verilator %Warning-WIDTH false positive

[veripoolbot]

---

Author Name: Tim Allen
Original Redmine Issue: 1613 from https://www.veripool.org

---

excerpt of file.sv...

logic   [31:0]          distance;
logic   [31:0]          bfield;
logic   [3:0]           cur_bit_len;
typedef struct packed {         // 10 bits total
     bit [8:0]          max_tbl_size;
     bit [15:0][31:0]   dec_len;
     bit [15:0][31:0]   dec_pos;
     bit [31:0]         q_bits;
} decode_table_t;

distance <= bfield - o_dec_table.dec_len[cur_bit_len - 4'd1];
    ^          ^                     ^         ^         ^
//32bits    32bits                32bits    4bits     4bits

yields the following false positive warnings
%Warning-WIDTH: file.sv:123: Operator SUB expects 32 or 6 bits on the LHS, but LHS's VARREF 'cur_bit_len' generates 4 bits.
%Warning-WIDTH: file.sv:123: Operator SUB expects 32 or 6 bits on the RHS, but RHS's CONST '4'h1' generates 4 bits.

I hope this is clear enough. AFAIK 

[veripoolbot]

---

Original Redmine Comment
Author Name: Wilson Snyder (@wsnyder)
Original Date: 2019-11-22T17:21:39Z

---

The rightmost index of dec_len is 31:0, meaning it needs 6 bits to index into. Your cur_bit_len as you indicate is only 4 bits. Perhaps I'm confused, but seems the warning is correct.

[veripoolbot]

---

Original Redmine Comment
Author Name: Tim Allen
Original Date: 2019-11-22T19:44:46Z

---

does bit [15:0][31:0] dec_len; not mean 16, 32 bit dec_lens for a total of 512 bits? So if I say something like ```dec_len[3]

[veripoolbot]

---

Original Redmine Comment
Author Name: Tim Allen
Original Date: 2019-11-22T20:35:12Z

---

I think what I said above is consistent with IEEE 1800-2017 7.4.5 which says (among other things), "A subarray is an array that is an element of another array. As in the C language, subarrays are referenced by omitting indices for one or more array dimensions, always omitting the ones that vary most rapidly..." I find that a bit obtuse but in the preceding paragraph it describes "rapidity" of various dimensions which, I think, went something like

logic [2nd][1st] var [4th][3rd]

[veripoolbot]

---

Original Redmine Comment
Author Name: Wilson Snyder (@wsnyder)
Original Date: 2019-11-22T23:13:10Z

---

You're right, an easy cross check:

  $display("%d", $size(o_dec_table.dec_len));

  16

What looks to be going on is for packed structs Verilator runs these checks twice, once on the code as written then once after the unpacking, the warning is on the second one which makes no sense for the user. Will look into fixing.