Multishell support and add assertion on min. timeout (bugfix)

[Hook25]

Description

In the launchpad builder for Noble, the shell selected is sh. This is how I discovered that sh's kill doesn't actually support the tree kill (-PID semantic). This adds a fallback mechanism (that tries to use known working shells instead of the default one), then tries the default one and if it fails tries one last time to at least kill the direct child.

This also fixes a race condition, when join timeout is 0, in very unlucky circumstances the subprocess may not be up after the join, leading to the kill failing and the subprocess leaking free. Also, a timeout of 0 doesn't make sense.

Resolved issues

Fixes: CHECKBOX-1701

Documentation

N/A

Tests

Update tests

[codecov]

Codecov Report

Attention: Patch coverage is 81.81818% with 2 lines in your changes missing coverage. Please review.

Project coverage is 48.97%. Comparing base (aac167f) to head (c3972d9).
Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
...eckbox-support/checkbox_support/helpers/timeout.py	81.81%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1654   +/-   ##
=======================================
  Coverage   48.96%   48.97%           
=======================================
  Files         371      371           
  Lines       40296    40306   +10     
  Branches     6808     6811    +3     
=======================================
+ Hits        19730    19738    +8     
- Misses      19846    19847    +1     
- Partials      720      721    +1     

Flag	Coverage Δ
checkbox-support	61.10% <81.81%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[fernando79513]

Quite interesting discovery.
LGTM+1!