Chris Michael [Wed, 7 Aug 2013 06:48:06 +0000 (07:48 +0100)]
Require up-to-date wayland-client library for ecore_imf wayland module
support.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 7 Aug 2013 06:46:43 +0000 (07:46 +0100)]
Require up-to-date versions of the wayland libraries for evas engine
support.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 7 Aug 2013 06:41:28 +0000 (07:41 +0100)]
Add wayland library version checks to configure.ac for EFL 1.8
NB: As of today, EFL Head will require an up-to-date version of the
wayland libraries.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 6 Aug 2013 10:26:26 +0000 (11:26 +0100)]
Add missing EINA_UNUSED to function parameters.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 6 Aug 2013 10:18:29 +0000 (11:18 +0100)]
Add functions for output done and output scale events.
NB: Apparently, setting these functions to NULL in the listener Does
Not Work anymore ! :( (used to work just fine).
NB: This makes EFL Apps work again inside Weston with latest from git.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 6 Aug 2013 10:18:11 +0000 (11:18 +0100)]
Bump compositor interface version to 3.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Tue, 6 Aug 2013 08:47:08 +0000 (09:47 +0100)]
Get the 'fd' from the handler First (regardless of read or write) so
that it can be used in the below event callback.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Tom Hacohen [Tue, 6 Aug 2013 08:28:38 +0000 (09:28 +0100)]
Evas textblock: Use max ascent/descent at the edges of the textblock.
Thanks to Youngbok Shin for reporting the difference in behaviour between
textblock and text object.
Carsten Haitzler (Rasterman) [Tue, 6 Aug 2013 07:13:03 +0000 (16:13 +0900)]
eina-file - adapt to bsd version of MADV_REMOVE
Cedric Bail [Tue, 6 Aug 2013 06:45:07 +0000 (15:45 +0900)]
evas: fix build with debugging rendering trace.
Cedric Bail [Tue, 6 Aug 2013 06:44:43 +0000 (15:44 +0900)]
evas: fix comment to actually compile if you use them.
Cedric Bail [Tue, 6 Aug 2013 06:44:22 +0000 (15:44 +0900)]
eina: fix build on FreeBSD.
Cedric Bail [Tue, 6 Aug 2013 00:17:46 +0000 (09:17 +0900)]
efreet: let's pass memset parameter in the right order.
Chris Michael [Mon, 5 Aug 2013 14:32:12 +0000 (15:32 +0100)]
Fix memleak in data_thread_script of edje_cc_out
NB: If we are going to exit data_thread_script function due to malloc
failure or fread failure, then we should close the previously fdopen'd
file so that we do not leak 'f'.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 13:24:46 +0000 (14:24 +0100)]
If efreet_menu_parse returns an actual menu, let's free the memory
allocated to it before we return.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 13:07:11 +0000 (14:07 +0100)]
Remove the directory from the font_first hash First, before calling
object_text_font_cache_dir_del as the latter function will actually
Free the fd.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 13:03:13 +0000 (14:03 +0100)]
Do not try to append a NULL async client.
NB: Above, _ethumb_client_async_free Could have been called, which
will free the async client. Add valid check for async before we append.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:58:13 +0000 (13:58 +0100)]
efreet_menu_move_free will free "move" here so remove the menu from
the parent's list Before we free the menu.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:52:06 +0000 (13:52 +0100)]
_ecore_con_dns_free already frees the timer and whole dns structure,
so setting dns->timer = NULL here b0rks.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:48:23 +0000 (13:48 +0100)]
Fix potential double-free of so->answer.
NB: dns_so_submit calls dns_so_reset twice (once at start, once at end
on error). As so_reset will free so->answer the first time through,
add validity check of so->answer for the second iteration.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:16:44 +0000 (13:16 +0100)]
Init fd to zero (removes compiler warning).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:13:19 +0000 (13:13 +0100)]
Remove Eet_Connection from hash Before we call eet_connection_close
(as that frees the connection).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 12:01:52 +0000 (13:01 +0100)]
Fix eina_file_virtual_map_new function (/me glares at Frenchie).
- Spank Cedric !!!!!
NB: How about we actually fill in "map" after allocation ??
NB: Previously we would malloc "map" and immediately exit without
filling it in, without adding it to the hash....nothing. Just allocate
and get out. Bad Frenchie !!!
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 11:42:31 +0000 (12:42 +0100)]
Avoid potential divide by zero in INF call.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 11:39:19 +0000 (12:39 +0100)]
Do not call _ecore_con_client_flush with a NULL client.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 11:32:00 +0000 (12:32 +0100)]
Check for Image_Entry being valid before calling evas_cache_image_drop.
