profile/ivi/weston-ivi-shell.git
11 years agotests: list available tests if an invalid test name is given
Peter Hutterer [Wed, 11 Sep 2013 06:08:47 +0000 (16:08 +1000)]
tests: list available tests if an invalid test name is given

11 years agotests: support -h/--help for the tests
Peter Hutterer [Wed, 11 Sep 2013 06:08:30 +0000 (16:08 +1000)]
tests: support -h/--help for the tests

Including listing the tests available in that binary

11 years agotests: include config.h in weston-test-runner
Peter Hutterer [Wed, 11 Sep 2013 06:08:04 +0000 (16:08 +1000)]
tests: include config.h in weston-test-runner

11 years agotests: use variable for test name in weston-tests-env
Peter Hutterer [Wed, 11 Sep 2013 05:58:08 +0000 (15:58 +1000)]
tests: use variable for test name in weston-tests-env

Slightly more readable and makes it easier to switch to use $2 for something
in the future (if that's ever needed).

11 years agotests: always build tests
Peter Hutterer [Wed, 11 Sep 2013 05:58:07 +0000 (15:58 +1000)]
tests: always build tests

check_PROGRAMS and friends are only built during make check. Which is a
great way of introducing compiler errors in tests. Always build them, TESTS
defines what's being run during make check.

11 years agoshell: Use wl_resource_for_each() for sending out workspace status
Kristian Høgsberg [Wed, 11 Sep 2013 19:00:47 +0000 (12:00 -0700)]
shell: Use wl_resource_for_each() for sending out workspace status

11 years agoinput: Use new wl_resource_for_each for sending updated seat caps
Rob Bradford [Fri, 6 Sep 2013 16:48:19 +0000 (17:48 +0100)]
input: Use new wl_resource_for_each for sending updated seat caps

11 years agovaapi-recorder: Don't leak drm fd
Ander Conselvan de Oliveira [Fri, 6 Sep 2013 14:49:38 +0000 (17:49 +0300)]
vaapi-recorder: Don't leak drm fd

Make vaapi_recorder take onwership of the fd and close it at destroy
time.

11 years agovaapi-recorder: Encode frames in a separate thread
Ander Conselvan de Oliveira [Fri, 6 Sep 2013 14:49:37 +0000 (17:49 +0300)]
vaapi-recorder: Encode frames in a separate thread

Previously, vaapi_recorder_frame() would wait until the encoded
contents for a frame is written to the output file descriptor. This
delayed the repainting of the next frame, and affected frame rate
when capturing with high resolutions. Instead, wait only if there is
and attempted to encode two frames at the same time.

Increases framerate from 30 to 60 fps when capturing at 1920x1200 on
my SandryBridge system, although there are periodic slowdowns due to
disk writes.

11 years agoadd libhybris support
Adrian Negreanu [Fri, 6 Sep 2013 12:16:09 +0000 (15:16 +0300)]
add libhybris support

it uses the Android fbdev HAL[1] (through libhybris[2])
and the libhybris implementation of wayland-egl.

Configure flags:
 cairo:
   --enable-glesv2=yes  --enable-egl=yes
 weston:
   --with-cairo-glesv2 --enable-fbdev-compositor
 hybris:
   --enable-wayland --enable-arch=x86
   --with-android-headers=<android-headers> --enable-alinker=jb

The android headers are extracted from an AOSP tree,
using hybris/utils/extract-headers.sh

[1]:
https://github.com/android/platform_hardware_libhardware/blob/master/include/hardware/fb.h

[2]: https://github.com/libhybris/libhybris

Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
11 years agoterminal: Stop complaining about OSC escape code 7
Kristian Høgsberg [Wed, 11 Sep 2013 18:52:56 +0000 (11:52 -0700)]
terminal: Stop complaining about OSC escape code 7

This is the current working directory of the shell as an URI.  We
don't use that for anythign at the moment, but let's not complain
about it at least.

11 years agocompositor-x11: Copying xkb_info when creating a seat causes problems
Andrew Wedgbury [Fri, 6 Sep 2013 08:29:12 +0000 (08:29 +0000)]
compositor-x11: Copying xkb_info when creating a seat causes problems

Sorry, I missed updating use of xkb_info in compositor-x11.c.
I've updated the patch.

11 years agofbdev: Support of TrueColor and Directcolor of video hardware
Marc Chalain [Tue, 3 Sep 2013 14:47:43 +0000 (16:47 +0200)]
fbdev: Support of TrueColor and Directcolor of video hardware

backend check the type of the framebuffer and accept DirectColor and TrueColor
I use a "switch case" to implement other cases in the future.

