Summary:
Coverity reports that va_end is never called for p_list when we error
out of this function. This patch adds a missing va_end before we error
out of here.
Fixes Coverity CID1399080
@fix
Depends on D8315
Reviewers: raster, cedric, q66, zmike, bu5hm4n, stefan
Reviewed By: q66
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D8316
Getter are usually not modifying there object. This is going to put a strong
limit on what a getter property for MVVM is, as it will prevent any side
effect on getting a property from a View.
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D7969
Summary:
the DFS tree walk was accidently stopped by a too early return
statement. We should only return if we found a reflection entry, if not,
then we should continue our search
Depends on D7996
Reviewers: cedric, zmike, q66, segfaultxavi
Reviewed By: cedric, zmike
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7997
this is done inorder to ensure that noone ever thinks of creating theire
own app/loop object.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7982
Summary:
efl_alive_get should be a method defined in efl_object.eo, protected by
EFL_BETA_API_SUPPORT while that class is still @beta.
Instead, it is defined in Eo.h, so the BETA guard has to be defined manually.
This was causing some unsighly warnings during "make" and "make check"
(implicit declaration of function ‘efl_invalidating_get’)
Test Plan: make && make check do not throw that warning anymore
Reviewers: zmike, cedric
Reviewed By: zmike
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7958
Summary:
Instead of surrounding all the #include "*.eo.h" lines in Efl.h
with #ifdef EFL_BETA_API_SUPPORT, include these files unconditionally, but mark
all classes as @beta in the eo files.
This will allow taking them out of beta one by one as we deem them stable enough.
Otherwise, the current procedure involves moving the #include line out of the
ifdef block, which is cumbersome and messes include order.
Depends on D7950
Fixes T7692
Test Plan: Nothing changes
Reviewers: zmike, bu5hm4n, cedric
Subscribers: #reviewers, #committers
Tags: #efl
Maniphest Tasks: T7692
Differential Revision: https://phab.enlightenment.org/D7951
Summary:
This declares the main part of Eo.h as stable, except the domain api for
now. We don't have a user of that API for now, nor do we have bindings
supporting threads, lets wait with this bit until we have bindings for
this, so we can test it.
Further more, this does not stabilize the composition API of
efl_object.eo.h, reason for this is, that we might want to overthink the
solution we have with events, as we might want to forward events per
default.
Depends on D7931
Reviewers: stefan_schmidt, cedric, zmike, segfaultxavi
Reviewed By: cedric
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7932
As in the previous commit explained, we want to get rid of class
functions in eo, and make them just c functions right away.
This commit removes the class parameter from the eo_class_function_set
call, and adjusts the tests to not depend on class functions anymore.
Class functions are now not tested anymore, tests that used them as a
way to test *things* are adjusted to test them now with object
functions, tests that just tested the working of class functions are
dropped.
This fixes T7675.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7902
Until this commit eo did class functions as part of the vtable, which
enabled those functions to be overwritten in classes inheriting another
class. However in task T7675 we decided that this is not really good for
bindings, as most OOP languages do not support this sort of feature.
After this commit eolian realizes class function completly outside of
the vtable, the c-symbol that is the class funciton is now just directly
redirecting to a implementation, without the involvement of the vtable.
This also means a change to the syntax created by eo:
Calling before:
class_function(CLASS_A);
Calling after:
class_function();
Implementation before:
class_function(const Eo *obj, void *pd) { ... }
Implementation after:
class_function(void) { ... }
This fixes T7675.
Co-authored-by: lauromauro <lauromoura@expertisesolutions.com.br>
Reviewed-by: Daniel Kolesa <daniel@octaforge.org>
Differential Revision: https://phab.enlightenment.org/D7901
this checks that eo-dbg is also working in the same manner as normal eo
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D7911
this is important, as basically every single class etc. can be a
EFL_OVERRIDE_CLASS. This makes efl_override work with eo_debug.
Differential Revision: https://phab.enlightenment.org/D7910
this while loop is actaully iterating obj->xref, therefore we should not
take the pointer of data_xrefs.
Differential Revision: https://phab.enlightenment.org/D7908
its not required anymore, our wonderfull new syntax solves the original
problem it was trying to solve.
fixes T7565
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7893
this adds support in eo to generate a reflection API. To get the actaul
reflection to the klass, the API efl_class_reflection_table_set needs to
be called, the table in the end can be generated by eolian. Reflection
API is inherited by the extended class. This means, if you have two
reflection tables, first, the most upperst is called, then the next
lower one is called.
