Jasper St. Pierre [Sat, 1 Feb 2014 23:36:41 +0000 (18:36 -0500)]
xdg-shell: Add set_margin request
This is used to figure out the size of "invisible" decorations, which we'll
use to better know the visible extents of the surface, which we can use for
constraining, titlebars, and more.
Jasper St. Pierre [Sat, 1 Feb 2014 23:17:34 +0000 (18:17 -0500)]
xdg-shell: Add a "delete" event
This is equivalent to WM_DELETE_WINDOW request under X11, or equivalent
to pressing the "close" button under CSD. Weston currently doesn't have
a compositor-side way to close the window, so no new code is needed on
its side.
Kristian Høgsberg [Thu, 6 Feb 2014 06:00:59 +0000 (22:00 -0800)]
build: Add missing AM_CFLAGS to client target CFLAGS
Kristian Høgsberg [Thu, 6 Feb 2014 05:36:04 +0000 (21:36 -0800)]
xdg-shell: Drop edges arg from xdg_surface configure event
Jason Ekstrand [Thu, 6 Feb 2014 03:38:27 +0000 (21:38 -0600)]
nested-client: Fix build error
Ander Conselvan de Oliveira [Wed, 29 Jan 2014 16:47:54 +0000 (18:47 +0200)]
compositor: Move view repositioning logic into shell
Remove the listener for output destroy from weston_view and instead
iterate views owned by the shell in its own output destroy listener.
This simplifies the code a bit since keeping the view listening for the
destroy on the right output was a bit complicated. This also removes the
function pointer output_destroyed from weston_view. The only user for it
was desktop shell, but now this is all handled in shell.c.
Ander Conselvan de Oliveira [Wed, 29 Jan 2014 16:47:52 +0000 (18:47 +0200)]
compositor: Remove weston_output::move_signal
Since that signal is per output, it is necessary to track in which
output a view is in so that the signal is handled properly.
Instead, add a compositor wide output moved signal, that is handled by
the shell. The shell iterates over the layers it owns to move views
appropriately.
Emilio Pozuelo Monfort [Thu, 30 Jan 2014 12:49:39 +0000 (13:49 +0100)]
dim-layer: fix dimming for unfocused surfaces
Unfocusing a surface should dim it when dim-layer is enabled,
but this got broken in commit 83ffd9.
Emilio Pozuelo Monfort [Mon, 3 Feb 2014 15:57:27 +0000 (16:57 +0100)]
compositor-rpi: Fix input initialization
The input initialization code assumes the outputs have already
been initialized; thus create the outputs first. This fixes a
segfault upon startup. It is also what the drm and fbdev backends
do.
Bryce W. Harrington [Mon, 3 Feb 2014 21:37:06 +0000 (21:37 +0000)]
README: Fix typo; add link to testing docs
Signed-off-by: Bryce Harrington <b.harrington@samsung.com>
Jason Ekstrand [Wed, 5 Feb 2014 03:36:39 +0000 (21:36 -0600)]
gl-renderer: Use eglSwapBuffersWithDamageEXT when available
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Jason Ekstrand [Wed, 5 Feb 2014 03:36:38 +0000 (21:36 -0600)]
gl-renderer: Track border damage and only repaint borders on an as-needed basis
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Kristian Høgsberg [Thu, 6 Feb 2014 00:53:43 +0000 (16:53 -0800)]
window: Delay scheduled redraws if we start maximize or fullscreen protocol
When we set the fullscreen flag, we have to wait for the corresponding
configure event and then attach a buffer of that size to indicate
that we've successfully gone fullscreen/maximized.
Without this patch, we can schedule a redraw and go through with it after
setting maximize/fullscreen and end up attaching a buffer of the wrong size.
In practice, what happens is that pressing the maximize button triggers
setting maximized, but also triggers a redraw to paint the maxmize button.
Without this change, repainting the button triggers a repaint that attaches
the same size buffer immediately.
https://bugs.freedesktop.org/show_bug.cgi?id=71927
Kristian Høgsberg [Wed, 5 Feb 2014 22:44:04 +0000 (14:44 -0800)]
terminal: Only reset title if we put the window size there
When resizing the terminal, it shows the grid size in the titlebar.
We reset the title next time we get an enter event. This patch makes
sure we only reset the title the first time we enter after a resize.
Kristian Høgsberg [Wed, 5 Feb 2014 22:25:18 +0000 (14:25 -0800)]
input: Run touch binding before calling touch grab
We don't want to send events if the binding is going to handle the touch
event. Also, this restricts touch bindings to only trigger on touch down.