11 years agogl_renderer: Use EGL_WAYLAND_Y_INVERTED_WL to query wl_buffer's orientation
Stanislav Vorobiov [Thu, 29 Aug 2013 07:36:44 +0000 (11:36 +0400)]
gl_renderer: Use EGL_WAYLAND_Y_INVERTED_WL to query wl_buffer's orientation

11 years agoCopying xkb_info when creating a seat causes problems
Andrew Wedgbury [Thu, 5 Sep 2013 13:31:40 +0000 (13:31 +0000)]
Copying xkb_info when creating a seat causes problems

Hi Kristian,

Here's a new patch for ref counting weston_xkb_info, as suggested.
So a seat created with a NULL keymap will now point to the global xkb_info.

11 years agoinput: Initialize data source to NULL
Kristian Høgsberg [Wed, 11 Sep 2013 16:45:03 +0000 (09:45 -0700)]
input: Initialize data source to NULL

This way we do the right thing, when we get a NULL wl_resource for self-dnd.

11 years agocompositor: Return 0 on success
Kristian Høgsberg [Wed, 11 Sep 2013 16:42:26 +0000 (09:42 -0700)]
compositor: Return 0 on success

Add missing return statement.

11 years agodnd: Support dropping as text/plain;charset=utf-8
Kristian Høgsberg [Mon, 9 Sep 2013 22:03:27 +0000 (15:03 -0700)]
dnd: Support dropping as text/plain;charset=utf-8

11 years agoxwm: Fix fd leak in weston_wm_send_data()
Kristian Høgsberg [Thu, 5 Sep 2013 05:32:50 +0000 (22:32 -0700)]
xwm: Fix fd leak in weston_wm_send_data()

The call to source->send(source, mime_type, p[1]); dups the fd, and we have
to close p[1] to not leak it.

11 years agoxwm: Try a non-blocking write before setting up an fd watch for property data
Kristian Høgsberg [Thu, 5 Sep 2013 05:12:28 +0000 (22:12 -0700)]
xwm: Try a non-blocking write before setting up an fd watch for property data

Typically we can write it immediately without blocking, so save the overhead
of setting up an fd watch and writing the data in a callback.  For the
case where the immediate write doesn't write all data, we fallback and
set up the fd watch as usual.

This patch also consolidates setting up the async write a bit.

11 years agoxwm: Add support for bridging Xdnd to wayaland DnD
Kristian Høgsberg [Thu, 5 Sep 2013 04:12:26 +0000 (21:12 -0700)]
xwm: Add support for bridging Xdnd to wayaland DnD

11 years agoxwm: Only handle selection notify events for CLIPBOARD in selection code
Kristian Høgsberg [Thu, 5 Sep 2013 04:09:24 +0000 (21:09 -0700)]
xwm: Only handle selection notify events for CLIPBOARD in selection code

11 years agoxwm: Fix fd leak in selection code
Kristian Høgsberg [Thu, 5 Sep 2013 03:48:46 +0000 (20:48 -0700)]
xwm: Fix fd leak in selection code

The file descriptors we receive in the wayland protocol callbacks are
already O_CLOEXEC, so don't dup them to set O_CLOEXEC.`

11 years agocompositor: Split dnd setup up into weston_seat_start_drag()
Kristian Høgsberg [Thu, 5 Sep 2013 03:44:26 +0000 (20:44 -0700)]
compositor: Split dnd setup up into weston_seat_start_drag()

This makes the drag-and-drop code available to in-weston data sources,
similar to how we can set a selection data source internally.  The
wl_data_device.start_drag entry point now calls this function after
validating protocol arguments.

11 years agoterminal: Accept utf-8 text drop
Kristian Høgsberg [Thu, 5 Sep 2013 03:41:06 +0000 (20:41 -0700)]
terminal: Accept utf-8 text drop

11 years agodnd: Only accept dnd with the application/x-wayland-dnd-flower mime-type
Kristian Høgsberg [Thu, 5 Sep 2013 02:36:49 +0000 (19:36 -0700)]
dnd: Only accept dnd with the application/x-wayland-dnd-flower mime-type

Currently, the dnd sample client accepts all mime-types and assumes they
are the custom flower mime-type.  Only accept if the offer has the right
mime-type.

