Fix vector type lowering

[zerkawei]

This PR fixes #2427 by adding lowered types for the vector types defined in Numerics. I don't know if this is the right approach but it seems to work in my testing.

[zerkawei]

I think there may be a bigger or at least different change to SIMD types and intrinsics that is needed. When trying to compute sqrt for a float4 with SIMD, you can either :

• use SSE.sqrt_ps. Fails with :

undefined reference to 'bf::System::Numerics::X86::SSE::sqrt_ps(bf::System::Numerics::v128)'

• extend float4 to add the sqrt intrinsic. Fails with :

Call parameter type does not match function signature!
  %mul = alloca <4 x float>, align 4
 <4 x float>  %21 = call <4 x float> @llvm.sqrt.v4f32(ptr %mul), !dbg !51

I'll close this PR as this fixes only part of the problem and does not address it properly.

[bfiete]

It's probably good to note here that I would probably SIMD support currently as "proof of concept". I believe that auto-vectorization should work about the same as in Clang, however.