Header (line 6, 11): "(IBUFDS, BUFG, IDDR)" → "(IBUFDS, BUFIO, BUFG,
MMCME2_ADV)"; "IDDR Q2 (falling-edge) data capture in
SAME_EDGE_PIPELINED mode" → "Falling-DCO-edge IOB-packed IFF capture
(post-AUDIT-C4 SDR shape; no IDDR, no rise/fall demux)".
$display banner: "(IBUFDS, BUFG, IDDR)" → "(IBUFDS, BUFIO, BUFG, MMCME2)".
Group 8 inline comment: "captures Q2 of the IDDR (falling-edge…)"
→ "captures the IOB-packed IFF on the falling DCO edge".
Test Group 4 retired. The block drove 0xAA on rising DCO and 0x55 on
falling DCO. Post-AUDIT-C4 the IFF only samples the falling edge —
every captured value was 0x55 — so the assertion
`saw_aa > 0 || saw_55 > 0` was trivially true and `cap_count > 0`
duplicates Group 5 / Group 8's stronger checks. Block replaced with
a tombstone comment; group numbering preserved for git-blame
continuity.
Two "IDDR" references intentionally retained inside negations
(line 12, line 186) — they explicitly contrast the current SDR
topology against the broken pre-C-4 shape so a reader who finds the
old vocabulary in git history understands what changed.
Verification: remote Vivado 2025.2 xsim → 15 / 15 PASS
(was 17 / 17; the lost 2 are Group 4's trivial-pass and existence
checks, both now subsumed by Groups 5 / 8).
Jason
ce869e9e20