11 years agoconfigure.ac: Only enable enable_simple_egl_clients if enable_egl is yes
Kristian Høgsberg [Wed, 4 Sep 2013 05:59:17 +0000 (22:59 -0700)]
configure.ac: Only enable enable_simple_egl_clients if enable_egl is yes

This simplifies a few checks and makes sure we only build the
subsurfaces client if EGL is enabled.

https://bugs.freedesktop.org/show_bug.cgi?id=68727

11 years agokeyboard: Add rudimentary touch support to keyboard.
Brian J Lovin [Tue, 27 Aug 2013 17:49:42 +0000 (10:49 -0700)]
keyboard: Add rudimentary touch support to keyboard.

At this time there is no way to have a key be activated when
touch_up is called, so all this patch does is activate they
key on touch_down.

Signed-off-by: Brian J Lovin <brian.j.lovin@intel.com>
11 years agoconfigure.ac: Only build libva recorder if we found libva
Kristian Høgsberg [Wed, 4 Sep 2013 05:46:02 +0000 (22:46 -0700)]
configure.ac: Only build libva recorder if we found libva

The automatic detection failed to disable the build if it didn't
find libva.

https://bugs.freedesktop.org/show_bug.cgi?id=68908

11 years agocompositor-drm: Pass the wl_resource to gbm_bo_import()
Kristian Høgsberg [Wed, 4 Sep 2013 05:27:08 +0000 (22:27 -0700)]
compositor-drm: Pass the wl_resource to gbm_bo_import()

Another silent regression from the wl_resource opaquify effort.  This was
causing our pageflip-to-client-buffer and sprites optimizations to
not kick in.

11 years agoxwm: Unmap frame window when client window is unmapped
Kristian Høgsberg [Tue, 3 Sep 2013 23:38:51 +0000 (16:38 -0700)]
xwm: Unmap frame window when client window is unmapped

We used to destroy the frame window and reparent the client window to
wm_window.  That means that we lose the destroy_notify event when the
client window is destroyed later, since we don't select for
substructure_notify on wm_window.

Instead of destroying and reparenting, just unmap the frame window.

11 years agoxwm: Split creation of frame window out into its own function
Kristian Høgsberg [Tue, 3 Sep 2013 23:19:18 +0000 (16:19 -0700)]
xwm: Split creation of frame window out into its own function

11 years agoxwm: Set window->surface to NULL when surface goes away
Kristian Høgsberg [Tue, 3 Sep 2013 23:15:42 +0000 (16:15 -0700)]
xwm: Set window->surface to NULL when surface goes away

Otherwise we may end up trying to remove the destroy listener if we
get the surface_destroy callback before the unmap_notify.

11 years agoxwm: Detect WM_NORMAL_HINTS fullscreen attempts
Kristian Høgsberg [Sat, 31 Aug 2013 07:12:25 +0000 (00:12 -0700)]
xwm: Detect WM_NORMAL_HINTS fullscreen attempts

11 years agoxwm: Detect legacy fullscreen modes
Kristian Høgsberg [Sat, 31 Aug 2013 07:08:27 +0000 (00:08 -0700)]
xwm: Detect legacy fullscreen modes

11 years agoxwm: Update size when override redirect windows change size
Kristian Høgsberg [Sat, 31 Aug 2013 07:00:57 +0000 (00:00 -0700)]
xwm: Update size when override redirect windows change size

11 years agoconfigure: Require wayland-server 1.2.90
Kristian Høgsberg [Fri, 30 Aug 2013 22:00:01 +0000 (15:00 -0700)]
configure: Require wayland-server 1.2.90

Make sure we look for a recent enough version of wayland-server for
the new wl_seat release requests.

11 years agowindow: Use new wl_pointer/keyboard_release request
Rob Bradford [Tue, 13 Aug 2013 19:11:03 +0000 (20:11 +0100)]
window: Use new wl_pointer/keyboard_release request

Since we bump the version we ask for from the compositor it is also
necessary to implement the new "name" event in the seat listener.

11 years agocompositor: Implement release request for input interfaces
Kristian Høgsberg [Tue, 13 Aug 2013 19:11:02 +0000 (20:11 +0100)]
compositor: Implement release request for input interfaces

v2 (Rob Bradford): Update the version numbering for this change

11 years agoconfigure.ac: Auto-enable vaapi recorder, but allow overrides
Kristian Høgsberg [Fri, 30 Aug 2013 21:28:22 +0000 (14:28 -0700)]
configure.ac: Auto-enable vaapi recorder, but allow overrides

If libva is available and recent, we'll enable the vaapi recorder feature,
but it can also be forced on or off by --enable/disable-vaapi-recorder.

