If an idler created a recursive main loop (just called
ecore_main_loop_begin()), then this recursive main loop should
continue to process idlers from there and on, thus idler_current was
added. When going back from recursion, the current iterator should be
updated properly.
This patch also fixes the deletion of idlers from recursive main loops
by reference counting them. This way, the node will not be free()d
inside inner loop cleanups and then crash when going back to outer
loop.
The following test case used to crash but not anymore:
#include <Ecore.h>
#include <Eina.h>
static int _log_dom;
#define INF(...) EINA_LOG_DOM_INFO(_log_dom, __VA_ARGS__)
static Ecore_Idler *handle;
static int idler(void *data)
{
INF("idler");
return 1;
}
static int cb2(void *data)
{
INF("cb2 - delete cb1 handle");
ecore_idler_del(handle);
ecore_main_loop_quit(); /* quits inner main loop */
return 0;
}
static int cb1(void *data)
{
INF("cb1: begin");
INF(" add cb2");
ecore_idler_add(cb2, NULL);
INF(" inner main loop begin (recurse)");
ecore_main_loop_begin(); /* will it crash due ecore_idler_del(handle)
* inside cb2()? It used to!
*/
INF("cb1: end");
ecore_main_loop_quit(); /* quits outer main loop */
return 0;
}
int main(void)
{
ecore_init();
_log_dom = eina_log_domain_register("test", EINA_COLOR_CYAN);
/*
* Creating a new main loop from inside an idler callback, and inside
* this new (inner) main loop deleting the caller callback used to
* crash since the handle would be effectively free()d, but when the
* recursion is over the pointer would be used.
*/
INF("main: begin");
handle = ecore_idler_add(cb1, NULL);
ecore_idler_add(idler, NULL);
ecore_main_loop_begin();
INF("main: end");
return 0;
}
SVN revision: 46406
We should start doing unit-test for ecore, accumulating these
problems. Follows the test case:
#include <Ecore.h>
#include <Eina.h>
static int _log_dom;
#define INF(...) EINA_LOG_DOM_INFO(_log_dom, __VA_ARGS__)
static int quiter(void *data)
{
INF("quit!");
ecore_main_loop_quit();
return 1;
}
static int idler(void *data)
{
INF("idler");
return 1;
}
static int cb1(void *data)
{
INF("cb1");
ecore_timer_add(0.0, quiter, NULL);
return 0;
}
int main(void)
{
ecore_init();
_log_dom = eina_log_domain_register("test", EINA_COLOR_CYAN);
/*
* Create a main loop with just idlers, there is a special case
* for just idlers without timers in ecore.
*
* From idler, add a timer that quits the application. It should
* always quit.
*
* If it does not quit, then there is a bug of new timers not
* being immediately detected and system never exits idle.
*/
INF("main: begin");
ecore_idler_add(cb1, NULL);
ecore_idler_add(idler, NULL);
ecore_main_loop_begin();
INF("main: end");
return 0;
}
SVN revision: 46405
if using ecore_main_loop_begin() multiple times (reentrant/recursive)
with glib doing threads, then it would deadlock since the same thread
would get the lock it already have.
multiple ecore_main_loop_begin() is required to implement WebKit's
alert/confirm/prompt dialogs since there is no async reply with
callbacks, rather one must return the value.
By: Lucas de Marchi
SVN revision: 46361
Subject: [E-devel] [PATCH] evas: compile out unused dither tables
Don't include the rather large dither tables if small dithering code
is enabled during ./configure stage.
SVN revision: 46353
* OpenGL ES : 2.0
* OpenGL desktop : >= 1.4
If OpenGL ES 3.* is out, this code must be fixed
If there are problems with the minimal version on desktop,
we will have to increase the minimal version.
Question: should we check GLSL version on desktop ?
SVN revision: 46309
* must disconnect connected callbacks, particularly those to
canvas. The object we previously connect will die anyway, but
canvas continues alive, dispatching its
EVAS_CALLBACK_CANVAS_FOCUS_IN and EVAS_CALLBACK_CANVAS_FOCUS_OUT,
causing nasty segmentation faults!
* must call _edje_clean_objects() *AFTER* the entry is shut
down. Otherwise ed->evas will be NULL and
evas_event_callback_del_full() will fail. I left extra checks on
those, since this call will return the given data (in our case
"ed") and NULL when callback connection was not found.
* flag existence of entries and if they were already initalized and
shutdown before, avoid redoing such work.
This fixes a stupid crash that bugged editje for a while now.
SVN revision: 46263