This change tracks, for each wlr_seat_client, the most recent serial
numbers which were sent to the client. When the client makes a
selection request, wlroots now verifies that the serial number
associated with the selection request was actually provided to that
specific client. This ensures that the client that was most
recently interacted with always has priority for its copy selection
requests, and that no other clients can incorrectly use a larger serial
value and "steal" the role of having the copy selection.
Also, the code used to determine when a given selection is superseded
by a newer request uses < instead of <= to allow clients to make
multiple selection requests with the same serial number and have the
last one hold.
To limit memory use, a ring buffer is used to store runs of sequential
serial numbers, and all serial numbers earlier than the start of the
ring buffer are assumed to be valid. Faking very old serials is
unlikely to be disruptive.
Assuming all clients are correctly written, the only additional
constraint which this patch should impose is that serial numbers
are now bound to seats: clients may not receive a serial number
from an input event on one seat and then use that to request
copy-selection on another seat.
This commit makes sure surface->mapped is true when the unmapped event is
emitted. This is necessary because listeners can only damage surfaces that are
mapped. This is similar to the fact that the destroy event is emitted before
any destruction is actually made.
Fixes https://github.com/swaywm/sway/issues/3568
Since xwm only manipulates the stack when focusing a window, newly
mapped windows should be stacked below the focused window. This prevents
the newly mapped window from stealing focus due to being on the top of
the stack.
This supersedes f24e17259e and
04c9ca4198. These commits were manually removing
wlr_data_source destroy handlers when starting a new drag. This is error-prone.
Instead, this commit destroys the previous source whenever we start a new drag.
This makes compositors able to block and/or customize set_selection requests
coming from clients. For instance, it's possible for a compositor to disable
rich selection content (by removing all MIME types except text/plain). This
commit implements the design proposed in [1].
Two new events are added to wlr_seat: request_set_selection and
request_set_primary_selection. Compositors need to listen to these events and
either destroy the source or effectively set the selection.
Fixes https://github.com/swaywm/wlroots/issues/1138
[1]: https://github.com/swaywm/wlroots/issues/1367#issuecomment-442403454
This has been causing troubles for some of our users and only been there
for legacy reasons, we trust Xwayland just as much as your next program
and weston doesn't take any such care when starting it.
This is a common interface that can be used for all primary selection
protocols, as discussed in [1]. A new function wlr_seat_set_primary_selection
is added to set the primary selection for all protocols.
The seat now owns again the source, and resets the selection to NULL when
destroyed.
[1]: https://github.com/swaywm/wlroots/issues/1367#issuecomment-442403454
This commits completely refactors wlr_gtk_primary_selection. The goal is to
remove gtk-primary-selection state from the seat and better handle inert
resources where it makes sense.
wlr_seat_client.primary_selection_devices has been removed and replaced by
wlr_gtk_primary_selection_device. This allows us to make offers inert when the
current selection is replaced.
wlr_seat_set_primary_selection has been removed because it relied on wlr_seat
instead of wlr_gtk_primary_selection_device_manager. A new function,
wlr_gtk_primary_selection_device_manager_set_selection (candidate for the
longest function name in wlroots) has been added. It doesn't take a serial
anymore as serial checking only makes sense for set_selection requests coming
from Wayland clients (serial checking is now done in the Wayland interface
implementation).
Since wlr_gtk_primary_selection_device_manager is now required to set the
selection, a new function wlr_xwayland_set_gtk_primary_selection_device_manager
(candidate number two for longest function name) has been added.
Devices are now made inert when the seat goes away.
Future work includes removing the last primary selection bits from the seat,
mainly wlr_seat.primary_selection_source and wlr_seat.events.primary_selection,
replacing those with new fields in wlr_gtk_primary_selection_device. Or maybe
we could keep those in the seat and replace them with a re-usable interface
(for future zwp_primary_selection_v1 support). We need to think how we'll sync
these three protocols (GTK, X11 and wayland-protocols).
See https://github.com/swaywm/wlroots/issues/1388
An X11 client can leave the hints->input WM hint unspecified,
by not setting the XCB_ICCCM_WM_HINT_INPUT flag in hints->flags.