11 years agocompositor-drm: Add hardware accelerated capture of screen using libva
Ander Conselvan de Oliveira [Fri, 23 Aug 2013 14:15:48 +0000 (17:15 +0300)]
compositor-drm: Add hardware accelerated capture of screen using libva

This patch adds a feature to the DRM backend that uses libva for
encoding the screen contents in H.264. Screen recording can be
activated by pressing mod-shift-space q. A file named capture.h264
will be created in the current directory, which can be muxed into
an MP4 file with gstreamer using

gst-launch filesrc location=capture.h264 ! h264parse ! mp4mux ! \
           filesink location=file.mp4

This is limitted to the DRM compositor in order to avoid a copy when
submitting the front buffer to libva. The code in vaapi-recorder.c
takes a dma_buf fd referencing it, does a colorspace conversion using
the video post processing pipeline and then uses that as input to the
encoder.

I'm sending this now so I get comments, but this is not ready for
prime time yet. I have a somewhat consistent GPU hang when using
i915 with SandyBridge. Sometimes a page flip never completes. If you
want to try this anyway and your system get stuck, you might need to
run the following:

  # echo 1 > /sys/kernel/debug/dri/0/i915_wedged

After that, alt-sysrq [rv] should work.

Once that's fixed it would also be good to make the parameters used by
the encoder more flexible. For now the QP parameter is hardcoded to 0
and we have only I and P frames (no B frames), which causes the
resulting files to be very large.

11 years agoautotools: Fix AC_PATH_PROG call
Quentin Glidic [Sat, 24 Aug 2013 18:36:10 +0000 (20:36 +0200)]
autotools: Fix AC_PATH_PROG call

Signed-off-by: Quentin Glidic <sardemff7+git@sardemff7.net>
11 years ago.gitignore: Reflect recent client binary rename
Bryce W. Harrington [Fri, 23 Aug 2013 19:57:04 +0000 (19:57 +0000)]
.gitignore: Reflect recent client binary rename

Commit 77ab1721 renamed the client binaries; this change updates
.gitignore to follow this change.

Signed-off-by: Bryce Harrington <b.harrington@samsung.com>
11 years agowindow.c: Fix SIGSEGV in pointer_handle_motion
Paul Winwood [Thu, 29 Aug 2013 09:52:54 +0000 (10:52 +0100)]
window.c: Fix SIGSEGV in pointer_handle_motion

Move null pointer check and assignments before first window dereference.

11 years agoevdev: Don't parse config file again, use compositor->config
Kristian Høgsberg [Thu, 29 Aug 2013 20:49:35 +0000 (13:49 -0700)]
evdev: Don't parse config file again, use compositor->config

11 years agoshell: Make sure we have seat->pointer and seat->touch before dereferencing
Kristian Høgsberg [Thu, 29 Aug 2013 06:16:20 +0000 (23:16 -0700)]
shell: Make sure we have seat->pointer and seat->touch before dereferencing

Either of these may not be available when we handle wl_shell_surface.move,
and we need to not crash when that's the case.

https://bugs.freedesktop.org/show_bug.cgi?id=68649

11 years agogl-renderer: Use GL_UNSIGNED_SHORT for index array type
Kristian Høgsberg [Thu, 29 Aug 2013 06:05:29 +0000 (23:05 -0700)]
gl-renderer: Use GL_UNSIGNED_SHORT for index array type

GL_UNSIGNED_INT is only supported when GL_OES_element_index_uint is
available (mesa implements that extension).  We don't need 32-bit
indices, so just use GL_UNSIGNED_SHORT.

11 years agosubsurfaces: fix rendering problem with cairo-glesv2
Stanislav Vorobiov [Wed, 28 Aug 2013 06:14:35 +0000 (10:14 +0400)]
subsurfaces: fix rendering problem with cairo-glesv2

when using cairo-glesv2 subsurface rendering code
implicitly calls eglMakeCurrent(dpy, NULL, NULL, ctx)
(since EGL_KHR_surfaceless_context is used), thus,
triangle_frame_callback:eglSwapBuffers returns EGL_BAD_SURFACE
error for all invocations other that the first one

