Guillaume Friloux [Fri, 24 May 2013 12:06:08 +0000 (14:06 +0200)]
cl->start_time isnt initialized, so ecore_con_client_uptime_get() just doesnt work. This patch fixes it. Bug found by asdfuser.
Guillaume Friloux [Fri, 24 May 2013 11:53:07 +0000 (13:53 +0200)]
I missed a timer_reset for the case we did set a server timeout and no client timeout. Thanks again Cravix for the awesome reports.
Daniel Willmann [Fri, 24 May 2013 13:42:11 +0000 (14:42 +0100)]
tests/ecore_con: Sadly IPv4 is still more prevalent than IPv6
Switch to IPv4 loopback - especially since our build slaves don't have
IPv6 support...
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Juyung Seo [Fri, 24 May 2013 11:22:07 +0000 (20:22 +0900)]
Evas_Common_Header.h: fixed documentation.
Daniel Willmann [Fri, 24 May 2013 10:47:47 +0000 (11:47 +0100)]
ecore_con: Reenable ecore_con tests - they seem fine (but don't use port 8080)
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Guillaume Friloux [Fri, 24 May 2013 09:51:09 +0000 (11:51 +0200)]
Fixes another bug found by Cravix (IRC) where even if data is going over the socket, we still timeout the connection if ecore_con_client_timeout_set and ecore_con_server_timeout_set are used! This is even a worst bug than the previous one. This patch should also be applied and tested on 1.8. Thx Cravix!
Guillaume Friloux [Fri, 24 May 2013 08:20:33 +0000 (10:20 +0200)]
Fix a bug introduced on 2011-12-10 05:42:32 (GMT) and nobody ever noticed because there isnt enough people that cares about ecore_con (and the lower level libs in general). Big thanks to Cravix (IRC) for finding this issue!
ChunEon Park [Fri, 24 May 2013 09:03:49 +0000 (18:03 +0900)]
evas - removed inherit mode in display mode.
it's not necessary at this moment and i don't think it's useful neither.
if you need, then please add it at the moment.
Daniel Juyung Seo [Thu, 23 May 2013 11:26:44 +0000 (20:26 +0900)]
eina_list.h: fixed documentation source formatting while reading the code.
Daniel Juyung Seo [Wed, 22 May 2013 11:40:56 +0000 (20:40 +0900)]
ecore_pipe.c: fixed fd handler increasing issue.
Call _ecore_pipe_unhandle() when you return from _ecore_pipe_read() or the fd will never be closed.
This fixed increasing numbers of fd handler issue when you call ecore_pipe_add/del repeatedly.
In that case, reusing ecore_pipe is recommended though.
Mike Blumenkrantz [Wed, 22 May 2013 09:14:43 +0000 (10:14 +0100)]
move ecore-x fd handler creation to after ecore-x init has finished so we can ignore more x events during startup
ChunEon Park [Wed, 22 May 2013 09:04:34 +0000 (18:04 +0900)]
edje/textblock - keep the styles if the new font by the text class is applied.
text is already did. for the consistency textblock should be kept the same way.
ChunEon Park [Wed, 22 May 2013 07:53:31 +0000 (16:53 +0900)]
edje - oops remove not intended lines
ChunEon Park [Wed, 22 May 2013 07:49:43 +0000 (16:49 +0900)]
edje/text - trivial change (typo, indent)
David Walter Seikel [Wed, 22 May 2013 02:56:28 +0000 (12:56 +1000)]
Typo--
David Walter Seikel [Wed, 22 May 2013 02:56:15 +0000 (12:56 +1000)]
Function prototype not needed.
David Walter Seikel [Wed, 22 May 2013 02:55:41 +0000 (12:55 +1000)]
Stop leaking evas objects in Lua.
Bug and test case reported by Leif Middelschulte.
Tom Hacohen [Tue, 21 May 2013 15:31:39 +0000 (16:31 +0100)]
Revert "Evas eo compat: Freeze object events when deleting."
