Add builtins for basic float operations (#3897)

Adds support for unary `-` and binary `+`, `-`, `*`, `/` for floating
point types.

Real literals are now transformed to `llvm::APFloat`s during the check
phase into the `FloatLiteral` instruction.

This PR likely collides a bit with #3892 and might need to be updated
when that one is merged.

toolchain/check/context.cpp

@@ -978,6 +978,7 @@ class TypeCompleter {
       case SemIR::Converted::Kind:
       case SemIR::Deref::Kind:
       case SemIR::FacetTypeAccess::Kind:
+      case SemIR::FloatLiteral::Kind:
       case SemIR::FieldDecl::Kind:
       case SemIR::FunctionDecl::Kind:
       case SemIR::ImplDecl::Kind:
@@ -1055,6 +1056,7 @@ class TypeCompleter {
 
       case SemIR::AssociatedEntityType::Kind:
       case SemIR::BindSymbolicName::Kind:
+      case SemIR::FloatType::Kind:
       case SemIR::IntType::Kind:
       case SemIR::PointerType::Kind:
       case SemIR::UnboundElementType::Kind: