Power Manager applet shows hex/null bytes for battery vendor/model from UPower

[sporteka2]

The Power Manager applet shows garbled text as the battery device name. This is because UPower returns vendor/model as raw SBS hex strings with null bytes instead of decoded text.

UPower output:

• vendor: 0x53 0x4D 0x50 0x00 0x32 0x30 0x31 0x39 0x00 0x00 0x00 0x00 (should be "SMP")
• model: 0x4C 0x31 0x39 0x4D 0x33 0x50 0x46 0x34 (should be "L19M3PF4")

Full output: https://files.catbox.moe/ga748g.txt

The vendor and model strings from UPower contain raw hex values and null bytes that should be stripped/decoded before display. The applet should sanitize these strings (e.g., filter out non-printable characters and trim null bytes).

See also: the popup content does not fit on screen (issue #13855) which compounds the problem — users cannot even see the garbled name properly.

[mtwebster]

What is the output from upower -d?

Don't ignore the rest of the issue template details, more info please.

[sporteka2]

Here is the output of upower -d:

Device: /org/freedesktop/UPower/devices/battery_BAT1
  vendor:               0x53 0x4D 0x50 0x00 0x32 0x30 0x31 0x39 0x00 0x00 0x00 0x00
  model:                0x4C 0x31 0x39 0x4D 0x33 0x50 0x46 0x34
  serial:               5650
  ...
  technology:          lithium-ion

The hex values decode to:

• vendor: 0x53 0x4D 0x50 = SMP (Samsung SDI), rest is binary junk with null terminators
• model: 0x4C 0x31 0x39 0x4D 0x33 0x50 0x46 0x34 = L19M3PF4

So UPower itself is reporting vendor/model as raw hex bytes rather than readable UTF-8 strings. This is what the Power Manager applet is trying to display directly.

[sporteka2]

Also noticed another power-related issue: pressing the power button shuts down the laptop instead of suspending, even though gsettings get org.cinnamon.settings-daemon.plugins.power button-power returns suspend.

Root cause: systemd defaults to HandlePowerKey=poweroff in /etc/systemd/logind.conf. Cinnamon places a block inhibitor via csd-media-keys to handle the power key itself, but if csd fails to process the event (race condition, crash, etc.), systemd falls through to its default — which powers off.

Workaround: setting HandlePowerKey=suspend in /etc/systemd/logind.conf. But ideally Cinnamon should either:

• Ensure the inhibitor is always active before the desktop starts
• Or set systemd's HandlePowerKey explicitly during session startup

Not sure if this is a separate bug or related to power management.