define scope keywords 'protected' and 'public'
disallow scope keywords as property or method name
use move_ts action to move token start after scope modifier if found
use end_property_scope and end_method_scope actions to handle it
Since rg_etc1 now outputs proper BGRA data, the shaders should not
swizzle the colors around. Stick to the normal fragment shaders.
Note: This is not tested.
this fixes and eo2 problem where when callstack grows (or shrinks)
and realloc nas to relocate memory, the frame ptrs like fptr become
invalid and all sorts of hell ensues.
this uses mmap so blowing the stack will segv, not scribble over
memory, also its separated from malloc heap, and now big enough to not
need to size ... ever (1024 entries).
Evas uses BGRA data while rg_etc1 uses RGBA data, so there
were incompatibilities between the two.
Now, rg_etc1 will take BGRA data as input and output.
Since the introduction of new colorspaces for GL_X11
(GRY8, AGRY88 and ETC1), stride_get() would return
an invalid value and data_get() would just abort().
Add proper support for these functions.
ETC1 data will NOT be returned from data_get() and
stride_get() will return 0. This is to avoid people from
messing up badly with encoded color spaces.
So I must have been a bit tired last Friday when "fixing" some
code producing artifacts, as I was just basically disabling part
of the code without realizing it :)
Let's just disable it then.
@fix
Summary:
In emotion_engine_instance_new(), if given module name is wrong or if loding is failed, fallback module loading should be triggered.a (previously, it check just whether name is NULL or not)
In case of T1025, "vlc" was given name. Therefore, fallback was not triggered.
Fixes T1025
Test Plan: enlightenment -> efm -> mouse over multimedia files -> check whether preview is shown properly or not
Reviewers: raster, cedric
CC: seoz, cedric, zmike
Maniphest Tasks: T1025
Differential Revision: https://phab.enlightenment.org/D773
Quality should not default to 100 unless specified in the
command line. In particular, we don't want to save ETC1 at
high quality by default since it can take hours (literally).
Add some comments as I'm reading and understanding the code.
Fix some rare artifacts happening mostly with medium quality
encoding, where a few pixels (2x2, 2x4 or 4x2) will have a
horrible contrast with their surroundings (eg. pink over black).
Add a new flag in EDC files to specify ETC1 compression
should be enabled. It follows the same rules as the
current LOSSY flag for JPEG compression.
@feature
The ETC1 encoder is expected to write all 8 bytes of the
output data. But in case of a solid color block, it was writing
only 1 of the first 3 bytes (R, G, B). So lots of solid blocks
were containing invalid data (for instance: R + dR < 0 or > 255).
There were a few critical issues:
- Invalid pointer arithmetics on the input data (char vs. int)
- Invalid logic in the pixel duplication code
All of these due to bad copy and paste :(
Also, use LZ4HC instead of LZ4 when compression is enabled.
ETC1 encoding is so damn slow you won't see the difference between
LZ4 and LZ4HC compression times.
Summary:
Changed uses of EINA_MAIN_LOOP_CHECK_RETURN for EINA_MAIN_LOOP_CHECK_RETURN_VAL
for functions that doesn't have void return types.
These only error out when compiling with --with-profile=debug
@fixed
Reviewers: raster, cedric, smohanty
CC: cedric
Differential Revision: https://phab.enlightenment.org/D765
Summary:
Due to unnecessary magic checking, there was freezing in /proc in efm.
Proc file's st_size is zero, but, it's readable. therfore, it takes unnecessary time in magic checking. And, there is no need to check magic in case of 0 sized regular files as well.
Therefore, skip magic check in case of st_size is zero.
Fixes T1173
Test Plan: enlightenment -> file browser (efm) -> get int /proc --> check whether efm freezes or not
Reviewers: raster, cedric, zmike
CC: seoz, cedric
Maniphest Tasks: T1173
Differential Revision: https://phab.enlightenment.org/D764
Signed-off-by: Cedric Bail <cedric.bail@free.fr>
The _ecore_wl_cb_handle_data callback only has the flag for ECORE_FD_READ
set while the callback funtion also monitors and acts on writes. I wonder
if and how that worked before.
