probably a leak fix in ecor-eglib support - doesnt cause any issues in

my testing, but documentation for g_static_mutex's isnt that great to
let us know what to do with a static mutex... considering it's...
static... pending respons from reporter for backport



SVN revision: 66360
This commit is contained in:
Carsten Haitzler 2011-12-20 05:57:57 +00:00
parent bf9467a712
commit 71f88a1a13
2 changed files with 9 additions and 2 deletions

View File

@ -421,3 +421,7 @@
2011-12-16 Mike Blumenkrantz
* Fix possible 0 byte allocation in ecore-x
2011-12-20 Carsten Haitzler (The Rasterman)
* Fix probable leak for g_static_mutex's on some architectures

View File

@ -190,7 +190,9 @@ _ecore_glib_select(int ecore_fds,
int ret;
if (g_main_context_acquire(ctx))
g_mutex_lock(mutex);
{
if (mutex) g_mutex_lock(mutex);
}
else
{
if (!_ecore_glib_cond)
@ -203,8 +205,9 @@ _ecore_glib_select(int ecore_fds,
ret = _ecore_glib_select__locked
(ctx, ecore_fds, rfds, wfds, efds, ecore_timeout);
g_mutex_unlock(mutex);
if (mutex) g_mutex_unlock(mutex);
g_main_context_release(ctx);
g_static_mutex_free(&lock);
return ret;
}