LLVM Has Opinions on Float Constants

The following results in an error while compiling the LLVM code (with default pass pipeline on main):

#[entry]
fn test(x: f32) -> f32 {
  return x + 0.3;
}

The line clang has problems with is

  %v3 = bitcast float 0.3 to float

with error floating point constant invalid for type.

assigned to @rarbore2

mentioned in merge request !170 (merged)

mentioned in commit 38be2e7c

closed with merge request !170 (merged)