Run C regression tests using clang and gcc when both are available

[tautschnig]

Our C front-end seeks to conform to whatever interpretation of the C
standard the underlying host C compiler has. To avoid incompatibilities
as documented in #2370, run regression tests using both Clang and GCC,
if available on a particular system.

• Each commit message has a non-empty body, explaining why the change was made.
• n/a Methods or procedures I have added are documented, following the guidelines provided in CODING_STANDARD.md.
• n/a The feature or user visible behaviour I have added or modified has been documented in the User Guide in doc/cprover-manual/
• Regression or unit tests are included, or existing tests cover the modified code (in this case I have detailed which ones those are in the commit message).
• n/a My commit message includes data points confirming performance improvements (if claimed).
• My PR is restricted to a single feature or bugfix.
• n/a White-space or formatting changes outside the feature-related changed lines are in commits of their own.

[codecov]

Codecov Report

Merging #6670 (272fe76) into develop (1154f13) will increase coverage by 0.20%.
The diff coverage is 86.66%.

@@             Coverage Diff             @@
##           develop    #6670      +/-   ##
===========================================
+ Coverage    76.92%   77.13%   +0.20%     
===========================================
  Files         1583     1582       -1     
  Lines       183314   182625     -689     
===========================================
- Hits        141018   140868     -150     
+ Misses       42296    41757     -539     

Impacted Files	Coverage Δ
src/cpp/cpp_typecheck_conversions.cpp	85.14% <0.00%> (+0.10%)	⬆️
src/goto-instrument/dump_c.cpp	80.49% <0.00%> (+0.28%)	⬆️
src/goto-instrument/dump_c_class.h	100.00% <ø> (ø)
src/goto-instrument/goto_program2code.cpp	68.84% <0.00%> (+0.06%)	⬆️
src/goto-programs/xml_expr.cpp	51.21% <0.00%> (ø)
src/util/expr_util.h	100.00% <ø> (ø)
src/util/pointer_predicates.cpp	93.10% <ø> (-0.45%)	⬇️
src/util/pointer_predicates.h	66.66% <ø> (ø)
src/analyses/custom_bitvector_analysis.cpp	55.96% <44.44%> (+0.07%)	⬆️
src/solvers/smt2/smt2_conv.cpp	67.57% <66.66%> (-0.04%)	⬇️
... and 39 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a5092cb...272fe76. Read the comment docs.

[chris-ryder]

Seems like an improvement, so approving as is - but just wondering slightly what the reasoning behind enabling these test profiles only when there is both clang and gcc available? Could the different profiles not just run when based on the availability of the single compiler needed? (e.g. ansi-c-clang running only if clang is available, ansi-c-gcc only if gcc is available, etc? rather than having a semperate 'else' block for running if only one of those two is available?

[tautschnig]

Seems like an improvement, so approving as is - but just wondering slightly what the reasoning behind enabling these test profiles only when there is both clang and gcc available? Could the different profiles not just run when based on the availability of the single compiler needed? (e.g. ansi-c-clang running only if clang is available, ansi-c-gcc only if gcc is available, etc? rather than having a semperate 'else' block for running if only one of those two is available?

Yes, I now cannot think of a good reason why I restricted it to both having to be available. I have now changed this as suggested.