Juan Zhao [Mon, 19 Mar 2012 17:48:46 +0000 (01:48 +0800)]
weston_surface: initialize pitch
Initializing pitch to 1 to avoid xxx/pitch errors
This won't influence the valid texture coordinate calculation, because
in that case buffer_attach will provide the correct value.
Bill Spitzak [Sat, 17 Mar 2012 22:22:03 +0000 (15:22 -0700)]
weston: don't hang in wait()
Pausing weston (with ^Z in the shell) and putting it in the background
made it hang.
Bill Spitzak [Sat, 17 Mar 2012 20:43:52 +0000 (13:43 -0700)]
weston: Don't ask for EGL depth buffer if not needed
This makes the compositor and demo clients work on the current nouveau
nvfx driver. Obviously does not fix any clients that actually want a
depth buffer, but this does allow more people to at least try wayland.
Tiago Vignatti [Fri, 16 Mar 2012 20:33:03 +0000 (17:33 -0300)]
evdev: use mtdev for multitouch devices
mtdev library translates all multitouch based devices to the slotted evdev
protocol. It provides an uniform interface for Weston, which eases mt
implementation when dealing with a big variety of devices.
Weston on drm now directly depends on such library.
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Ander Conselvan de Oliveira [Fri, 16 Mar 2012 15:25:11 +0000 (17:25 +0200)]
compositor: properly restore keyboard_focus in notify_keyboard_focus()
Commit
f992b2fc removed the saved keyboard focus logic to fix a crash
when the saved surface is destroyed. However, setting keyboard focus to
the first surface on the list ends up trying to set the focus to the
cursor surface most of the time. The end result is a NULL keyboard
focus.
This patch restores the saved keyboard focus logic and fixes the crash
mentioned above using a destroy listener.
Ander Conselvan de Oliveira [Fri, 16 Mar 2012 15:25:10 +0000 (17:25 +0200)]
compositor: allow multiple input_device_attach()'s with a nil buffer
Without this change, weston would crash whenever a nil buffer was
passed to input_device_attach() if the cursor sprite was not mapped.
While at it, change the unmapping code to use weston_surface_unmap().
Ander Conselvan de Oliveira [Fri, 16 Mar 2012 15:25:09 +0000 (17:25 +0200)]
compositor-wayland: use image loading code from shared/
Function weston_load_image() was deleted in
f02a649a but the wayland
backend was not adapted to the new interface. This probably went
unoticed because the prototype for the missing function was not deleted
from compositor.h so the backend would compile without warnings.
Ustun Ergenoglu [Wed, 14 Mar 2012 20:07:58 +0000 (22:07 +0200)]
cairo-util: load_cairo_surface returns NULL if loading the file fails.
Otherwise a non-existent file in the configuration crashes the desktop shell.
Signed-off-by: Ustun Ergenoglu <ego@ustun.fi>
Tiago Vignatti [Mon, 12 Mar 2012 22:40:09 +0000 (19:40 -0300)]
compositor: use smoother range for backlight control
now it goes from 0 to 255.
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Tiago Vignatti [Mon, 12 Mar 2012 22:40:08 +0000 (19:40 -0300)]
tests: add backlight test
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Alex Wu [Mon, 12 Mar 2012 08:06:01 +0000 (16:06 +0800)]
shell: Fix 2 switcher crashes.
In case of no regular window, switcher will crash in switch_next()
due to "next == NULL" and in switcher_destroy() due to
"switcher->current == NULL".
Eric Anholt [Thu, 9 Feb 2012 02:02:00 +0000 (18:02 -0800)]
automake: Don't fail the build if we can't setuid.
The configure default is to setuid root the weston compositor.
However, if installing as non-root (say, to your prefix in homedir),
the install fails anyway, even if you didn't need setuid to run weston
in your configuration.
Eric Anholt [Thu, 9 Feb 2012 01:46:31 +0000 (17:46 -0800)]
configure: Fix build on debian by avoiding bashisms in the test command.
