Spell out exactly when a client may re-use a wl_buffer or its backing
storage. Mention the optimization for GL-compositor with wl_shm-clients.
Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
This change breaks the protocol.
The current protocol is racy in that updates to surface content and
surface state (e.g. damage, input and opaque regions) are not guaranteed
to happen at the same time. Due to protocol buffering and handling
practices, the issues are very hard to trigger.
Committing damage to a surface at arbitrary times makes it hard to
track when the wl_buffer is being read by the server, and when it is
safe to overwrite (the case of wl_shm with a single buffer reused
constantly).
This protocol change introduces the concept of double-buffered state.
Such state is accumulated and cached in the server, unused, until the
final commit request. The surface will receive its new content and apply
its new state atomically.
A wl_surface.commit request is added to the protocol. This is thought to
be more clear, than having wl_surface.attach committing implicitly, and
then having another request to commit without attaching, as would be
required for a GL app that wants to change e.g. input region without
redrawing.
When these changes are implemented, clients do not have to worry about
ordering damage vs. input region vs. attach vs. ... anymore. Clients set
the state in any order they want, and kick it all in with a commit.
The interactions between wl_surface.attach, (wl_surface.commit,)
wl_buffer.release, and wl_buffer.destroy have been undocumented. Only
careful inspection of the compositor code has told when a wl_buffer is
free for re-use, especially for wl_shm and wrt. wl_surface.damage.
Try to clarify how it all should work, and what happens if the wl_buffer
gets destroyed.
An additional minor fix: allow NULL argument to
wl_surface.set_opaque_region. The wording in the documentation already
implied that a nil region is allowed.
Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
The only way to make the global object listener interface thread safe is to
make it its own interface and make different listeners different wl_proxies.
The core of the problem is the callback we do when a global show up or
disappears, which we can't do with a lock held. On the other hand we can't
iterate the global list or the listener list without a lock held as new
globals or listeners may come and go during the iteration.
Making a copy of the list under the lock and then iterating after dropping
the lock wont work either. In case of the listener list, once we drop the
lock another thread may unregister a listener and destroy the callbackk
data, which means that when we eventually call that listener we'll pass it
free memory and break everything.
We did already solve the thread-safe callback problem, however. It's what
we do for all protocol events. So we can just make the global registry
functionality its own new interface and give each thread its own proxy.
That way, the thread will do its own callbacks (with no locks held) and
destroy the proxy when it's no longer interested in wl_registry events.
This makes the scanner generate the code and meta data to send the
interface name and version when we pass a typeless new_id. This way, the
generic factory mechanism provided by wl_display.bind can be provided by
any interface.
My vim spell checker is able to find typos of xml files after adding "syn spell
toplevel" to ~/.vim/after/syntax/xml.vim
aah, and Wayland is capital letter :)
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Change the description of pointer.set_cursor() so that a client without
pointer focus but that currently owns the pointer surface can update the
hotspot using it.
The set_transient and set_popup requests take a wl_shell_surface as
the parent surface argument. We don't need the parent surface to be
a shell surface and this restricts the types of surfaces we can use
the transient and popup surface types on.
With this event, the compositor generates one canonical keymap for all
clients, and then sends an mmap()able fd over the wire, rather than all
the clients generating possibly disparate keymaps.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
This event sends the current keyboard modifier/group state from the
compositor to the client, allowing all clients to have a consistent view
of the keyboard state (e.g. current layout, Caps Lock, et al). It
should be sent after a keyboard enter event, and also immediately after
any key event which changes the modifier state.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Properly handle a drag with no data source, i.e., don't crash and send
events only to the client that initiated the drag. This way a client can
do self drag and drop without offering anything to other clients.
wl_input_device has been both renamed and split. wl_seat is now a
virtual object representing a group of logically related input devices
with related focus.
It now only generates one event: to let clients know that it has new
capabilities. It takes requests which hand back objects for the
wl_pointer, wl_keyboard and wl_touch interfaces it exposes which all
provide the old input interface, just under different names.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
These events let us track when a surface enters or leaves the scanout region of
an output. This way if a surface moves to another output and that output is on
a different card, we can suggest the toolkit reallocate its buffers
appropriately.
Signed-off-by: Casey Dahlin <cdahlin@redhat.com>
The wayland protocol, as X, uses timestamps to match up certain
requests with input events. The problem is that sometimes we need to
send out an event that doesn't have a corresponding timestamped input
event. For example, the pointer focus surface goes away and new
surface needs to receive a pointer enter event. These events are
normally timestamped with the evdev event timestamp, but in this case,
we don't have a evdev timestamp. So we have to go to gettimeofday (or
clock_gettime()) and then we don't know if it's coming from the same
time source etc.
However for all these cases we don't need a real time timestamp, we
just need a serial number that encodes the order of events inside the
server. So we introduce a serial number mechanism that we can use to
order events. We still need real-time timestamps for actual input
device events (motion, buttons, keys, touch), to be able to reason
about double-click speed and movement speed so events that correspond to user input carry both a serial number and a timestamp.
The serial number also give us a mechanism to key together events that
are "logically the same" such as a unicode event and a keycode event,
or a motion event and a relative event from a raw device.
There's a big cost to setting up and tearing down a mmap and faulting in
the pages to back it. For cases where we're continuously reallocating
shm wl_buffers (resizing a surface, typically) it is a big performance
improvement to be able to reuse a mmap area. This change makes the shm
buffer allocation a two step process: first allocate a wl_shm_pool, then
allocate a buffer from the pool. The wl_shm_pool encapsulate the shared
memory pool, and lets clients allocate wl_buffers backed by chunks of that
memory. Buffers are allocated at an offset into the pool, so it's possible
to create multiple buffers from one pool, for example for icons or cursor
images.
