Remove builtin types

[Kronos3]

This PR removes the special handling of "builtin" types, which are type aliases that were previously handled inside FpConfig.h by C-level typedefs.

Some notes:

• I removed the #include <FpConfig.h[pp]> from generated alias headers since it's redundant now. It was previously there in case there was an alias of a built-in
• I removed the tests that tested builtin types since they are no longer relevant as there is no additional code coverage over the simple type alias tests.
• I found a bug in the #define PRI code that was not referencing built-in PRIs properly
  • We had not hit this yet because C++ code did not use an alias type's #define in generated code. With builtin's removed now the old builtin PRIs were being hit.
  • The common pattern was like this:

// Queue priority type
typedef PlatformQueuePriorityType FwQueuePriorityType;
#define PRI_FwQueuePriorityType PRI_PlatformQueuePriorityType

// The type used to serialize a size value
typedef U16 FwSizeStoreType;
#define PRI_FwSizeStoreType PRIu16

Notice that builtin types like U16 have PRIs defined PRI[typeName.lower() where-as aliases are PRI_[typeName]. That underscore was always assumed in previously generated code (see diffs for what I'm talking about).

Closes #649

[Kronos3]

I've opened a sister PR in nasa/fprime#3417 that updates the FpConfig.h which this PR's check-cpp will depend on.

[Kronos3]

Future note to self, the PRIs for scalars are defined twice with slightly different names:
https://github.com/nasa/fprime/pull/3422/files#diff-4a4e7f315f68d77a9f724564407a9a79e284175fbd14fd0e81a428a00dca0293R51-R88

[bocchino]

Notice that builtin types like U16 have PRIs defined PRI[typeName.lower() where-as aliases are PRI_[typeName].

I think there are additional macros in there now so that PRI_U16 should work. If that's true, then maybe we should update FPP to use that form. That way the PRIu16 form would be used once to define PRI_U16, which is the F Prime standard, and PRI_U16 would be used everywhere else.

[LeStarch]

@bocchino should be ready for re-review, and once finished a point-release!

[bocchino]

@LeStarch. It looks good so far. Just a couple of comments:

1. I think there may be a missing header include for Fw/FPrimeBasicTypes.hpp. I noted this in an inline comment.

2. Let's be consistent and write #include "Fw/FPrimeBasicTypes.hpp" everywhere. Sometimes it is written #include <Fw/FPrimeBasicTypes.hpp>. I started noting this in comments, but then there were several cases. Let's sweep through and make everything consistent. grep -r "#include <Fw/FPrimeBasicTypes.hpp>" in the fpp-check/test directory should come up clean.

Where is the corresponding F Prime branch, so I can run check-cpp and verify that it passes?

[bocchino]

It looks like the corresponding F Prime branch is here: nasa/fprime#3422.

[bocchino]

@LeStarch I'm not seeing Fw/FPrimeBasicTypes.hpp on that branch.

[LeStarch]

@bocchino, I have fixes the issues, except for the one I left a comment about. I have also pushed to the FPrime branch so you can run cpp-check.

I have:

1. Fixed suggestions
2. Updated refs
3. Confirmed No angle bracket includes for the new header
4. Ran cpp-check

[bocchino]

Looks good! I added a few comments explaining what is going on with the headers in the alias code gen. I confirmed that check-cpp runs cleanly.

[bocchino]

Before merging this branch into fpp/main, I would like to get nasa/fprime#3434 reviewed and merged. It fixes problems in FppTest, which we need for testing the integration of this branch with F Prime.

[bocchino]

I confirmed that everything works with the updated FppTest. Merging now.