Refactor ABI wrappers to not rely on --wrap

[gbournou]

b/330648201

Test-On-Device: true

[codecov]

Codecov Report

Attention: Patch coverage is 47.82609% with 24 lines in your changes are missing coverage. Please review.

Project coverage is 58.86%. Comparing base (60884cb) to head (fd69425).
Report is 10 commits behind head on main.

Files	Patch %	Lines
...ular/starboard_layer_posix_pthread_abi_wrappers.cc	0.00%	15 Missing ⚠️
...modular/starboard_layer_posix_time_abi_wrappers.cc	0.00%	4 Missing ⚠️
.../modular/starboard_layer_posix_time_abi_wrappers.h	0.00%	4 Missing ⚠️
...modular/starboard_layer_posix_mmap_abi_wrappers.cc	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2665      +/-   ##
==========================================
- Coverage   58.87%   58.86%   -0.02%     
==========================================
  Files        1908     1908              
  Lines       94447    94449       +2     
==========================================
- Hits        55604    55595       -9     
- Misses      38843    38854      +11     

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

[andrewsavage1]

Is there a good way for us to ensure that this would not conflict with wrapping at the starboard layer in some way? I.e. some commit on top of this that we won't submit that tests wrapping a symbol at the starboard layer?

[gbournou]

Is there a good way for us to ensure that this would not conflict with wrapping at the starboard layer in some way? I.e. some commit on top of this that we won't submit that tests wrapping a symbol at the starboard layer?

My plan was to follow up with a PR to convert the existing macro-based cases to use --wrap. Not sure what other kind of commit you were thinking of

[gbournou]

The single Raspi failure is from an unrelated issue and should be fixed in #2719