The window module re-entered the same shared_mutex while refreshing IPC state:
update() took the lock and then called queryActiveWorkspace(), which tried to
lock it again. That is undefined behavior for std::shared_mutex and could
manifest as a deadlock.
Remove the recursive lock path and reset the derived window state before each
IPC refresh. That keeps solo/floating/swallowing/fullscreen classes from
sticking around when the client lookup fails or a workspace becomes empty.
Signed-off-by: Austin Horstman <khaneliman12@gmail.com>
The first crash occurs when trying to parse the
ID of a workspace as an uint, since named
workspaces has negative IDs. This is fixed by
using ints for workspace IDs instead of uints.
The second crash occurs when converting a
workspace name that isn't a number to an integer.
This is fixed by wrapping std::stoi in a try
block and only sorting by number, when both names
can successfully be converted to integers.
The AButton class is designed as full a substitute to ALabel. The
GtkButton attribute 'button_' is initialized with a label. This
label can the be referenced by the subsequent inheritors of AButton
instead of the GtkLabel attribute 'label_' of ALabel.
For convenience a GtkLabel* 'label_' attribute is added to AButton.
If the button cannot be clicked it is disabled, effectively acting
like its label predecessor.
GtkButton seems to catch one-click mouse events regardless of the
flags set on it. Therefore, 'signal_pressed' is connected to a
function creating a fake GdkEventButton* and calling 'handleToggle'
(for details on this possible bug in GTK see:
https://stackoverflow.com/questions/45334911 )
In accordance with other GtkButtons (i.e. the sway/workspace ones)
set_relief(Gtk::RELIEF_NONE) is called on the 'button_' instance.