In that case, we should assume a sane default.
Make the hint default to true, so that clients which do not specify
the hint, like mupdf, still get keyboard focus.
This should fixswaywm/sway#2231
It's not used (XCB_IMAGE_FORMAT_Z_PIXMAP comes from xproto.h) and we
don't even have a pkg-config dependency on xcb-image, making the build
to fail on that inclusion on systems without the package.
153f37bdf5 (#1145) removed the
wlr_xwayland_is_unamanged function while fixing OR, because it was
belieived that it's supposed to work around the broken OR handling.
This was a misunderstanding. is_unmanaged is (while sort of a hack)
intended to work around inherent differences between "real" X sessions
and our Xwayland/wayland situation.
The main reason it exists is to support applications like rofi and dzen,
while not handing focus to other OR windows (which should *not* be
required).
Traditionally, these applications just grabbed input from X and didn't
need to be focused by any logic in the WM. Which of course doesn't work
in wayland compositors. So we have to give them focus in some way.
Giving *every* OR window focus, breaks other applications that don't
expect focus to change.
A testcase that was pointed out to me where wlr_xwayland_is_unamanged was
breaking things is https://github.com/swaywm/sway/issues/2128 (syncplay,
gitk, gitgui)
Supposedly it broke using keyboard to navigate the menus.
I can't reproduce this with this patch. The popups can be navigated as
long as the parent has focus.
The override_redirect flag can change on configure notify and
on map notify. This adds an event to know when it changes.
This removes wlr_xwayland_surface_is_unmanaged which was wrongly
using the window type to decide whether the view should be
unmanaged.
A similar patch was proposed to Weston, but has never been
merged upstream [1].
[1]: https://patchwork.freedesktop.org/patch/211161/
Happens when e.g. closing gimp.
==24039==ERROR: AddressSanitizer: heap-use-after-free on address 0x6150001a7a78 at pc 0x7f09b09f1bb2 bp 0x7ffcf0237bf0 sp 0x7ffcf0237be0
WRITE of size 8 at 0x6150001a7a78 thread T0
#0 0x7f09b09f1bb1 in wl_list_remove ../util/signal.c:55
#1 0x7f09b094cf03 in xwayland_surface_destroy ../xwayland/xwm.c:295
#2 0x7f09b0950245 in xwm_handle_destroy_notify ../xwayland/xwm.c:717
#3 0x7f09b095304a in x11_event_handler ../xwayland/xwm.c:1149
#4 0x7f09b0c68f01 in wl_event_loop_dispatch src/event-loop.c:641
#5 0x7f09b0c67601 in wl_display_run src/wayland-server.c:1260
#6 0x40a2f4 in main ../sway/main.c:433
#7 0x7f09b011018a in __libc_start_main (/lib64/libc.so.6+0x2318a)
#8 0x40b749 in _start (/opt/wayland/bin/sway+0x40b749)
0x6150001a7a78 is located 120 bytes inside of 496-byte region [0x6150001a7a00,0x6150001a7bf0)
freed by thread T0 here:
#0 0x7f09b2b58880 in __interceptor_free (/lib64/libasan.so.5+0xee880)
#1 0x7f09b094d1a1 in xwayland_surface_destroy ../xwayland/xwm.c:315
#2 0x7f09b0950245 in xwm_handle_destroy_notify ../xwayland/xwm.c:717
#3 0x7f09b095304a in x11_event_handler ../xwayland/xwm.c:1149
#4 0x7f09b0c68f01 in wl_event_loop_dispatch src/event-loop.c:641
#5 0x7f09b0c67601 in wl_display_run src/wayland-server.c:1260
#6 0x40a2f4 in main ../sway/main.c:433
#7 0x7f09b011018a in __libc_start_main (/lib64/libc.so.6+0x2318a)
#8 0x40b749 in _start (/opt/wayland/bin/sway+0x40b749)
previously allocated by thread T0 here:
#0 0x7f09b2b58e50 in calloc (/lib64/libasan.so.5+0xeee50)
#1 0x7f09b094b585 in xwayland_surface_create ../xwayland/xwm.c:119
#2 0x7f09b0950151 in xwm_handle_create_notify ../xwayland/xwm.c:706
#3 0x7f09b0953032 in x11_event_handler ../xwayland/xwm.c:1146
#4 0x7f09b0c68f01 in wl_event_loop_dispatch src/event-loop.c:641
#5 0x7f09b0c67601 in wl_display_run src/wayland-server.c:1260
#6 0x40a2f4 in main ../sway/main.c:433
#7 0x7f09b011018a in __libc_start_main (/lib64/libc.so.6+0x2318a)
#8 0x40b749 in _start (/opt/wayland/bin/sway+0x40b749)
Makes the xwayland startup process two phased.