11 years agoconfigure.ac: fix broken compilation when configure with --disable-egl option
Samuel Iglesias Gonsalvez [Wed, 28 Aug 2013 09:22:41 +0000 (11:22 +0200)]
configure.ac: fix broken compilation when configure with --disable-egl option

Fix bug 67561: "configure with --disable-egl option breaks fbdev backend
compilation"

https://bugs.freedesktop.org/show_bug.cgi?id=67561

Signed-off-by: Samuel Iglesias Gonsalvez <siglesias@igalia.com>
11 years agoconfigure.ac: Bump configure.ac version to 1.2.90
Kristian Høgsberg [Thu, 29 Aug 2013 05:16:10 +0000 (22:16 -0700)]
configure.ac: Bump configure.ac version to 1.2.90

We usually do this when we branch off the stable branch, better late than
never.

11 years agoevdev: Don't transform device->abs.x/y in place
Kristian Høgsberg [Thu, 29 Aug 2013 05:12:24 +0000 (22:12 -0700)]
evdev: Don't transform device->abs.x/y in place

We don't always get both an X and an Y event in a SYN report, so we end
up transforming the coordinate we don't get twice.  For example, if we
only receive an ABS_X event, we transform the already transformed
device->abs.y again in transform_absolute() when applying the calibration.

11 years agodesktop-shell: Add support for touch unlock
Brian J Lovin [Mon, 26 Aug 2013 22:58:22 +0000 (15:58 -0700)]
desktop-shell: Add support for touch unlock

Add support for unlocking on touch up using the
green button on the unlock dialog

Signed-off-by: Brian J Lovin <brian.j.lovin@intel.com>
11 years agoAdd a --shell option to override the default desktop-shell
Jason Ekstrand [Thu, 22 Aug 2013 22:28:03 +0000 (17:28 -0500)]
Add a --shell option to override the default desktop-shell

Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
11 years agosimple-egl: Handle missing EGL_EXT_buffer_age better
Kristian Høgsberg [Thu, 22 Aug 2013 19:16:51 +0000 (12:16 -0700)]
simple-egl: Handle missing EGL_EXT_buffer_age better

Including src/weston-egl-ext.h breaks distcheck, so lets just copy the
one EGL_EXT_buffer_age into simple-egl.c.

11 years agocompositor.h: Don't include config.h
Pier Luigi Fiorini [Fri, 23 Aug 2013 21:43:23 +0000 (23:43 +0200)]
compositor.h: Don't include config.h

Public headers should not include config.h.

11 years agoAdd more missing config.h #includes
Kristian Høgsberg [Mon, 26 Aug 2013 21:43:17 +0000 (14:43 -0700)]
Add more missing config.h #includes

Now that we use AC_SYS_LARGEFILE, we need to pull in config.h at least
whereever we use mmap().  Fixes at least the test-suite and simple-shm
on 32 bit systems.

11 years agoevdev: Initliaze device->link so we can wl_list_remove() without crashing
Kristian Høgsberg [Mon, 26 Aug 2013 21:35:32 +0000 (14:35 -0700)]
evdev: Initliaze device->link so we can wl_list_remove() without crashing

We were testing for wl_list_empty() on a { NULL, NULL } list (which
returns false) and then wl_list_remove()ing the device (which crashes).

11 years agoxwm: Use window->x/y for override redirect surfaces
Kristian Høgsberg [Thu, 22 Aug 2013 23:24:15 +0000 (16:24 -0700)]
xwm: Use window->x/y for override redirect surfaces

window->x/y is the coordinate of the top-level surface (whether that's
the frame window or an override-redirect window) and the wayland surface
should be placed there, without the t->margin offset.

11 years agoxwm: Fix transform listener
Kristian Høgsberg [Thu, 22 Aug 2013 23:20:44 +0000 (16:20 -0700)]
xwm: Fix transform listener

The coordinate transformation was broken (worked for first output where
output->x/y was 0,0, broke on all other outputs).  We can just use
surface->geometry.x/y directly.  We can't use the full transformation,
the best we can do is to move the X window to the geometry.x/y location.

Get rid of the static old_sx/sy hack as well.

11 years agoxwm: Fix configure notify handler
Kristian Høgsberg [Thu, 22 Aug 2013 23:18:17 +0000 (16:18 -0700)]
xwm: Fix configure notify handler

We only get configure notify for toplevel (frame or override-redirect window)
and those are the cases where we want to update window->x/y.  The way the
code worked, we'd exit immeidately in those cases and window->x/y would
not be updated.

