file != NULL does not mean it's valid. Since Eina_File is
a basic eina type a magic check is still better than nothing.
It can avoid doing eina_file_dup() on a closed file for instance.
This "fixes" a crash in eina_file_close with invalid files.
Now I can go hunt the root cause...
evas_render_method_lookup calls evas_module_find_type that
in turn actually goes and loads the module. All we wanted to
know was whether the render_method corresponded to one of the
wayland engines.
See 453770137f
The frame object requires a theme of version 119 or more. In fact
I think until we are totally happy with the window API (for EO) we
might want to bump that version regularly. That would indeed disallow
theme customization for border.edc until it's done.
This patch uses a pretty brute force way to set the theme file to
the default file from EFL installation. elm_config is not reliable
here.
This is very custom made and there may be a more generic way to force
a widget to use a minimum theme version. Yes that could mean ugly
widgets if we change the theme API but at least that would make them
work. Note that the border theme contains no visual elements, so the
colors of the background, etc... should all depend on the user
selected theme. But of course CSD (in Wayland) will have to use the
default theme -- and look grey.
Fixes D4976
Summary:
If "text" is NULL, it is set as empty string in the above code.
So, the removed line is unreachable.
Test Plan: N/A
Reviewers: raster, cedric, herdsman, woohyun, jpeg, Hermet
Reviewed By: Hermet
Differential Revision: https://phab.enlightenment.org/D4462
Summary:
When _access_init was called, initted will be increased always even though failed to initialize.
This patch will fix this problem.
Signed-off-by: Jonghee Choi <joi.choi@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>
Reviewers: cedric, Hermet
Reviewed By: Hermet
Subscribers: joi.choi, jpeg
Differential Revision: https://phab.enlightenment.org/D4466
Summary:
- old_file_ids is freed but not set as NULL.
If it goes to error code, old_file_ids will be freed again.
Reviewers: jpeg, cedric, Hermet
Reviewed By: Hermet
Subscribers: conr2d
Differential Revision: https://phab.enlightenment.org/D4467
eglGetProcAddress should be queried with dlsym unconditionally. What we
had could query it with other extended forms of eglGetProcAddress, which
is probably not what anyone wants.
Also, throwing away the weird extended forms because there's a good chance
our other gl bits don't run on any stacks that don't support normal
eglGetProcAddress.
In case a warning gets printed from lcov some internal function changes the dir
to / and thus fails to create files in the given file later on. Using an
absolute path here is a workaround to avoid this problem. The fix is in lcoc
already but not yet in any release so we better keep this around here.
Lcov fix reference:
632c25a0d1
I have seen hangs with gcov on Jenkins and locally where the processing just
keeps spinning in an infinite loop. From what I have found out this boils down
to using gcov --all-blocks which is what lcov does with branch coverage enabled.
It is supposed to be fixed in gcc 4.8+ but I see this here with 5.3.1. So its
either a regression or not completely fixed. In any case we will ignore branch
coverage for now. I hoped it would work well but it did only for a while and
having line and function coverage is better than having nothing.
Commit fcef8d8392
breaks any evas engine that frees/NULLs its own outbuf before
calling evas_render_engine_software_generic_update()
We should unconditionally set the outbuf, we only need to
do the free conditionally.
This fixes an issue where efl_isa() wouldn't work for extensions or
ancestors of extensions of a class.
Example:
Class A implements interface F2
F2 inherits from interface F1
obj is of class A
Before this patch efl_isa(obj, F1) would return false, now it returns
true as expected.
This is just one example, there is a whole array of variations to this
issue that are now fixed.
Thanks to Gustavo for reminding me of this.
@fix
This changes the string in Efl_Class_Description to use the real
class name (with namespaces) instead of the C class name. The
reason for this is that this string is generic, not C-related.
Like 79d76fb25e, this is useful when
debugging a core dump.
It accepts a valid pointer to an object, for example as returned from
$eo_resolve, and a name of a class or mixin, and returns a pointer to
the private data. Essentially the same as efl_data_scope_get(), but also
works on core dumps, and accepts a class name instead of a class
pointer.
Usage:
Print the pointer:
(gdb) print $eo_data_get($eo_resolve(obj), "Efl_Canvas_Object")
$1 = (void *) 0x555555eb9290
Use it directly (e.g. to print a value):
(gdb) print ((Evas_Object_Protected_Data *) $eo_data_get($eo_resolve(obj),
"Efl_Canvas_Object"))->last_event_type
$2 = EVAS_CALLBACK_MOUSE_UP
@feature
evas_render_engine_software_generic_update(0 definitely is wrong where
it wants to always free the outbuf even if the buffer passed in is the
same one and thus it ends up being freed and now invalid. fix it
@fix
Summary: Due to implicit declaration, evas_vg_container_child_get casts to int and on 64-bit can return wrong pointer
Reviewers: cedric, NikaWhite, myoungwoon, jpeg, an.kroitor, Hermet
Reviewed By: Hermet
Subscribers: t.naumenko
Differential Revision: https://phab.enlightenment.org/D4464
Summary:
- ptrag is set NULL and is allocated only when prop is rotated.
but *ptrag = 0xFF00 | ptr[0]; without checking rotation.
Reviewers: jpeg, cedric, Hermet
Reviewed By: Hermet
Subscribers: conr2d
Differential Revision: https://phab.enlightenment.org/D4463
Because we async render into buffers before the compositor has told us
we can use them, we can end up kicking over to fallback while still
rendering into a buffer.
Refcount the manager to let us clean up properly without crashing when
this happens.
If we render fast enough we can use more than one buffer before the
compositor assigns us buffer ids. We need to be careful to clear the
busy bit on all but the most recent one.
Enlightenment needs to know if a specific dmabuf format is supported
before it lets clients use it. This lets E test commit a wayland
dmabuf object without assigning it a buffer.
Previous patch to not destroy Outbuf on resize should not have removed
this line else software generic tilebuffers will not be in sync with
the updated size.
ref 5ebba4463
Signed-off-by: Chris Michael <cp.michael@samsung.com>