To reproduce, invoke distcheck from an out-of-source tree.
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
Distribute all source files that we need for buildling.
Plus, remove the html file on make clean.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Remove the absolute coordinate fields from the pointer motion and
pointer_focus events. Clients are not supposed to see any global
coordinates.
Fix wayland-server code accordingly. wayland-client code is unaffected.
Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
In the effort to make everything a regular surface, remove
data_device.attach request. To maintan the functionality, add
an icon surface parameter to data_device.start_drag.
Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Map the surface as a fullscreen surface. On the output the
surface is assigned to. The client can use different fulllscreen
method to fix the size mismatch issue: default, scale, driver
and fill.
Hints to indicate compositor how to deal with this fullscreen surface.
"default" means the client has no preference on fullscreen
behavior, policies are determined by compositor.
"scale" means the client prefers scaling by the compositor.
Scaling would always preserve surface's aspect ratio.
And the surface is centered.
"driver" means the client wants to switch video mode to the
smallest mode that can fit the client buffer. If the
sizes do not match, black borders are added. And the
framerate parameter is used for "driver" method,
to indicate the preferred framerate. framerate=0 means
that the app does not care about framerate
"fill" means the client wants to add blackborders to the
surface. This would be preferring 1:1 pixel mapping
in the monitor native video mode. The surface is
centered.
A request from the client to ask the compositor to maximize the surface.
The compositor will reply with a configure event telling
the expected new surface size. The operation is completed on the
next buffer attach to this surface.
A maximized client will fill the fullscreen of the output it is bound
to, except the panel area. This is the main difference between
a maximized shell surface and a fullscreen shell surface.
On Wed, 18 Jan 2012 12:29:37 -0800
"Kristensen, Kristian H" <kristian.h.kristensen@intel.com> wrote:
> Yeah, that looks good. I was thinking of a separate <description> tag
> to avoid stuffing too much into an attribute.
How does this look? It adds a summary attribute to atomic elements,
and a <description> tag with a summary for others. Spits out enum
documentation like this:
/**
* wl_display_error - global error values
* @WL_DISPLAY_ERROR_INVALID_OBJECT: server couldn't find object
* @WL_DISPLAY_ERROR_INVALID_METHOD: method doesn't exist on the specified interface
* @WL_DISPLAY_ERROR_NO_MEMORY: server is out of memory
*
* These errors are global and can be emitted in response to any server request.
*/
enum wl_display_error {
WL_DISPLAY_ERROR_INVALID_OBJECT = 0,
WL_DISPLAY_ERROR_INVALID_METHOD = 1,
WL_DISPLAY_ERROR_NO_MEMORY = 2,
};
and structure documentation like this:
/**
* wl_display - core global object
* @bind: bind an object to the display
* @sync: (none)
*
* The core global object. This is a special singleton object. It is used for
* internal wayland protocol features.
*/
struct wl_display_interface {
void (*bind)(struct wl_client *client,
struct wl_resource *resource,
uint32_t name,
const char *interface,
uint32_t version,
uint32_t id);
void (*sync)(struct wl_client *client,
struct wl_resource *resource,
uint32_t callback);
};
Requests like 'move' and 'set_toplevel' are really methods of a surface,
not methods of a global shell object. Move all these methods to a new
interface, wl_shell_surface.
The global object wl_shell will contain only 'get_shell_surface'
request, which creates and associates a wl_shell_surface object to a
given wl_surface object.
This will also give the shell plugin (if you look at the demo
compositor) means to store per-surface private data in a natural way.
Due to a limitation in delete_id event handling on client side, the
client must destroy its wl_shell_surface object before destroying the
wl_surface object. Otherwise it may just leak an id.
Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
This commit brings a big change to the DND and copy/paste interfaces.
Most importantly the functionality is now independent of wl_shell.
The wl_shell interface is intended for desktop style UI interaction and
an optional and experimental interface.
The new interface also allows receiving the DND data multiple times or
multiple times during the drag, and the mechanism for offering and receiving
data is now shared between DND and selections.
We need to make sure the client doesn't reuse an object ID until the
server has seen the destroy request. When a client destroys an ID
the server will now respond with the display.delete_id event, which lets
the client block reuse until it receives the event.
The visual interface was meant to be a generic mechanism for
specifying the content of a buffer. It goes back to before we had the
buffer factory interfaces (like wl_drm and wl_shm) and we wanted to
keep it open-ended enough that yuv, png or even svg buffer or so would
be possible.
Now that we have the buffer abstraction, we can add different buffer
types by introducing new interfaces that create buffers. It only
makes sense to leave it to those interfaces to specify the contents of
the buffers.
For wl_shm, this means that we now just specify the pixel format using
an enum. For EGL buffers, the exact pixel formats are controlled by
the implementation (part of wl_drm and similar), and from the client
point of view, everything is controlled using EGLConfigs.
So obvious in retrospect. The object system can do all the work for us
and keep track of pending calls as regular objects and we don't need to
abuse the resource system to get them cleaned up on client exit. We
don't need the custom key management or (broken) lookup, we just sue
object IDs. And last but not least, anybody can receive the callback,
not just display listeners.
Change 4453ba084a disallows using
post_global with objects not on the global list. Therefore selection
and drag offers have to be added to the global list from now on.
However these may often get replaced by a newer object and thus need a
way to remove a global from the global list.
The new map_toplevel() request no longer specifies a position and takes
the size from the attached buffer. The attach request now takes a
position relative to the top-left corner of the old buffer to let
clients specify the relative position of the new buffer.