11 years agoxwm: Set _NET_WM_CM_S0 instead of relying on the xwayland module to do it
Kristian Høgsberg [Thu, 22 Aug 2013 05:14:58 +0000 (22:14 -0700)]
xwm: Set _NET_WM_CM_S0 instead of relying on the xwayland module to do it

Not sure why I made xwayland claim this selection, seem a little
awkward in retrospect.

11 years agocompositor-drm: Zero out create_arg when creating dumb framebuffers
Kristian Høgsberg [Thu, 22 Aug 2013 05:14:14 +0000 (22:14 -0700)]
compositor-drm: Zero out create_arg when creating dumb framebuffers

The create_arg struct has a flags member that we didn't properly set to 0.

11 years agoterminal: Include config.h instead of defining _XOPEN_SOURCE
Kristian Høgsberg [Thu, 22 Aug 2013 05:13:17 +0000 (22:13 -0700)]
terminal: Include config.h instead of defining _XOPEN_SOURCE

11 years agoconfigure.ac: Add AC_SYS_LARGEFILE for mmap()ing dumb GEM buffers
Kristian Høgsberg [Wed, 21 Aug 2013 22:44:57 +0000 (15:44 -0700)]
configure.ac: Add AC_SYS_LARGEFILE for mmap()ing dumb GEM buffers

We're mmap()ing buffers with 64 bit offsets coming from the
DRM_IOCTL_MODE_MAP_DUMB ioctl and need to make sure mmap accepts
those.

11 years agodata-device: Only bail out if we failed to allocate offer
Kristian Høgsberg [Tue, 20 Aug 2013 22:58:25 +0000 (15:58 -0700)]
data-device: Only bail out if we failed to allocate offer

It's valid to have a NULL data source (self-dnd) in which case we send
a NULL offer on enter.

11 years agocompositor: Only check buffer for NULL tried to allocate something
Kristian Høgsberg [Tue, 20 Aug 2013 18:30:54 +0000 (11:30 -0700)]
compositor: Only check buffer for NULL tried to allocate something

In particular, if we get a NULL buffer attach, it's ok for buffer to be NULL.

11 years agoautotools: Don't use wayland-scanner.m4
Daiki Ueno [Tue, 20 Aug 2013 07:56:52 +0000 (09:56 +0200)]
autotools: Don't use wayland-scanner.m4

11 years agoset RDP output enabled by default (stable 1.2)
Hardening [Fri, 16 Aug 2013 22:30:31 +0000 (00:30 +0200)]
set RDP output enabled by default (stable 1.2)

This patch fixes a bug found by Marek Romanowic: the RDP peer output must
be enabled by default, or we have to unfocus/focus the RDP client window to
have disable/enable output messages sent (and finally receive updates).

11 years agoFixes CJK wide character display
Peng Wu [Mon, 19 Aug 2013 02:59:21 +0000 (10:59 +0800)]
Fixes CJK wide character display

By jumping two columns when wide character prints,
and draw wide background under wide character.

11 years agoAdd touch support for wl_shell_surface_move
Rusty Lynch [Thu, 15 Aug 2013 16:10:08 +0000 (09:10 -0700)]
Add touch support for wl_shell_surface_move

11 years agonested: Fix skipping frames due to texture update without a context
Ander Conselvan de Oliveira [Tue, 16 Jul 2013 11:24:04 +0000 (14:24 +0300)]
nested: Fix skipping frames due to texture update without a context

Calls into cairo-gles may change the current context, so it was only by
chance that sometimes we had the proper one as current and updated the
correct texture in surface_attach().

In order to fix this, calling display_acquire_window_surface() before
binding the texture for setup is necessary. However this call has the
side effect of allocating a cairo surface for the window. At flush time,
the existence of this surface will cause an eglSwapBuffers(), even
if no rendering was done to it, leading to undefined contents on the
screen. This happens when the idle redraw task runs while there is a
pending frame callback.

Workaround this by moving the texture setup from surface_attach() to the
redraw handler, so that the cairo surface is only allocated when
redering is done.

11 years agonested: The wl_resouce is the buffer, not the wl_resource user data
Kristian Høgsberg [Thu, 15 Aug 2013 22:17:57 +0000 (15:17 -0700)]
nested: The wl_resouce is the buffer, not the wl_resource user data

11 years agoAdd more missing config.h includes
Daniel Stone [Thu, 15 Aug 2013 00:10:24 +0000 (01:10 +0100)]
Add more missing config.h includes

config.h includes were missing in a few files, including input.c, the
lack of which caused the X11 backend to segfault instantly due to not
having an xkbcommon context.