Kristian Høgsberg [Mon, 12 Mar 2012 05:05:25 +0000 (01:05 -0400)]
Consolidate image loading code and move to shared/
Kristian Høgsberg [Mon, 12 Mar 2012 01:05:57 +0000 (21:05 -0400)]
Add an option parser
On one hand, getopt (in particular the -o suboption syntax) sucks on the
server side, and on the client side we would like to avoid the glib
dependency. We can roll out own option parser and solve both problems
and save a few lines of code total.
Kristian Høgsberg [Mon, 12 Mar 2012 05:18:24 +0000 (01:18 -0400)]
x11: Move the xcb fd to the input loop
This way, we handle input just before repaint, similar to the drm backend.
Kristian Høgsberg [Sun, 11 Mar 2012 20:35:16 +0000 (16:35 -0400)]
Fix picking crashes
The surface list is only valid after repaint, so we need to do picking and
input delivery there.
Rob Clark [Mon, 12 Mar 2012 00:48:44 +0000 (19:48 -0500)]
drm: fix issue copying supported plane formats
Originally on the first 1/4th of the plane formats were copied, leading
to problems if AR24/XR24 appears later in the list.
Rob Clark [Mon, 12 Mar 2012 00:48:41 +0000 (19:48 -0500)]
clients/window: fix build issues w/ gles
David Herrmann [Sun, 11 Mar 2012 19:05:22 +0000 (20:05 +0100)]
compositor-drm: use udev_device_get_property_value() helper
Udev provides a convenient helper. Use it instead of working with the
property-list directly.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
David Herrmann [Sun, 11 Mar 2012 19:05:21 +0000 (20:05 +0100)]
compositor-drm: ignore hotplug-events from other devices
If we have multiple video devices on the system (card0, card1), we should
ignore hotplug events for cards that we do not use. This avoids calling
update_outputs() if the event was not generated by our device so we avoid
refreshing the DRM information if it didn't change.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Christopher Michael [Wed, 7 Mar 2012 19:55:21 +0000 (14:55 -0500)]
drm: Fix drmModeRes leak on error paths
When creating outputs in the drm compositor, if allocating crtcs
fails, then free the drm resources. Also, if the base output list is
empty, free drm resources
Benjamin Franzke [Wed, 7 Mar 2012 18:42:10 +0000 (19:42 +0100)]
compositor: Fix compilation without xserver launcher
Ander Conselvan de Oliveira [Mon, 5 Mar 2012 13:39:23 +0000 (15:39 +0200)]
shell: don't assign output for surface of type none
If map is called with a surface of type none it will call
weston_surface_assign_output, even though the surface will
not be mapped.
This change was lost with the changes for using wl_layers.
Alex Wu [Mon, 5 Mar 2012 06:13:13 +0000 (14:13 +0800)]
shell: Fix segment fault when clicking to activate a NULL pointer_focus
This can happen for instance running a toytoolkit application, launching
the popup menu and then clicking on a surface that does not belong to
the client that create the popup surface.
Alex Wu [Mon, 5 Mar 2012 03:01:40 +0000 (11:01 +0800)]
shell: Fix not destroy black surface when fullscreen surface exit abnomally.
When a client with fullscreen surface displayed was aborted by Ctrl-C, the
black surface still be there. Destroy the black surface in
destroy_shell_surface().
Kristian Høgsberg [Tue, 6 Mar 2012 01:51:34 +0000 (20:51 -0500)]
shell: Put popups and transients on top of their parent
In other words, don't map the panel menu under the panel.
Kristian Høgsberg [Tue, 6 Mar 2012 01:38:08 +0000 (20:38 -0500)]
window: Compute frame boundary for all windows except fullscreen
We were only computing it for toplevel type windows, which broke
the unlock dialog size allocation and repainting.
Kristian Høgsberg [Tue, 6 Mar 2012 01:37:51 +0000 (20:37 -0500)]
desktop-shell: Simplify unlock dialog redraw function
Kristian Høgsberg [Tue, 6 Mar 2012 00:57:35 +0000 (19:57 -0500)]
drm: Disable sprites if they're broken
Current sprite implementation has a color format bug. Disable sprites
if they don't work and avoid spamming stderr and syslog.
Kristian Høgsberg [Tue, 6 Mar 2012 00:50:08 +0000 (19:50 -0500)]
compositor: Only process input once per frame
When we're repainting, there's no point in polling for input events.
