The missing free was only for one of the paths to the end label. The other one
freed it before already. Take this into account to avoid the double free.
To avoid people getting confused over this code, like myself, add a comment
to clarify that only the elements need to be set to zero while the payload
does not matter here.
If a slave is still alive during shutdown, then it will be killed
and the dead callback should be called. This would then trigger
a request failed callback on invalid data.
But there's no point in calling the dead_cb since we're shutting
down already.
glibc double free happens when shutting down cserve, while
requests are being processed. this is because the entry is
being deleted from the hash but the cancel request fails
(so it should be deleted again from the hash)
Another patch going to fix the double free.
Test 23: load lots of scaled images and draw them
Problem: cs2 fails to load more images at some point. (root cause to be fixed!)
After that, client expects some data, but there is none, so it segfaults.
The "compose" component of the ecore input events must be translated to
the correct value when CTRL keys are pressed. This implementation was
copy & pasted from the xcb backend.
Also changes the default window type to ECORE_WL_WINDOW_TYPE_TOPLEVEL as it is
the most used case. If someone needs another type, set it manually.
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
extn_data_size is not equal to extn_data_off,
current class data size and data offset must be substracted first
elementary_test bubble peak usage goes from 13.7 MiB to 12.5 MiB
This reverts commit 9473c4a9a5.
This commit is not correct. It just goes through every font in the
fontset, which is usually all the fonts in the system and tries to get
their ascent/descent and by that getting the max ascent/descent. This
won't work nicely.
The solution is to properly fix textblock, if you think there's
something lacking (I.e in the same way I did text).
However, my changes did not change previous behaviour, but were just
wrappers, so I don't see why extra changes would be needed. Please
elaborate.
I'm reverting this because:
1. I don't think it's correct.
2. It severly broke e in some cases (reference:
http://www.enlightenment.org/ss/e-51b5d5e98cd387.04568822.png ).
- Add code to allow Async Render with Wayland_Egl
(NB: Currently disabled. Evas_Textgrid borks when this is enabled)
- Create ecore_wl_window first so we can get surface and pass to evas.
- During show, we no longer need to pass the surface to evas.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
AND get the surface for it...so that we can pass the surface directly
to the evas engine.
NB: Works. nothing broken.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
- In common_render_updates_process, calls to window_damage will
Accumulate the damage, so add a call After to push damages.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
- Setenv for egl_platform before we init gl/egl, etc
- Ignore check for NULL surface to indicate destroying (this is all
not need now).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
When destroying window, also destroy animator callback.
Change ecore_wl_window_damage function to Only do damage.
Add separate ecore_wl_window_commit function to push those accumulated
damages.
In ecore_wl_window_surface_create, if the window already has a
surface, return that ... Not Null.
During window show, only create a shell_surface if it does not Already
exist.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
keypress, but rather Only on they actual modifier event.
Add some printf debugging for Non-Printables like Ctrl, etc
Signed-off-by: Chris Michael <cp.michael@samsung.com>
If this function is called recursively, it will free the list of post
callbacks before the list stops being used, which will cause a segfault.
The only place where this seems to happen is on
https://phab.enlightenment.org/T124, probably due to the extensive mouse
events which are launched in an unexpected way.
This bug started happening after commit
164cc07237395f8fe6efa465e4f0c0c4863f78ed, where the
_elm_scroll_page_x_get() started being called by a post_event callback,
and forcing an edje recalc. This recalc triggered another post_event
callback, thus causing the mentioned segfault.
If there's a better way to prevent this from happening, please change
the mentioned code.
Additional fixes to cc4cf77865.
There were two problems:
1. Usage of max_ascent instead of ascent.
2. Initialization of the ascent value according to the first font,
instead of 0 (as the first font might not even be used).
comparisons.
Add comment about NULL surface being passed in.
Fix segfault when we close windows...uartie, this one's for you ;)
Signed-off-by: Chris Michael <devilhorns@comcast.net>
Account for the framespace offset on perspective fx and fy of each map
point.
This is just a workaround because I could not understand exactly why it
is needed. When no perspective is used, the viewport is set to the size
of the surface, and each map point seems to be used as is, with no
adjustment being needed.
However, when the map is not flat (perspective is used), a more complex
math is used to find the glViewport size. It ends up being bigger, but
there are some offsets used to compensate that, gc->shared->ax/y, which
are added to each vertex before pushing them. Thus, it seems to me that
the framespace offset should not be added to them, but things are weird
if this is not done.
Anyway, I just added those offsets, and it should not impact on gl_x11
since it's not using framespace, and software engines don't seem to
implement perspective anyway, so it all should be fine.
If anyone understands better what's going on here, please make a proper
fix or at least contact me to explain the problem, and maybe I can fix
it by myself.
It seems some images are fully loaded even when we request
only preload. The load handler will take care of both cases anyway.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
This is an astonishing bug, I wonder since how long it has been there. It
is basically due to the use of void * and a wrong cast. Type checking is
clearly useful, let's use it more !
Because arc doesn't like me much at the moment here's a simple patch to
fix the gif loader for the latest libgif version. DGifOpen now takes a
error param as its final argument, if the param is not NULL it will
return a error code if it fails. Similar to another patch i sent a few
months back. The patch has been tested on OBS for old and new versions.
or tripple clicked.
selection should be allowed when select_allow is true whatever
select_mode is.
Reproduce step is below.
1. ELM_THEME=default-desktop ELM_PROFILE=standard elementary_test
--test-win-only "Entry Scrolled
2. double click on password entry.
This causes the textblock layouting problem.
It should be checked when does it really render_pre() called.
If someone knows the reason of this, please ping me.
eina_str_convert() is GUARANTEED to break when doing any such encoding (eg. UTF16->UTF8). I don't know who added the original function, but this is very bad, and we should almost certainly deprecate eina_str_convert() so people are not surprised when they are unable to convert strings as expected.
eldbus_address_connection_get() and eldbus_private_address_connection_get()
are similar to the respective _connection_get() counterparts, but enables
users to connect to buses other than system or session.
The new type introduced for those connections is ELDBUS_CONNECTION_TYPE_ADDRESS
and they require an additional address parameter, which will be passed to
dbus_connection_open().
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
This is actually a split on the _eldbus_connection_unref() function, that
will be called either when the refcount reaches 0 or to force the deletion
of an Eldus_Connection.
The second use case will be contemplated in a following patch.
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
Set surface_user_data in attach function to be the buffer we are
attaching (NB: Expected by some compositors like weston).
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Daniel Juyung Seo
Carsten Haitzler
Chris Michael
Stefan Schmidt
Jean-Philippe Andre
Cedric Bail
Rafael Antognolli
Chris Michael
Jiyoun Park
Daniel Willmann
Jérémy Zurcher
Eduardo Lima (Etrunko)
Tom Hacohen
José Roberto de Souza
Boris Faure
Daniel Willmann
ChunEon Park
Ederson Desouza
Daniel Zaoui
Simon
Ryuan Choi
Mike Blumenkrantz
Jihoon Kim
