Switch from absolut microseconds since epoch to a monotonic clock with
realtive values. Switch from unsigned long long to double.
This aligns it with how we present time in efl. ecore_time_get is used
when possible. For the tizen modules we convert the the epoch timestamp
we get into a double. This is still a working monotone clock source.
As this will be released the first time with 1.8 we don't have any API
break here.
Eeze sensor and Tizen types can't be in sync as the order have been changed
several times in Tizen already. We need to keep the order stable for ABI
compliance though after 1.8.
While we did the translation in one direction we did not in the other. Fix
this bug that shows now.
The real canvas size contains enough space for rendering the frame, but
ecore_evas_geometry_get() should be consistent accross all the engines
and report only the client area.
Maybe we should switch to use -Wno-unused-parameter in our default cflags?
I mean unused parameters are really common and I do not see their harm.
Using -Wno-unused-parameter would let us remove all the cluttered EINA_UNUSED
and __UNUSED__ and also be safe when a paramter gets is use later as the
compiler could optimize it out when tagged with unused. Still we would avoid
the warnings to find the important ones.
The main eeze sensor log domain was EAPI as it was used in the modules as well.
(Gives some nice runtime error when loading the shared object). Switch to different
domains for the module and make the main domain private.
Spotted by zmike.
Jenkins as well as klocwork complained here as data is already used for
the parameter of the function and we shadow it here locally. Does not look
like it would really harm us here but better avoid it.
The event callback will take care of freeing the object.
This issue was found thanks to clangs -fsanitize=address option - coming
to a CI job near you.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
When creating windows directly with Ecore_Evas, instead of using
Elementary, a default smart object is provided for the frame (client
side window decorations).
Now this smart object inherits from the clipped smart object, which
means that all its member objects will be clipped to its main clipper,
being skipped when the framespace clipper checks for objects without a
clipper.
Fix a ticket, which I don't remember the number because trac doesn't
work.
At least for wayland, eglGetProcAddress seems to fail to get the
eglQueryString symbol. Not sure if it's the right fix, but brings
wayland_egl back to working again.
Fixes#2301.
It is a very tricky things to get header order right on windows. Having that
order only in .c files simplify the work a lot. So let's try to do it with
Ecore_Evas after it rewrite and split into modules.
Using the wb->buffer pointer to compare if a buffer was already sent is
bad, since the buffer pointer can be changed in a way that that
comparison is true, even if different buffers are used.
When requesting a new buffer to the wl_shm_pool, it can return the
buffer address of a previously freed buffer. This reused buffer pointer
can be the same as the last buffer sent to the compositor, and the
comparison will be true, skipping a new wl_buffer_attach.
Resource contexts/surfaces are used for creating resources within Evas_GL.
In oder to handle Evas_GL runnig from different thread than the main one,
a resource context/surface pool was used. This turned out to be unnecssary
as they are not used very frequently. So, I got rid of the pool and
made the resources create as needed.
This will update the edges info inside the Evas_Engine_Info, and make it
correctly resize to the top/left if needed.
Also call ecore_wl_window_buffer_attach() on the egl engine, since it's
needed after the ecore_wl_window_update_size().
This will force the window to correctly update its size when the event
is received by the compositor, rather than just after an
ecore_evas_resize().
It fixes the window resizing of non-elementary applications, since
the elementary window already deals with such resize by calling
ecore_evas_resize.
This can be used to reconfigure a swapper to another size, without the
need to destroy the swapper itself.
Although the shm pool is not being reused even when reconfiguring to a
smaller size, it could easily be.
This change is done right now only to keep the dx and dy offsets of a
previously requested swapper, which were not still used.
When doing double/triple buffering, and we go to merge the rectangles,
if we are triple buffering then we should not use the double-buffer
rectangles as a valid check for triple-buffer rectangles.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
When using swapping (double/triple), and we go to merge rectangles,
then we should check for a valid triple buffer (not double) before
trying to merge the 3rd buffer rectangles.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
downscaling. This makes quality much better and "at best"
equates to a 16 point sample (2x2 linear interpolation samples,
where a linear interpolation sample equates to a 2x2 sample).
This will have perfomance impact, but the quality is worth it and
makes it closer to software downscaling in quality. It supports
2x2, 2x1 and 1x2 oversampling. YUV not done, nor image mask
(font shaders not needed).
Using the server_allocation/allocation to determine the resize offset
was not completely precise, and causing the window to not always resize
correctly.
Additionally, calling evas_engine_info_set() on every resize step caused
the window content to blink and resize very slow, because the swap
buffer, swapper, and everything were being destroyed and recreated. Now
only the swapbuf_reconfigure is being called during the resize, which is
way faster.
