Zcf/Zcd RVTEST_CASE macros inconsistent

[jordancarlin]

The new Zcf and Zcd tests introduced in #497 and #587 are not consistent in their use of the RVTEST_CASE isa regex macro.

For rv32 and rv64 Zcd, c.fsdsp uses check ISA:=regex(.*I.*D.*C.*) while all of the other tests use check ISA:=regex(.*I.*F.*D.*C.*). D implies F so this one isn't a big deal, but they should be consistent.

More significantly, for rv32 Zcf, c.fswsp uses check ISA:=regex(.*I.*F.*C.*) while all of the other tests use check ISA:=regex(.*I.*F.*Zcf.*). Either approach is acceptable, but using Zcf requires modifying the ISA string that is passed to RISCOF when selecting tests.

All of the Zcf and Zcd tests should use the same approach for this, either using the Zcf and Zcd extensions or the C and F/D extensions.

[jordancarlin]

Related and even more problematic, the c.fldsp test for rv32i_m specifies rv64 instead of rv32 (RVTEST_ISA("RV64IFDC"))

[jordancarlin]

The RVTEST_CASE check ISA regex is also missing 32/64 for rv32 and rv64 respectively.

[jordancarlin]

@anuani21 @UmerShahidengr Is there a preference for one approach to the ISA string over another? I'm happy to open a PR to fix this, but we need to decide if we want that ISA string to use C or Zcf/Zcd.

[UmerShahidengr]

Good catch @jordancarlin , I didnt check the ISA string for all tests, I think it should be consistent with Zcf/Zcd strings instead of C as these tests belong to the respective Zcf and Zcd extensions and not all C-extension enabled designs supports these tests

[UmerShahidengr]

Would be great if you open up a PR to resolve this issue, we can do one thing, I can merge Zhix tests too, and @anuani21 and @jordancarlin can make all tests consistent via same PR. I am seeing some Zhinx tests also have the same issue