fix(gc): implement better way to handle root overload in inc_non_root_count

[shruti2522]

It changes the following:

saturating at NON_ROOTS_MAX pins the counter instead of panicking. pinned counter returns true for is_rooted so the object is kept alive rather than freed, a leak is safer than a UAF anyways. The limit ~2^31 heap references to one object is unreachable in practice

[github-actions]

Test262 conformance changes

Test result	main count	PR count	difference
Total	52,963	52,963	0
Passed	49,932	49,932	0
Ignored	2,210	2,210	0
Failed	821	821	0
Panics	0	0	0
Conformance	94.28%	94.28%	0.00%

Tested main commit: 5ee4764a3f2b6c8d82a9cc7e9fd5206cfddde65d
Tested PR commit: 0035829c40cc3cfe7f3dc1f426db0e29074de7bd
Compare commits: 5ee4764...0035829

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 59.06%. Comparing base (6ddc2b4) to head (0035829).
⚠️ Report is 841 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #4936       +/-   ##
===========================================
+ Coverage   47.24%   59.06%   +11.82%     
===========================================
  Files         476      563       +87     
  Lines       46892    62365    +15473     
===========================================
+ Hits        22154    36837    +14683     
- Misses      24738    25528      +790     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[jedel1043]

Well that would make it so that we underflow after all references get dropped. right? Then we're patching one panic with another one

[shruti2522]

Well that would make it so that we underflow after all references get dropped. right? Then we're patching one panic with another one

oh yes, missed that 😅 let me look into this more carefully

[shruti2522]

I have capped non_root_count so it can never exceed ref_count. This guarantees is_rooted will never return false just because the counter overflowed, so we won't accidentally free a live object and cause a UAF

[jedel1043]

Looks reasonable. Thank you for fixing the TODO

[jedel1043]

Ping me after you address the CI errors so that I can merge

[shruti2522]

Ping me after you address the CI errors so that I can merge

@jedel1043 fixed the CI, ready for merge