This reverts commit
3c53b824d861ef91662fbe523e4457b6a6d3c8b7.
This is a bad idea for now, as we need to mark some more events as hot
(including compat generated ones) before doing it.
Tom Hacohen [Tue, 21 May 2013 14:07:34 +0000 (15:07 +0100)]
Evas eo compat: Freeze object events when deleting.
In the past events wouldn't be called for deleted objects, retain that
behaviour.
Tom Hacohen [Tue, 21 May 2013 14:06:05 +0000 (15:06 +0100)]
Eo: Made del event an hot event (we always want to get it).
Chris Michael [Tue, 21 May 2013 13:29:37 +0000 (14:29 +0100)]
Fix memleak in ecore_x_randr_crtc_settings_set.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Daniel Juyung Seo [Tue, 21 May 2013 10:21:33 +0000 (19:21 +0900)]
evas_object_textblock.c: fixed a typo in the documentation.
Chris Michael [Tue, 21 May 2013 09:51:15 +0000 (10:51 +0100)]
Set the surface user data to be the current buffer.
NB: This is needed because some compositors (Weston) are expecting
wl_surface_get_user_data to return a buffer.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 21 May 2013 09:49:40 +0000 (10:49 +0100)]
Use ecore_wl_window_surface_find to get the Ecore_Wl_Window.
NB: This is needed because some compositors like Weston are expecting
the wl_surface user_data to be a buffer, not some arbitrary data.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 21 May 2013 09:46:04 +0000 (10:46 +0100)]
Add function to get a Ecore_Wl_Window from a wl_surface.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Cedric Bail [Mon, 20 May 2013 10:23:36 +0000 (19:23 +0900)]
evas: always set degre to zero before using it.
Ryuan Choi [Tue, 21 May 2013 00:38:19 +0000 (09:38 +0900)]
ecore_evas/x: Fixed that cursor object was not disappear
When moved mouse among two objects having cursor set,
EVAS_CALLBACK_MOUSE_OUT and EVAS_CALLBACK_MOUSE_IN are fired.
I observed that evas_object_del in callback of EVAS_CALLBACK_MOUSE_OUT
calls EVAS_CALLBACK_MOUSE_IN
so that newly added cursor object was detached.
Rafael Antognolli [Mon, 20 May 2013 17:19:45 +0000 (14:19 -0300)]
evas/gl: Use GL_FRAMEBUFFER_BINDING, which is available on both.
Instead of a define from GL_DRAW_FRAMEBUFFER_BINDING to the above one,
just use it directly since it's available on gles and full gl.
Chris Michael [Mon, 20 May 2013 17:18:09 +0000 (18:18 +0100)]
With antognolli's framespace changes, we no longer need to subtract
framespace from the output/canvas to get proper x/y coordinates.
Signed-off-by: Chris Michael <devilhorns@comcast.net>
Mike Blumenkrantz [Mon, 20 May 2013 10:12:33 +0000 (11:12 +0100)]
fix eldbus generator memleak when failing to fetch args
Cedric Bail [Mon, 20 May 2013 09:59:43 +0000 (18:59 +0900)]
ecore: animator can get called with the same value due to rounding and range limit.
Cedric Bail [Mon, 20 May 2013 09:17:01 +0000 (18:17 +0900)]
evas: make canvas event to hot to be frozen.
Cedric Bail [Mon, 20 May 2013 09:16:39 +0000 (18:16 +0900)]
eo: some event should not be frozen.
Carsten Haitzler (Rasterman) [Mon, 20 May 2013 09:09:50 +0000 (18:09 +0900)]
remove about 16m mem for mesa for extra evgl context
only creating the extra evasgl context on demand when/if needed and
thus save 16m of memory allocations corresponding to the extra context
Chris Michael [Sun, 19 May 2013 06:13:03 +0000 (07:13 +0100)]
Fix antognolli build break when --with-opengl=es.