For now this API accepts NULL setter or getter, and will ignore them
silently when they are called.
fix T7681
Differential Revision: https://phab.enlightenment.org/D7879
The next commit will bring support for something like reflection. This
commit prepares the whole tree for getting another argument in
efl_class_functions_set.
ref T7681
Differential Revision: https://phab.enlightenment.org/D7882
remove the whitespaces from this file. Every editor in this world
removes unneccessay whitespaces at the end of a line. So every time i am
touching this header, i need to get rid of them, which is super
annoying. I know we are used to *not* doing such patches, but this is
just a waste of time, getting them out of the header every and every
single time.
Differential Revision: https://phab.enlightenment.org/D7880
the function has changed, but the macro was not adjusted.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7878
this enables an app or a platform to add an override for a given class,
then return a different object when that class is created. the benefit is
that a class can be internally customized by the app without needing to
modify upstream versions of that class
@feature
fix T7516
Reviewed-by: Marcel Hollerbach <marcel-hollerbach@t-online.de>
Differential Revision: https://phab.enlightenment.org/D7702
mro only contains special interfaces, not in general all, in order to
have a working function, while NOT having a bad performance. In order to
achive that we just copy the code that is done on efl_isa for a object.
which is a linear walk of a list
ref D7857
Reviewed-by: SangHyeon Jade Lee <sh10233.lee@samsung.com>
Differential Revision: https://phab.enlightenment.org/D7860
Memory allocation is checked for the given domain's table data like
other cases for main and shared table data.
Differential Revision: https://phab.enlightenment.org/D7810
This reduce the amount of time efl_event_callback_call(EFL_EVENT_NOREF) is triggered by a factor of 10.
Reviewed-by: Derek Foreman <derekf@osg.samsung.com>
Differential Revision: https://phab.enlightenment.org/D7744
EFL_EVENT_SPECIAL_SKIP has been defined to 1 for a long time and there
wasn't a way to disable it at all. There is no reason to not use it at
all, so here we go. This was default, now it is without useless code.
Reviewed-by: Derek Foreman <derekf@osg.samsung.com>
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D7743
efl_noref is unused, but will still trigger the full eo call for nothing.
This was the most expensive NOOP. EFL_EVENT_NOREF does the same thing and
we can even filter generating that call as we know when someone is listening.
So removing the function.
Reviewed-by: Derek Foreman <derekf@osg.samsung.com>
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D7742
this commit is a first step in order to add free functions to all the
types that are owned.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7627
Their was two different way to implement this, either like this with
a simple function that work on Efl_Class, or by a function on
Efl.Object. I leaned on the first one, but I could easily be convinced
it should be done on Efl.Object actually.
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D7441
So it can be references from the later coming definition of
callback_forwarder
ref D7532
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7542
Note: Their isn't any ability to do something like a static array of
events at the moment. It might lead to large memory being used when it
wouldn't be necessary. If that was the case, we could fix it, but it
would require a lot of dynamic hash operation I think.
Reviewed-by: Marcel Hollerbach <marcel-hollerbach@t-online.de>
Differential Revision: https://phab.enlightenment.org/D7482
If you are to trigger an event and just cancelled the one being processed, it would just
not propagate any event while it should. The code looks like more like it was forgotten line
or something.
Reviewed-by: Marcel Hollerbach <marcel-hollerbach@t-online.de>
Differential Revision: https://phab.enlightenment.org/D7481
This reverts commit 9b5155c9f1.
For now lets revert this, this breaks copy and paste, further more it
has the potential to break a lot more things, as eio_model tends to use
efl_loop_promise new, and then eina_promise_data_set, which is
explicitly forbidden.
This fixes crashing terminology instances.
I am not sure this is the right way to do it as binding would have to likely
to bind it manually.
Reviewed-by: Lauro Neto <Lauro Moura <lauromoura@expertisesolutions.com.br>>
Differential Revision: https://phab.enlightenment.org/D7492
This build was never complete and also was not maintained probebly.
It is also dropped in favour of meson which is cool, merged, works & is fast.