Signed-off-by: Daniel Stone <daniel@fooishbar.org>
11 years agowindow: Use xzalloc() instead of silently failing to add output
Kristian Høgsberg [Thu, 15 Aug 2013 21:28:25 +0000 (14:28 -0700)]
window: Use xzalloc() instead of silently failing to add output

11 years agoadd [xwayland] path weston.ini option
Maksim Melnikau [Wed, 14 Aug 2013 19:33:10 +0000 (22:33 +0300)]
add [xwayland] path weston.ini option

It sets the path to the xserver to run.

Signed-off-by: Maksim Melnikau <maxposedon@gmail.com>
11 years agoweston-info: Use xmalloc() in weston-info
Kristian Høgsberg [Thu, 15 Aug 2013 21:20:53 +0000 (14:20 -0700)]
weston-info: Use xmalloc() in weston-info

11 years agowindow: Use xzalloc() instead of failing to add input device
Kristian Høgsberg [Thu, 15 Aug 2013 21:17:13 +0000 (14:17 -0700)]
window: Use xzalloc() instead of failing to add input device

11 years agowindow: Fix logic for looking up widget default cursor
Kristian Høgsberg [Thu, 15 Aug 2013 21:15:18 +0000 (14:15 -0700)]
window: Fix logic for looking up widget default cursor

We may deref a NULL pointer if there is no grab and no focus widget.

11 years agocompositor: Handle OOM in weston_buffer_from_resource()
Kristian Høgsberg [Thu, 15 Aug 2013 19:26:42 +0000 (12:26 -0700)]
compositor: Handle OOM in weston_buffer_from_resource()

Handle NULL-return in call-sites as well.

11 years agowindow: Don't check for subcompositor in window_add_subsurface()
Kristian Høgsberg [Thu, 15 Aug 2013 18:56:41 +0000 (11:56 -0700)]
window: Don't check for subcompositor in window_add_subsurface()

Client should check for subcompositor support up front.  Now
window_add_surface() can never return NULL.

11 years agowindow: Add API to check for subcompositor, use in subsurface example
Kristian Høgsberg [Thu, 15 Aug 2013 18:54:03 +0000 (11:54 -0700)]
window: Add API to check for subcompositor, use in subsurface example

This way we can fail with a nice error message.

11 years agosimple-touch: Handle OOM when allocating struct touch
Kristian Høgsberg [Thu, 15 Aug 2013 18:40:30 +0000 (11:40 -0700)]
simple-touch: Handle OOM when allocating struct touch

11 years agosimple-shm: Handle OOM when allocating struct display
Kristian Høgsberg [Thu, 15 Aug 2013 18:39:52 +0000 (11:39 -0700)]
simple-shm: Handle OOM when allocating struct display

11 years agokeyboard: Don't use next_utf8_char() to go back where we came from
Kristian Høgsberg [Thu, 15 Aug 2013 18:37:09 +0000 (11:37 -0700)]
keyboard: Don't use next_utf8_char() to go back where we came from

We go one utf8 char back and then one forward.  Just remember the original
position instead, which also avoids a warning about potentially
dereferencing a NULL return value from next_utf8_char().

11 years agoautotools: Add configure summary
Armin K [Wed, 14 Aug 2013 01:13:53 +0000 (03:13 +0200)]
autotools: Add configure summary

11 years agoterminal: Make utf-8 state machine assemble unicode code point value
Kristian Høgsberg [Wed, 14 Aug 2013 00:23:54 +0000 (17:23 -0700)]
terminal: Make utf-8 state machine assemble unicode code point value

11 years agocompositor: ref-count weston_surface instances
Giulio Camuffo [Tue, 13 Aug 2013 21:10:14 +0000 (23:10 +0200)]
compositor: ref-count weston_surface instances

This allows a surface to live on after its resource has been
destroyed.  The ref-count can be increased in a resource destroy signal
listener, to keep the surface around for a destroy animation, for example.

11 years agoevdev: Ignore joystick devices
Kristian Høgsberg [Tue, 13 Aug 2013 21:55:39 +0000 (14:55 -0700)]
evdev: Ignore joystick devices

We don't handle them in any way now and having your steering wheel move
the cursor isn't useful.  Applications can still open evdev devices and
access them directly like they already do.