For gesture bindings we want something similar to the motion signal we
have for the pointer.
Kristian Høgsberg [Wed, 5 Feb 2014 21:43:44 +0000 (13:43 -0800)]
window: Ask xkbcommon whether or not to repeat a key
Kristian Høgsberg [Wed, 5 Feb 2014 06:12:29 +0000 (22:12 -0800)]
input: Transform to surface coordinates outside resource for loop
When we send the pointer motion event, the transform from compositor to
surface coordinates doesn't depend on the resource. Transform the
coordinates up front instead of everytime we send to a resource.
Kristian Høgsberg [Wed, 5 Feb 2014 06:44:57 +0000 (22:44 -0800)]
shell: Remove elaborate if-statement with empty body
Kristian Høgsberg [Mon, 3 Feb 2014 23:50:38 +0000 (15:50 -0800)]
shell: Handle wl_shell surfaces created by xwayland correctly
When xwayland creates a shell surface we don't have a resource. The
recently added shell_surface_is_wl_shell/xdg_surface() tests don't
handle that very well.
For now, we assume that a surface without a resource is created from
xwayland and is a wl_shell surface. We'll want to modify that to be a
xdg surface eventually, but for now this stops weston from crashing.
Kristian Høgsberg [Mon, 3 Feb 2014 19:05:41 +0000 (11:05 -0800)]
build: Pass pkg-config CFLAGS as target_CFLAGS, not CPPFLAGS
Kristian Høgsberg [Mon, 3 Feb 2014 18:55:51 +0000 (10:55 -0800)]
build: Add a rule for creating git-version.h when we don't have a git repo
For dist tarballs we ship git-version.h but if you do a git archive or
similar to check out a source tree, there's no git-version.h and no
way to make one.
https://bugs.freedesktop.org/show_bug.cgi?id=74459
Pekka Paalanen [Mon, 3 Feb 2014 11:12:34 +0000 (13:12 +0200)]
build: fix weston-info CFLAGS
CC clients/weston-info.o
clients/weston-info.c:31:28: fatal error: wayland-client.h: No such file or directory
compilation terminated.
make[1]: *** [clients/weston-info.o] Error 1
Only triggerable if libwayland is only in a custom prefix.
Fix by adding CLIENT_CFLAGS.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk>
Pekka Paalanen [Mon, 3 Feb 2014 11:12:33 +0000 (13:12 +0200)]
build: pass client flags to test programs
Fix build failures of the kind:
CC tests/bad-buffer-test.o
In file included from tests/weston-test-client-helper.h:28:0,
from tests/bad-buffer-test.c:28:
./protocol/wayland-test-client-protocol.h:35:28: fatal error: wayland-client.h: No such file or directory
compilation terminated.
make[1]: *** [tests/bad-buffer-test.o] Error 1
These are only triggerable if libwayland has not been installed
system-wide, but only in a custom prefix.
Since the Makefile already uses AM_CPPFLAGS, simply add
TEST_CLIENT_CFLAGS to test programs instead of dropping AM_CPPFLAGS.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk>
Kristian Høgsberg [Sun, 2 Feb 2014 06:16:43 +0000 (22:16 -0800)]
.gitignore: Add .dirstamp
Guillem Jover [Sat, 1 Feb 2014 17:48:43 +0000 (18:48 +0100)]
build: Set a default aux dir
This moves all the auxiliary build scripts into a build-aux directory,
and fixes an issue with configure being unable to find scripts because
it tries to change to an empty directory to get the absolute path,
which results in getting the path to the user's home directory instead.
,--
checking whether build environment is sane... yes
/bin/bash: /home/user/missing: No such file or directory
configure: WARNING: 'missing' script is too old or missing
`---
Signed-off-by: Guillem Jover <guillem@hadrons.org>
Kristian Høgsberg [Sun, 2 Feb 2014 05:52:17 +0000 (21:52 -0800)]
build: Don't dist generated sources
Quentin Glidic [Sat, 1 Feb 2014 20:39:12 +0000 (21:39 +0100)]
Makefile.am: Fix protocol source files usage
Signed-off-by: Quentin Glidic <sardemff7+git@sardemff7.net>
Kristian Høgsberg [Sun, 2 Feb 2014 04:44:54 +0000 (20:44 -0800)]
xwayland: Fix typo in open flags for Xorg log file
Kristian Høgsberg [Sat, 1 Feb 2014 19:02:41 +0000 (11:02 -0800)]
build: Add -Wsign-compare to GCC flags
Jasper St. Pierre [Mon, 9 Dec 2013 20:20:16 +0000 (15:20 -0500)]
Revert "stacking: Remove transient window support"
This reverts commit
4c1a11074af2c2221d50b0c35d2d0d883647bc15.
