Basic floating point support.

[vext01]

This gives us just enough floating point support to print a float value in a test.

We only support the common floating point types: float and double.

Requires https://github.com/ykjit/ykllvm/pull/175

[ltratt]

This looks good but could we quickly add:

• syntax tests in jit_ir::parser::test::all_jit_ir_syntax
• well-formedness checks in well_formedness.rs (i.e. should_panic tests)

Both well make it less likely we accidentally break things in the future :)

[vext01]

ah, sorry, wasn't aware of those tests. Yep, will add some.

[vext01]

I just realised that i accidentally re-used an old branch unintentionally! Never mind, it should be ok.

[vext01]

https://github.com/ykjit/yk/pull/1258/commits/33b8901951a78019c98cb95d76b1eee065d29102 adds tests (and strengthens some checks).

OK to squash.

[ltratt]

Please squash.

[vext01]

squashed.

[vext01]

This is a strange error:

11:36:58      Running `target/debug/xtask cfmt --check`
11:36:58 ./tests/c/double.newcg.c:51:46: error: code should be clang-formatted [-Wclang-format-violations]
11:36:58     fprintf(stderr, "%d -> %f\n", i, (double) i);
11:36:58                                              ^
11:36:58 
11:36:58 ./tests/c/float.newcg.c:53:45: error: code should be clang-formatted [-Wclang-format-violations]
11:36:58     fprintf(stderr, "%d -> %f\n", i, (float) i);
11:36:58                                             ^
11:36:58 

I am unable to reproduce it, but also the error is really generic and doesn't actually say what the problem is.

Any ideas?

[vext01]

Sorry, I got it. I was on the wrong commit. It wants me to remove the space between the cast.

[vext01]

Force pushed the fix.