11 years agoxwm: do not crash when activating a NULL surface
Giulio Camuffo [Tue, 13 Aug 2013 09:42:02 +0000 (11:42 +0200)]
xwm: do not crash when activating a NULL surface

11 years agoevdev-touchpad: Set some options using weston.ini
Armin K [Tue, 30 Jul 2013 23:41:03 +0000 (01:41 +0200)]
evdev-touchpad: Set some options using weston.ini

This patch adds 3 new options to weston.ini to allow
the user to change default constant_accel_factor,
min_accel_factor and max_accel_factor. If no options
are set, it falls back using defaults as it did before.

v2: create weston_config_section_get_double and use it
instead of manualy converting string to double.

v3: add default values in weston_config_get_double
instead of using conditionals.

v4: don't pass diagonal as pointer.

11 years agoweston-launch: Only pass non-NULL value into setenv()
Rob Bradford [Fri, 9 Aug 2013 10:30:38 +0000 (11:30 +0100)]
weston-launch: Only pass non-NULL value into setenv()

getenv() can return NULL is the key is not set, passing NULL into
setenv() is an error

11 years agoevdev: write EV_SYN input_event along with led events to device
Rolf Morel [Fri, 9 Aug 2013 14:32:17 +0000 (16:32 +0200)]
evdev: write EV_SYN input_event along with led events to device

Other clients of an evdev device need to have the events they receive
be separated, in moment in time, from other events by an EV_SYN/
SYN_REPORT. This is the responsibility of the client who writes events
into the stream.

11 years agonested: Use a void * to avoid warnings with EGL_WL_bind_wayland_display
Kristian Høgsberg [Tue, 13 Aug 2013 05:21:19 +0000 (22:21 -0700)]
nested: Use a void * to avoid warnings with EGL_WL_bind_wayland_display

11 years agobuild: Allow more control over cairo use in the clients
Rob Bradford [Fri, 9 Aug 2013 15:52:46 +0000 (16:52 +0100)]
build: Allow more control over cairo use in the clients

Previously the configure script would silently disable the use of
accelerated cairo in the clients if cairo-gl could not be found (or
cairo-glesv2 if that was requested.) Conversely the use of cairo-gl
would be automatically enabled if it was found with no way to disable
that feature

This change adds --with-cairo which takes one of "image", "gl" or
"glesv2" (defaulting to "image"). If "gl" or "glesv2" is specified
cairo-egl is checked for as well as the specified renderer. If the check
fails then the configure process errors out.

11 years agoscreenshooter: Add missing newline and use program_invocation_short_name
Kristian Høgsberg [Tue, 13 Aug 2013 05:15:38 +0000 (22:15 -0700)]
screenshooter: Add missing newline and use program_invocation_short_name

11 years agoscreenshooter: Only link to $(CLIENT_LIBS) not libtoytoolkit.la
Kristian Høgsberg [Tue, 13 Aug 2013 04:45:19 +0000 (21:45 -0700)]
screenshooter: Only link to $(CLIENT_LIBS) not libtoytoolkit.la

This application doesn't use toytoolkit, so don't link to it.  We have to
add a copy of xmalloc, to link without toytoolkit.

11 years agoInstall shared/zalloc.h
Giulio Camuffo [Sat, 10 Aug 2013 12:41:04 +0000 (14:41 +0200)]
Install shared/zalloc.h

Plugins need to have it available, as it is included in compositor.h.

https://bugs.freedesktop.org/show_bug.cgi?id=67948

11 years agofix compile error when EGL_BUFFER_AGE_EXT is missing
Adrian Negreanu [Mon, 12 Aug 2013 13:30:11 +0000 (16:30 +0300)]
fix compile error when EGL_BUFFER_AGE_EXT is missing

Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
11 years agoremove dependency on EGL_KHR_surfaceless_context
Adrian Negreanu [Mon, 12 Aug 2013 07:07:59 +0000 (10:07 +0300)]
remove dependency on EGL_KHR_surfaceless_context

cairo_egl_device_create(), which is called next,
already checks if EGL_KHR_surfaceless_context is
available. If not, it fallbacks to pbuffer.

Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
11 years agoinput: For serial generation get the display from the compositor
Rob Bradford [Mon, 22 Jul 2013 16:31:38 +0000 (17:31 +0100)]
input: For serial generation get the display from the compositor

This removes the use of wl_client_get_display() where the client is
derived from the focussed resource. This starts the removal of the
assumption of a single resource on a client that would be notified about
events on the focussed surface.