Use the new window_set_transient_for / window_get_transient_for
and xdg-shell support for this...
Jasper St. Pierre [Mon, 9 Dec 2013 20:26:25 +0000 (15:26 -0500)]
window: Add wrappers for xdg_surface_set_transient_for
Jasper St. Pierre [Mon, 9 Dec 2013 19:58:00 +0000 (14:58 -0500)]
westoy: Port the toy toolkit over to xdg-shell
Jasper St. Pierre [Tue, 12 Nov 2013 00:16:11 +0000 (19:16 -0500)]
westoy: Remove MAXIMIZED and FULLSCREEN as separate window types
xdg_shell changes this around so that they are flags on the remote
object itself, not separate surface types. Move to a system where
we calculate the state from the flags ourselves and set the appropriate
wl_shell_surface type.
When we port to xdg_shell, we'll drop these flags and simply sync
on the client.
Jasper St. Pierre [Tue, 12 Nov 2013 00:04:29 +0000 (19:04 -0500)]
westoy: Remove fullscreen methods
These aren't supported under xdg_shell, at least not right now.
If xdg_shell ever gets support for them, we can revert this commit
and adapt it.
Jasper St. Pierre [Tue, 12 Nov 2013 00:42:23 +0000 (19:42 -0500)]
westoy: Use subsurfaces for tooltips instead of transient windows
Transient windows, at least not as they are today, don't exist in
xdg_shell. Subsurfaces allow for specially placed surfaces relative
to a window, so use these instead.
Jasper St. Pierre [Tue, 12 Nov 2013 01:07:33 +0000 (20:07 -0500)]
westoy: Schedule a full resize when we add a subsurface
If a client adds a subsurface, we need to make sure it's allocated
properly, so queue a resize and redraw on the parent window.
Jasper St. Pierre [Mon, 9 Dec 2013 19:52:33 +0000 (14:52 -0500)]
stacking: Remove transient window support
This will be replaced with xdg_surface_set_transient_for at some point...
Jasper St. Pierre [Mon, 9 Dec 2013 22:36:28 +0000 (17:36 -0500)]
weston: Send xdg_surface_send_focused_[un]set when focusing surfaces
Ander Conselvan de Oliveira [Wed, 29 Jan 2014 16:47:51 +0000 (18:47 +0200)]
input: Remove exported function weston_pointer_verify()
Instead, add a compositor signal that an output has been destroyed and
handle that case locally in input.c.
Emilio Pozuelo Monfort [Tue, 28 Jan 2014 12:54:16 +0000 (13:54 +0100)]
desktop-shell: initialize children link
Since commit 9046d2, when destroying a surface, we remove all the
links from its children. But when the child surfaces are destroyed,
those links will be removed again, but since they were not properly
initialized, weston will crash.
Call shell_surface_set_parent instead which removes the link and
sets parent while also initializing the link, thus avoiding this
crash.
Emilio Pozuelo Monfort [Wed, 29 Jan 2014 10:11:12 +0000 (11:11 +0100)]
desktop-shell: maximize the surface with the kbd focus
We don't have focus-follows-mouse, so it makes more sense to
maximize or fullscreen the surface that has the keyboard focus,
not the one behind the pointer.
Guillem Jover [Fri, 31 Jan 2014 19:41:21 +0000 (20:41 +0100)]
compositor: Clarify XDG_RUNTIME_DIR misconfiguration message
If only the mode or the owner are wrong, do not say both are wrong.
Change the text to state that there's a problem and the current
values, and let the user figure it out.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
Ander Conselvan de Oliveira [Fri, 31 Jan 2014 15:35:45 +0000 (17:35 +0200)]
input: Fix weston_seat_init_keyboard() error path
The pointer seat->keyboard was set before some possible error returns.
That pointer was left unchanged in case of failure, pointing to an
uninitialized keyboard struct (that was also leaked). If a client sent
a wl_seat::get_keyboard request, that would cause Weston to crash.
Fix this by setting the seat->keyboard pointer only after the keymap
initialization is done and there is no more possibilities for failure.
