This is a preparation step for (experimental) textblock support.
Textblock objects won't have a single filter, and the buffer's
geometry wouldn't be that of of the object itself. Thus a few
internal APIs need to be reworked first.
i found evas_common_draw_context_apply_cutouts() was procsessing 300+
cutouts and as it's O(n^2)/2 to try and merge adjacent rects for
cutouts this really performs like complete junk. we apply cutout rects
a LOT. this is not the best solution, but it's quick and much faster
than doing the clipouts which drop framerate to like 1-2fps or so in the
nasty case i say (tyls -m of photos in a dir with a 2160 high
terminal).
this figures out the target area to limit the count of rects
significantly so O(n^2) is far far better when n is now < 10 most of
the time. and for the few operations where it's a high value this now
uses qsort to speed up merges etc. etc.
@optimize
This patch introduces the possibility to set the pointer mode and
query other properties like current position per pointer device.
The old API will still works, however it will only act on the default seat.
preparing an object is a good idea. especially with gl. you want to do
texture uploads BEFORE using textures all in one batch. otherwise this
may mean the gl implementation has to make a copy of your data in a
tmp location then copy it in later when texture becomes "unused" as it
may be in use at the moment, or it may have to stall and wait.
i have seen somewhere around 7-10% speedups on nvidia and intel
drivers with this on given a very special test case i brewed up (1000
32x32 images where i change 1 pixel every frame). this should have
impact really when we are modifying textures a lot. this is all i've
implemented for now, but this should/would/could do much more like
re-order map, proxy renders to render FIRST in a pre-render list
instead of inline and to pre-render fbo/buffer content for complex
objects like text or textblock etc.
so i have been doing some profiling on my rpi3 ... and it seems
memcmp() is like the number one top used function - especially running
e in wayland compositor mode. it uses accoring to perf top about 9-15%
of samples (samples are not adding up to 100%). no - i cant seem to
get a call graph because all that happens is the whole kernel locks up
solid if i try, so i can only get the leaf node call stats. what
function was currently active at the sample time. memcmp is the
biggest by far. 2-3 times anything else.
13.47% libarmmem.so [.] memcmp
6.43% libevas.so.1.18.99 [.] _evas_render_phase1_object_pro
4.74% libevas.so.1.18.99 [.] evas_render_updates_internal.c
2.84% libeo.so.1.18.99 [.] _eo_obj_pointer_get
2.49% libevas.so.1.18.99 [.] evas_render_updates_internal_l
2.03% libpthread-2.24.so [.] pthread_getspecific
1.61% libeo.so.1.18.99 [.] efl_data_scope_get
1.60% libevas.so.1.18.99 [.] _evas_event_object_list_raw_in
1.54% libevas.so.1.18.99 [.] evas_object_smart_changed_get
1.32% libgcc_s.so.1 [.] __udivsi3
1.21% libevas.so.1.18.99 [.] evas_object_is_active
1.14% libc-2.24.so [.] malloc
0.96% libevas.so.1.18.99 [.] evas_render_mapped
0.85% libeo.so.1.18.99 [.] efl_isa
yeah. it's perf. it's sampling so not 100% accurate, but close to
"good enough" for the bigger stuff. so interestingly memcmp() is
actually in a special library/module (libarmmem.so) and is a REAL
function call. so doing memcmp's for small bits of memory ESPECIALLY
when we know their size in advance is not great. i am not sure our own
use of memcmp() is the actual culprit because even with this patch
memcmp still is right up there. we use it for stringshare which is
harder to remove as stringshare has variable sized memory blobs to
compare.
but the point remains - memcmp() is an ACTUAL function call. even on
x86 (i checked the assembly). and replacing it with a static inline
custom comparer is better. in fact i did that and benchmarked it as a
sample case for eina_tiler which has 4 ints (16 bytes) to compare
every time. i also compiled to assembly on x86 to inspect and make sure
things made sense.
the text color compare was just comparing 4 bytes as a color (an int
worth) which was silly to use memcmp on as it could just cast to an
int and do a == b. the map was a little more evil as it was 2 ptrs
plus 2 bitfields, but the way bitfields work means i can assume the
last byte is both bitfields combined. i can be a little more evil for
the rect tests as 4 ints compared is the same as comparing 2 long
longs (64bit types). yes. don't get pedantic. all platforms efl works
on work this way and this is a base assumption in efl and it's true
everywhere worth talking about.
yes - i tried __int128 too. it was not faster on x86 anyway and can't
compile on armv7. in my speed tests on x86-64, comparing 2 rects by
casting to a struct of 2 long long's and comparing just those is 70%
faster than comapring 4 ints. and the 2 long longs is 360% faster than
a memcmp. on arm (my rpi3) the long long is 12% faster than the 4 ints,
and it is 226% faster than a memcmp().
it'd be best if we didnt even have to compare at all, but with these
algorithms we do, so doing it faster is better.
we probably should nuke all the memcmp's we have that are not of large
bits of memory or variable sized bits of memory.
