GitLab

A surface can get destroyed while a shell grab is active, which can
for example happen if the command running in weston-terminal exits.

When a surface gets destroyed, grab->shsurf is reset to NULL by
destroy_shell_grab_shsurf(), but otherwise the grab remains active and
its callbacks continue to be called. Thus, dereferencing grab->shsurf
in a callback without checking it for NULL first can lead to undefined
behavior, including crashes.

Several functions were already properly checking grab->shsurf for NULL,
move_grab_motion() being one example. Others, however, were not, which
is what this commit fixes.

Related to wayland/weston#192Signed-off-by: Sergey Bugaev <bugaevc@gmail.com>

Changing the focused surface did remove the surface_destroy_listener from the wl_signal list,
but destroying the focus state did not. As a result, sometimes the same listener would be added
to two surfaces, which would join their wl_signal lists together, which would cause infinite
loops and use-after-frees when closing desktop surfaces.

This crash was happening when *releasing* a pointer button on a window
that was being resized and got destroyed during the grab.

[@daniels: Cosmetic fixes; apply same fix to grab cancel.]

This crash was happening *during resizing* of an xwayland window that was destroyed.

Discovered by: John Good @archiesix

[@daniels: Moved tests below declarations.]

If xwayland is disabled, compositor/weston is built without
compositor/xwayland.c, which defines wet_load_xwayland.

  compositor/fb12c4d@@weston@exe/main.c.o: In function `main':
  ../weston-5.0.0-169-g2d4cc4f4/compositor/main.c:3103: undefined reference to `wet_load_xwayland'

Provide an empty stub for wet_load_xwayland if xwayland is disabled.
With that we also have to remove xwayland.c from the autotools build
if xwayland is disabled, to avoid a multiple definition error.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

If the 'renderer-gl' option is enabled, ENABLE_EGL is defined, and
libweston/pixel-formats.c includes EGL/egl.h. This requires an egl
dependency, as X11-less platforms need the MESA_EGL_NO_X11_HEADERS
define from egl.pc cflags:

  In file included from /usr/include/EGL/egl.h:39:0,
                   from ../libweston/pixel-formats.c:42:
  /usr/include/EGL/eglplatform.h:124:10: fatal error: X11/Xlib.h: No such file or directory
  compilation terminated.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

If Weston is not installed, running ivi-layout test would fail on lots of image
files not found which presumably causes the creation of some ivi surfaces to
fail, leading to an assert failure.

Looking at the test setup in weston-tests-env, these IVI plugin tests are
supposed to run with --no-config instead.

Fixes: wayland/weston#195Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

It seems WESTON_DATA_DIR was missed. If you have already installed Weston, then
the files will be found in the install location, but if not, they were not
found at all.

This caused the xwayland test to SEGV the compositor in
weston_wm_window_create_frame() when frame_crate() returned NULL.

This patch fixes the test suite only.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

At least the following commits break the libweston ABI, strictly speaking:
- a5630eaf
- ce62cb3d
- 195dadeb
- c18ffd39
- ea54c2fd
- 27d7c395
- acff29b3
- 67629674
- c9c24773
- 65e1be12

Furhtermore, 64fbd0f4 break the weston
executable exported ABI.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Better to excercise the current rather than outdated protocol.

Pekka:
- split the patch, rewrote commit message
- rename xdg_shell_ping to xdg_wm_base_ping
- rename xdg_shell_listener to wm_base_listener
- fix continued line alignment
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Now that Weston supports the stable revision, use it. Better to excercise the
current rather than outdated protocol.

Pekka:
- split the patch, rewrote commit message
- rename xdg_shell_ping to xdg_wm_base_ping
- rename xdg_shell_listener to wm_base_listener
- rename shell to wm_base
- fix continued line alignment
- drop unrelated change of adding parentheses around bit-wise and
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Some clients like the mpv video player now request the xdg_shell
protocol so these will fail if the compositor only provides the
xdg_shell_unstable_v6 protocol. Compositors like mir and gnome provide
both protocols.

The two protocols are very similar therefore the code in xdg-shell-v6.c
has been refactored to work with the new xdg_shell protocol and now
resides in xdg-shell.c.

Pekka:
- split the patch
- fix continued line alignment

Daniel
- allow anchor_rect to initially have zero dimensions
- account for get_popup allowing NULL parent surface
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>

Autotools is going away. Break the autotools build so that people are
guaranteed to notice before it is gone. If they have problems with Meson, they
can still use --enable-autotools to build with autotools, but we really want to
hear about any problems.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Explain how to use with Meson, as autotools is going away.