Also plug the memory leaks on the error path.
https://bugs.freedesktop.org/show_bug.cgi?id=74035
Ander Conselvan de Oliveira [Fri, 31 Jan 2014 14:07:51 +0000 (16:07 +0200)]
input: Empty the current input region when configuring pointer surfaces
The input region of the cursor surface is set to empty in
pointer_cursor_surface_configure(). Since during the commit process
this function is called before the pending input region is made
current, it empties surface->pending.input instead of surface->input.
But pointer_cursor_surface_configure() is also called from
pointer_set_cursor() in order to map the cursor even if there isn't a
subsequent attach and commit to the cursor surface. In that case,
surface->input is never emptied, since the configure function emptied
only the pending input region and there wasn't a commit that made it
effective.
Fix this by emptying both pending and current input regions. The latter
shouldn't cause problems since the surface can't have a role prior to
being assigned the cursor role, so it shouldn't be mapped in the first
place.
Also change toytoolkit so that it triggers the bug.
https://bugs.freedesktop.org/show_bug.cgi?id=73711
Kristian Høgsberg [Sat, 1 Feb 2014 08:38:15 +0000 (00:38 -0800)]
build: Stop using xmllint to validate protocol files
The scanner needs to be good enough. If it crashes or fails to report
invalid input, that needs to get fixed.
Kristian Høgsberg [Sat, 1 Feb 2014 08:33:04 +0000 (00:33 -0800)]
build: Drop wayland-scanner.mk
We now only have one Makefile.am so we don't need to keep these rules
in a separate file for sharing.
Kristian Høgsberg [Sat, 1 Feb 2014 00:33:48 +0000 (16:33 -0800)]
build: Only regenerate src/git-version.h when .git/logs/HEAD changes
We rely on .git/logs/HEAD to be a file that changes when we commit to HEAD.
The first idea is to make the makefile rule depend on .git/HEAD, but that's
a symbolic ref that points to the current ref in refs/heads. However,
.git/logs/HEAD changes whenever we commit to HEAD, so we can use that in the
makefile rule.
Kristian Høgsberg [Tue, 28 Jan 2014 06:02:41 +0000 (22:02 -0800)]
build: Move shared/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Sat, 1 Feb 2014 00:15:11 +0000 (16:15 -0800)]
build: Move clients/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Fri, 31 Jan 2014 22:03:09 +0000 (14:03 -0800)]
build: Move tests/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 07:05:52 +0000 (23:05 -0800)]
build: Move src/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 06:33:04 +0000 (22:33 -0800)]
build: Move man/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 06:26:29 +0000 (22:26 -0800)]
build: Move protocol/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 05:59:55 +0000 (21:59 -0800)]
build: Move xwayland/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 05:56:12 +0000 (21:56 -0800)]
build: Move desktop-shell/Makefile.am into toplevel Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 05:46:30 +0000 (21:46 -0800)]
build: Move wcap/Makefile.am into top-level Makefile.am
Kristian Høgsberg [Tue, 28 Jan 2014 05:40:28 +0000 (21:40 -0800)]
build: Move data/Makefile.am into Makefile.am
Kristian Høgsberg [Sat, 1 Feb 2014 07:49:33 +0000 (23:49 -0800)]
build: Use subdir-objects automake option
This makes automake place the object files in the same subdir as the
source file. For a recursive build system as we have now, there's
no difference, but with a non-recursive build system it means that
the object files don't all end up in the toplevel directory.
Kristian Høgsberg [Fri, 31 Jan 2014 22:39:01 +0000 (14:39 -0800)]
clients: Drop pdf viewer sample client
This doesn't really add anything over the existing sample clients
and only complicates the build system with extra checks for poppler.
Kristian Høgsberg [Tue, 28 Jan 2014 07:02:35 +0000 (23:02 -0800)]
compositor: Fix a couple of #includes to use "" instead of <>
Kristian Høgsberg [Sat, 1 Feb 2014 05:56:21 +0000 (21:56 -0800)]
build: Drop use of rsvg-convert
Just include the png in the repo.
Jonas Ådahl [Fri, 3 Jan 2014 18:46:51 +0000 (19:46 +0100)]
input: Unlink saved kbd focus listener when releasing seat
Not doing this would leave a invalid list item in the view's destroy
signal listener list if destroying a seat that had previously lost
keyboard focus.
