Call object's function to get the private engine_data (here, the
image object). Thanks Dongyeon for your patch which inspired me to
do that instead of forcing pre_render.
This will currently optimize most of the masks when using the
GL engine[1].
This is a very special case that adds a highly optimized path
for masking in GL. It works by creating a virtual image, containing
a pointer to the original image and a new geometry[2].
Instead of creating a new FBO-based surface (image_map_surface),
we refer to the original image and adjust the mask geometry on
the fly.
KNOWN BUGS:
- masking a map with such a scaled image is now broken.
[1] Right now all masks are simple Evas Object Image, so that means
all cases of masking, except masks of masks, or masks of maps,
will be optimized with this new method.
[2] This virtual image mechanism is still quite hackish and may
be improved (for memory usage, refcounting, etc...)
when dealing with non-kbd devices, the seat can be iterated to locate a keyboard
this may or may not accurately set depressed, latched, locked, group values
Summary:
For example of a bug, part of .obj file:
vn 0.5536 -0.7200 -0.4185\n
vn -0.5536 -0.7200 -0.4185\n
\# 239 vertex normals\n
\n
vt 0.4998 0.2618 0.0000\n(lines like this were ignored)
vt 0.5205 0.2550 0.0000\n
vt 0.5249 0.2618 0.0000\n
@fix
Test Plan: Run colorpick example. Before and after this update. ("M15.obj" has fixed places.)
Reviewers: cedric, Hermet, raster
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D2049
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
With this commit I'm finally able to use -j10 for make install on my machine.
During install libtool does some relinking which can result in to broken linking
if the dependencies are not handled correctly. Sadly automake has a problem with
the automatic dependency handling during install with LTLIBRARIES which we use
for all our modules. For the details please see this 4.5 years old bug report:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=7328
We are now setting the dependency manually to force automake to the right decision
during install relinking.
Speed improvement itself is not that high (make -j 1 compared to -j10):
real 0m21.410s vs. real 0m17.066s
The bigger benefit is the unified use of MAKEOPTS or normal -j X in all our
build targets. I have seen quite some bug reports where -j was used for install
target when it was used in the build target. Last but not least it helps me to
unify some parts of the jenkins jobs and finally allows me to run distcheck
with -j Which uses install internally and failed before. Which goes down from
real 12m50.349s to real 5m52.120s.
if you fork and even if you do ecore_fork_reset() a thread calling
ecore_main_loop_thread_safe_call_async(0 for example eill end up
resetting the mainloop thread id to itself (a non mainlopo thread) via
calling eina_main_loop_is() since pid changed. there is little point
in doing this so remove the pid tracking from eina and ensure mainloop
thread id is updated in ecore's fork reset.
@fix
Summary:
Added test case for eina_memdup function in eina_str test module
Signed-off-by: vivek <vivek.ellur@samsung.com>
Reviewers: cedric
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D2026
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
While install-exec-hook gets normally executed after install and
thus we would have this we need to ensure it here when we want to
be safe regarding parallel install.
While we used different variation of mkdir -p all over we also had spots
where we did not use the option. This is one step in trying to make our
build system ready for parallel install. Using something like -j 10 even
for the install should help to speed up our jenkins jobs as well as distcheck.
When destroying a GLES 1.1 surface, it is necessary to also
destroy and remove the main surface from the list.
This issue probably never really showed up because people
don't:
- use GLES 1.1
- constantly create & destroy new Evas GL surfaces
- but mostly no one cares about 1.1 anymore :)
@fix
Most of the time the style string will come from the eet file directly, so
thanks to the dictionnary build in they should be pointing to the same string.
We still need to keep strcmp case for Edje_Edit case, but that shouldn't be
a real issue as the worst case is when it match. When it doesn't match strcmp
should return quite fast on average.
Summary:
When textblock styles have text_classes, all edjes in the files were added
to text_class_member_hash even if the edjes didn't use the textblock styles. It
makes time long to update text_class.
This will add the edje using the textblock style which has a text_class to
text_class_member_hash.
Reviewers: cedric, raster
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D2035
This requires a special context that matches the configuration
required for GLES 1.1. Otherwise eglMakeCurrent() would fail
miserably with EGL_BAD_MATCH in case of indirect rendering
(at least on some drivers).
Summary:
- Implement glGetString() wrapper func in the same way as gles2.x.
- Small bug fix glGetString() for gles2.x.
Reviewers: cedric, raster, jpeg
Subscribers: cedric, mythri, wonsik, spacegrapher
Differential Revision: https://phab.enlightenment.org/D2033
Those 2 new values are here to avoid using environment variables
that have side effects on the whole application.
I'm actually wondering if we shouldn't just kill off the env
vars altogether. Also, direct override is a terrible option that
should never be used.
Memory optimization can make sense (needs more testing tho).
Summary:
"far" and "near" are keywords on windows and can't be used as names of variables.
@fix
Reviewers: cedric, Hermet, raster, perepelits.m
Subscribers: reutskiy.v.v, cedric
Differential Revision: https://phab.enlightenment.org/D2037
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
Oops!
According to Coding Convention it should be like that:
...
>>> function forward declaration/prototype should be a single line;
>>> function definition should have the return at one line, then function name starts at next line, column 0;
...
Summary:
To distinguish supported extension name from not supported.
This patch can be solution to the problem, glGetString() returns non-supported extention name.
Test Plan: Local tests
Reviewers: raster, jpeg, Hermet, cedric
Subscribers: cedric, spacegrapher, wonsik
Differential Revision: https://phab.enlightenment.org/D1981
Signed-off-by: Jean-Philippe Andre <jp.andre@samsung.com>
Summary: When we raise an event for an output, also include the output
id in the event structure. This will allow us to better identify which
output the event occured on.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Summary:
Textures for evas-3d are standardized 256*256.
.tga for parallax occlusion aren't changed to save alpha channel.
-50Mb for efl without obvious lost of quality
Reviewers: cedric, Hermet, raster
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D2032
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
This affects eo_do() and eo_add() that used to use the ({}) GCCism.
Following a discussion with Peter de Ridder after my talk at FOSDEM,
we've decided to reopen the GCCism (works with other gcc compatible
compilers like clang and intelc) discussion, and after a bit of back and
forth it was decided to make things more portable, at the cost of ease
of use.
For example:
if (eo_do(obj, visible_get()))
is no longer allowed, the portable alternative
Eina_Bool tmp;
if (eo_do_ret(obj, tmp, visible_get()))
is to be used instead.
However:
eo_do(obj, a = a_get(), b = b_get(), bool_set(!bool_get))
are still allowed and OK.
eo_do(obj, if (a_get()) return;);
is no longer allowed, but:
eo_do(obj, if (a_get()) something());
is still allowed.
For clarity, this commit only incorporates the Eo changes, and not the
EFL changes to make the efl conform with this change.
Thanks again to Peter de Ridder for triggering this important discussion
which led to this change.