NB: Image_Entry passed in to this function Could be NULL (hence the
previous if ie check above. Do not call evas_cache_image_drop on a
NULL Image_Entry as evas_cache_image_drop does not validity checking.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 11:17:26 +0000 (12:17 +0100)]
Actually declare fd at the top of function.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 10:46:04 +0000 (11:46 +0100)]
ecore_main_fd_handler_fd_get can return -1 so we need to check for a
valid fd before calling read().
Reuse returned fd variable in the callback to save extra function call
to ecore_main_fd_handler_fd_get (since we already acquired the fd
above).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 10:40:02 +0000 (11:40 +0100)]
Check for valid fd returned from ecore_main_fd_handler_fd_get before
potentially passing -1 to the read() function.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 10:36:24 +0000 (11:36 +0100)]
Don't pass negative values to close:
ecore_main_fd_handler_fd_get can return -1, so check return value
before passing to close().
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 10:34:34 +0000 (11:34 +0100)]
open() can fail for various reasons, so check for valid 'fd' return
before calling fdopen.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 10:24:11 +0000 (11:24 +0100)]
ecore_main_fd_handler_fd_get can return -1 so check fd validity before
calling read() with potentially -1.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 09:56:37 +0000 (10:56 +0100)]
ecore_main_fd_handler_fd_get Can return -1 so check fd validity before
calling close() on it.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 09:41:16 +0000 (10:41 +0100)]
fopen can fail. check file is open before trying to write to it.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 09:37:15 +0000 (10:37 +0100)]
strchr can return NULL, so check validity before tryint to pass NULL
to ecore_file_cp.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 09:26:38 +0000 (10:26 +0100)]
_edje_real_part_recursive_get can return NULL so check validity before
calling _edje_child_remove (with potentially NULL as the real part).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Mon, 5 Aug 2013 09:13:57 +0000 (10:13 +0100)]
Don't call fcntl functions if we do not HAVE_FCNTL support built in.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Cedric Bail [Mon, 5 Aug 2013 08:43:17 +0000 (17:43 +0900)]
evas: limit resize to always be of the right size.
This should fix T157.
Cedric Bail [Mon, 5 Aug 2013 08:41:09 +0000 (17:41 +0900)]
edje: do not move object until all positionning information is known.
Sebastian Dransfeld [Mon, 5 Aug 2013 08:43:25 +0000 (10:43 +0200)]
efreet: Fix allocated data size
Since we check for count < 10 and then add 1, we could get count == 11.
Also set array memory to 0 in one command.
Sebastian Dransfeld [Tue, 25 Jun 2013 02:42:45 +0000 (04:42 +0200)]
efreet: Formatting
Chris Michael [Mon, 5 Aug 2013 06:44:50 +0000 (07:44 +0100)]
Patch from Phoenix Kayo <kayo@illumium.org> to fix
ecore_x_selection_parser functions for mozurl.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Yury Usischev [Fri, 2 Aug 2013 09:06:55 +0000 (18:06 +0900)]
Add neon for upscaling and map routines in evas.
Daniel Juyung Seo [Fri, 2 Aug 2013 04:27:37 +0000 (13:27 +0900)]
edje: added BLOCK_HANDLE description and vim syntax.
Cedric Bail [Fri, 2 Aug 2013 01:25:26 +0000 (10:25 +0900)]
evas: change mapping policy to be less agressive into loading file in memory.
Should fix T271.
Chris Michael [Thu, 1 Aug 2013 14:18:01 +0000 (15:18 +0100)]
Add #ifdef ECORE_XRANDR around the edid functions.
NB: Cannot really happen for 1.8 (as --disable-ecore-x-randr has no
effect), but really these should have been ifdef'd from the start.
NB: IMO, Don't think a ChangeLog is really needed for this....
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 1 Aug 2013 11:48:17 +0000 (12:48 +0100)]
Remove calls to ecore_wl_init and ecore_wl_shutdown.
NB: These calls are NOT needed here (they setup and tear down the
actual wayland display fd connections, and are not necessary in order
to use ecore_wl functions).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 1 Aug 2013 11:45:14 +0000 (12:45 +0100)]
Check for valid windows hash before searching it.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Mike Blumenkrantz [Thu, 1 Aug 2013 11:17:58 +0000 (12:17 +0100)]
ecore-evas-x no longer tries making x calls on deleted windows when trying to free its sync counter
Cedric Bail [Thu, 1 Aug 2013 10:39:58 +0000 (19:39 +0900)]
evas: let's not leak temporary list left over.
Cedric Bail [Thu, 1 Aug 2013 10:35:21 +0000 (19:35 +0900)]
edje: fix memory leak introduced by map_colors support.
Cedric Bail [Thu, 1 Aug 2013 02:40:13 +0000 (11:40 +0900)]
edje: another small speedup for elemines startup time (-2%).