Signed-off-by: Jonas Ådahl <jadahl@gmail.com>
Jonas Ådahl [Fri, 3 Jan 2014 18:46:50 +0000 (19:46 +0100)]
simple-egl: Fix -Wsign-compare compiler warning
Fixes the following compiler warning:
simple-egl.c:434:36: warning: comparison between signed and unsigned
integer expressions [-Wsign-compare]
Signed-off-by: Jonas Ådahl <jadahl@gmail.com>
Kristian Høgsberg [Mon, 27 Jan 2014 20:07:19 +0000 (12:07 -0800)]
Document WESTON_VERSION_AT_LEAST() macro behavior
Adding this comment to explain the behavior:
This macro may not do what you expect. Weston doesn't guarantee any
stable API between 1.X and 1.Y, and thus this macro will return
FALSE on any WESTON_VERSION_AT_LEAST(1,X,0) if the actualy version
is 1.Y.0 and X !=Y). In particular, it fail if X < Y, that is,
1.3.0 is considered to not be "at least" 1.4.0.
If you want to test for the version number being 1.3.0 or above or
maybe in a range (eg 1.2.0 to 1.4.0), just use the WESTON_VERSION_*
defines above directly.
Version number testing is the one thing we can't break in the weston API,
so we'll have to settle for documenting the behavior and recommending
using the version number macros directly.
https://bugs.freedesktop.org/show_bug.cgi?id=74023
Kristian Høgsberg [Fri, 24 Jan 2014 04:51:40 +0000 (20:51 -0800)]
configure.ac: Bump version to 1.4.0
Kristian Høgsberg [Fri, 24 Jan 2014 00:25:06 +0000 (16:25 -0800)]
compositor-drm: Set cursor surface to NULL when pageflip fails
If we VT switch away between picking a cursor surface and actually doing
the pageflip in drm_output_repaint(), we never set output->cursor_view to
NULL. Then we unplug all the input devices and as the last pointer device
goes away we destroy the cursor surface. Then when we switch back, we
call drm_output_set_cursor() with an invalid surface and crashes.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=73566
Kristian Høgsberg [Wed, 22 Jan 2014 20:25:20 +0000 (12:25 -0800)]
shell: Properly track the focus state surface
We have to move the surface destroy listener around as we track the
currently focused surface. Introduce a helper function,
focus_state_set_focus() for this and use throughout.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=73768
Kristian Høgsberg [Wed, 22 Jan 2014 19:03:50 +0000 (11:03 -0800)]
evdev: Remove output destroy notifier on device destroy
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=73895
Kristian Høgsberg [Wed, 22 Jan 2014 18:45:26 +0000 (10:45 -0800)]
configure.ac: Check for libsystemd-login >= 198
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=73870
Ricardo Vieira [Sat, 18 Jan 2014 16:30:50 +0000 (16:30 +0000)]
shell: Don't move maximized window
We'll want to ask the client to unmaximize once we get support for that in
xdg-shell, but for now, just refuse moving a maximized window.
Kristian Høgsberg [Mon, 20 Jan 2014 18:35:26 +0000 (10:35 -0800)]
shell: Add mod+shift left-click as a resize binding
This lets us use compositor-initiated resize on laptops with touchpads and
no middle buttons.
Kristian Høgsberg [Mon, 20 Jan 2014 06:56:14 +0000 (22:56 -0800)]
configure.ac: Bump version to 1.3.93
This is the release candidate.
U. Artie Eoff [Fri, 10 Jan 2014 18:15:17 +0000 (10:15 -0800)]
drm: prepend stamp space to output mode logging
Use the STAMP_SPACE to make the output mode logging
a little nicer looking.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
Reviewed-by: Bryce Harrington <b.harrington@samsung.com>
Hardening [Fri, 10 Jan 2014 10:33:06 +0000 (11:33 +0100)]
Fix compilation with FreeRdp 1.1 and master v2
The API to use remoteFx encoding has changed between master and stable 1.1
branch. This patch should fix compilation for both.
This new version adds checks for the freerdp/version.h file
U. Artie Eoff [Sat, 18 Jan 2014 00:28:15 +0000 (16:28 -0800)]
rpi: call eglTerminate if we're bailing
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 21:56:41 +0000 (13:56 -0800)]
rpi-backend: ensure pixels is destroyed in resource_update
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 23:08:51 +0000 (15:08 -0800)]
animation: don't NULL deref if _run() returns NULL
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 22:44:05 +0000 (14:44 -0800)]
shared/frame: NULL check before attempting to deref
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 22:00:18 +0000 (14:00 -0800)]
rpi: free renderer if we bail early during create
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 21:19:01 +0000 (13:19 -0800)]
compositor: free memory from str config sections after done using it
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 20:36:58 +0000 (12:36 -0800)]
compositor-wayland: free 'theme' config option value
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 20:22:50 +0000 (12:22 -0800)]
compositor-wayland: don't free unallocated memory
Assigning a string constant (i.e. memory that we didn't allocate)
to a char* pointer and then freeing that pointer is bad news.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 19:19:46 +0000 (11:19 -0800)]
shared/frame: fix potential memory leak in frame_create
In frame_create, we need to destroy any frame buttons created
in preceding calls to frame_button_create during the function
execution if any of the successive calls to frame_button_create
fail.