Differential Revision: https://phab.enlightenment.org/D7010
Summary:
In the case when you have multiple future in flight related to one object, you
couldn't use the previous version of efl_future_then. Now all function calls
take a void* pointer that allow multiple future to have their private data
request data accessible in all the callback.
This should not break released API as Eo.h is not released yet and so
was efl_future_Eina_FutureXXX_then.
Depends on D7332
Reviewers: felipealmeida, segfaultxavi, vitor.sousa, SanghyeonLee, bu5hm4n
Reviewed By: segfaultxavi
Subscribers: #reviewers, #committers
Tags: #efl
Maniphest Tasks: T7472
Differential Revision: https://phab.enlightenment.org/D7379
This new syntax separates the parent class from extensions, in
a familiar way to similar to e.g. Java. Since changing everything
at once is a lot of effort, implement it alongside for the time
being.
Summary: This is done to keep consistency with all the reference counting functions in Eo.
Reviewers: Hermet, smohanty
Reviewed By: Hermet
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7274
Summary:
This commit removes some clashes (i.e. names as classes and namespaces
at the same time). It'll avoid nested items that are either forbidden
(C#) or problematic (Python) in some languages.
Reviewers: segfaultxavi, bu5hm4n, felipealmeida
Reviewed By: segfaultxavi
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7260
Summary:
Check if the Eo to be replaced is NULL before calling efl_unref.
The documentation implies that the replaced object can be NULL, so this
check avoids unnecessary function calls and warning logs.
Add an EINA_SAFETY check in order to properly print an error message
when the function is used with an NULL storage pointer.
The documentation specifies that the first parameter can not be NULL.
So printing an error message should be better than silently returning.
Add a boolean return to the function that signalizes if the content of
the storage was changed. It is NOT an success/error flag, it is just a
simple helper to quickly test for a change in content.
This feature was inspired by eina_stringshare_replace that is used in
similar ways around the code.
Change the documentation to match the changes and to be more specific
about what is expected and how the arguments are treated.
Reviewers: raster, bu5hm4n, cedric, felipealmeida
Reviewed By: bu5hm4n
Subscribers: #reviewers, #committers, larryolj
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7256
This uses the meson/ninja depfile functionality + eolian to make
sure proper dependencies between generated files and .eo files
are managed, to ensure consistent re-generation of all generated
files that are affected upon .eo file modification.
For custom rules with multiple outputs, Ninja currently does not
support depfiles. Therefore, split those into two custom rules
so that the depfiles functionality can be enabled. While this
is ugly and slows down the process a little by having to invoke
Eolian twice instead of once, it has to be done and it's still
better than what we had in Autotools anyway.
Differential revision: D7187
Fixes T6700.
a new shiny buildtool that currently completes in the total of ~ 4 min..
1 min. conf time
2:30 min. build time
Where autotools takes:
1:50 min. conf time
3:40 min. build time.
meson was taken because it went quite good for enlightenment, and is a traction gaining system that is also used by other mayor projects. Additionally, the DSL that is defined my meson makes the configuration of the builds a lot easier to read.
Further informations can be gathered from the README.meson
Right now, bindings & windows support are missing.
It is highly recommented to use meson 0.48 due to optimizations in meson
that reduced the time the meson call would need.
Co-authored-by: Mike Blumenkrantz <zmike@samsung.com>
Differential Revision: https://phab.enlightenment.org/D7012
Depends on D7011
Summary:
This commit add null check on _efl_object_invalidate
In EO_OBJ_POINTER, if obj_id is null, obj can also be null.
Test Plan: N/A
Reviewers: cedric, zmike
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7068
The function simply checks if the object is finalized and not
started to be invalidated or invalidated.
Differential Revision: https://phab.enlightenment.org/D6723
Summary:
there is now invalidated & invalidating.
invalidated returns true when all children are invalidated, and the
object is / was requested to be invalidated.
invalidating return true when the object is called to be invalidated but
not all children are invalidated yet. However, the object is garanteed
to be invalidated in near future.
Reviewers: zmike
Reviewed By: zmike
Subscribers: cedric
Tags: #efl, #do_not_merge
Differential Revision: https://phab.enlightenment.org/D6722
Summary:
efl_new(CLASS), i.e., without initializers, never worked before.
The problem is the dangling comma: efl_add_ref(klass, NULL, );
Fortunately the C preprocessor concatenation operator # # has a special
case just for this occasion: When you do A ## B, and A is a comma and B
is empty, it removes also A.
https://gcc.gnu.org/onlinedocs/cpp/Variadic-Macros.html
This trick is already in use in several other EFL headers using variadic macros.