Since the introduction of color spaces other than RGBA8888 in
the GL engines, there was an issue with border images scaled in
GL. The left and right edges were simply not properly copied.
This would then show artifacts when scaling very thin images
(typically 2px wide).
The default quality is 80 and high quality is horribly slow,
so default to medium quality instead (it's already very slow).
If you really want to shoot yourself in the foot and use high
quality, just set quality > 95 and go make some coffee.
Also, disable dithering, it creates horrible artifacts on real
life pictures (gradients and flat surfaces especially).
the comments say it all. this stops adding massive full window updates
when clippers lose all children (and are visible) or gain a first
child while visible as all code hides these clip rects (or shows them)
in these cases and other usages just make no sense.
due to recent changes a lot of objects are now NULL (correctly) and eo
complains on access of them. it's simply too noisy adding too many
if's all through code, so let's just make eo be sensible here.
LZ4HC has a higher compression ratio than LZ4 but basically the
same decompression speed.
The performance cost during encoding is actually still pretty low
considering how expensive ETC1 compression can be (even at medium
quality).
@fix: There is no use for an input cursor surface if the seat does not
have the pointer ability (for kiosk cases, or touch-only cases where
a pointer is not created).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
@fix: As it turns out, we cannot just blindly set the regions here
during resize. Elementary apps will set the opaque region to account
for any edj frames, and having the region_set calls Here was
causing issues....
Signed-off-by: Chris Michael <cp.michael@samsung.com>
this fixes the png loader code to use png_read_row properly with the
number of passes needed to load aninterlaced image as well as handling
this right with scale ratio scaledown set.
Summary:
This problem occurred due to xkb_keysym_t value of libxkbcommon by e wl server.
e wl server should pass keycode from evdev input device on to wl client.
In order that e wl server receives valid keycode Ecore_Event_Key should have
an extended data member. This patch should be applied with server side patch.
@fix
Test Plan: run e wl server -> create wl client -> type keys
Reviewers: raster, devilhorns, zmike
CC: cedric
Differential Revision: https://phab.enlightenment.org/D712
It's now completely valid to do:
a = eo_do(obj, a_get());
or:
b = eo_do(obj, a_set(1), b_get());
Also, the default return value for eo2 functions is now also returned
when the object is invalid, not just when the object does not match
class.
It's a small refactor that fixed both issues at once.
@feature
@fix
They are not needed on normal build. Two flags have been added to
enable/disable INF and DBG.
Two warnings have been removed since they are not printing interesting
and relevant information.
num was changed to size_t which makes it unsigned while count is a normal
(signed) int. Comparing them does not always work as expected but in this
case we already checked if count is < 0 before and would have returned if
that would be true. Thus we can safely cast count to unsigned here as it
will be greater 0 anyway.
Eo2 is the new object system for the EFL, and the replacement for Eo. It
improves Eo1 in many ways that will be listed on the Wiki and discussed
on the ML before.
Another change this merge includes is using the slightly different API
eo2 offers in actual code.
For example:
eo_do(obj, elm_object_text_get(&text));
becomes:
eo_do(obj, text = elm_object_text_get());
@feature
create/destroy tls key (_eo2_call_stack_key) at eo_init()/eo_shutdown().
use _eo2_call_stack_get() to allocate the stack when required.
register _eo2_call_stack_free() as eina_tls_cb_new() delete callback.
Summary:
2 threads run 'eo2_do(o, a(), b());'
- A goes first, creates an object, enters 'eo2_do(o, a(), b());'
in a() call, it blocks, releases B and waits for it.
- B when released, creates an object, enters 'eo2_do(o, a(), b());'
in a() call, it joins and releases A, then blocks.
- A returns from a(); and enters b() using current call stack frame,
which is the one pushed by B! then pop the frame and releases B.