The _pixel_alpha_get() function used in evas_object_image_is_inside won't
work with engines other than software - since it relies on engine data
being *always* RGBA_Image * - which is wrong for OpenGL backend that uses
Evas_GL_Image * for "engine_data" pointer.
On system like 64bits windows it is not a guaranty that a
sizeof (long) == sizeof (void*), we are better using the right type.
This generate a warning that we could have seen with just cross compilation.
I add new example related with this. (ecore_evas_extn_socket & plug example)
ecore extn use this infrasturcture, server app and client app can communicate each other
later, this can be used to contorl access message
SVN revision: 83942
Again they are already used but nobody did bother to declare or define them.
I have choosen the name with what came into my head first and version seem
to be 1 in other engines, too. Evas engine gods please review.
SVN revision: 83826
The suported atom gets allocated in the ecore_x call but we leave
the scope here without freeing the resource again. Free it when
going out of scope.
SVN revision: 83790
Previously whenever evas_software_xlib_outbuf_new_region_for_update was
called for images that were rotated (!= 0) we created a new
evas_cache_image. This resulted in (quite severe) memory spikes whenever
an image was rotated.
Now we try to get the original image first and only if that fails allocate
a new one.
TDevilhorns is already working on the port to the xcb backend.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
Signed-off-by: Stefan Schmidt <s.schmidt@samsung.com>
SVN revision: 83789
This will allow it to report to Ecore_Evas that the window has changed
its state. Elementary uses this to update its maximized/fullscreen/other
window states internal information.
The code that uses this callback is also added to Ecore_Evas.
SVN revision: 83625
It's needed to set the edge where the middle click is being done in
order to allow Evas know which direction the resize should take.
SVN revision: 83610
Comment out idle_flush (for now) as it is causing some segfaults with
elm_win_util_standard_add for some strange reason.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 83436
This avoids blocking in eglSwapBuffers and has the side effect of
avoiding doing unnecessary work - painting where a frame won't be
presented.
We do this by using the event that the wayland compositor will send us
to tell us that the frame has been presented. Due to the fact that
evas_render_updates() could do no work and not cause a eglSwapBuffers we
must always have a frame callback listener setup.
Original patch by: Rob Bradford <rob@linux.intel.com>
(I just adjusted the patch to the single efl tree)
SVN revision: 83387
Remove dead commented out code
Do not call wl_surface_attach if the buffer is the same as the one
already attached.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 83293
If we do not have a private palette here, we free the buffer and
return NULL, However we never free'd the xob that was allocated
previously.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 83163
- If the output buffer has a 'onebuf' and we are going to return
that, then we do not need to allocate an Outbuf_Region.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 83161
Always returning -1 for the frame duration does not have the desired
effect. Fix the logic checking whether eina_file_open has succeeded.
Animated GIFs play way too fast without this.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
SVN revision: 83100
for direct rendering optimization. For direct rendering in EvasGL,
it falls back to FBO rendering if the conditions are not met. Before,
the fallback resources were created in the beginning but now they are
created and destroyed on need base.
SVN revision: 83057
Before this commit, when using any of the wayland backends, if the user
requested an ecore_evas window of width x height, the reported size of
the window would include the frames width and height.
Now the requested size is stored on the ecore_evas and reported when
ecore_evas_geometry_get() is used, although the ecore_evas is still
created with additional space for the frame area.
This is one of the reasons causing bug #1673.
SVN revision: 83042
This new engine function will only be used in software generic for
now - since it's the only engine used with the async render.
This function has been introduced in order to avoid growing thread
command queue too much to draw a text_props at a time on render calls
from textgrid objects.
Patch by: Paulo Alcantara <pcacjr@profusion.mobi>
SVN revision: 82832
Instead of -I$(top_srcdir)... -I$(top_builddir)... and then do it for
the .la, use the EFL_ macros to generate the contents to be used in
automake files.
There is a nasty bit that libtool will parse Makefile*.am and will not
get _DEPENDENCIES from _LIBADD and _LDADD if these are in
@REPLACEMENT@. To solve this we must explicitly set _DEPENDENCIES. The
contents of this is almost the same as _LIBADD or _LDADD with the
"_INTERNAL_" replacement name.
I hope the code will be result will be shorter and consistent as there
is less places to change when we add/remove dependencies.
Statistics are quite impressive (diffstat):
{{{
37 files changed, 663 insertions(+), 1599 deletions(-)
}}}
SVN revision: 82785
fixed to be byte array instead of void pointer, did this based on the
code of convert... but what I found very strange is that math is "*
32"... really? Shouldn't it be "* 4"?
SVN revision: 82733
This one was a painful bitch. The edbus2 port was quite broken, mainly
leaking eina_stringshare and also not adding the '\0' to the strings
that are represented as bytearray (paths cannot be utf8 to avoid
translations).
