avoid using non initialized value for cursor position in X backend.

CID 1195445.

10 years agoeo: do not risk dereferencing NULL.
Cedric Bail [Thu, 17 Apr 2014 19:10:34 +0000 (21:10 +0200)]
eo: do not risk dereferencing NULL.

CID 1199563.

10 years agoecore: Fixed se of EINA_MAIN_LOOP_CHECK_RETURN macros with functions with return...
Felipe Magno de Almeida [Thu, 17 Apr 2014 19:08:43 +0000 (21:08 +0200)]
ecore: Fixed se of EINA_MAIN_LOOP_CHECK_RETURN macros with functions with return types

Summary:
Changed uses of EINA_MAIN_LOOP_CHECK_RETURN for EINA_MAIN_LOOP_CHECK_RETURN_VAL
for functions that doesn't have void return types.

These only error out when compiling with --with-profile=debug

@fixed

Reviewers: raster, cedric, smohanty

CC: cedric
Differential Revision: https://phab.enlightenment.org/D765

10 years agoedje_edit: image_del fixed
Andrii Kroitor [Thu, 17 Apr 2014 17:40:34 +0000 (19:40 +0200)]
edje_edit: image_del fixed

Summary: fixed segfault on broken refs in edje object

Reviewers: cedric, seoz, raster

CC: reutskiy.v.v, cedric
Differential Revision: https://phab.enlightenment.org/D759

Signed-off-by: Cedric Bail <cedric.bail@free.fr>
10 years agoeo: fix broken children iterator, remove redundant fields.
Lukasz Stanislawski [Thu, 17 Apr 2014 17:36:10 +0000 (19:36 +0200)]
eo: fix broken children iterator, remove redundant fields.

@fix

Summary: Tests added.

Reviewers: raster, JackDanielZ, tasn

CC: cedric
Differential Revision: https://phab.enlightenment.org/D763

Signed-off-by: Cedric Bail <cedric.bail@free.fr>
10 years agoefreet: check magic only for file which has size
Wonguk Jeong [Thu, 17 Apr 2014 17:34:33 +0000 (19:34 +0200)]
efreet: check magic only for file which has size

Summary:
Due to unnecessary magic checking, there was freezing in /proc in efm.
Proc file's st_size is zero, but, it's readable. therfore, it takes unnecessary time in magic checking. And, there is no need to check magic in case of 0 sized regular files as well.

Therefore, skip magic check in case of st_size is zero.

Fixes T1173

Test Plan: enlightenment -> file browser (efm) -> get int /proc --> check whether efm freezes or not

Reviewers: raster, cedric, zmike

CC: seoz, cedric
Maniphest Tasks: T1173

Differential Revision: https://phab.enlightenment.org/D764

Signed-off-by: Cedric Bail <cedric.bail@free.fr>
10 years agoecore_wl: Actually watch out for ECORE_FD_ERROR
Stefan Schmidt [Thu, 17 Apr 2014 15:01:45 +0000 (17:01 +0200)]
ecore_wl: Actually watch out for ECORE_FD_ERROR

10 years agowcore_wl: Handle errors on wayland display fd
Stefan Schmidt [Thu, 17 Apr 2014 14:16:57 +0000 (16:16 +0200)]
wcore_wl: Handle errors on wayland display fd

Cancel the callback if we receive an error on the wayland display
file descriptor.

10 years agoecore_wl: Add FD_WRITE flag to handler as we monitor read and write
Stefan Schmidt [Thu, 17 Apr 2014 14:37:44 +0000 (16:37 +0200)]
ecore_wl: Add FD_WRITE flag to handler as we monitor read and write

The _ecore_wl_cb_handle_data callback only has the flag for ECORE_FD_READ
set while the callback funtion also monitors and acts on writes. I wonder
if and how that worked before.

10 years agoEolian/Generator: remove Eo1 generation
Daniel Zaoui [Thu, 17 Apr 2014 12:50:22 +0000 (15:50 +0300)]
Eolian/Generator: remove Eo1 generation

10 years agoEolian/Lexer: support of complex types
Daniel Zaoui [Wed, 16 Apr 2014 13:17:27 +0000 (16:17 +0300)]
Eolian/Lexer: support of complex types

Complex types are now supported in .eo files:
Eina_List * @own <Eo *>

The parser is in charge of creating a list of formatted basic types.

10 years agoformatting fix (spacing)
Carsten Haitzler (Rasterman) [Thu, 17 Apr 2014 07:35:43 +0000 (16:35 +0900)]
formatting fix (spacing)

10 years agoEvas gl: Fix texture border drawing
Jean-Philippe Andre [Thu, 17 Apr 2014 05:59:52 +0000 (14:59 +0900)]
Evas gl: Fix texture border drawing

Since the introduction of color spaces other than RGBA8888 in
the GL engines, there was an issue with border images scaled in
GL. The left and right edges were simply not properly copied.

This would then show artifacts when scaling very thin images
(typically 2px wide).