Cedric Bail [Thu, 1 Aug 2013 02:22:52 +0000 (11:22 +0900)]
edje: let's not stupiddly create list all over the place when we can avoid it.
Cedric Bail [Thu, 1 Aug 2013 02:22:29 +0000 (11:22 +0900)]
edje: fix a git merge issue that did slip in.
Cedric Bail [Thu, 1 Aug 2013 01:51:52 +0000 (10:51 +0900)]
edje: improve loading time of Edje_Object with Evas_Object_Textblock and multiple styles.
This bring you a 10% speedup to the first frame of Elementary_Test.
Cedric Bail [Wed, 31 Jul 2013 09:42:13 +0000 (18:42 +0900)]
emotion: fix backend order.
Tom Hacohen [Wed, 31 Jul 2013 13:31:23 +0000 (14:31 +0100)]
Evas Text object: Improved the docs explaining font selection.
Cedric Bail [Wed, 31 Jul 2013 08:16:20 +0000 (17:16 +0900)]
eina: add eina_file_refresh().
Carsten Haitzler (Rasterman) [Wed, 31 Jul 2013 06:13:40 +0000 (15:13 +0900)]
edje - Fix efl to use key not keyname (so xmodmap works).
Cedric Bail [Wed, 31 Jul 2013 02:53:39 +0000 (11:53 +0900)]
evas: use eina_file_virtualize() in evas_object_memfile_set().
NOTE: I am tempted to mark evas_object_memfile_set() as deprecated due
to the forced memcpy it introduced and the fact that it is now a 3 lines
of code to do the same in a more efficient way.
Cedric Bail [Wed, 31 Jul 2013 02:51:51 +0000 (11:51 +0900)]
eina: add support for in memory only Eina_File.
Guillaume Friloux [Tue, 30 Jul 2013 13:39:12 +0000 (15:39 +0200)]
Job--;
Sung W. Park [Tue, 30 Jul 2013 10:45:40 +0000 (19:45 +0900)]
Evas: Fix Evas_GL engine shutdown error.
It was trying to use access a struct for egl_disp after the struct
was freed..
Rajesh P S [Tue, 30 Jul 2013 05:52:06 +0000 (14:52 +0900)]
evas: unset the right backend when unregistering module.
Adrian Negreanu [Mon, 29 Jul 2013 12:20:49 +0000 (13:20 +0100)]
configure: fix build when disabling ecore_con and crypto.
configure command:
./configure --disable-ecore-con --disable-curl --with-glib=no --with-crypto=none
EFL_CRYPTO_DEPEND macro appends the crypto libraries to the ECORE_CON dependency list.
In case the ECORE_CON dependency list and the build_crypto are both empty
the resulting list will only contain " ", resulting in an error
configure: error: Package requirements ( ) were not met:
Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
Cedric Bail [Mon, 29 Jul 2013 09:54:14 +0000 (11:54 +0200)]
evil: fix access value to shared memory.
Cedric Bail [Mon, 29 Jul 2013 07:28:07 +0000 (16:28 +0900)]
eina: seems like all the compiler we support do support __typeof__.
Ryuan Choi [Sun, 28 Jul 2013 11:16:51 +0000 (20:16 +0900)]
Moved brace of extern "C" to fix build break in c++ applications
ChunEon Park [Sun, 28 Jul 2013 11:24:46 +0000 (20:24 +0900)]
evas/proxy - render the children of the proxy's source object when ther visibility is true.
This is not the perfect solution at this moment. This doesn't consider the cached clipper's visibility at all.
But at least this would be better than exist works. Should be improved.
Chris Michael [Fri, 26 Jul 2013 06:06:58 +0000 (07:06 +0100)]
Set Crtc info->outputs and info->possible to NULL initially.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Ryuan Choi [Fri, 26 Jul 2013 05:00:16 +0000 (14:00 +0900)]
eina: use __typeof__ for public headers to build applications with c++0x.
José Roberto de Souza [Thu, 25 Jul 2013 17:22:10 +0000 (14:22 -0300)]
eldbus: Check if property_get_all() returned a error before read properties
Chris Michael [Thu, 25 Jul 2013 14:48:40 +0000 (15:48 +0100)]
During a call to ecore_evas_show (using wayland_shm), we can avoid
erroneous calls to engine_setup code if we check surface validity.
NB: Short Version: Don't reassign surface if we don't need to.
NB: Prior to this, during a call to ecore_evas_show we would always
Reset the evas engine data (which was basically triggering code to
tear down and reconstruct all the engine internals including the
buffers and the swapper). By checking surface validity first, we can
potentially avoid all that teardown and reconstruction Thus resulting
in faster "show" speed.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 14:45:55 +0000 (15:45 +0100)]
Check that we have both Rects and Count is > 0 during buffer attach.