Basically, EGL does not define a GL_DRAW_FRAMEBUFFER_BINDING so we
test for that and if not found then define it to be
GL_FRAMEBUFFER_BINDING
Signed-off-by: Chris Michael <devilhorns@comcast.net>
ChunEon Park [Sat, 18 May 2013 18:21:03 +0000 (03:21 +0900)]
edje/vim - syntax updated
Chris Michael [Sat, 18 May 2013 16:48:13 +0000 (17:48 +0100)]
Revert this, there is no build break apparently. Give cookie back to
antognolli
Revert "include evas_gl_common.h"
This reverts commit
81cee894e13257c6cedfdddbcab3e453cc5cd01a.
Chris Michael [Sat, 18 May 2013 14:53:55 +0000 (15:53 +0100)]
include evas_gl_common.h
(Bad antognolli, No cookie !!)
NB: Fixes build break !!! If we are going to use defines from glext.h,
let's be sure to have the necessary header included Yes ?? ;)
Signed-off-by: Chris Michael <devilhorns@comcast.net>
Carsten Haitzler (Rasterman) [Sat, 18 May 2013 01:04:37 +0000 (10:04 +0900)]
disable all the waiting on gl - squeeze some more performance out of gl.
Rafael Antognolli [Fri, 17 May 2013 23:28:07 +0000 (20:28 -0300)]
evas/gl: Restore framebuffer after texture creation.
If we don't do this, the framebuffer object used on shader_array_flush()
will be wrong.
NOTE: shader_array_flush() should take care of setting the fbo
correctly.
José Roberto de Souza [Fri, 17 May 2013 21:44:24 +0000 (18:44 -0300)]
edje: Fix edje box crash
When a signal change a state of a edje box to a state that was already active
and we add/del a item of edje box a crash happen.
Example:
test.edc:
collections {
group {
name: "box_text";
parts {
part {
name: "elm.box.buttons";
type: BOX;
description {
state: "default" 0.0;
min: 0 100;
max: 9999 100;
align: 0 0;
box {
layout: "horizontal";
min: 1 1;
}
visible: 0;
}
description {
state: "show" 0.0;
inherit: "default" 0.0;
visible: 1;
}
}
}
programs {
program {
signal: "show,buttons";
source: "gui";
action: STATE_SET "show" 0.0;
transition: ACCELERATE 1;
target: "elm.box.buttons";
}
program {
signal: "hide,buttons";
source: "gui";
action: STATE_SET "default" 0.0;
transition: ACCELERATE 1;
target: "elm.box.buttons";
}
}
}
}
test.c:
//Compile with:
//gcc -g test.c -o test `pkg-config --cflags --libs elementary edje ecore`
static int count = 0;
static Eina_Bool _fill(void *data)
{
int i;
Evas_Object *layout = data;
Evas_Object *edje = elm_layout_edje_get(layout);
edje_object_part_box_remove_all(edje, "elm.box.buttons", EINA_TRUE);
elm_layout_signal_emit(layout, "show,buttons", "gui");
printf("_fill()\n");
for (i = 0; i < 5; i++, count++)
{
Evas_Object *btn;
char buf[50];
btn = elm_button_add(layout);
snprintf(buf, sizeof(buf), "button %d", count);
elm_object_text_set(btn, buf);
edje_object_part_box_append(edje, "elm.box.buttons", btn);
evas_object_show(btn);
}
return EINA_TRUE;
}
EAPI_MAIN int
elm_main(int argc, char **argv)
{
Evas_Object *win, *bg, *layout;
win = elm_win_add(NULL, "test", ELM_WIN_BASIC);
elm_win_title_set(win, "Test");
elm_policy_set(ELM_POLICY_QUIT, ELM_POLICY_QUIT_LAST_WINDOW_CLOSED);
elm_win_autodel_set(win, EINA_TRUE);
