Add new API to cleanup OpenSSL threads.

[ashman-p]

• Does this PR change the input/output behaviour of a cryptographic algorithm (i.e., does it change known answer test values)? (If so, a version bump will be required from x.y.z to x.(y+1).0.)
• Does this PR change the list of algorithms available -- either adding, removing, or renaming? Does this PR otherwise change an API? (If so, PRs in fully supported downstream projects dependent on these, i.e., oqs-provider will also need to be ready for review and merge by the time this is merged.)

'run_tests' memory leak tests fails when build options enables OQS_USE_PTHREADS and OQS_USE_OPENSSL.

valgrind tests/test_sig SPHINCS+-SHA2-256s-simple
==158054== Memcheck, a memory error detector
==158054== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==158054== Using Valgrind-3.15.0 and LibVEX; rerun with -h for copyright info
==158054== Command: tests/test_sig SPHINCS+-SHA2-256s-simple
==158054== 
Testing signature algorithms using liboqs version 0.10.2-dev
Configuration info
==================
Target platform:  x86_64-Linux-5.4.0-126-generic
Compiler:         gcc (9.4.0)
Compile options:  [-Wa,--noexecstack;-O3;-fomit-frame-pointer;-fdata-sections;-ffunction-sections;-Wl,--gc-sections;-Wbad-function-cast]
OQS version:      0.10.2-dev
Git commit:       f47817d8cb0cd621dada5276f30f99934f3132a9
OpenSSL enabled:  Yes (OpenSSL 3.3.3.8.3.0-dev )
AES:              NI
SHA-2:            OpenSSL
SHA-3:            OpenSSL
OQS build flags:  OQS_DIST_BUILD OQS_LIBJADE_BUILD OQS_OPT_TARGET=generic CMAKE_BUILD_TYPE=Release 
CPU exts active:  AES AVX AVX2 BMI1 BMI2 PCLMULQDQ POPCNT SSE SSE2 SSE3
================================================================================
Sample computation for signature SPHINCS+-SHA2-256s-simple
================================================================================
verification passes as expected
==158054== 
==158054== HEAP SUMMARY:
==158054==     in use at exit: 240 bytes in 1 blocks
==158054==   total heap usage: 7,654 allocs, 7,653 frees, 844,712 bytes allocated
==158054== 
==158054== LEAK SUMMARY:
**==158054==    definitely lost: 240 bytes in 1 blocks**
==158054==    indirectly lost: 0 bytes in 0 blocks
==158054==      possibly lost: 0 bytes in 0 blocks
==158054==    still reachable: 0 bytes in 0 blocks
==158054==         suppressed: 0 bytes in 0 blocks
==158054== Rerun with --leak-check=full to see details of leaked memory
==158054== 
==158054== For lists of detected and suppressed errors, rerun with: -s
==158054== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

Fixes #1941

Previous PR attempt is ... PR1942

[SWilson4]

I added one suggestion for some additional documentation. Code-wise, this looks good to me; thanks @ashman-p for the fix.

I also confirmed that the leak no longer appears in my setup when this branch is used.

[ashman-p]

Please, 1 more reviewer? @cothan ?