We just read input events once before each repaint and send out events
as needed. The input events come with an accurate timestamp, so this
doesn't affect the timing information and client should always look at
the event timestamps if they're trying to determine pointer motion
speed or double click speed. If we go idle (stop repainting) we add the
input devices back into the primary main loop and wait for the next event.
This avoids waking up the compositor separately (one or more times per
frame) to handle input events. We also avoid updating cursor position
and other compositor state after the client has rendered its new
frame, reducing lag between what the client renders and the pointer
position.
Kristian Høgsberg [Mon, 5 Mar 2012 22:49:30 +0000 (17:49 -0500)]
compositor: Move input devices to their own event loop
First step towards only handling input at repaint time.
Kristian Høgsberg [Mon, 5 Mar 2012 22:47:15 +0000 (17:47 -0500)]
compositor: Swallow binding key for bindings that don't set up a grab
This prevents passing the key press and release to the keyboard focus client.
Kristian Høgsberg [Mon, 5 Mar 2012 04:47:56 +0000 (23:47 -0500)]
window: Only set opaque and input regions if they change
Setting these regions damages the entire window or causes a repick, which
we don't want to trigger if we don't need to.
Kristian Høgsberg [Mon, 5 Mar 2012 03:35:47 +0000 (22:35 -0500)]
Use wl_array_for_each() where possible
Pekka Paalanen [Fri, 2 Mar 2012 15:33:17 +0000 (17:33 +0200)]
compositor-drm: fix libbacklight build
Bad inclusion of DRM headers, missing some system headers.
Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
Kristian Høgsberg [Mon, 5 Mar 2012 02:57:37 +0000 (21:57 -0500)]
Use typesafe server-side stubs
Kristian Høgsberg [Sun, 4 Mar 2012 19:53:40 +0000 (14:53 -0500)]
compositor: Add a repaint debugging mechanism
Toggle on/off with Super+SPACE.
Kristian Høgsberg [Fri, 2 Mar 2012 13:14:40 +0000 (08:14 -0500)]
Merge branch 'layers'
I was integrating patches on the 'layers' branch and had a fixed version of
the fullscreen patch there. My master branch had the fullscreen patch with
the resize bug, and when I pushed I meant to push the layers branch to master.
Conflicts:
src/shell.c
Kristian Høgsberg [Thu, 1 Mar 2012 22:11:36 +0000 (17:11 -0500)]
shell: Don't launch a new screensaver if the old one is still running
Kristian Høgsberg [Thu, 1 Mar 2012 22:10:25 +0000 (17:10 -0500)]
compositor: Damage sprite surface on attach
Normally the client is responsible for damaging the surface after attach,
but the sprite surface is an internal surface.
Kristian Høgsberg [Thu, 1 Mar 2012 22:09:37 +0000 (17:09 -0500)]
compositor: Post keyboard leave before destroying surface
We may want to move this into wayland-server.
Ander Conselvan de Oliveira [Thu, 1 Mar 2012 12:09:44 +0000 (14:09 +0200)]
compositor: split drag surface update into smaller functions
Ander Conselvan de Oliveira [Thu, 1 Mar 2012 12:09:44 +0000 (14:09 +0200)]
compositor: fix crash when a drag surface is destroyed during the drag
This can happen for instance if the client that started the drag
crashes. Weston would crash because of the invalid surface pointed by
device->drag_surface.
Fix this by reseting the drag surface to nil on a destroy listener.
Kristian Høgsberg [Wed, 29 Feb 2012 17:42:35 +0000 (12:42 -0500)]
compositor: Organize surface stack using new struct weston_layer
The surface data structure is now a list of list of surfaces. The core
compositor defines the fade and cursor layer, and it's up to the shell to
provide more layers for the various surface types it implements.
Kristian Høgsberg [Wed, 29 Feb 2012 04:07:47 +0000 (23:07 -0500)]
compositor: Just damage compositor->damage in weston_output_damage()
Kristian Høgsberg [Wed, 29 Feb 2012 03:47:14 +0000 (22:47 -0500)]
compositor: Implement damage_below by using the new compositor->damage
We can now clip the surface bounding box against the previous frame
opaque clip, and then just union the result (visible damage) into
compositor->damage immediately.