i set breakpoints for memcmp and found at least a chunk in efl. but
also it seems the vc4 driver was also doing it too. i have no idea how
much memory it was doing this to and it may ultimately be the biggest
culprit here, BUT we may as well reduce our overhead since i've found
this anyway. less "false positives" when hunting problems.
why am i doing this? i'm setting framerate hiccups. eg like we drop 3,
5 or 10 frames, then drop another bunch, then go back to smooth, then
this hiccup again. finding out WHAT is causing that hiccup is hard. i
can only SEE the hiccups on my rpi3 - not on x86. i am not so sure
it's cpufreq bouncing about as i've locked cpu to 600mhz and it still
happens. it's something else. maybe something we are polling? maybe
it's something in our drm/kms backend? maybe its in the vc4 drivers or
kernel parts? i have no idea. trying to hunt this is hard, but this is
important as this is something that possibly is affecting everyone but
other hw is fast enough to hide it...
in the meantime find and optimize what i find along the way.
@optimize
Summary:
Visual position of ellipsis item should be set according to
its bidi direction. But, by setting visual position in same way
as logical position, the end ellipsis could be put opposite side.
Also, start ellipsis must placed on left side of RTL text.
@fix T3187
Test Plan: Test an sample on T3187
Reviewers: tasn, woohyun, herdsman
Subscribers: raster, Blackmole, z-wony, cedric, jpeg, minudf
Maniphest Tasks: T3187
Differential Revision: https://phab.enlightenment.org/D3769
Summary:
We can't assume the given font is same with current fdesc by comparing string.
Since Evas starts to supporting "auto" for language,
the given font string should be parsed once before comparing it.
@fix
Test Plan: N/A
Reviewers: tasn, raster, cedric, herdsman
Subscribers: jpeg, minudf, z-wony, Blackmole, woohyun
Differential Revision: https://phab.enlightenment.org/D4227
Summary:
BiDi text is truncated because the way we find the last visual item in text object is wrong.
This patch is similar with _line_native_last_visual_get function in textblock.
Reviewers: cedric, tasn, herdsman
Subscribers: id213sin, cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3947
Those are marked as EINA_DEPRECATED since they are for legacy only,
and EDC is the preferred method.
Note: these are new in EFL 1.18.
I would rather have not added them as legacy APIs. This patch is
here to align to Tizen which added them by mistake. Also, remove
the "name" argument from filter_program_set as it's not required,
and also not there in Tizen's code base.
Sorry for the mess.
Mostly unused vars following the removal of eo_do_ret().
However, there are some cases where the migration script got some things
wrong, and I had to manually fix them.
I just ran my script (email to follow) to migrate all of the EFL
automatically. This commit is *only* the automatic conversion, so it can
be easily reverted and re-run.
Those should now be considered stable, even if their internals
may change. Also, these APIs are in Tizen so adding these will
help merging Tizen EFL and upstream.
- Remove @beta flags,
- Update @since to match stabilization,
- Change methods to properties with keys,
- Use eo_prefix and add filter_ prefix to all properties since
they use very generic names,
The filter API stays under Efl.Gfx since there are other kinds of
filters, and this one is the particular "graphical filter" or
"effect" API.
The EO API mostly not change from an application point of view,
except for "source_get" which now returns a string directly. Also,
state and data can now be queried.
Summary:
When only ellipsis is changed from 0.0~1.0 to -1.0 without resize,
the text is never updated. Because, previous state for ellipsis is never kept
and used properly to check when Evas Text needs to be updated.
It does not have any effect when ellipsis is changed from -1.0 to 0.0~1.0.
Because, Evas text always resize itself according to its text size.
So, necessarily, Evas text object has to be resized to the smaller size.
Commonly, Edje will handle its size if Evas text needs to be ellipsized.
@fix
Test Plan: Test case is included.
Reviewers: tasn, woohyun, herdsman
Subscribers: cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3448
Summary:
_render_pre() function could be called for an object which is
going to be deleted. According to state changes of the object,
text could be recalculated with free'd pointers. It caused an
invalid read and crash.
@fix
Test Plan:
1. Apply D1747.
2. Run elementary_test.
3. Put any character in elm_entry and change paragraph direction.
4. Put any character again.
5. It can cause a crash which is caused by invalid read in Evas Text.
Reviewers: herdsman, woohyun, tasn, raster
Subscribers: cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3445
Summary:
Evas Text, Textblock, Textgrid keeps own language information.
This language information could be vary from the result of setlocale().
Especially, Evas Textblock supports <lang> tag. The language could be
changed in the middle of text. All of these language has to be used
for harfbuzz shaping.
@fix
Test Plan: N/A
Reviewers: herdsman, raster, woohyun, tasn
Reviewed By: tasn
Subscribers: cedric, jpeg
Differential Revision: https://phab.enlightenment.org/D3628
Summary:
The configuration files for Fontconfig can describe
how font list is made according to language information.
EFL also set the language for each Evas textblock styles
and used for loading font list.
