The variable '__environ' seems to be libc implementation specific, and
not avaible on Android.

Use the POSIX standard variable 'environ', which also luckily happens to
be available on Android, which is not POSIX.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

evdev.c: In function 'evdev_led_update':
evdev.c:57:9: warning: ignoring return value of 'write', declared with
attribute warn_unused_result

Useless in this case.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Android backend needs mtdev now, and evdev-touchpad.c, too.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Weston's notify_keyboard_focus_*() assume that a keyboard is present, if
they are called. With evdev, there might not always be a keyboard.

Also clean up the variable definition in evdev_notify_keyborad_focus().
I read that function through many times and finally had to grep where
does 'all_keys' come from.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

All input devices are re-added during enter VT, so these messages in the
log will clarify what happened with input.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

If the sysname of the udev device did not start with "event", the
function returned without unreferencing udev_device.

The function is refactored to have a common exit path that unrefs
udev_device. The return value semantics are not changed.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Write information about found input devices into the log. Also fetch and
record the device name.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Move all udev-related and now drm backend specific code into
compositor-drm.c.

This makes evdev.c free of udev and launcher-util, and allows it to be
used on Android.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

This API does not depend on udev or launcher-util.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

in preparation of removing evdev_seat

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

in preparation of removing evdev_seat

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

in preparation of removing evdev_seat

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

We are phasing out struct evdev_seat.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

This makes the generic evdev code (i.e. the functions not relying on
udev) independent of launcher-util too. The aim is to allow re-using the
generic evdev code in the Android backend, where neither udev nor
launcher-util are available.

evdev_input_device_create() signature is changed:
- add the opened device file descriptor
- remove wl_display as unused

Also add a bit of failure logging.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

evdev_input_device_destroy() will completement the API of
evdev_input_device_create(), both being independent from udev.

Since the udev-specific device_removed() would only call
evdev_input_device_destroy() and do nothing else, device_remove() calls
are simply replaced with evdev_input_device_destroy().

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Since the time values are unsigned integers we can check whether the msec is
smaller than spring->timestamp by checking if the subtraction overflows into a
value greater than half the maximum unsigned integer range (ie. top bit set)

isalpha() and isdigit() don't like huge values.
Noticed with ascii-art in the shell prompt.

Like how we handle overlays, we split drm_output_set_cursor() into
drm_output_prepare_cursor_surface() that will assign the hw cursor to
a suitable surface and drm_output_set_cursor() that will do the actual
work of updating cursor contents and moving it.  This is more in line
with how we handle everything else, and lets us update cursor contents
based on changes to the cursor plane damage.

This works when the cursor content changes but also accumulates
damage when the cursor changes output.  In that case we move the
cursor surface to the cursor plane in the new output and
weston_surface_move_to_plane() will generate the damage that triggers
download of sprite contents there.

When we analyze and accumulate damage prior to repainting, we need to
accumulate damage per plane, so that whatever damage a surface
contributes is accumulated in the plane that it's assigned to.  Before,
we would always accumulate damge in the primary plane, which caused
repaints in the primary plane whenever a surface in a sprite or
framebuffer was damaged.  Eliminating this repaint is a big win for
cases where we pageflip to a client surface or use a sprite overlay.

This also prepares for fixing the missing cursor updates, since we
now track damage to the cursor surface in a dedicated sprite plane.

Fixes visual bell in screen.

This backend has not seen even build testing for months, presumably does
not even compile, and is starting to hinder development a little.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>
Cc: Benjamin Franzke <benjaminfranzke@googlemail.com>

This chunk of code is what uses the move grab and used to come right
after it.  At some point it got split up, but move it back now.

We're trying to keep the API exposed by the core compositor in compositor.h

This allows specifying a modeline in the config for the 'mode' key
in the output section, such as one you would get from cvt.

This fix xwayland.so that was complaining about jpeg symbols from the
libshared.

Note that xwayland is not using the jpeg ones, but now that symbols are read
up front we need to link anyway.. or break apart that library.

Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>

Should reduce update overhead.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Fix the off by one error in checking whether we can draw the marker
without exceeding buffer dimensions.

Fixes a segfault.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

Without this, it won't get any input anymore.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>

In the wl_seat conversion, struct wl_touch got fields for the focused
surface and the client resource for the input device being focused.
However, the conversion was incomplete: the old fields
weston_seat::touch_focus* we still used by the event dispatching code,
but the new code never set them. Therefore no touch events were ever
sent.

From weston_seat, remove the fields touch_focus, touch_focus_listener,
touch_focus_resource, and touch_focus_resource_listener. They are
replaced by the corresponding fields and listeners from struct
wl_touch.

While doing this, fix touch_set_focus().

If touch_set_focus() was called first with surface A, and then with
surface B, without being called with NULL in between, it would corrupt
the destroy_signal list. It was equivalent of calling wl_signal_add()
for different signal sources with the same listener without removing in
between.

Now, touch_set_focus() first removes focus and listeners, and then
attempts to assign focus if requested. If the target client has not
subscribed for touch events, the touch focus will now be NULL.
Before this patch, the touch focus was left to the previous surface.

NOTE: this patch depends on the patch "server: add lose_touch_focus()"
for Wayland.

Signed-off-by: Pekka Paalanen <ppaalanen@gmail.com>