Reviewers: zmike, bu5hm4n, devilhorns, herdsman
Reviewed By: zmike
Subscribers: cedric, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6553
Summary:
a common use case is for a class to be constructed during a thread+mainloop
sync (e.g., ecore_thread_main_loop_begin() ecore_thread_main_loop_end())
and then naturally destroyed in the main thread during shutdown
ref 023a9ca2ee
Reviewers: bu5hm4n, devilhorns
Reviewed By: bu5hm4n
Subscribers: cedric, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6472
recent changes added unlocks for these mutexes in some other places, so only
do the unlocks where necessary
fix T7020
Depends on D6356
Differential Revision: https://phab.enlightenment.org/D6357
Summary:
unreffing too often is not good. However, its definitly not a critical
error. Most of the wikipedia articles (*) that are speaking about ciritcal or
fatal errors are highlighting that no reasonable recovery can be done.
In our case a recovery is not even needed, due to our pointer safety we
dont need to recover anything since the refcounting safed us from doing
something stupid.
ref T6987.
Depends on D6384
Reviewers: zmike, segfaultxavi, devilhorns
Reviewed By: zmike
Subscribers: cedric, #committers
Tags: #efl
Maniphest Tasks: T6987
Differential Revision: https://phab.enlightenment.org/D6385
Summary:
The 'if' block should only be executed when the string contains the colon and
something else behind, but sscanf cannot be used in this case.
If the string contained no colon, the following line with strchr(s, ':')
returns NULL and everything explodes.
Test Plan: eo_debug -l now works for me without segfaulting.
Reviewers: bu5hm4n, zmike, devilhorns, q66
Subscribers: cedric, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6347
Summary:
calling the constructor and deconstructor from different threads
causes issues with mempool deallocation, so ensure that the class is
always initialized in the main thread for safety, e.g., call the
SOME_NAMED_CLASS macro during init to instantiate the class
fix T7003
Reviewers: bu5hm4n, devilhorns
Reviewed By: bu5hm4n
Subscribers: cedric, #committers
Tags: #efl
Maniphest Tasks: T7003
Differential Revision: https://phab.enlightenment.org/D6332
Summary:
this is no functinal change, we safe the calling to efl_parent_set.
Nothing else, the base implementation is initializated with NULL as
parent, and efl_parent_set with NULL would have returned in the first
few lines.
fix T7032
Reviewers: devilhorns, zmike
Reviewed By: zmike
Subscribers: cedric, #committers, zmike
Tags: #efl
Maniphest Tasks: T7032
Differential Revision: https://phab.enlightenment.org/D6327
The class's Eo constructor can return a different object, which makes
the efl_add return that object instead. However, a bug was not
initializing the internal reference count when a different object was
returned.
Summary:
ensure that this is always called for the first time from the
main thread to avoid triggering asserts during shutdown
ref T7003
Reviewers: bu5hm4n, devilhorns
Reviewed By: bu5hm4n
Subscribers: cedric, #committers
Tags: #efl
Maniphest Tasks: T7003
Differential Revision: https://phab.enlightenment.org/D6272
this check caused a leaked reference.
Eo objects are having two reference counters a internal and external
one. The external one can be manipulated via efl_ref / efl_unref. The
internal one can be manipulated via _efl_ref _efl_unref.
The external reference counter is keeping a internal reference by the
time the external counter is > 0. When the external counter reaches ==
0 this internal reference is given up with the _efl_unref call in
eo.c:1928.
However, checking unref_compensate in the block in line 1950 leads that
to the leak that this internal reference is not given up at the pointer
user_refcount reaches exactly 0. This check also does not prevent
anything, the object is kept alive anywayys as the efl_unref method
keeps its private internal reference.
This lead to leaks in efl_device_* classes, parts have not been
destructed correctly.
Differential Revision: https://phab.enlightenment.org/D6252
Summary:
this function calls _efl_ref in the beginning, so _efl_unref them in the
end. This fixes bugs where wrong evas_object_del is called to often,
which would lead to the internal refcount rising instead of getting to
0.