The instructions have been copied from
https://wayland.freedesktop.org/building.html .
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

The in-tree clients can access the functions via libshared, but they are
currently not available for external clients, such as custom shell helper
applications similar to weston-desktop-shell or
weston-ivi-shell-user-interface.
The functions to read the content of the config are already exported.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Since ivi-shell now supports xdg-protocol, the surface_created listener
can be removed and the desktop_surface_configured listener is needed.
ivi-layout: libweston-desktop api is used to send configure event to
application.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Since the surface_destroy_listener is only registered for ivi-shell
applications, it should only be removed for ivi-shell applications.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

The libweston-desktop api is used to send configure
event to application.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Introduced surface create and configure function for xdg-apps.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Introduced signal configure_desktop_changed with interface. This signal
should be sent, when a desktop-surface is configured.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

This interface enables an id-agent to change the surface ids of an
ivi-layout-surface once.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Introduction of IVI_INVALID_ID for xdg-shell applications.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Removed assert, that checks if ivi-surface is not available, since this
can now happen with xdg-shell support.
Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Signed-off-by: Michael Teyfel <mteyfel@de.adit-jv.com>

Make it official that libweston will export the weston_config API, as requested
in wayland/weston!29 .

There is no other way third party helper clients could access the API.

The autotools build has been accidentally exporting it all the time, but the
Meson build needed fixing.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Support drawing a mandelbrot set in the fragment shader, rendering it
with separate draw calls, one for each cell in a virtual 4x4 grid. This
more complex and heavy drawing will potentially help us to visually
discover any present or future explicit synchronization issues.

The mandelbrot set rendering is enabled with the -m/--mandelbrot
command-line switch.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Changes in v7:
  - Merge acquire fence and release fence code blocks in redraw().
  - Use 1 << n to define option bitflags.
  - Remove redundant statement involving OPT_IMPLICIT_SYNC.

Changes in v6:
  - Add option for window size.
  - Add option for enabling/disabling explicit sync.

Changes in v5:
  - Meson support.

Add tests to check that the zwp_buffer_release_v1 events are emitted per
surface commit.

To be able to test this we need to use a renderer that holds the buffer
until the next buffer is committed, hence we use the pixman renderer.

Changes in v7:
  - Remove references to obsolete noop-hold renderer.

Changes in v5:
  - Meson support.

Changes in v4:
  - Use the pixman renderer instead of the (now gone) noop-hold
    renderer.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Implement the get_release request of the zwp_surface_synchronization_v1
interface.

This commit implements the zwp_buffer_release_v1 interface. It supports
the zwp_buffer_release_v1.fenced_release event for surfaces rendered by
the GL renderer, and the zwp_buffer_release_v1.immediate_release event
for other cases.

Note that the immediate_release event is safe to be used for surface
buffers used as planes in the DRM backend, since the backend releases
them only after the next page flip that doesn't use the buffers has
finished.

Changes in v7:
  - Remove "partial" from commit title and description.
  - Fix inverted check when clearing used_in_output_repaint flag.

Changes in v5:
  - Use the new, generic explicit sync server error reporting function.
  - Introduce and use weston_buffer_release_move.
  - Introduce internally and use weston_buffer_release_destroy.

Changes in v4:
  - Support the zwp_buffer_release_v1.fenced_release event.
  - Support release fences in the GL renderer.
  - Assert that pending state buffer_release is always NULL after a
    commit.
  - Simplify weston_buffer_release_reference.
  - Move removal of destroy listener before resource destruction to
    avoid concerns about use-after-free in
    weston_buffer_release_reference
  - Rename weston_buffer_release_reference.busy_count to ref_count.
  - Add documentation for weston_buffer_release and ..._reference.

Changes in v3:
  - Raise NO_BUFFER for get_release if no buffer has been committed,
    don't raise UNSUPPORTED_BUFFER for non-dmabuf buffers,
    so get_release works for all valid buffers.
  - Destroy the buffer_release object after sending an event.
  - Track lifetime of buffer_release objects per commit, independently
    of any buffers.
  - Use updated protocol interface names.
  - Use correct format specifier for resource ids.

Changes in v2:
  - Raise UNSUPPORTED_BUFFER at commit if client has requested a
    buffer_release, but the committed buffer is not a valid linux_dmabuf.
  - Remove tests that are not viable anymore due to our inability to
    create dmabuf buffers and fences in a unit-test environment.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.