Emotion plugin was also quite bogus and the video thumbnail as edje
(animated) is not working yet due bug in Edje_Edit api -- someone
needs to investigate this, seems strange.
Emotion plugin also had a bug that it was deleting the object from
inside object callback.
Now it seems to work. Please report if it does not.
SVN revision: 82675
This patch should make us get a reference on images, maps and glyphs
which are sent in a command to the render thread. Before we were doing
some useless ref and unref operations.
SVN revision: 82666
This is intended to preserve old behavior now that we have
evas_common_font_draw_cb() to handle both sync and async callbacks.
However, we need to check where why we end up with no glyphs in a
text_props even after calling evas_common_font_draw_prepare().
SVN revision: 82664
buffer is lightweight and dependency for many engines, merge it back
into core.
extn is a module on its own, and it's the only one linking to
ecore_ipc, no need to add that to ecore_evas.
minor cosmetic changes to configure to make output consistent.
SVN revision: 82648
Change the module infra and allow one to register using the
description structure itself, saves some work.
Replace module_open and module_close with add() and del(), that are
more descriptive.
Move stuff to init/shutdown where it was a better fit.
Since add() now receives the first parameter being the owning
structure we can create one structure per generic player and they will
merge transparently with other modules, then we can use a simple
algorithm that considers the plugin priority. ALL DYNAMIC!
Future work could be done to load the plugin priority specified by
user from configuration file. But this is not in my plans right now.
SVN revision: 82602
no - not planning a backport of this at this stage. small thing. maybe
backport (and chlog/news) if this doesnt cause any new issues?
SVN revision: 82549
we were using "" for locale that made eina_prefix think that we did
not had a common prefix.
take the time and add a magic file "checkme" to be checked at runtime
to validate stuff.
also made the binaries be searched in MODULE_ARCH namespace.
SVN revision: 82507
this one was quite a huge work, but hopefully it's correct.
NOTES:
* removed vlc generic module, it should go into a separate package.
* gstreamer is enabled by default (see --disable-gstreamer)
* xine is disabled by default (see --enable-gstreamer)
* generic is always built statically if supported
* gstreamer and xine can't be configured as static (just lacks command line options, build system supports it)
* v4l2 is enabled by default on linux if eeze is built (see --disable-v4l2)
* emotion_test moved to src/tests/emotion and depends on EFL_ENABLE_TESTS (--with-tests), but is still installed if enabled.
TODO (need your help!):
* fix warnings with gstreamer and xine engine
* call engine shutdown functions if building as static
* remove direct usage of PACKAGE_*_DIR and use eina_prefix
* add eina_prefix checkme file as evas and others
* add support for $EFL_RUN_IN_TREE
* create separate package for emotion_generic_modules
* check docs hierarchy (doxygen is segv'in here)
SVN revision: 82501
We must check whether source image has been loaded or not before
enqueueing draw commands. Otherwise, cserve2 may not have a specific
loader and then it does not load the source image and the enqueued draw
commands dereference the NULL pointer "data" from source image.
Patch by: Paulo Alcantara <pcacjr@profusion.mobi>
SVN revision: 82483
evas_gl_shader.c file and made an internal generic caching api in
evas_gl_common.h for use in other places ie. evas_gl.
Then implemented evas_gl surface cap. caching code in gl backend to
accelerate the engine creation.
SVN revision: 82321
it is "const char * const *", not "const char **", and it was triggering a warning in our code.
it's just constness and will not trigger an error in our user's code, just an warning that he should fix.
SVN revision: 82278
evas_common_font_draw() should be called with glyphs array for now,
instead of intl_props.
This patch should fix ticket #2149.
Signed-off-by: Paulo Alcantara <pcacjr@profusion.mobi>
Patch by: Paulo Alcantara <pcacjr@profusion.mobi>
SVN revision: 82229
Since we are already inside wayland_egl code, force this env var to
avoid the EGLNativeDisplayType autodetection (done by EGL code), which
seems to be incorrect.
This commit should fix ticket #1972.
Note: revert this as soon as EGL autodetection gets fixed.
SVN revision: 82183
this is still in progress, mostly the multisense stuff is pending.
it seems that when we merge ecore_audio in edje the libremix and
similar are gone, at least from Edje, and will be in ecore_audio
itself (or pulseaudio).
Changes:
* __UNUSED__ to EINA_UNUSED
* binaries (epp, embryo_cc, edje_cc) now consider EFL_RUN_IN_TREE and
will assume the binaries are still not installed, running from
build tree location (needs more testing, maybe doesn't work with
srcdir != builddir, still doesn't solve cross compile builds)
SVN revision: 82139
Changes also in this commit:
* fix missing EAPI in symbols used by modules
* removed old libudev and libmount support as agreed by discomfitor/zmike
* replaced __UNUSED__ with EINA_UNUSED
* fixed docs hierarchy
SVN revision: 82100
- EVAS_CALLBACK_FREE and EVAS_CALLBACK_DEL were doing the same thing
at different stages, causing a segv due double free.