bg = elm_bg_add(win);
elm_bg_color_set(bg, 255, 255, 255);
evas_object_size_hint_weight_set(bg, EVAS_HINT_EXPAND, EVAS_HINT_EXPAND);
elm_win_resize_object_add(win, bg);
evas_object_show(bg);
layout = elm_layout_add(win);
elm_layout_file_set(layout, "test.edj", "box_text");
evas_object_size_hint_weight_set(layout, EVAS_HINT_EXPAND, EVAS_HINT_EXPAND);
evas_object_show(layout);
_fill(layout);
ecore_timer_add(5, _fill, layout);
evas_object_resize(win, 800, 320);
evas_object_show(win);
elm_run();
elm_shutdown();
return 0;
}
ELM_MAIN()
(gdb) bt
0x00000000 in ?? ()
0xb7add287 in _edje_box_layout_calculate_coords ()
from /usr/lib/libedje.so.1
0xb7add5cd in _edje_box_layout () from /usr/lib/libedje.so.1
0xb7c98ff7 in _evas_object_box_smart_calculate ()
from /usr/lib/libevas.so.1
0xb7c97a36 in evas_object_smart_calculate () from /usr/lib/libevas.so.1
0xb7ae1b12 in _edje_part_recalc_single () from /usr/lib/libedje.so.1
0xb7ae3380 in _edje_part_recalc () from /usr/lib/libedje.so.1
0xb7ae59ff in _edje_recalc_do () from /usr/lib/libedje.so.1
0xb7b3603a in _edje_smart_calculate () from /usr/lib/libedje.so.1
0xb7c97be2 in evas_call_smarts_calculate () from /usr/lib/libevas.so.1
0xb7cc7d88 in evas_render_updates_internal () from /usr/lib/libevas.so.1
0xb7ba7c97 in _ecore_evas_x_render () from /usr/lib/libecore_evas.so.1
0xb7ba27c6 in _ecore_evas_idle_enter () from /usr/lib/libecore_evas.so.1
0xb7c27212 in _ecore_idle_enterer_call () from /usr/lib/libecore.so.1
0xb7c28d3d in _ecore_main_loop_iterate_internal ()
from /usr/lib/libecore.so.1
0xb7c2941f in ecore_main_loop_begin () from /usr/lib/libecore.so.1
0xb7e8def8 in elm_run () from /usr/lib/libelementary.so.1
0xb7f6f8ee in appcore_efl_main () from /usr/lib/libappcore-efl.so.1
0xb7644bb5 in app_efl_main () from /usr/lib/libcapi-appfw-application.so.0
0x08074c5a in main (argc=1, argv=0xbffffb64) at src/main_tizen.c:155
==2036== Jump to the invalid address stated on the next line
==2036== at 0x0: ???
==2036== by 0x44C35CC: _edje_box_layout (in /usr/lib/libedje.so.1.7.99)
==2036== by 0x42E8FF6: _evas_object_box_smart_calculate (in /usr/lib/libevas.so.1.7.99)
==2036== by 0x42E7A35: evas_object_smart_calculate (in /usr/lib/libevas.so.1.7.99)
==2036== by 0x44C7B11: _edje_part_recalc_single (in /usr/lib/libedje.so.1.7.99)
==2036== by 0x44C937F: _edje_part_recalc (in /usr/lib/libedje.so.1.7.99)
==2036== by 0x44CB9FE: _edje_recalc_do (in /usr/lib/libedje.so.1.7.99)
==2036== by 0x451C039: _edje_smart_calculate (in /usr/lib/libedje.so.1.7.99)
==2036== by 0x42E7BE1: evas_call_smarts_calculate (in /usr/lib/libevas.so.1.7.99)
==2036== by 0x4317D87: evas_render_updates_internal (in /usr/lib/libevas.so.1.7.99)
==2036== by 0x4465C96: _ecore_evas_x_render (in /usr/lib/libecore_evas.so.1.7.99)
==2036== by 0x44607C5: _ecore_evas_idle_enter (in /usr/lib/libecore_evas.so.1.7.99)
==2036== Address 0x0 is not stack'd, malloc'd or (recently) free'd
We called anim->start aftwards just to be sure to stick to something if
all fail.