The implementation uses the acquire fence in two ways:

1. If the associated buffer is used as GL render source, an
   EGLSyncKHR is created from the fence and used to synchronize
   access.
2. If the associated buffer is used as a plane framebuffer,
   the acquire fence is treated as an in-fence for the atomic
   commit operation. If in-fences are not supported and the buffer
   has an acquire fence, we don't consider it for plane placement.

If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.

Issues for discussion
---------------------

a. Currently, a server-side wait of EGLSyncKHR is performed before
   using the EGLImage/texture during rendering. Unfortunately, it's not clear
   from the specs whether this is generally safe to do, or we need to
   sync before glEGLImageTargetTexture2DOES. The exception is
   TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
   and then glBindTexture for any changes to take effect.

Changes in v5:
  - Meson support.
  - Make explicit sync server error reporting more generic, supporting
    all explicit sync related interfaces not just
    wp_linux_surface_synchronization.
  - Fix typo in warning for missing EGL_KHR_wait_sync extension.
  - Support minor version 2 of the explicit sync protocol (i.e., support
    fences for opaque EGL buffers).

Changes in v4:
  - Introduce and use fd_clear and and fd_move helpers.
  - Don't check for a valid buffer when updating surface acquire fence fd
    from state.
  - Assert that pending state acquire fence fd is always clear
    after a commit.
  - Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
    renderer.
  - Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
  - Dup the acquire fence before passing to EGL.

Changes in v3:
  - Keep acquire_fence_fd in surface instead of buffer.
  - Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
    renderer.
  - Move comment about non-ownership of in_fence_fd to struct
    drm_plane_state definition.
  - Assert that we don't try to use planes with in-fences when using the
    legacy KMS API.
  - Remove unnecessary info from wayland error messages.
  - Handle acquire fence for subsurface commits.
  - Guard against self-update in fd_update.
  - Disconnect the client if acquire fence EGLSyncKHR creation or wait
    fails.
  - Use updated protocol interface names.
  - User correct format specifier for resource ids.
  - Advertise protocol for X11 backend with GL renderer.

Changes in v2:
  - Remove sync file wait fallbacks.
  - Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
    fence, but the committed buffer is not a valid linux_dmabuf.
  - Don't put buffers with in-fences on planes that don't support
    in-fences.
  - Don't advertise explicit sync protocol if backend does not
    support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Introduce an internal API for dealing with linux sync files,
and use it in the codebase to replace ad-hoc sync file management.

The linux_sync_file_is_valid function is not currently used, but will be
utilized in upcoming commits to implement the
zwp_linux_explicit_synchronization_unstable_v1 protocol.

Changes in v5:
  - Meson support.

Changes in v3:
  - Use parameter name in function documentation.
  - Move kernel UAPI to separate header file.

Changes in v2:
  - Add function documentation
  - Remove linux_sync_file_wait()
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>

Introduce support for the zwp_linux_explicit_synchronization_unstable_v1
protocol with an implementation of the zwp_linux_explicit_synchronization_v1
interface.

Explicit synchronization provides a more versatile notification
mechanism for buffer readiness and availability, and can be used to
improve efficiency by integrating with related functionality in display
and graphics APIs.

In addition, the per-commit nature of the release events provided by
this protocol potentially offers a solution to a deficiency of the
wl_buffer.release event (see
wayland/wayland#46).

Support for this protocol depends on the capabilities of the backend, so
we don't register it by default but provide a function which each
backend will need to call. In this commit only the headless backend when
using the noop renderer supports this to enable testing.

Note that the zwp_surface_synchronization_v1 interface, which contains
the core functionality of the protocol, is not implemented in this
commit. Support for it will be added in future commits.

Changes in v7:
  - Added some information in the commit message about the benefits of
    the explicit sync protocol.

Changes in v6:
  - Fall back to advertising minor version 1 of the explicit sync protocol,
    although we support minor version 2 features, until the new
    wayland-protocols version is released.

Changes in v5:
  - Meson support.
  - Advertise minor version 2 of the explicit sync protocol.

Changes in v4:
  - Enable explicit sync support in the headless backend for all
    renderers.

Changes in v3:
  - Use wl_resource_get_version() instead of hardcoding version 1.
  - Use updated protocol interface names.
  - Use correct format specifier for resource id.
  - Change test name to 'linux-explicit-synchronization.weston'
    (s/_/-/g).

Changes in v2:
  - Move implementation to separate file so protocol can be registered
    on demand by backends.
  - Register protocol in headless+noop backend for testing purposes.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>