The first phase just initialises the X11 sockets.
The second phase starts the Xwayland server itself.
When starting xwayland lazily the second phase will be postponed until
a client has connected to the X11 socket.
Changes in behaviour:
The DISPLAY environment is now set immediately after the X11 sockets
are created.
When the Xwayland server is killed or crashes, the sockets will not be
recreated, but reused.
Fixes#849: Start up Xwayland lazily
Some comboboxes (e.g. in chrome://flags) are advertized as…
Notifications of course! Yeah, notifications, the thing that
tells you you have mail, your battery is low, or the dog has
eaten your carpet. This isn't the first time we notice Chromium's
X11 backend is pretty shit.
Anyway, added notifications and splash screens to the list of
unmanaged windows. Also removed utility windows because those
should be managed, but maybe I'm wrong and I'll revert this.
This unbreaks the build on armhf that otherwise fails like
../xwayland/selection/incoming.c: In function 'xwm_data_source_write':
../include/wlr/util/log.h:34:17: error: format '%ld' expects argument of type 'long int', but argument 6 has type 'ssize_t {aka int}' [-Werror=format=]
_wlr_log(verb, "[%s:%d] " fmt, wlr_strip_path(__FILE__), __LINE__, ##__VA_ARGS__)
^
../xwayland/selection/incoming.c:34:2: note: in expansion of macro 'wlr_log'
wlr_log(L_DEBUG, "wrote %zd (chunk size %ld) of %d bytes",
^~~~~~~
../xwayland/selection/incoming.c:34:44: note: format string is defined here
wlr_log(L_DEBUG, "wrote %zd (chunk size %ld) of %d bytes",
~~^
%d
xwm.h was meant to be private, so move it to include/xwayland/xwm.h
We had an ifdef WLR_HAS_XCB_ICCCM in xwayland.h which was easy to move
to xwm, it is not safe to use the WLR_HAS_* in the public headers.
I checked a few of our current users and none rely on xwm.h being
public as expected (rootston, sway, hsroots)
Now message can look like:
[xwayland/xwm.c:991] xcb error: op ChangeProperty (no minor), code Window (no extension), value 6291465
instead of this one when the lib is not available:
[xwayland/xwm.c:999] xcb error: op 18:0, code 3, sequence 103, value 6291465
The value in case of Window is the window id, so we can tell what
function applied on which window which is a good start.
The sequence ought to be able to tell us more precisely which
invocation it was, but we never log it when calling functions
so is useless in practice and no longer logged.
gcc 6.3.0 (at least) complains about &values:
expected ‘const uint32_t * {aka const unsigned int *}’ but argument
is of type ‘uint32_t (*)[1] {aka unsigned int (*)[1]}’
Reported by thorwil on irc
This adds `wlr_xwayland_surface_is_unamanged`, to allow compositors more
fine grained control over XWayland focus.
A surface that is unmanaged should not receive focus, while other
windows that are just override redirect may want it (dmenu).
The way unamanged is determined is taken from wlc.
This is more robust than trying to recopy a few fields.
Note: wlr_xwayland->events.ready now resets after each trigger, so one can
hook on it again and be called when restart is ready
Xwayland uses SIGUSR1 to signal readiness.
With direct(-freebsd) session and Xwayland, wlroots was confusing the
Xwayland readiness signal with a vt switch signal, freezing the desktop.
Same problem was found in Weston in 2014:
https://phabricator.freedesktop.org/T7080