This has minimal severity since most, if not all, cases in
frame_button_create that result in a fail (i.e. NULL result) means
a program is OOM and the program will have to exit/abort anyway.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Fri, 17 Jan 2014 18:08:25 +0000 (10:08 -0800)]
shell: assert get_shell_surface() != NULL as appropriate
Various functions that operate on a weston_surface assume the
surface has a shell_surface. That is, they unconditionally
deref the get_shell_surface() result. Hence, if for some reason
the call to get_shell_surface() returned NULL to those functions then
a segmentation fault would occur and the program would crash. So,
adding an assert(...) on the get_shell_surface() return value adds an
extra sanity check and does not change this behavior. The assert also
adds an extra benefit to the programmer by documenting that the function
expects and requires the weston_surface to have a shell_surface and
would be a program logic error, otherwise.
The assert() also silences some static analyzers about the possible
NULL deref.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 22:45:59 +0000 (14:45 -0800)]
shell: validate create_focus_surface results
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 22:26:31 +0000 (14:26 -0800)]
shell: validate get_animation_type parameter
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 22:03:56 +0000 (14:03 -0800)]
input-panel: validate return value of get_default_view
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 21:42:18 +0000 (13:42 -0800)]
exposay: remove redundant NULL check in highlight_surface
exposay_highlight_surface() is called from exposay_pick(),
exposay_layout(), and exposay_maybe_move() where the esurface
parameter is already validated prior to the call. This makes
the 'esurface' NULL check redundant. This assumes any future
calls to exposay_highlight_surface() will also validate the
'esurface' parameter prior to the call.
This fixes the logic in exposay_highlight_surface so static
analyzers don't complain about the possibility that 'view'
might be NULL deref'd when a 'esurface' == NULL condition is
true.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 21:38:51 +0000 (13:38 -0800)]
clients/window: fail if NULL returned from frame_create
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 21:02:28 +0000 (13:02 -0800)]
weston-info: free allocated memory and check OOM conditions
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 19:40:37 +0000 (11:40 -0800)]
terminal: use xstrdup instead to handle OOM
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 19:13:38 +0000 (11:13 -0800)]
use xzalloc wrapper instead
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 18:59:50 +0000 (10:59 -0800)]
clients/desktop-shell: check OOM conditions
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 17:36:00 +0000 (09:36 -0800)]
screenshooter: check for NULL malloc and fix memleaks
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 16:52:07 +0000 (08:52 -0800)]
wcap-decode: remove dead code
Avoids leaking memory we never use.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 16:24:40 +0000 (08:24 -0800)]
log: print dummy timestamp if localtime call fails
Handle the case where localtime fails (NULL) and print
something else to indicate localtime is erroneous.
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
U. Artie Eoff [Wed, 15 Jan 2014 16:12:19 +0000 (08:12 -0800)]
backlight: fix backlight_path memory leak
Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
Kristian Høgsberg [Mon, 20 Jan 2014 05:48:10 +0000 (21:48 -0800)]
shell: Don't start a fade animation if we don't have an output
When the last window of the X11 compositor is closed during a fade or
while locked, we'll try to start a fade back to the lock screen. However,
if we closed the last window, there are no outputs left and the animation
will try to run with surface->output == NULL.
https://bugs.freedesktop.org/show_bug.cgi?id=73665
Kristian Høgsberg [Sat, 18 Jan 2014 00:39:17 +0000 (16:39 -0800)]
Kristian Høgsberg [Sat, 18 Jan 2014 00:22:50 +0000 (16:22 -0800)]
shell: Cancel popups when triggering screensaver
This is part of the fix for bug 72540. We cancel the popup grab when the
screensaver kicks in so that the screen unlock dialog can get input events.
The bigger problem is in mesa however, where we try to allocate new buffers
as cairo-gles2 does a gratuituous (but valid) eglMakeCurrent() as we
remove the tooltip or popup-menu.
Since we removed the weston_layer with the regular surfaces, EGL blocks
waiting for a frame event that never comes.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=72540