But, this is inconvenient to use if we want to apply language
for loading font list according to system-wide locale information.
This patch will apply locale information for font list if there is
no specific language in description.
And it also add [lang=auto] for Evas Textblock.
auto - It loads locale for language.
none - It disables language.
@feature
Test Plan: N/A
Reviewers: woohyun, herdsman, tasn
Subscribers: jpeg, cedric
Differential Revision: https://phab.enlightenment.org/D3344
Summary:
Evas Text only concerns about a advance of each text item.
When a width of last character is bigger than its advance, the last character can be truncated.
And the different line size calculation caused different aligning between Evas Text and Evas Textblock.
So, the width of last character will be considered in Evas Text just like Evas Textblock.
@fix
Test Plan:
The following text shows how the size calculation is different between Evas Textblock and Text.
Get native size from Evas Textblock and get width(geometry) of Evas Text.
You can see the width of Evas Text is bigger than native size of Evas Textblock.
(adv > width)
こんにちは。
The following text will be truncated without this patch.
(adv < width)
ନୂଁ
Reviewers: woohyun, tasn, herdsman
Subscribers: jpeg, cedric
Differential Revision: https://phab.enlightenment.org/D3004
Summary:
If text object has outline and shadow effect (OUTLINE_SHADOW or OUTLINE_SOFT_SHADOW), and has shadow direction except BOTTOM_RIGHT then outline is not displayed.
So fix logic of style checking.
OUTLINE_SOFT_SHADOW effect of text object is different with text block, so fix it.
Test Plan: Make text obect with OUTLINE_SHADOW or OUTLINE_SOFT_SHADOW effect with any direction except BOTTOM_RIGHT
Reviewers: cedric, woohyun, herdsman, tasn
Subscribers: id213sin
Differential Revision: https://phab.enlightenment.org/D3047
Summary:
It adds evas_object_paragraph_direction_set, get APIs.
The APIs set or get paragraph direction to/from the given object.
It changes BiDi calculations and affect the direction and aligning of text.
It doesn't have any effect to text without Fribidi library.
The default paragraph direction is EVAS_BIDI_DIRECTION_INHERIT.
If dir is EVAS_BIDI_DIRECTION_INHERIT, paragraph direction is changed
according to smart parent object. If there is no smart parent object,
paragraph direction works as EVAS_BIDI_DIRECTION_NEUTRAL.
@feature
Test Plan:
Test cases included to the following files.
- evas_test_textblock.c
- evas_test_text.c
- evas_test_object_smart.c
Run "make check".
Reviewers: woohyun, raster, herdsman, tasn
Subscribers: c, raster, cedric
Differential Revision: https://phab.enlightenment.org/D1690
Summary:
Fix memory leak
Position is getting returned without
freeing the temporary logical item list.
@fix
Test Plan: NA
Reviewers: tasn, raster
Reviewed By: raster
Subscribers: raster, cedric
Differential Revision: https://phab.enlightenment.org/D3345
Summary:
Remove fixme marker
Algorithm used is good enough to sort
very few number of items, usually 2 to 6
items.
Test Plan: NA
Reviewers: herdsman, cedric, tasn
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D3312
This is really just a safety measure, as we always just check if the value
is true. However, with char the potential values are -1 and 0, and with
Eina_Bool they are 1 and 0, so fixing that.
Thanks to the "sparse semantic parser" for detecting that, and to Stefan
for reporting.
Summary:
Fix Segmentation Fault when TEXT part getting resized and when it has
some params (for example ellipsis).
Fix T2640
@fix
Test Plan: Refer to T2640
Reviewers: raster, Hermet, seoz, herdsman, cedric, reutskiy.v.v, NikaWhite
Subscribers: stefan_schmidt, tasn, cedric
Maniphest Tasks: T2640
Differential Revision: https://phab.enlightenment.org/D2944
Summary:
Fix memory leak
Delimiter string is being saved using
eina_stringshare_replace without any del or free
when object is deleted.
@fix
Test Plan: NA
Reviewers: cedric, tasn, herdsman
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D3201
Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
Otherwise there would be conflicts in certain circumstances.
This also requires adding const on many existing functions,
and similar work is necessary in Elementary.
@fix
Deep down internally there was already a name, but no API could
really set it properly.
Here Edje will set the name of the filter based on the part name
or the data item name if relevant.
This creates the new interface
Efl.Gfx.Filter
And the implementation is a mixin (evas_filter_mixin.c):
Evas.Filter
All the filter rendering code has now been moved to this
new file. TODO: Merge image filtering.
This should preserve ABI stability with earlier versions of
edje_cc while still providing more advanced control over
proxy bindings for evas filters from EDC.
Also fix proxy binding for filters.
@feature
Jean-Philippe Andre
Cedric BAIL
Carsten Haitzler
Chris Michael
Guilherme Iscaro
Youngbok Shin
Tom Hacohen
Minwoo, Lee
Subodh Kumar
Vitalii Vorobiov
Daniel Kolesa