Issue #1072: add LSAN_DO_LEAK_CHECK macro

[jiridanek]

• Embrace shutdown leaks and adjust leak checking for the resource cleanup strategy currently in use #1072

This is just a proposal, but the general ideas are ready do be reviewed:

• Orderly router shutdown and resource cleanup #293

The motivation in the decision to close issue 293 is to call lsan explicitly and early. As a possible future extension, lsan could be runnable upon request from Python system_tests (via skmanage, for example)?

The disadvantage of this is that it will become impossible to start an lsan-enabled router multiple times in the same process (which is what (currently disabled) cpp_system tests do, (when run with lsan)). Whether such ability is (or might be in the future) valuable regarding improved testability is up to a discussion. From what I observed so far, it is not very promising avenue.

Other projects (like CPython) usually have some special leak-checking mode that is enabled with sanitizers. For example, python has an allocation list that collects all allocations so that they can be then orderly freed upon exit, if lsan is watching. This seems a bit too much like a silly workaround and waste of effort on first sight, but it gives them the ability to use this to fix leaks that happen during regular runtime (if you fix all leaks, then you will naturally fix also the important ones).

[codecov]

Codecov Report

Merging #1073 (a9bea85) into main (32286a2) will increase coverage by 0.1%.
Report is 16 commits behind head on main.
The diff coverage is n/a.

Additional details and impacted files

@@           Coverage Diff           @@
##            main   #1073     +/-   ##
=======================================
+ Coverage   77.7%   77.8%   +0.1%     
=======================================
  Files        247     247             
  Lines      63999   64000      +1     
  Branches    5896    5896             
=======================================
+ Hits       49738   49831     +93     
+ Misses     11564   11493     -71     
+ Partials    2697    2676     -21     

Flag	Coverage Δ
pysystemtests	87.8% <ø> (+<0.1%)	⬆️
pyunittests	54.6% <ø> (ø)
systemtests	71.4% <ø> (+0.2%)	⬆️
unittests	21.9% <ø> (+<0.1%)	⬆️

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

Components	Coverage Δ
unittests	25.9% <ø> (+<0.1%)	⬆️
systemtests	78.5% <ø> (+0.1%)	⬆️