Kristian Høgsberg [Wed, 29 Feb 2012 03:31:58 +0000 (22:31 -0500)]
compositor: Store opaque clip for previous frame in weston_surface
Alex Wu [Thu, 1 Mar 2012 04:57:46 +0000 (12:57 +0800)]
shell: Add implementation of fullscreen.
Undo fullscreen in shell_unset_fullscreen(), do all the stacking order
in shell_stack_fullscreen(), and configure black surface, method in
shell_configure_fullscreen().
Signed-off-by: Alex Wu <zhiwen.wu@linux.intel.com>
Signed-off-by: Juan Zhao <juan.j.zhao@linux.intel.com>
Alex Wu [Thu, 1 Mar 2012 04:57:46 +0000 (12:57 +0800)]
shell: Add implementation of fullscreen.
Undo fullscreen in shell_unset_fullscreen(), do all the stacking order
in shell_stack_fullscreen(), and configure black surface, method in
shell_configure_fullscreen().
Signed-off-by: Alex Wu <zhiwen.wu@linux.intel.com>
Signed-off-by: Juan Zhao <juan.j.zhao@linux.intel.com>
Kristian Høgsberg [Wed, 29 Feb 2012 21:15:53 +0000 (16:15 -0500)]
window.c: Drop support for rgb contexts
Kristian Høgsberg [Wed, 29 Feb 2012 21:14:22 +0000 (16:14 -0500)]
wscreensaver: Use the argb config, rgb is not compatible with window.c
Fixes the config/surface mismatch, spotted by Scott Moreau.
Kristian Høgsberg [Wed, 29 Feb 2012 19:23:51 +0000 (14:23 -0500)]
compositor-drm: Just pull in libbacklight source for now
libbacklight is 300 lines of code in one .c file, and we're relying on
udev changes that aren't yet upstream. For now, let's just keep a
copy in weston and if the Xorg DDX drivers start using libbacklight and
it becomes more widely available, we'll make it an external dependency.
Tiago Vignatti [Wed, 29 Feb 2012 17:53:50 +0000 (19:53 +0200)]
compositor: add dpms and backlight support
DPMS kicks in only when wscreensaver is launched, in the moment that shell
call lock() for the second time. Backlight control internals are managed by
libbacklight:
http://cgit.freedesktop.org/~vignatti/libbacklight/
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Kristian Høgsberg [Wed, 29 Feb 2012 04:12:52 +0000 (23:12 -0500)]
zoom: Don't damage_all here
Setting geometry.dirty = 1 is enough.
Kristian Høgsberg [Wed, 29 Feb 2012 04:38:53 +0000 (23:38 -0500)]
compositor: Insert dnd icon surface below device sprite surface
Kristian Høgsberg [Wed, 29 Feb 2012 17:22:38 +0000 (12:22 -0500)]
compositor: Set an empty input region for the fade surface
Otherwise we don't set the cursor until the fade is done.
Kristian Høgsberg [Wed, 29 Feb 2012 02:52:12 +0000 (21:52 -0500)]
window: Set opaque region after child widget adjusts size
Rodney Lorrimar [Wed, 29 Feb 2012 16:31:03 +0000 (17:31 +0100)]
weston: Change path for default terminal launcher to ${prefix}/bin
If you don't have anything at ~/.config/weston-desktop-shell.ini and
have weston installed somewhere other than /usr, then this patch will
help.
Cheers,
Signed-off-by: Rodney Lorrimar <rodney@rodney.id.au>
Kristian Høgsberg [Tue, 28 Feb 2012 19:29:19 +0000 (14:29 -0500)]
clients: Use wl_display_disconnect()
Ander Conselvan de Oliveira [Tue, 28 Feb 2012 15:59:33 +0000 (17:59 +0200)]
compositor: make the input region of drag surfaces empty
With the input region changes, drag surfaces were being picked and
receiving events which led to problems on the client side.
Kristian Høgsberg [Tue, 28 Feb 2012 14:20:21 +0000 (09:20 -0500)]
shell: Mark geometry dirty when we change surface alpha
We need to recompute the opaque region when the alpha changes.
