Method compile_sub

crates/jit/src/instructions.rs:814–818

(&mut self, a: Value, b: Value)

    }

    fn compile_sub(&mut self, a: Value, b: Value) -> Value {
        let (out, carry) = self.builder.ins().ssub_overflow(a, b);
        self.builder.ins().trapnz(carry, TrapCode::INTEGER_OVERFLOW);
        out
    }

    /// Creates a double–double (DDValue) from a regular f64 constant.
    /// The high part is set to x and the low part is set to 0.0.