fix: Polars thread pool was not used properly in various functions

[nameexhaustion]

Found these by running py.test with a rayon fork that panicked on attempts to use the global thread pool.

In theory it might be possible to initialize rayon's global thread pool with a start_handler that panics for debug builds, but I couldn't try it because I don't know where to put the call to build the thread pool - I can't call it outside of function contexts because it wasn't a const fn.

[codecov]

Codecov Report

Attention: 4 lines in your changes are missing coverage. Please review.

Comparison is base (df5dd43) 77.94% compared to head (29e9052) 77.94%.

Files	Patch %	Lines
crates/polars-ops/src/frame/join/asof/groups.rs	87.50%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #14583      +/-   ##
==========================================
- Coverage   77.94%   77.94%   -0.01%     
==========================================
  Files        1326     1326              
  Lines      173058   173050       -8     
  Branches     2448     2448              
==========================================
- Hits       134892   134880      -12     
- Misses      37722    37726       +4     
  Partials      444      444              

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

[ritchie46]

Nice! I will tag this as performance as well. As this will save a lot unexpected contention.