Kristian Høgsberg [Tue, 28 Feb 2012 14:19:39 +0000 (09:19 -0500)]
compositor: Dont set the opaque region if we have an overall surface alpha
Kristian Høgsberg [Tue, 28 Feb 2012 14:17:32 +0000 (09:17 -0500)]
terminal: Don't queue a resize after setting fullscreen
We end up doing an attach at the non-fullscreen size before resizing to
fullscreen, causing the terminal to jump to the center for a frame before we
render the fullscreen image.
Alex Wu [Tue, 28 Feb 2012 10:07:54 +0000 (18:07 +0800)]
compositor: Export functions and add weston_suface::force_configure for fullscreen
This will facilitate the implementation of fullscreen.
Signed-off-by: Alex Wu <zhiwen.wu@linux.intel.com>
Scott Moreau [Fri, 24 Feb 2012 05:28:37 +0000 (22:28 -0700)]
Restructure output zoom key handling.
This effectively eats the keybinding events, as we don't want them sent
to clients.
Kristian Høgsberg [Fri, 24 Feb 2012 16:03:27 +0000 (11:03 -0500)]
drm: Remove redundant cursor region math
We already check this in weston_output_set_cursor().
Kristian Høgsberg [Fri, 24 Feb 2012 04:07:45 +0000 (23:07 -0500)]
drm: Remove redundant weston_surface_update_transform() call
Kristian Høgsberg [Fri, 24 Feb 2012 04:06:13 +0000 (23:06 -0500)]
compositor: Remove surface overlapped attribute
We need to compute this as we assign surfaces to overlays, so that
if we assign a surface to an overlay, it isn't considered overlapping.
Kristian Høgsberg [Fri, 24 Feb 2012 04:03:14 +0000 (23:03 -0500)]
Move weston_output_set_cursor into compositor-drm.c
First step towards moving hw cursor handling into compositor-drm.c a its
general overlay handling.
Kristian Høgsberg [Fri, 24 Feb 2012 02:45:32 +0000 (21:45 -0500)]
compositor-drm: Destroy sprites on shutdown, turn off on vt switch
Kristian Høgsberg [Fri, 24 Feb 2012 02:29:55 +0000 (21:29 -0500)]
compositor: Remove unused compositor vfunc create_cursor_image
Jesse Barnes [Thu, 23 Feb 2012 14:45:49 +0000 (09:45 -0500)]
compositor-drm: Add sprite support
Add support for assigning surfaces to overlay sprites using the new
assign_planes hook.
Jesse Barnes [Thu, 9 Feb 2012 21:12:57 +0000 (13:12 -0800)]
compositor: add an assign_planes hook to the output
This allows each output back end to optimize drawing using overlay planes
and cursors (yet to be integrated). If a surface is assigned to a
plane, the back end should clear its damage field so that the later
repaint code won't look at it.
Scott Moreau [Wed, 22 Feb 2012 20:57:51 +0000 (13:57 -0700)]
Fix pointer position clipping.
Scott Moreau [Wed, 22 Feb 2012 21:21:41 +0000 (14:21 -0700)]
Introduce output zoom.
Ideally, we would want to use <modifier>+Scroll binding but that will have
to wait for axis events. For now we just use keybindings. Zoom in/out with
Super+Up/Down.
Kristian Høgsberg [Thu, 23 Feb 2012 22:38:33 +0000 (17:38 -0500)]
compositor: Initialize destroy_listener_list for internal surfaces too
We use this to look up user data on the surface, so make sure the list is
always valid.
Kristian Høgsberg [Thu, 23 Feb 2012 22:30:45 +0000 (17:30 -0500)]
window: Track and report input and opaque regions
We just set the input region to the bounding box of the window frame
and set the opaque region to be the opaque rectangle inside the window
if the child widget is opaque.
Kristian Høgsberg [Thu, 23 Feb 2012 21:20:38 +0000 (16:20 -0500)]
simple-touch: Update to enter/leave changes
Kristian Høgsberg [Thu, 23 Feb 2012 21:11:59 +0000 (16:11 -0500)]
compositor: Implement opqaue and input regions
Ander Conselvan de Oliveira [Thu, 23 Feb 2012 11:29:26 +0000 (13:29 +0200)]
compositor: assign output to drag surfaces
Otherwise we endup with a surface that is mapped but with outtup nil.
