Remove options memory leak during consent setting

[stima]

Do not do early exit loop early, that leads to a incorrect reference counting of options during consent setting.

[stima]

Not sure if it worth adding some testing around that.

[codecov]

Codecov Report

Merging #922 (a1dbb0f) into master (0b17731) will increase coverage by 8.54%.
The diff coverage is 89.47%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #922      +/-   ##
==========================================
+ Coverage   74.10%   82.64%   +8.54%     
==========================================
  Files          52       53       +1     
  Lines        7170     7376     +206     
  Branches     1150     1186      +36     
==========================================
+ Hits         5313     6096     +783     
+ Misses       1310     1171     -139     
+ Partials      547      109     -438     

[supervacuus]

Thanks, @stima, nice catch!

By the way, the correct approach to exit these synchronization scopes early would be to use continue instead of break. I looked through the code to see if there was another occurrence of such an early exit, but there was none. From my side, it is okay to add the nesting in this function.

Not sure if it worth adding some testing around that.

It is 💯 worth it, given that triggering this bug in a test is trivial (you only need to consent twice in a row), and our unit tests are run through ASAN in CI, which would have revealed this earlier if the branch had been covered. Not adding this case as a regression test would be careless.

Please give another consent here

sentry-native/tests/unit/test_consent.c

Line 30 in a51ccbb

sentry_user_consent_give();

and add a short comment that explains why it was done with a reference to this PR (you can also write a separate test if you want, but that should be enough). Please also add the fix to the changelog. Thx!

[supervacuus]

Can you please fix the formatting of your comment https://github.com/getsentry/sentry-native/actions/runs/7165776324/job/19508507795?pr=922#step:3:90 accordingly? Thx!