NB: When trying to attach a buffer to a surface, there was a corner
case that could fail in that if we just had 'rects' but count was 0,
then the surface would not get damaged (essentially a call to
wl_surface_damage would get passed 0,0 as the size). This fixes that
problem in that if count is <= 0 now, then we'll damage the whole
surface.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 12:32:06 +0000 (13:32 +0100)]
Fix typo
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 12:09:34 +0000 (13:09 +0100)]
Guard against allocating 0 outputs and 0 possible outputs.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 12:08:18 +0000 (13:08 +0100)]
Guard against trying to allocate 0 outputs.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 11:32:49 +0000 (12:32 +0100)]
Trap for number of crtcs < 1 so we do not try to allocate space for
zero crtcs.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Thu, 25 Jul 2013 08:12:17 +0000 (09:12 +0100)]
When we destroy the shm_pool, reset the swapper used_size to zero.
Remove FIXME and resize the the shm_pool if it is too small for the
new buffer.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
ChunEon Park [Thu, 25 Jul 2013 07:05:56 +0000 (16:05 +0900)]
evas - found a inccorect check while reading the code.
ChunEon Park [Thu, 25 Jul 2013 07:01:49 +0000 (16:01 +0900)]
evas - skip the map rendering if all points are transparent.
ChunEon Park [Thu, 25 Jul 2013 06:57:53 +0000 (15:57 +0900)]
efl - correct Changelog date.
Jaehyun Cho [Thu, 25 Jul 2013 04:00:37 +0000 (13:00 +0900)]
edje: use eval coordinate to show TEXT part correctly.
Mike Blumenkrantz [Wed, 24 Jul 2013 12:58:57 +0000 (13:58 +0100)]
ecore-evas-x no longer uses X windows after receiving a destroy event
Carsten Haitzler (Rasterman) [Wed, 24 Jul 2013 10:22:31 +0000 (19:22 +0900)]
ecore-x: fix randr query when noutputs is 0.
Chris Michael [Wed, 24 Jul 2013 08:15:43 +0000 (09:15 +0100)]
Check for valid text_input object before trying to call wayland
functions on it.
NB: Wayland (internal) functions typically do not do Any object
validity checking, so lets be proactive and do it here first.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 24 Jul 2013 08:10:42 +0000 (09:10 +0100)]
Fix indentation of some if statements
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 24 Jul 2013 08:08:50 +0000 (09:08 +0100)]
Fix formatting so I can actually read this code in 80 columns ;)
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Daniel Willmann [Wed, 24 Jul 2013 06:55:08 +0000 (08:55 +0200)]
ecore example: Disable ecore_audio examples while Eo is in flux
There is no legacy API we could use here, so just disable building for
now.
Daniel Willmann [Wed, 24 Jul 2013 06:52:44 +0000 (08:52 +0200)]
ecore examples: Use old API as long as Eo is in flux
This avoids build errors while Eo/Eo2 is being moved around.
Cedric Bail [Wed, 24 Jul 2013 06:49:39 +0000 (15:49 +0900)]
evas: let's return the correct file name.
Cedric Bail [Wed, 24 Jul 2013 06:46:20 +0000 (15:46 +0900)]
edje: fix rounding fill issue to take the information after adjusting it not before.
Chris Michael [Wed, 24 Jul 2013 06:23:58 +0000 (07:23 +0100)]
Check for valid returns of ecore_wl_registry_get and
ecore_wl_globals_get before trying to use them.
After we have bound a text input manager, there is no need to keep
looping the globals, so break out.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 24 Jul 2013 06:15:44 +0000 (07:15 +0100)]
Add some checking/trapping for valid wayland display.
Remove dead commented lines
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 24 Jul 2013 06:11:32 +0000 (07:11 +0100)]
Fixup formatting.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Chris Michael [Wed, 24 Jul 2013 06:06:49 +0000 (07:06 +0100)]
Add wayland to ecore_imf "in tree" modules
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Jérémy Zurcher [Tue, 23 Jul 2013 19:41:37 +0000 (19:41 +0000)]
remove double def of Eo_Callback_Priority
Carsten Haitzler (Rasterman) [Tue, 23 Jul 2013 12:19:13 +0000 (21:19 +0900)]
reformat some of changelog to be consistent/neater
Yury Usischev [Tue, 23 Jul 2013 12:15:19 +0000 (21:15 +0900)]
evas - fix long standing c "fallback" code "simd" (2 register) multiply
Daniel Juyung Seo [Tue, 23 Jul 2013 09:59:41 +0000 (18:59 +0900)]
edc.vim: updated edc syntax files.