- extn->file.updates and its Ipc_Data_Update were leaking.
PS: I can't backport this to 1.7, but the problem is still
there. Could someone look into those?
SVN revision: 81304
This is in preparation for threaded render landing: the render thread will
hold a reference to a text object's glyphs while it hasn't been rendered
yet (and will drop that reference after drawing). This changes the internal
API a little bit (evas_common_font_rgba_draw() now takes an Evas_Glyph_Array
instead of an Evas_Text_Props).
SVN revision: 81183
line position is slightly different between gl drivers.
I have no idea why it is. So added to work differently based on the manufacturers.
This work may be based on the renderer. If you can test it with much drivers then please test and fix.
Also changed the ENV name from EVAS_GL_LINE_NO_OFFSET_HACK to EVAS_GL_LINE_OFFSET_HACK_DISABLE.
SVN revision: 81016
I tested gl line drawing on a few devices and found the x line start position was 1.
On the other hand, our evas draws the line on start position 0.
So it needs to shift by 1 pixel if evas is working on gl backcned.
SVN revision: 80734
Implementing support for loadables modules. It makes the engines been
loaded when they are needed. It not breakes the api, so each engine
still has its own api.
The implementation basically is:
* Functions that creates Ecore_Evas, for example
ecore_evas_software_x11_new, request to load its module and then get
the module's function to create the Ecore_Evas.
* The other functions such as \(.*\)_window_get from the Ecore_Evas
its interface and then call the appropriate method.
* As there is no unified interface to communicate with the engines
(not break api problem), all interfaces were declared in
ecore_evas_private.h
* Now the data necessary for each module is not declared in the
Ecore_Evas_Engine structure, instead of this, the struct has a void
pointer that is used by the modules.
* In this first moment engines as software_x11 and gl_x11 were put
together in the same module, but obviously exporting all the things
necessary.
SVN revision: 80280
After agreement in the mail list, core developers agree to remove this
engine that was not being supported for a long time.
Given that most operations Evas uses are not accelerated in DirectFB,
or at least hardware that exclusively supports DirectFB, it's better
for those people to just use Evas/Ecore software (buffer) rendering
and expose DirectFB's framebuffer as destination surface.
SVN revision: 80232
direct rendering and then another image object using Native
Surface rendering, there was a potential for it to fall into
the same direct rendering path.
Also, fixed some minor Evas GL extension bugs that came from refactoring.
SVN revision: 79532
EGL 1.4 spec Section 3.5.1: If there is already an
EGLSurface associated with win (as a result of a previous
eglCreateWindowSurface call), then an EGL_BAD_ALLOC error is
generated.
So that this eglCreateWindowSurface() will fail if the egl driver is
a strict conformance to the spec.
SVN revision: 79505
weren't set properly where if a program uses evas_gl to do GL rendering
in direct rendering mode and then use a pixmap to do native
GL rendering in the same program, native pixmap rendering would
also fall into the direct rendering path and not render anything for
image object. It's been fixed.
SVN revision: 79493
x1, x2 shadow something in the math library.
Would probably be better to turn off -Wshadow, but for some
reason people think this there's some value in it...
Signed-off-by: Mike McCormack <mikem@atratus.org>
SVN revision: 79446
Unfortunately setting unused variable to zero still produces
a warning about variables being set but not used (on gcc 4.6.3).
Signed-off-by: Mike McCormack <mikem@atratus.org>
SVN revision: 79445
This was removed on latest mesa, and seems to don't belong to GLES. See
the specification at http://www.khronos.org/registry/gles/.
Additionally, it wasn't being used anywhere.
SVN revision: 79400
Stefan Schmidt
Igor Murzov
Jiyoun Park
Carsten Haitzler
Jihoon Kim
Sung W. Park
Cedric Bail
Cedric BAIL
Rafael Antognolli
Daniel Willmann
Tom Hacohen
Chris Michael
Henrique Dante de Almeida
Ivan Briano
Paulo C. A. Cavalcanti Jr
Cedric BAIL
Jiyoun Park
Stefan Schmidt
Rafael Antognolli
Sung Park
Gustavo Sverzut Barbieri
Leandro Pereira
Ulisses Furquim
Paulo Alcantara
Jihoon Kim
Lucas De Marchi
Mike McCormack
ChunEon Park
Flavio Vinicius Alvares Ceolin
Vincent Torri
Christopher Michael