Reviewers: cedric, zmike, stefan_schmidt, q66
Reviewed By: zmike
Subscribers: cedric, #committers, zmike
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6246
Summary:
this brings down the first frame time from 2.0 sec. to something at 0.7
sec.
ref T6983
Reviewers: zmike
Reviewed By: zmike
Subscribers: YOhoho, cedric, #committers, zmike
Tags: #efl
Maniphest Tasks: T6983
Differential Revision: https://phab.enlightenment.org/D6244
Summary:
the object is invalidated AFTER the parent is lost, (thats what the doc
says). Returning invalidated as true when the parent is still present is
thus not correctly.
Depends on D6222
Reviewers: cedric, zmike, JackDanielZ
Subscribers: cedric, #committers, zmike
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6223
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Making wref disapear before the last event require additional work to use
them during the FREE event. As I think we should introduce a wref that
disapear after the invalidate state, it makes sense to move that to a later
stage. Especially because their is no specific event related to them
being set to NULL.
Allowing override of efl_del allow for messing with object life cycle which can create
problem for binding where an object was expected to be invalidated, but isn't really.
The interface efl_part_get should not be directly called from C, but the efl_part
wrapper should. It rely on efl_noref to properly destroy the object. Binding can
control the lifecycle of the reference the way they want by either calling the
wrapper or efl_part_get directly. It also means that the ugly ___efl_auto_unref_set
doesn't need to be exposed outside of EFL anymore.
Differential Revision: https://phab.enlightenment.org/D6098
Summary:
All events must have a type now, otherwise bindings don't know how to handle
the event_info field.
Most of the missing event types were actually "void" (no event_info present).
Some struct definitions had to be moved to eo instead of h files, so they
are available to bindings. Some have not, and are marked with FIXME.
Some namespaces have been fixed (like Efl_Event_Cb -> Efl.Event_Cb).
In general, there are hundreds of changed files, but mostly to add a type which
was not present before, so there's no harm done.
Also, A lot of FIXMEs have been added which should be, like, fixed.
For example, some events can send different types of event_info, which is
very inconvenient (and error prone).
Test Plan: make with c# bindings works, make check and make examples work too.
Reviewers: cedric, q66, lauromoura
Subscribers: zmike
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6169
These events have been removed from efl_object.eo. Bindings can
still use the EFL_EVENT_CALLBACK_ADD and EFL_EVENT_CALLBACK_DEL
events from C, but they won't be available to the bound language.
The rationale is that bound languages probably will have their
own way to handle callbacks and C function pointers will mean
very little to them. For example, the binding code could route
the native callback through a dispatcher so the received
function ptr would always be that of the dispatcher method.
Closes D6179.
Summary:
This is just a wrapper around efl_add_ref() with a NULL parent.
It only simplifies a bit object creation when no parent is desired, but it
simplifies a lot the explanation of the refcounting concepts (and tutorials).
Reviewers: cedric, bu5hm4n
Reviewed By: cedric, bu5hm4n
Subscribers: bu5hm4n, zmike
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D6024
This changes a lot of things all across the EFL. Previously,
methods tagged @const had both their external prototype and
internal impl generated with const on object, while property
getters only had const on the external API. This is now changed
and it all has const everywhere.
Ref T6859.
Summary:
This commit add null check on __efl_auto_unref_set.
In EO_OBJ_POINTER, if obj_id is null, obj can also be null.
Test Plan: N/A
Reviewers: woohyun, kimcinoo, cedric
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D5869
Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
Summary:
This will avoid infinite loops and errors when the parent tries
to orphan an invalidated child.
Fixes T6780
Test Plan: Run `make check`
Reviewers: cedric
Maniphest Tasks: T6780
Differential Revision: https://phab.enlightenment.org/D5839
Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
This is just a first step. All user of destructor should be updated to
move the code that rely on their efl_parent and on efl_provider_find to
invalidate. Then we will be able to change the way efl_add and efl_del
work to properly refcount things. efl_noref won't be triggered at the
moment until both efl_parent_set(obj, NULL) and the last user ref are
set to NULL. This is not what we want, but due to how user refcount is
accounting parent at the moment, until all the code is move to rely
on invalidate we can not fix this.
so the MAIN loop is actually an efl.app object. which inherits from
efl.loop. the idea is that other loops in threads will not be efl.app
objects. thread on the creator side return an efl.thread object.
inside the thread, like the mainloop, there is now an efl.appthread
object that is for all non-main-loop threads.