- B does as above using the stack pushed by A!
eo_class_new() returns NULL on error in _eo2_class_funcs_set().
it covers: NULL API func, API redefined, dich func override,
overriding non-existing fct.
as op descs are sorted, we can't output fct indexes in error msgs, but
as they are sorted using api_fct as key, we can detect multiple usage of
the same api_fct which leads to an unpredictable call to whatever is
returned by _eo2_api_op_id_get()->_eo2_api_desc_get().
it is still possible to instanciate an object of a not well defined class.
"void func(void)" != "void func()" !
The former is a func that accepts 0 parameters.
The latter is a func that accepts variable number of parameters.
Watch out.
a mixin class must not inherit
- _eo2_api_desc_get()
accept NULL klass param
EO_CLASS_TYPE_REGULAR_NO_INSTANT is an acceptable extension class type for
- _eo2_class_funcs_set() do not shout if parent is NULL
there is no more difference in class or regular functions prototypes and definitions
- eo2_api_op_id_get() uses _eo_is_a_class() at runtime
- add 'void *class_data EINA_UNUSED' parameter to eo2_base class functions
- Eo2_Op_Call_Data.klass is kept only for eo2_hook_call_pre end eo2_hook_call_post,
but could be removed easily
in the EPAI, don't show that Eo* could be IDs.
in the implementation, use klass_id and obj_id
if you know or want it to be a class or an object,
use eo_id in general cases.
We want to have normal functions as non-default constructors, not va_arg
ones. What we should do is split the object creation to two parts again.
The creation, the constructing (changes using the macro) and the
verification/end part that checks the constructor has been called.
This reverts commit 2ff2ce1894f173b306a896bda595e1a7768c074d.
to protect us against bad use of break, goto, return ... in eo2_do,
we use __attribute__((cleanup(eo2_do_end))) to ensure that eo2_do_end()
is called whatever.
at class elaboration, sort the op descriptions using the function pointer.
when calling a function, do a dichotomic search in the
class op descriptions to find the corresponding OP_ID,
then keep it in a static variable.
This patch adds support for protected functions.
In the .eo file, the scope (public by default) has to be added before
the function name e.g:
protected foo ...
To access the protected APIs, #define (CLASS)_PROTECTED is needed e.g:
#define ELM_BUTTON_PROTECTED
Summary:
plug image object was not cleaned after server disconnection.
specifically, image data was cleaned (NULL), but it was not updated for real on screen
--> set dirty of plug image object
@fix
Test Plan: elementary_test -> open window socket -> open window plug -> close window socket -> plug should be cleaned up
Reviewers: raster, cedric
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D703
Summary:
Since plug sends message by using deleted server handle (dangling pointer),
I could see glorious error message as below:
*** ECORE ERROR: Ecore Magic Check Failed!!!
*** IN FUNCTION: ecore_con_server_send()
ERR<842>:ecore lib/ecore/ecore.c:729 _ecore_magic_fail() Input handle pointer is NULL!
ERR<842>:ecore lib/ecore/ecore.c:740 _ecore_magic_fail() *** NAUGHTY PROGRAMMER!!!
*** SPANK SPANK SPANK!!!
*** Now go fix your code. Tut tut tut!
Therefore, I initialized server handle on server deletion
@fix
Test Plan: create socket -> create plug -> destroy socket -> mouse move on plug area
Reviewers: raster, cedric
Reviewed By: raster
CC: seoz, cedric
Differential Revision: https://phab.enlightenment.org/D701
Jérémy Zurcher
Carsten Haitzler
Jean-Philippe Andre
Daniel Zaoui
Mike Blumenkrantz
Wonguk Jeong
Cedric Bail
Felipe Magno de Almeida
Andrii Kroitor
Lukasz Stanislawski
Stefan Schmidt
Matvey Konovalov
Chris Michael
Gwanglim Lee
Tom Hacohen
Yossi Kantor
Cedric Bail
Daniel Kolesa