10 years agoEvas TGV saver: Use medium quality by default
Jean-Philippe Andre [Wed, 16 Apr 2014 09:28:56 +0000 (18:28 +0900)]
Evas TGV saver: Use medium quality by default

The default quality is 80 and high quality is horribly slow,
so default to medium quality instead (it's already very slow).

If you really want to shoot yourself in the foot and use high
quality, just set quality > 95 and go make some coffee.

Also, disable dithering, it creates horrible artifacts on real
life pictures (gradients and flat surfaces especially).

10 years agofix overdraw issue in evas when clips change
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 07:14:16 +0000 (16:14 +0900)]
fix overdraw issue in evas when clips change

the comments say it all. this stops adding massive full window updates
when clippers lose all children (and are visible) or gain a first
child while visible as all code hides these clip rects (or shows them)
in these cases and other usages just make no sense.

10 years agoremove unused code
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 06:22:42 +0000 (15:22 +0900)]
remove unused code

10 years agoevas rects - missing update rect del if rect solid && not changed
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 06:22:08 +0000 (15:22 +0900)]
evas rects - missing update rect del if rect solid && not changed

small improvement

10 years agoeo - let's make NULL objects simply debug warnings, not errors.
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 03:31:44 +0000 (12:31 +0900)]
eo - let's make NULL objects simply debug warnings, not errors.

due to recent changes a lot of objects are now NULL (correctly) and eo
complains on access of them. it's simply too noisy adding too many
if's all through code, so let's just make eo be sensible here.

10 years agoRevert "eo null object complaints - silence."
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 03:27:40 +0000 (12:27 +0900)]
Revert "eo null object complaints - silence."

This reverts commit 1678d272af676a17e4608de655f231171c5a5af2.

... actually... on second thought... i think eo just needs to sensibly
handle null objects.

10 years agoeo null object complaints - silence.
Carsten Haitzler (Rasterman) [Wed, 16 Apr 2014 03:18:06 +0000 (12:18 +0900)]
eo null object complaints - silence.

10 years agoRevert "build: Temporary disable debug log output during coverage"
Stefan Schmidt [Tue, 15 Apr 2014 13:24:51 +0000 (15:24 +0200)]
Revert "build: Temporary disable debug log output during coverage"

This reverts commit 02a5a6bfea50e722bbb71b9cd813d32e453517ae.

Daniel fixed the underlaying problem in eolian debug statements which should
make this band aid unnecessary.

10 years agoevas - improve speed of unpremul with redundant pixel data
Matvey Konovalov [Tue, 15 Apr 2014 10:22:33 +0000 (19:22 +0900)]
evas - improve speed of unpremul with redundant pixel data

10 years agoEvas ETC1: Use LZ4HC for higher compression ratios
Jean-Philippe Andre [Tue, 15 Apr 2014 02:41:16 +0000 (11:41 +0900)]
Evas ETC1: Use LZ4HC for higher compression ratios

LZ4HC has a higher compression ratio than LZ4 but basically the
same decompression speed.

The performance cost during encoding is actually still pretty low
considering how expensive ETC1 compression can be (even at medium
quality).

10 years agoEvas: Use Evas_Colorspace to declare cspace
Jean-Philippe Andre [Tue, 8 Apr 2014 08:19:38 +0000 (17:19 +0900)]
Evas: Use Evas_Colorspace to declare cspace

Why use int when we have a proper type?
All these APIs are internal.

10 years agoecore-wl: Check for existance of data device manager
Chris Michael [Tue, 15 Apr 2014 09:38:51 +0000 (10:38 +0100)]
ecore-wl: Check for existance of data device manager

@fix: Don't segfault if the running compositor never creates a data
device manager.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
10 years agoecore-wl: Don't create an input cursor surface unless the input has pointer capability.
Chris Michael [Tue, 15 Apr 2014 09:36:37 +0000 (10:36 +0100)]
ecore-wl: Don't create an input cursor surface unless the input has pointer capability.

@fix: There is no use for an input cursor surface if the seat does not
have the pointer ability (for kiosk cases, or touch-only cases where
a pointer is not created).

Signed-off-by: Chris Michael <cp.michael@samsung.com>
10 years agoecore-evas-wl: Ooops, Fix opaque & input regions being set to incorrect values
Chris Michael [Fri, 11 Apr 2014 06:21:20 +0000 (07:21 +0100)]
ecore-evas-wl: Ooops, Fix opaque & input regions being set to incorrect values

@fix: As it turns out, we cannot just blindly set the regions here
during resize. Elementary apps will set the opaque region to account
for any edj frames, and having the region_set calls Here was
causing issues....

Signed-off-by: Chris Michael <cp.michael@samsung.com>
10 years agofix recent png loader break with etc1 support that broke interlaced imgs
Carsten Haitzler (Rasterman) [Tue, 15 Apr 2014 08:56:07 +0000 (17:56 +0900)]
fix recent png loader break with etc1 support that broke interlaced imgs

this fixes the png loader code to use png_read_row properly with the
number of passes needed to load aninterlaced image as well as handling
this right with scale ratio scaledown set.