Daniel Willmann [Fri, 17 May 2013 17:16:09 +0000 (18:16 +0100)]
tests/ecore_animators: Don't rely on consistent timing
This may break under heavy load so we shouldn't rely on it. Instead make
sure that time only moves forward.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Willmann [Thu, 16 May 2013 17:09:14 +0000 (18:09 +0100)]
ecore_animator: Ensure correct start/end values are returned.
Make sure pos_map(0) == 0 and pos_map(1) == 1
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Willmann [Thu, 16 May 2013 17:07:56 +0000 (18:07 +0100)]
tests/ecore: Add test for ecore_animator
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Willmann [Fri, 17 May 2013 13:43:19 +0000 (14:43 +0100)]
eina_fp: Fix eina_f32p32_mul rounding issues
Previous multiplication just truncated the fraction part to 16 bits. If
we round properly we can actually increase accuracy a bit.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Willmann [Fri, 17 May 2013 15:46:23 +0000 (16:46 +0100)]
eina_test_fp: Test multiplication and division as well
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Jérémy Zurcher [Fri, 17 May 2013 09:39:01 +0000 (11:39 +0200)]
eo ptr ind: fix comments
Jérémy Zurcher [Fri, 17 May 2013 08:35:16 +0000 (10:35 +0200)]
eo ptr ind: empty tables are freed except 1 kept as spare
- this reduces unused memory usage
- the spare table avoids the free/alloc corner case
Jérémy Zurcher [Fri, 17 May 2013 07:23:00 +0000 (09:23 +0200)]
eo ptr ind: current_table -> _current_table
Jihoon Kim [Fri, 17 May 2013 00:33:22 +0000 (09:33 +0900)]
eio: fix Eio.h author information
Eduardo Lima (Etrunko) [Wed, 15 May 2013 19:28:58 +0000 (16:28 -0300)]
evas/wayland-egl: Don't crash when re->win is NULL
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
ChunEon Park [Thu, 16 May 2013 13:34:03 +0000 (22:34 +0900)]
eio - fix header doc.
Jérémy Zurcher [Thu, 16 May 2013 12:56:41 +0000 (14:56 +0200)]
eo ptr ind: store partial Eo Id in the table
- very small speed up
- maybe later use this to avoid table indexes not NULL checks
in _eo_id_release and _eo_obj_pointer_get
Chris Michael [Thu, 16 May 2013 12:22:35 +0000 (13:22 +0100)]
Add missing NULL fields.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 16 May 2013 12:20:21 +0000 (13:20 +0100)]
Revert "ecore evas wayland: if(()) are used for assignments, add missing NULL fields"
This reverts commit
ab505d5d23299e658b39fa7d0f1aa73742e3e7fb.
Jérémy Zurcher [Thu, 16 May 2013 11:15:43 +0000 (13:15 +0200)]
eo ptr ind: use of mprotect when EINA_DEBUG_MALLOC is set
- to enable this feature, compile with --with-profile=debug
- the mid tables and tables are write protected after modifications,
you will segfault if you mess with them
Jérémy Zurcher [Thu, 16 May 2013 09:50:21 +0000 (11:50 +0200)]
eo ptr ind: set generation counter range to [1;max-1]
- so we never generate an Eo Id equal to 0,
thus don't need to skip the first top table index anymore
Jérémy Zurcher [Thu, 16 May 2013 09:46:20 +0000 (11:46 +0200)]
eo ptr ind: minimize amount of wasted memory
- because of mmap PAGE_SIZE alignement and added magic header, almost a
memory page is wasted per table and mid table allocation.
- reducing the number of tables per mid table and the number of entries
per table solves this.
