The memory usage graph was going up and to the right!
I was told this is always a good thing!
... maybe not this time :)
Hopefully I didn't forget a case. An intense session of
genlist scrolling with masks all over the place and masks
of masks didn't show any glitch, crash or memory leak.
The main difference between 1.12 and 1.13 memory foot print is actually
related to this two pointer to mask. I am wondering if there is not an
issue here also has we do have a duplicated pointer. We have prev_mask
and mask in both cur and prev state of an Evas_Object, but only mask
and prev_mask from the cur state seems to be accessed.
If we can remove two pointers from those state, we should have a decent
win in expedite benchmark. Hopefully 300KB to win there (Close to half
the additional cost in memory).
Thanks cedric for the report.
We'll need a new function (after 1.12) to detect if the current
cow pointer is the default data or not. Maybe eina_cow_default_get()
or maybe something just returning a bool.
Eina_Cow does memory comparison during standby time. So in expedite as we don't
have any standby time, we end up allocating some vast amount of memory that
will never be compared. It is way simpler to compare that the data are not
going to change before hand. It should also reduce the CPU consumed during
idle time.
This patch save about 1MB of data at peak time in expedite.
Before this change eo_add() used to create an object with 1 ref, and if
the object had a parent, a second ref.
Now, eo_add() always returns an object with 1 ref, and eo_add_ref()
preserves the old behaviour (for bindings).
eo_unref now un-parents if refcount is 0, and eo_del() is an alias for
eo_unref (will change to be a way to ensure an object is dead and goes
to zombie-land even if still refed).
Summary: Fix not to clean and re-render if the position of an object enabled map is changed.
Reviewers: raster, cedric, Hermet
Reviewed By: Hermet
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1220
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
const have been added in object parameter of two legacy APIs to fit
Eolian generated files.
Since these functions retrieve information from object, it is logic that
the object would be const.
Clipees can be cleared before the loop is finished because
evas_object_clip_unset calls smart function of clip_unset.
So, if we use EINA_LIST_FOREACH_SAFE, invalid next list pointer
can be kept and read after obj->clip.clipees is freed.
Thanks to Davide Andreoli for reporting.
@fix
EINA_LIST_FREE does eina_list_remove_list, and clip_unset does
the same thing to the same list pointer. So, EINA_LIST_FOREACH_SAFE
is proper for this case.
Proxy sources & objects were not properly unset.
This results either in crashes (especially in the Edje tests)
or dangling objects with tons of references.
Remove the refcount increase/decrease, as it is redundant.
Store pairs proxy+source instead of just the source in all hashes,
so we can unset the is_proxy flag on the proxy when there are no
sources anymore.
for more gurantee to update map properly,
we should reset the map changed flag after the map updation is performed.
this will fix a mapbuf bug that map is not updated.
when the map is changed without rendering but it's in the active object list,
the map updation couldn't be happened later that map is rendered. (if the map is not updated at this frame)
This means unref will really free the object instead of just let it
linger in memory once the object is unrefed.
This is also correct because once detached from the canvas, Evas has
nothing to do with the object and it shouldn't need the manual free
anymore.
This reverts commit eb6af1f1ff.
This commit was making the code much more complex than required. Let's make
eina_cow_free set the value back to its default.
many valgrind complaints on e shutdown are there regarding accessing
cow sections, lists and object elements during shutdown. this plugs
theses little holes to avoid the invalid accesses and thus avoids
potential crashes.
This reverts commit 1714fe93f4.
We actually want this type, it makes things clearer.
Conflicts:
src/tests/eo/function_overrides/function_overrides_inherit2.c
src/tests/eo/function_overrides/function_overrides_simple.c
src/tests/eo/suite/eo_test_class_simple.c
Jean-Philippe Andre
Cedric BAIL
ChunEon Park
Jaehyun Cho
Jaeun Choi
Tom Hacohen
Sebastian Dransfeld
Carsten Haitzler
Cedric BAIL
Cedric Bail
subhransu
Taekyun Kim
Yossi Kantor
Daniel Zaoui
WooHyun Jung
Jérémy Zurcher
Daniel Juyung Seo
Cedric Bail