every thread (main loop or child) when it spawns a thread is the
parent. there are i/o pipes from parnet to child and back. so parents
are generally expected to, if they want to talk to child thread, so
use the efl.io interfaces on efl.thread, and the main loop's elf.app
class allows you to talk to stdio back to the parent process like the
efl.appthread does the same using the efl.io interfaces to talk to its
parent app or appthread. it's symmetrical
no tests here - sure. i have been holding off on tests until things
settle. that's why i haven't done them yet. those will come back in a
subsequent commit
for really quick examples on using this see:
https://phab.enlightenment.org/F2983118https://phab.enlightenment.org/F2983142
they are just my test code for this.
Please see this design document:
https://phab.enlightenment.org/w/efl-loops-threads/
The cache had a generation count, removing this check breaks eo suite.
EO is designed to tolerate init/shutdown cycles (as long as
dlopen/dlclose isn't involved).
Note: the ugly goto are useless as GCC -O2 understands EINA_(UN)LIKELY
and deals with it as expected (just look at the asm produced).
See 34d9f20706
This makes eo print a WRN message in case a function is called on NULL.
efl_del is an exception to this rule (implemented in a hackish way, I
admit). I don't know any language or object model where using a null
object doesn't result in an exception or crash, except EO. In any case,
calls to null are invalid.
The next commits will resolve most warnings for EFL.
Ref T6326
Summary:
ecore_evas: remove debug
eina: unregister log level when done with
Fixes a constant memory leak.
eina: introduce EINA_HOT and EINA_COLD
These attributes respectivelly expand to __attribute__ ((hot)) and
__attribute__ ((cold)) when available. They allow to mark functions are
being hot/cold (frequently used or not) as well as to qualify labels
within a function (likely/unlikely branches).
eo: speed-up generated calls by removing call cache
The call cache needed to by thread-local, to avoid concurrency issues.
Problem with TLS is that is adds an extra overhead, which appears to be
greater than the optimization the cache provides.
Op is naturally atomic, because it is an unsigned integer. As such, it
cannot be tempered with while another thread is reading it. When
entering the generated function, the first operation done is reading
'op'. If we have concurrency, we will have access sequences returning
either EFL_NOOP or a VALID op, because 'op' is not set until the very
end of the function, when everything has been computed. As such, we now
use the 'op' atomic integer to instore a lock-free/wait-free mechanism,
which allows to drop the TLS nature of the cache, speeding up the access
to the cache, and therefore making functions execute faster.
We don't test anymore the generation count. This can be put as a
limitation. If means that if you call efl_object_shutdown() and
re-initialize it later with different data, opcodes will be invalid.
I am not sure there is any usecase for this to ever happen.
We could move all the caches in a dedicated section, that can be
overwritten after a call to efl_object_shutdown(), but I am not sure it
will be very portable.
Benchmark: mean over 3 executions of
ELM_TEST_AUTOBOUNCE=100 time elementary_test -to genlist
```
BEFORE AFTER
------------------------------------------------------------
time (ns) 11114111647.0 9147676220.0
frames 2872.3333333333335 2904.6666666666665
time per frame (ns) 3869364.6666666665 3149535.3333333335
user time (s) 11.096666666666666 9.22
cpu (%) 22.666666666666668 18.333333333333332
```
Ref T6580
Reviewers: raster, cedric
Subscribers: cedric, jpeg
Maniphest Tasks: T6580
Differential Revision: https://phab.enlightenment.org/D5738
we can't sensibly use things like massif to track memory if we bypass
itr with mmaping -1 fd anonymous memory... so if built with valgrind
support and running under valgrind, use malloc/calloc and free so
these tools actually do something useful for these bits of memory.
Triggered after (almost) complete destruction of the object.
Not called "deleted" because the other event is already "del".
I don't like "destruct" much but this follows the terminology of
"constructor" / "destructor".
@feature
Christopher Michael
Cedric BAIL
Marcel Hollerbach
Jaehyun Cho
Xavi Artigas
Mike Blumenkrantz
Lauro Moura
Cedric BAIL
Carsten Haitzler
Marcel Hollerbach
Daniel Kolesa
Vitor Sousa
Yeongjong Lee
junsu choi
Felipe Magno de Almeida
Daniel Kolesa
Jean Guyomarc'h
Vincent Torri
Jean-Philippe Andre