Jérémy Zurcher [Thu, 16 May 2013 08:55:32 +0000 (10:55 +0200)]
eo ptr ind: mostly cosmetic
- add and use SHIFT_* macros
- rename queue into fifo
- try to clarify the structure top table -> mid table -> table[entry]
Jérémy Zurcher [Wed, 15 May 2013 14:42:31 +0000 (16:42 +0200)]
ecore evas wayland: if(()) are used for assignments, add missing NULL fields
Tom Hacohen [Thu, 16 May 2013 10:04:01 +0000 (11:04 +0100)]
Edje textblock: Fixed issue with quoted formats. (Fixes T113).
Tom Hacohen [Thu, 16 May 2013 09:38:56 +0000 (10:38 +0100)]
Evas textblock: Removed double ;.
Tom Hacohen [Thu, 16 May 2013 08:56:17 +0000 (09:56 +0100)]
Evas language: Handle signed wchar_t in a nicer way.
Ryuan Choi [Wed, 15 May 2013 14:19:47 +0000 (15:19 +0100)]
cmake : find_package should find Ecore modules using underline. Accepted
Daniel Willmann [Wed, 15 May 2013 12:22:13 +0000 (13:22 +0100)]
evas: fix make distcheck
Follow font rename from DejaVuSans.eet to TestFont.eet in commit e7cd15.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Daniel Willmann [Wed, 15 May 2013 12:10:20 +0000 (13:10 +0100)]
emotion: Fix distcheck by referring to the correct file name
It's emotion_modules.h - not emotion_module.h
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Carsten Haitzler (Rasterman) [Wed, 15 May 2013 10:38:16 +0000 (19:38 +0900)]
remove rect leak in swapbuf flush in dri2 software path of evas
Stefan Schmidt [Tue, 14 May 2013 14:45:22 +0000 (15:45 +0100)]
evas: No need to finish a block with NULL statement.
Stefan Schmidt [Tue, 14 May 2013 12:54:43 +0000 (13:54 +0100)]
eo: Fix memleak in EO_DBG_INFO_APPEND macro
If List is NULL we would never make any use of the alloced Tmp. Better check
for NULL before doing the actual alloc and doing the work.
Chris Michael [Tue, 14 May 2013 10:30:37 +0000 (11:30 +0100)]
Ooops, this should return the allocated Ecore_Evas.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 09:48:40 +0000 (10:48 +0100)]
Fix 'return' with no value, in function returning non-void
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 09:46:21 +0000 (10:46 +0100)]
Add ecore_evas_drm private header.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 09:45:47 +0000 (10:45 +0100)]
Start on basic ecore_evas_drm code.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 07:16:45 +0000 (08:16 +0100)]
Add start of ecore_evas_drm code.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 07:16:21 +0000 (08:16 +0100)]
Add pc file for evas drm engine.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 14 May 2013 07:15:53 +0000 (08:15 +0100)]
Start on ecore_evas Drm code.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Jihoon Kim [Tue, 14 May 2013 07:14:38 +0000 (16:14 +0900)]
ecore_imf: fix crash when ecore_imf_context_del is called in ecore_imf_context_input_panel_callback_call
Jihoon Kim [Tue, 14 May 2013 00:33:19 +0000 (09:33 +0900)]
Edje entry: return surrounding string until the start position of selection
Tom Hacohen [Mon, 13 May 2013 14:13:25 +0000 (15:13 +0100)]
Fixed text/block tests: Added Korean UnDotum font and fixed some issues.
Tom Hacohen [Mon, 13 May 2013 12:51:14 +0000 (13:51 +0100)]
Fix broken test from previous commit.
Tom Hacohen [Mon, 13 May 2013 12:49:51 +0000 (13:49 +0100)]
Evas bidi test: Improved error output.
thiepha (Thiep Ha) [Mon, 13 May 2013 10:11:51 +0000 (19:11 +0900)]
[Edje_Entry] Add selection handlers to entry
Add selection handlers to entry.
https://phab.enlightenment.org/D117
Conflicts:
ChangeLog
NEWS
Sung W. Park [Mon, 13 May 2013 05:12:36 +0000 (14:12 +0900)]
evas: fix evas flushing its command queue to wrong gl context in gl-backends
For native surface rendering (glX, egl) in gl-backend, if pixel get callback
function is called for native rendering, there is a case where evas will
try to call evas_gl_common_context_flush() to flush its remaining commands
in its command queue while the context is set current to the native surface/
context that is called within the pixel get callback. So, I've added an
engine function that forces a flush before it enters the pixel get callback
function.
