[warning] module sway/workspaces: Disabling module "sway/workspaces", Unable to connect to the SYSTEM Bus!...
[warning] module sway/mode: Disabling module "sway/mode", Unable to connect to the SYSTEM Bus!...
[warning] module sway/scratchpad: Disabling module "sway/scratchpad", Unable to connect to the SYSTEM Bus!...
[warning] module custom/media: Disabling module "custom/media", Unable to connect to the SYSTEM Bus!...
[warning] module sway/window: Disabling module "sway/window", Unable to connect to the SYSTEM Bus!...
[warning] module cpu: Disabling module "cpu", Unable to connect to the SYSTEM Bus!...
[warning] module memory: Disabling module "memory", Unable to connect to the SYSTEM Bus!...
[warning] module temperature: Disabling module "temperature", Unable to connect to the SYSTEM Bus!...
[warning] module sway/language: Disabling module "sway/language", Unable to connect to the SYSTEM Bus!...
[warning] module battery: Disabling module "battery", Unable to connect to the SYSTEM Bus!...
[warning] module battery#bat2: Disabling module "battery#bat2", Unable to connect to the SYSTEM Bus!...
The segfaults were happening on GTK icon theme functions, which are
called via the C++ interface functions such as Gtk::IconTheme::has_icon.
There are multiple modules and threads using this functions on the default
icon theme by calling Gtk::IconTheme::get_default(), which returns the same
object for all callers, and was causing concurrent access to the same internal
data structures on the GTK lib. Even a seemingly read-only function such as
has_icon can cause writes due to the internal icon cache being updated.
To avoid this issues, a program wide global mutex must be used to ensure
a single thread is accessing the default icon theme instance.
This commit implements wrappers for the existing IconTheme function calls,
ensuring the global lock is held while calling the underling GTK functions.
After upgrading to the latest release of Waybar the bar will crash
whenever I close the laptop lid. After some debugging I believe it is
because the watching added by watch_name is not being correctly canceled
using unwatch_name. After the Tray object and Host object are destroyed,
additional callbacks will become use-after-free.
Looks like commit 3af1853260 removed the
unwatch_name. I'm not sure why it did that, but it seemed dangerous.
Additionally, bus_name_id_ is created by own_name. According to that
function's documentation, the correct inverse operation is unown_name.
