https://bugzilla.gnome.org/show_bug.cgi?id=538438

At least two places before 90b36abe commit
used screen coordinates. Update meta_window_get_titlebar_rect function to
return titlebar rectange in screen coordinates.

This also reverts first change in place.c that did not make any sense. x and
y in titlebar_rect always was 0 meaning that we always used CASCADE_FUZZ
value while previously visible border size was used.

https://bugs.launchpad.net/ubuntu/+source/metacity/+bugs/1509031

https://bugs.launchpad.net/ubuntu/+source/metacity/+bug/1592953

free_win function already has two error traps, but there is other
functions that could generate X errors. Replace existing traps with
one that is added for whole function.

It seems that picture might be destroyed with window if it is
created by passing window id as drawable to XRenderCreatePicture.

Add error trap around XRenderFreePicture to avoid Metacity crash
with RenderBadPicture error.

https://bugs.launchpad.net/ubuntu/+source/metacity/+bug/1582687

https://bugzilla.gnome.org/show_bug.cgi?id=436416

Owen W. Taylor authored Mar 21, 2011 and Alberts Muktupāvels committed May 23, 2016

We assume when applying a mask that it is a depth-1 pixmap that
cairo interprets as alpha. So ignore a non-depth-1 mask.

https://bugzilla.gnome.org/show_bug.cgi?id=641975

Owen W. Taylor authored Mar 21, 2011 and Alberts Muktupāvels committed May 23, 2016

Getting the contents of a depth-1 pixmap through cairo gives us
an alpha pixmap. We need to convert to a monochrome pixmap
as is expected by the ICCCM definition of WM_HINTS.

https://bugzilla.gnome.org/show_bug.cgi?id=641975

Jasper St. Pierre authored Feb 28, 2011 and Alberts Muktupāvels committed May 23, 2016

apply_mask made some wrong assumptions that became untrue
after rendering_cleanup landed, such as the number of channels
in the pixmap, causing corruption.

https://bugzilla.gnome.org/show_bug.cgi?id=641975

Jasper St. Pierre authored Jun 23, 2015 and Alberts Muktupāvels committed Apr 15, 2016

Going from fullscreen to unfullscreen involves a frame border size, so
in order to properly interpret the saved rect size, we need to make sure
that the frame borders are fully up to date.

When unmaximizing, we changed bits of window state, then called
out to code that used the frame extents *before* we cleared old
cached extents. Clear the cache up-front as soon as we change the
window state.

Based on mutter commit:
https://git.gnome.org/browse/mutter/commit/?id=8d5ab6b5b34a3dbeaba6033c31d06090579ed574

Remove window from compositor when calling meta_window_free and
make sure to re-add window to the compositor when we re-create
MetaWindow on MapRequest.

https://bugzilla.gnome.org/show_bug.cgi?id=504876
https://bugzilla.gnome.org/show_bug.cgi?id=751833

Owen W. Taylor authored Nov 15, 2013 and Alberts Muktupāvels committed Apr 13, 2016

Cache the computed border size so we can fetch the border size at
any time without worrying that we'll be spending too much time in
the theme code (in some cases we might allocate a PangoFontDescription
or do other significant work.)

The main effort here is clearing the cache when various bits of window
state change that could potentially affect the computed borders.

https://bugzilla.gnome.org/show_bug.cgi?id=707194

Florian Müllner authored Oct 05, 2015 and Alberts Muktupāvels committed Apr 11, 2016

As design patterns have evolved, dialogs that use CSD do use titlebar
buttons, so it's time to re-enable them for SSD as well.

https://bugzilla.gnome.org/show_bug.cgi?id=641630

https://bugzilla.gnome.org/show_bug.cgi?id=759310

This partially reverts commit d982d0a5.

Iain Lane authored Mar 24, 2016 and Alberts Muktupāvels committed Mar 24, 2016

As we opt out of GTK+/Clutter's HiDPI handling, we need to apply the
window scaling factor manually to decorations, both the geometry and
when drawing.

Backport of 643ce3e1,
7e5058da and
e6cf7fd3 with extra changes.