Investigate segfault

[virgesmith]

trying to track down cause of segfault in win tests... (worked 29/6, failed 6/7) pybind11 2.13 was released 26/6

Using asan it seems that all C++ exceptions thrown in neworder have some address violation associated with them, and this is not a recent change in pybind11.

(add extra_compile_args=["-fsanitize=address" ], run with e.g. LD_PRELOAD=$(gcc --print-file-name=libasan.so) python umr.py)

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 93.04%. Comparing base (396c162) to head (7b9c8e5).

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #87      +/-   ##
==========================================
+ Coverage   92.48%   93.04%   +0.55%     
==========================================
  Files          16       14       -2     
  Lines         679      719      +40     
==========================================
+ Hits          628      669      +41     
+ Misses         51       50       -1     

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[virgesmith]

failure disappeared resolved itself in Aug. assuming compiler bug that was fixed