Carsten Haitzler (Rasterman) [Sun, 12 May 2013 06:50:52 +0000 (15:50 +0900)]
fix cedric's image property code... that broke load opt downscaling...
CEDRIC... it WAS YOU!... YOU BROKE ETHUMB!... i was.. RIGHT! :) well
done. your borking skills are pretty good. :) you broke load opt
downscaling for jpegs in general.. it just happened to turn up in
ethumb.
Boris 'billiob' Faure [Sun, 12 May 2013 20:22:48 +0000 (22:22 +0200)]
eeze: do no try to match an unknown property
Chris Michael [Sat, 11 May 2013 15:09:39 +0000 (16:09 +0100)]
Add ecore_x_randr_config_timestamp_get to return config timestamp that
XRandr has currently.
NB: needed for some updates/fixes to randr dialog.
Signed-off-by: Chris Michael <devilhorns@comcast.net>
Jiyoun Park [Sat, 11 May 2013 13:40:50 +0000 (22:40 +0900)]
* Evas: Fix crash if app use native surface in wrong engine.
in sw engine, visual cannot be the image data.
Tom Hacohen [Fri, 10 May 2013 16:44:30 +0000 (17:44 +0100)]
Evas font/text: Fixed an issue with asc/desc calculation in text object.
Added font_int asc/desc getters in order to achieve that.
Thanks to MinSu Seo for reporting.
Carsten Haitzler (Rasterman) [Fri, 10 May 2013 07:59:33 +0000 (16:59 +0900)]
fix cedric build break
Cedric Bail [Fri, 10 May 2013 05:28:48 +0000 (14:28 +0900)]
emotion: oops, forgot to update the Makefile.
Cedric Bail [Fri, 10 May 2013 02:39:55 +0000 (11:39 +0900)]
edje: let's not reopen the file for nothing.
Carsten Haitzler (Rasterman) [Fri, 10 May 2013 04:35:02 +0000 (13:35 +0900)]
free strings in dri/drm swapper so they don't leak.
Cedric Bail [Thu, 9 May 2013 11:54:57 +0000 (13:54 +0200)]
emotion: rename Emotion_Module.h to emotion_modules.h
Raoul Hecky [Fri, 10 May 2013 01:43:33 +0000 (10:43 +0900)]
ecore_evas: backport svn r83875.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
Carsten Haitzler (Rasterman) [Fri, 10 May 2013 00:10:58 +0000 (09:10 +0900)]
note the pa abort and disble del of out. leak it for now.
Carsten Haitzler (Rasterman) [Fri, 10 May 2013 00:00:02 +0000 (09:00 +0900)]
eet_mmap uses reader/writers caches for lookup.. but never ADDs to
these caches! bad bad bad!
Carsten Haitzler (Rasterman) [Thu, 9 May 2013 23:54:50 +0000 (08:54 +0900)]
do not make eet files get refcounts of 2 on eet_mmap so they then never free.
we set references to 1 then on success references++... thus ALWAYS
being on 2.
Carsten Haitzler (Rasterman) [Thu, 9 May 2013 23:24:32 +0000 (08:24 +0900)]
slightly refactor eo usage in multishesne to ... make use of eo's
ability to call setup methods.. on setup... all at once.
Carsten Haitzler (Rasterman) [Thu, 9 May 2013 23:04:14 +0000 (08:04 +0900)]
multisense segv (free data that isnt malloced) fix...
ChunEon Park [Thu, 9 May 2013 16:51:42 +0000 (01:51 +0900)]
evas/gl - and don't try blend as possible.