evas will use all loaders (brute force) if it can't guess by extension
- this means loaders need to stay quiet about invalid file formats as
this is literally intended to probe if the file is really of that
format when nothng is known about it.
on arm when a is 0 ... per pixel rgb can be > 0 which violates premul
rgb leading to junk rendering. this now is worked around in the png
loader and forces rgb to 0 when a is 0 per pixel.
@fix
Summary:
Values different from numbers and percentages should be ignored
and the default values should be applied (zeros).
And set the min and max of the offset value to be 0, 1.
Also, this patch make that the offset is not input in the reverse order.
Test Plan:
Test SVG Image
```
<svg viewBox="0 0 200 200" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" >
<linearGradient id="grad" x1="0" y1="0" x2="1" y2="1">
<stop offset="10%" stop-color="white"/>
<stop offset="0.2" stop-color="red"/>
<stop offset="30% k" stop-color="blue"/>
<stop offset="40%" stop-color="yellow"/>
<stop offset="0.5m" stop-color="red"/>
<stop offset="0.6 " stop-color="green"/>
<stop offset="70%m" stop-color="black"/>
<stop offset="80%" stop-color="white"/>
</linearGradient>
<rect x="20" y="20" width="160" height="160" fill="url(#grad)"/>
</svg>
```
Result
{F4792365}
Reviewers: Hermet, raster, kimcinoo
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12318
Summary:
"none" is the default value of dasharray and can actually be used.
Currently using "none" causes a segfault. This patch prevents it.
Test Plan:
SVG image
```
<svg viewBox="0 0 30 10" xmlns="http://www.w3.org/2000/svg">
<line x1="0" y1="3" x2="30" y2="3" stroke="black" stroke-dasharray="none" />
</svg>
```
Reviewers: Hermet, raster, kimcinoo
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12317
Summary:
After finding no attributes but spaces, attrsLength could be negative.
This will cause a segfault in parser functions.
So, change the position of attrs_length to prevent this.
Test Plan:
Example SVG
```
<?xml version="1.0" encoding="UTF-8"?>
<svg><g ></g></svg>
```
Reviewers: Hermet, raster, kimcinoo
Reviewed By: Hermet
Subscribers: #reviewers, cedric, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12314
i see no good reason to not allow these loaders to run in threads.
they are isolated and i don't see a reason for the libs called to not
be threadsafe.
Summary: this backend code was for selecting XCB over Xlib. Now XCB is removed, this code is useless
Reviewers: raster
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12299
Summary:
egl images created using tbm surface for native surface set use
GL_TEXTURE_EXTERNA_OES as texture target, so we should bind to
this target when rendering. Or there is a GL_INVALID_OPERATION
error on glBindTexture in function _orig_shader_array_flush.
Thia patch follows logic of following commit;
7db0e20 evas/gl: Bind texture with external target for tbm surface
Reviewers: Hermet, raster, jsuya, herb
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12286
Summary:
Remove unnecessary point movement in rgb(255, 255, 255) case in svg parsing.
In svg parsing, move the pointer by 'rgb(' before calling
_color_parser() in the rgb(255, 255, 255) case.
In function, string pointer moved unnecessary, so parsing is incorrect.
Therefore, remove unnecessary point movement.
Test Plan:
svg sample code
```
<svg xmlns:svg="http://www.w3.org/2000/svg" viewBox="0 0 300 300">
<path d="M 0 0 h 200 v 200 z" style="fill:rgb(255, 155, 55);"/>
</svg>
```
before
{F4504779}
after
{F4504778}
Reviewers: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12284
Summary:
Change-Id: I17c79f89550e9a758a532babd55826b3c7ad2810
To show gif animation smoothly, we keep the current frame even if getting line is failed.
Currently, the frame image will be freed if getting gif line is failed and this makes the frame drop.
Reviewers: Hermet, kimcinoo, raster
Reviewed By: raster
Subscribers: raster, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12277
Summary:
Patch from a series of patches to rename EAPI symbols to specific
library DSOs.
= The Rationale =
EAPI was designed to be able to pass
`__attribute__ ((visibility ("default")))` for symbols with
GCC, which would mean that even if -fvisibility=hidden was used
when compiling the library, the needed symbols would get exported.
MSVC __almost__ works like GCC (or mingw) in which you can
declare everything as export and it will just work (slower, but
it will work). But there's a caveat: global variables will not
work the same way for MSVC, but works for mingw and GCC.
For global variables (as opposed to functions), MSVC requires
correct DSO visibility for MSVC: instead of declaring a symbol as
export for everything, you need to declare it as import when
importing from another DSO and export when defining it locally.
With current EAPI definitions, we get the following example
working in mingw and MSVC (observe it doesn't define any global
variables as exported symbols).
Example 1:
dll1:
```
EAPI void foo(void);
EAPI void bar()
{
foo();
}
```
dll2:
```
EAPI void foo()
{
printf ("foo\n");
}
```
This works fine with API defined as __declspec(dllexport) in both
cases and for gcc defining as
`__atttribute__((visibility("default")))`.
However, the following:
Example 2:
dll1:
```
EAPI extern int foo;
EAPI void foobar(void);
EAPI void bar()
{
foo = 5;
foobar();
}
```
dll2:
```
EAPI int foo = 0;
EAPI void foobar()
{
printf ("foo %d\n", foo);
}
```
This will work on mingw but will not work for MSVC. And that's why
EAPI is the only solution that worked for MSVC.
