TST/MAINT: fix broken tests, general cleanup

ZLLentz commented 10 months ago

Description

Fix unit test issue where an arbiter (with an entire hash table) was allocated on the stack in FB_Arbiter_Test.RemoveReq. This was done by moving the arbiter from VAR to VAR_INST to put it on the heap instead of the stack. The test suite can now run with the default stack memory size without crashing. https://github.com/pcdshub/lcls-twincat-pmps/commit/332ae4c5c14cfc161b5e6b132a9f98b515e0e4d2#diff-4c3956add6b24a881a2227bed9ad0c39f03d16c89fff372b911102d732947d12R334-R337
Fix unit test issue where FB_FastFaultTest.FFOvrdAndNextFault could last anywhere from 1hr to infinitely long if the PLC wasn't on the UTC time zone. This was done by retrieving and passing in the correct time zone information into the existing test suite function. https://github.com/pcdshub/lcls-twincat-pmps/commit/332ae4c5c14cfc161b5e6b132a9f98b515e0e4d2#diff-07237d9b483b4eb40c39b5866e02e8931e0184ae908e3bd3b8f29789c50f120eR160
Hide unused solution elements, such as Motion and C++
Pin 4024.35 (current vm latest)
Set target to the PMPS test PLC
Remove TC version numbers from project header files
pre-commit run --all

Motivation and Context

These two unit tests were very disruptive when provisioning our TwinCAT CI solutions:

The stack overflow crash also made the next unit test suite fail, even if it was from another repo
The long-running test makes the unit test run roll all the way up to the timeout every time

The other changes are minor and are mostly just "while I'm here" sort of standard-practice adjustments.

How Has This Been Tested?

The unit tests pass in about 10s now, both on the ci PLC and on the pmps test PLC, even with default stack sizes.

Where Has This Been Documented?

Only here

Pre-merge checklist

[x] Libraries are set to Always Newest version (Library, *)
[x] Code works interactively
[ ] Code contains descriptive comments
[x] Test suite passes locally
[x] Committed with pre-commit or ran pre-commit run --all-files
[x] Check that BP_IO parameters weren't modified unintentionally

ZLLentz commented 10 months ago

@ghalym I can break this up into smaller PRs if you want, I just remembered that you had a branch open for other work that might conflict with the aggressive changes here

ZLLentz commented 9 months ago

I'm definitely going to close this and submit just the part that makes the tests pass in an isolated PR following your upcoming TXI vetoes PR.

pcdshub / lcls-twincat-pmps