Ander Conselvan de Oliveira [Thu, 23 Feb 2012 11:29:25 +0000 (13:29 +0200)]
shell: don't assign output for surface of type none
If map is called with a surface of type none it will call
weston_surface_assign_output, even though the surface will
not be mapped.
Kristian Høgsberg [Thu, 23 Feb 2012 14:59:05 +0000 (09:59 -0500)]
Update to enter/leave events for pointer and keyboard
Alex Wu [Wed, 22 Feb 2012 06:50:46 +0000 (14:50 +0800)]
shell: Fix never map fullscreen/transient surface.
If client send set_fullscreen/set_transient request before the first attach,
compositor has no chance to map the surface due to "if (es->output == NULL)".
You can pull it from git://gitorious.org/wayland-for-krh/weston.git map-bug
Kristian Høgsberg [Mon, 20 Feb 2012 03:46:13 +0000 (22:46 -0500)]
compositor: Remove redundant weston_surface_damage_below() calls
In the attach case we always end up calling weston_surface_damage_below()
through weston_surface_update_transform() if and only if we need it.
Kristian Høgsberg [Mon, 20 Feb 2012 03:15:21 +0000 (22:15 -0500)]
compositor: Surface is unmapped when surface->output is NULL
surface->buffer can be NULL, which happens when a client destroys the
attached buffer. Fixes list corruption on resize.
Kristian Høgsberg [Sun, 19 Feb 2012 23:52:44 +0000 (18:52 -0500)]
shell: Revive super-tab application switcher
We can do this right, now that we have keyboard grabs.
Kristian Høgsberg [Sun, 19 Feb 2012 23:50:55 +0000 (18:50 -0500)]
compositor: Don't run keybindings if we have a key grab
Scott Moreau [Sat, 18 Feb 2012 12:05:30 +0000 (05:05 -0700)]
compositor: Setup for keyboard grabs.
Scott Moreau [Sat, 18 Feb 2012 12:05:29 +0000 (05:05 -0700)]
Track server changes.
Kristian Høgsberg [Sat, 18 Feb 2012 04:49:07 +0000 (23:49 -0500)]
compositor: Move fullscreen_output to shell_surface
Kristian Høgsberg [Fri, 17 Feb 2012 17:45:43 +0000 (12:45 -0500)]
compositor: Make weston_surface_damage_below_noupdate() the default
We never want to update the transform and then damage below. Damage
below is always used to trigger a repaint where the surface used to be
so we need to record the damage before updating the transform.
Kristian Høgsberg [Fri, 17 Feb 2012 17:43:56 +0000 (12:43 -0500)]
compositor: Move unmap logic to new function weston_surface_unmap()
Kristian Høgsberg [Fri, 17 Feb 2012 17:15:27 +0000 (12:15 -0500)]
compositor: Only delete surface texture if we have one
Kristian Høgsberg [Fri, 17 Feb 2012 15:45:16 +0000 (10:45 -0500)]
compositor: Get rid of surface->visual
Kristian Høgsberg [Fri, 17 Feb 2012 04:43:59 +0000 (23:43 -0500)]
shell: Set geometry directly in configure
We already set half of it directly, and overwrite it in other cases, so
just skip weston_surface_configure().
Kristian Høgsberg [Fri, 17 Feb 2012 04:38:14 +0000 (23:38 -0500)]
shell: Restore basic fullscreen functionality
This just fixes the default case.
Kristian Høgsberg [Fri, 17 Feb 2012 03:57:21 +0000 (22:57 -0500)]
window: Don't draw decorations in fullscreen mode
Kristian Høgsberg [Fri, 17 Feb 2012 03:56:12 +0000 (22:56 -0500)]
window: Set window type as it changes, not on every attach
It was always a bit sloppy, and the new fullscreen request doesn't work
that way.
Kristian Høgsberg [Fri, 17 Feb 2012 03:51:33 +0000 (22:51 -0500)]
view: Don't schedule resize after setting fullscreen
window_set_fullscreen will schedule a resize or wait for the configure
event as needed.