[BUG] Setting keyboard layout fails but returns success

Description
When the user applies a state to set a keyboard keymap that is not available,
the command fails, but salt reports a success.
I think I have found the cause and am working on a patch.

Setup
I think this would affect everyone using this module, but this is where
I encountered it:

• on-prem machine
• VM (libvirt, qemu, suse SLES15SP4 minion, provisioned with sumaform)
• VM running on a cloud service, please be explicit and add details
• container (Kubernetes, Docker, containerd, etc. please specify)
• or a combination, please be explicit
• jails if it is FreeBSD
• classic packaging
• onedir packaging
• used bootstrap to install

Steps to Reproduce the behavior

Run:
salt-call --local state.single keyboard.system ca.map.gz

[ERROR   ] Command 'localectl' failed with return code: 1
[ERROR   ] stdout: Failed to set keymap: Keymap ca.map.gz is not installed.
[ERROR   ] retcode: 1
[ERROR   ] Command 'localectl' failed with return code: 1
[ERROR   ] output: Failed to set keymap: Keymap ca.map.gz is not installed.
local:
----------
          ID: ca.map.gz
    Function: keyboard.system
      Result: True
     Comment: Set system keyboard layout ca.map.gz
     Started: 17:53:24.783471
    Duration: 214.29 ms
     Changes:   
              ----------
              layout:
                  ca.map.gz

Summary for local
------------
Succeeded: 1 (changed=1)
Failed:    0
------------
Total states run:     1
Total run time: 214.290 ms

Expected behavior
Result should be False, Succeeded should be 0, Failed should be 1.

Screenshots

Versions Report

Additional context
I think I've worked out why this happens and am working on a fix, over here:
https://github.com/saltstack/salt/compare/master...zzaimeche:salt:fix-keyboard-fails
I think Salt does not check whether the shell command was successful, but whether
the python calling the command ran successfully. So whether the shell command
succeeds or fails, Salt reports a success.
The issue is in https://github.com/saltstack/salt/blob/master/salt/modules/keyboard.py
and https://github.com/saltstack/salt/blob/master/salt/states/keyboard.py .
I think Salt needs to check for the retcode, so I used cmd.retcode based on other
examples in the codebase, though I couldn't find the documentation for cmd.retcode,
so maybe there's a better way. My WIP patch seems to fix
the issue on my machine (but my WIP patch will break things on other
distros at the moment). I'll make a PR shortly. I'm hoping someone from the Salt
team can take a look and advise if there's a preferred approach, since I suspect there
might have been similar issues with other modules and there may be a convention for
this kind of fix.

Read salt/modules/keyboard.py and salt/states/keyboard.py, then reproduce the failure with salt-call --local state.single keyboard.system ca.map.gz. Compare the behavior with the linked WIP patch and verify that a failed localectl command produces Result: False, Succeeded: 0, and Failed: 1 without breaking other distributions.