CVE-2024-31082

Fixes: 14205ade ("XQuartz: appledri: Fix byte swapping in replies")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <!1463>

CVE-2024-31081

Fixes: d220d690 ("Xi: add GrabButton and GrabKeysym code.")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <!1463>

CVE-2024-31080

Reported-by: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=69762


Fixes: 53e821ab ("Xi: add request processing for XIGetSelectedEvents.")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <!1463>

Compiler gives warning:

    ../config/wscons.c: At top level:
    ../config/wscons.c:262:1: warning: no previous prototype for ‘config_wscons_init’ [-Wmissing-prototypes]
      262 | config_wscons_init(void)
          | ^~~~~~~~~~~~~~~~~~
    ../config/wscons.c:270:1: warning: no previous prototype for ‘config_wscons_fini’ [-Wmissing-prototypes]
      270 | config_wscons_fini(void)
          | ^~~~~~~~~~~~~~~~~~

Problem is: commit bcee84c3 wanted to enable
wscons support on NetBSD, but only added compiling config/wscons.c, but didn't
enable setting the CONFIG_WSCONS symbol, so it won't be called at all.

Fixes: bcee84c3
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1450>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 11 months ago

When X11 isn't installed directly at /usr hierarchy (eg. NetBSD uses
/usr/X11R7/), build breaks:

../test/list.c:31:10: fatal error: X11/Xlib.h: No such file or directory
   31 | #include <X11/Xlib.h>
      |          ^~~~~~~~~~~~

Needs explicitly dependency on libX11, so the include path is added.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1442>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 11 months ago

Fix warning:

../hw/xfree86/drivers/modesetting/driver.c:1612:19: warning: unused variable ‘pEnt’ [-Wunused-variable]
 1612 |     EntityInfoPtr pEnt = ms->pEnt;
      |                   ^~~~

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1443>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 11 months ago

Fix warning:

../config/wscons.c:47:11: warning: initialization discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
   47 |  { KB_UK, "gb" }, \
      |           ^~~~
../config/wscons.c:58:5: note: in expansion of macro ‘KB_OVRENC’
   58 |     KB_OVRENC,
      |     ^~~~~~~~~
../config/wscons.c:48:11: warning: initialization discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
   48 |  { KB_SV, "se" }, \
      |           ^~~~
[...]

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <xorg/xserver!1444>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 11 months ago

Fix meson warning:

> WARNING: Project targeting '>= 0.56.0' but tried to use feature deprecated since '0.55.0': ExternalProgram.path. use ExternalProgram.full_path() instead

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1448>

The Xprintf() function is deprecated. Most of it's call sites had been moved
to asprintf(), so also do it for the remaining two ones.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1449>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

globals.h shouldn't be cluttered with extension specific things, that
aren't even used by external modules/drivers.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1328>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

Those fields are better off in the extension instead of the OS layer.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1328>

Was previously being shown as part of previous entry.

Fixes: 41277766 ("Disallow byte-swapped clients by default")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <!1439>

Fixes: 41277766 ("Disallow byte-swapped clients by default")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <xorg/xserver!1439>

Otherwise it causes the server to return BadDrawable giving a
byte-swapped resource id instead of the real id the client sent.

Reported-by: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=69762


Fixes: 397dfd9f ("Create/Destroy/Trigger/Reset/Query Fence Sync objs")
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Part-of: <!1437>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

fix warning:

> ../test/xi2/protocol-eventconvert.c:276:9: warning: variable 'buttons' set but not used [-Wunused-but-set-variable]
>     int buttons, valuators;
>         ^

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1429>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

fix warning:

> ../test/simple-xinit.c:58:1: warning: function 'usage' could be declared with attribute 'noreturn' [-Wmissing-noreturn]

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <xorg/xserver!1429>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

Tidy up the huge file a little bit.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <xorg/xserver!1433>

Enrico Weigelt, metux IT consult . authored 1 year ago and Marge Bot committed 1 year ago

fix warning:

> ../os/rpcauth.c:159:16: warning: result of comparison of constant -1 with expression of type 'XID' (aka 'unsigned int') is always false [-Wtautological-constant-out-of-range-compare]
>     if (rpc_id == ~0L)
>         ~~~~~~ ^  ~~~

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1431>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

It's much cleaner to always include directly what one needs,
instead of relying on very indirect including.

Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Part-of: <!1435>

Implement fractional scale with Xwayland rootful by scaling the content
to the desired fractional scale using a viewport.

For now this applies to Xwayland rootful only.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Fractional scaling may not be available, or not suitable for the current
configuration (e.g. if running rootless).

Add a helper function to tell whether fractional scaling should be used.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Add support for wp_fractional_scale_v1 protocol.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

This is needed to support the "wp-fractional-scale-v1" protocol.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

To support the fractional scale protocol, we need a viewport.

Rename the existing function xwl_window_enable_viewport() to avoid
confusion with the viewport we use for fullscreen XRandR emulation in
rootless mode.

No functional change.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Recompute and update the global screen scale based on the different
outputs the root window is placed on.

For backward compatibility, this functionality is however disabled by
default and can be enabled using a new command line option "-hidpi".

That option has no effect if Xwayland is running rootless.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Recompute the window scale each time the window enters or leaves an
output.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Add a list of outputs a window is placed on, adding an output whenever
the surface enters the output and removing it once it leaves the output.

Note that not all Wayland compositors actually send a leave surface
event on output removal, so we need to make sure to remove the output
from the list for each window, otherwise we might end up pointing to
freed memory.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

By using a sensible scale factor for input even when there is no
viewport enabled, no need to have xwl_window_has_viewport_enabled()
public anymore.

Small cleanup, no functional change.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

The viewport's scale_x/y is currently applied to the motion event only.

Apply the same viewport_scale_x/y to all relevant input coordinates.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

When the viewport is disabled, set the scale x/y back to 1.0 so that we
can apply the scale factor regardless of the viewport being enabled.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

The scale_x/y factor applies when a viewport is in use, rename the
fields to reflect that and distinguish these from the other scale
factors such as the core protocol surface scale and the fractional
scaling.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

Apply the scale factor to the root window and adjust the coordinates and
hotspot location for cursors.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>

For now, the global surface scale is always 1, no functional change.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-By: Kenny Levinsen <kl@kl.wtf>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <!1197>