Summary:
There is a crash in naviframe demo and the stack points to the clip set function where it tries to acess evas object
from a NULL layer , by going through the log found out raster has already added the NULL check in clip_unset function
so just added the check in clip_set.
Reviewers: seoz, raster
CC: cedric
Differential Revision: https://phab.enlightenment.org/D625
This tag indicates that the ownership of the parameter/return
value changes.
It is needed by generators (C++/LUA...) to determine if it has to be
freed or not, if it can be used as is or need to be copied...
Before eo_do invocation, generated legacy functions returning a value
initialize it to 0.
This change is needed in the case that eo_do fails to find some
function, which leads to an unitialized value and behavior.
Tokenizer's approach of looking back is horrible and breaks the
following simple case (bug I had that lead to this patch):
"string\\"
As the parser would get the end quote and check the previous character
if it was a backslash and it was, but it was not escaping the quote,
but being escaped by the previous backslash.
The best approach is to first check for escape and then go to
quote. Escape is simple and only the following byte, so we enter
escape, process the byte and then are back to regular mode (be it
quote or unquote).
Added testcase so we avoid breaking it again.
@bugfix cherry-pick
It is not necessary to dynamically link to glReadPixels since
this is not an extension. This code wouldn't even work on some
devices.
Also, the pixels returned are not premultiplied (yeah >_<)
And some devices (EGL) don't support GL_BGRA... so glReadPixels
would just fail and not fill in the pixels. Conversion is required.
When a parameter of a property is const for get but not for set, the
.eo file indicates it by setting a flag 'const' for this parameter.
The generation was checking this flag for C files generation but not for
H files.
If we are given a drm device name, tell ecore_drm to try and find That
device. If it is not found, we fallback to the default drm device.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
from drm.
This assigns a name to the ecore_evas. If no device name was passed
in, we will get the device name from the drm card.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
@feature: Add new API function to return the drm device name
NB: This will be used in the ecore_evas_drm code to set the
ecore_evas->name.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
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.
@fix: Fix building Enlightenment without X support
These changes are needed so that we can build Enlightenment without X
support. Many places in the E code reference
Ecore_X_Atom/Ecore_X_Icon. If we build E without X support, these end
up being undefined, causing build to fail, so we need to predefine them.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
@fix: Fix building Enlightenment without X support
These changes are needed so that we can build Enlightenment without X
support. Many places in the E code reference
Ecore_X_Atom/Ecore_X_Icon. If we build E without X support, these end
up being undefined, causing build to fail, so we need to predefine them.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Use two optimizable functions for BOX blur: vertical and horizontal.
These functions will run as many times as requested (from 1 to 6 max).
The horizontal case is pretty straightforward as the source is already
contiguous (nice in terms of cache hits). The only catch is to swap
src and dst without ever writing to the input buffer.
In case of vertical blur, we apply the same method as above, after
rotating the column into a horizontal (contiguous) span, and rotating
it back afterwards.
Now, the same needs to be done for RGBA :)
BOX blur is a lot faster (and easier to optimize, too)
than GAUSSIAN blur. Repeating 2x or 3x BOX blur will also
give similar results to GAUSSIAN blur (very smooth), but
in much less time.
Add a count parameter to the BOX blur instruction.
Actually, there is a very nice trick with BOX blur.
Pass BOX blur 3 times and you can approximate a GAUSSIAN
blur with up to 3% accuracy. This is way more than enough
for just a simple graphical effect.
So, despite the crappy quality of BOX blur, we should
optimize it a lot so we can replace large GAUSSIAN blurs
with series of BOX blurs instead.
Source: Wikipedia's page on box blur :)
This commit also moves around some duplicated definitions.
When a blur operation requires a copy-back to the source
buffer, then the render_op must be set to COPY instead of
BLEND. Otherwise the non blurred content will be visible.
@fix
Prepare optimization paths for blur operations, as they are VERY
costly. This simple change, when using gcc -O3 flag, boosts
horizontal blur performance by > 50%, because STEP is 1 (and
so, memory accesses, increments, etc... are all very simple)
The objective is to have support for NEON, MMX, SSE, too, with
runtime detection.