Directionless action arguments provided by the P4 program should behave like in

[jaehyun1ee]

The spec mentions in section 6.8. Calling convention that,

For an action, a directionless parameter indicates that it is "action data". See Section 14.1 for the meaning of action data, but its meaning includes the following possibilities:

• The parameter’s value is provided in the P4 program. In this case, the parameter behaves as if the direction were in. Such an argument expression need not be a compile-time known value.
• The parameter’s value is provided by the control plane software when an entry is added to a table that uses that action. See Section 14.1.

But test issue473.p4 is categorized under p4_16_errors, expecting the frontend to reject the program with:

control cIngress(inout Parsed_packet hdr,
                 inout mystruct1 meta,
                 inout standard_metadata_t stdmeta) {
    action a1(in bit<32> x) {
        meta.b = meta.b + (bit<4>) x;
        meta.b = (bit<4>) ((bit<32>) meta.b + x);
    }
    action b(inout bit<32> x, bit<8> data) {
        meta.a = meta.a ^ (bit<4>) x ^ (bit<4>) data;
    }
    table t1 {
        key = { hdr.ethernet.srcAddr : exact; }
        actions = {
            a1((bit<32>) meta.a);
            b(meta.c);
        }
        default_action = b(meta.c, (bit<8>) meta.d); // HERE
    }

    apply {
        t1.apply();
    }
}

issue473.p4(70): [--Werror=type-error] error: (bit<8>)meta.d: action argument must be a compile-time constant
        default_action = b(meta.c, (bit<8>) meta.d);
                                   ^^^^^^^^^^^^^^^

I believe this is not in line with the current spec, since b(meta.c, (bit<8>) meta.d) is written within a P4 program (not by the control plane), thus the directionless argument (bit<8>) meta.d need not be a compile-time known value.