Co-authored-by: João Paulo Taylor Ienczak Zanette <jpaulotiz@gmail.com>
Co-authored-by: Ricardo Campos <ricardo.campos@expertise.dev>
Co-authored-by: Lucas Cavalcante de Sousa <lucks.sousa@gmail.com>
Reviewers: vtorri, jptiz, woohyun, lucas, SPAM-smith78899
Reviewed By: vtorri, SPAM-smith78899
Subscribers: SPAM-smith78899, raster, SPAM-cabanacatalogs, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12228
this can happen if dmabuf is invalid (already destroyed) so handle
this so evas knows and falls back to rendering a rect and complains to
stderr so we know
@fix
Summary: and minor fix of some values in avif loader
Test Plan: entice still can read heif
Reviewers: raster
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12257
a 2x2 matrix visually here is about as good as 4x4. at least in what i
see, but on low end gpu's it can halve the cost. in fact i was watching the
gpu on my old i5-4200u drop down to 340-410mhz (no dithering is 320-360mhz).
i got to 630-660mhz with the original 4x4 code.
the 4x4 is still there ifdefed out. perhaps i can bring it back with a
high-quality dither option, but 2x3 i think is good enough.
this adds a
EVAS_GL_RENDER_DISABLE_DITHER=1
env var (set it to anything) to disable dithered rendering.p this is
an added cost, but normally not much at all, but it seems some gl
implementations are broken, and they can't generate correct code for
the dither shader, so this disables this if this env var above is set
Summary: libavif can be built without a AV1 decoder. Check this in the image loader
Test Plan: entice
Reviewers: raster
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12252
we need this to do formats and modifiers properly... and this is what
broke on the rpi4 - it started using interesting formats etc. ... new
feature to support a bug fix in e
@feat
Summary:
The tex->pt->references is descreased by
evas_gl_common_texture_free -> pt_unref
if tex->references is 0
And tex->pt->texture is removed by
evas_gl_common_texture_free -> pt_unref -> glDeleteTextures
if tex->pt->references is 0
The evas_gl_common_texture_free decreases tex->references only
if tex->references is bigger than 0. There is no chance to decrease
tex->pt->references at this point.
So if orient_set increases both references of tex and tex->pt, then
the tex->pt->reference is not decreased till tex->references is 0.
So do not increase tex->pt->references in eng_orient_set.
Reviewers: raster, cedric, Hermet
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12243
Summary: that supports images : *.heif, *hiec and *.avif I have disabled *.avif images, there is already a loader.
Reviewers: stefan_schmidt, raster
Subscribers: raster, vtorri, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12135
Summary:
There could be 2 evas_gl_image referencing 1 evas_gl_texture.
evas_object_image_orient_set could make this case.
In this case, when one evas_gl_image is removed(free), the evas_gl_texture
is not removed because its reference count.
After this point, if the other evas_gl_image is removed without drawing
(see function evas_gl_common_image_draw, line "im->tex->im = im")
then evas_gl_texture is reading invalid adress when it is removed.
Reviewers: Hermet, jsuya, herb, devilhorns
Reviewed By: devilhorns
Subscribers: devilhorns, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12229
when a file has no extension to guess by.. evas goes back to trying
every loader until one works. this leads to noise. loaders should not
printf/stder or eina log err anything if the file seems to not be valid.
@fix
Summary:
The strech region has paired information; total and strechable.
Refer to function _strech_region_load retrieving strech region info.
But if 9-patch information line ends with strechable,
png did not push the strechable information.
And it leads to devide by zero.
This patch is adding strechable info to the strech region,
if 9-patch information ends with strechable.
Test Plan:
[Code]
{F4219278}
{F4219280}
[Test]
ECORE_EVAS_ENGINE=opengl_x11 ./evas-image-9patch ./end_with_strech.9.png
Reviewers: Hermet, jsuya, herb, cedric
Reviewed By: Hermet
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12204
this adds a dither func (4x4 dither matrix) to experiment with higher
quality rendering in gl - this assumes you have a normal 8bit per
channel buffer for now (99% of people) and will approximate values in
between the 256 steps 8 bits provides by using the dither matrix based
on gl_FragCoord position. it's just a flag in the shader flags for now
so can be turned on/off in code. this definitely makes blurs look much
better... everything else seems basicall the same. let's see how this
goes.
@feat
Summary:
evas image might have a better quaility if scaling/transform is not necessary,
so we have a condition to check if image is axis-aligned transformed or not.
On the other hand sub-pixel(floating point coordinates unit) rendering necessary
if image has an effect such a zooming. This would result in a smoother effect
than integer coodinate system.
We need a more precise condition to confirm this,
so we intrduce "anti-alias" option to decide the condition.
now, anti-aliased objects will have a sub-pixel rendering always.
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12194
Summary:
The 9 patch is using image_stretch_region_get, but GL did not override it.
So the 9 patch did not work for GL engine at all.
Test Plan:
Evas_Object*img = evas_object_image_filled_add(evas);
evas_object_image_file_set(img, "test.9.png", 0);
evas_object_show(img);
Reviewers: Hermet, jsuya, herb, cedric
Reviewed By: Hermet
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12192
Carsten Haitzler
Vincent Torri
junsu choi
Vincent Torri
Shinwoo Kim
Taehyub Kim
Alastair Poole
Felipe Magno de Almeida
thierry1970
Hermet Park