electronics/essim
1
0
Fork 0
Code Issues 3 Pull Requests Packages Releases Wiki Activity

P3: BSDL completeness check (missing device power/ground pins)

Browse Source 
check_bsdl_completeness(System*): for each BSDL-attached part, re-parse the
.bsd and report the device power/ground ports with no matching pin on the
netlist part (matched by port name or physical pad) — a rail the schematic
symbol is missing. One aggregated BsdlPinMissing per part; restricted to
power/ground so unused I/O balls don't create noise. Surfaced as a 7th verify
pass and in the analyze/dashboard model counts. 76 cases / 327 assertions
green; the real 8-card system reports 0 (all FPGA rails present). This closes
out P3.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Foué
2026-06-03 16:21:02 +02:00
parent c9ac186a20
commit a914b9d7e8
9 changed files with 100 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
DESIGN.md
View File

@@ -131,7 +131,7 @@ The explore screen shows the type in the signal detail header.
**Unified application (`apply_model`)**: connector layout and BSDL are two implementations of one `PinModel` interface (`src/system/pin_model.{hpp,cpp}`: `spec_for(pin_name)`, `layout()`, `source()`). `ConnectorModel` wraps `pin_role`/`pin_layout`; `BsdlPinModel` wraps a parsed `BsdlModel`, indexed by both port name and physical pad. A single `apply_model(Part*, const PinModel&)` materialises the layout pins missing from the netlist, then sets each pin's `spec` **only where the model speaks** (`spec.source != None`). Sources are ranked (`spec_source_rank` in `pin_spec.hpp`: UserOverride > Bsdl > ConnectorModel > Inferred > Imported) and apply_model refuses to overwrite a spec owned by a higher-rank source — so one source never clobbers a more authoritative one, which is also the basis for `check_source_conflicts`. `set-connector-type` and `attach-bsdl` both funnel through it (the latter via the thin `apply_bsdl` adapter); `verify` stays agnostic of where a spec came from. A future SPICE/Modelica source would be a third `PinModel`.
**`verify` (six passes)**: (1) typed pins — local mismatch between each pin's `expected_signal_type()` (derived from its `PinSpec`) and the actual signal type; (2) bridged nets — Power↔GndShield inconsistencies; (3) orphan summary `N orphan pin(s) at import (X imported NC, Y dropped singleton)` (filters out pins bridged via any `Connection::pin_map` — typically `FillIdentityNCs`-materialised); (4) **model-driven pin checks** (`check_pin_specs`): `DriveContention` (≥2 push-pull `Out` on a net), `UndrivenNet` (a **fully-modelled** net with input(s) but no driver — nets with any Unknown-direction pin are skipped, so un-modelled drivers don't cause false positives), `NcWired` (a no-connect pin on a multi-pin net); (5) **JTAG chain** (`check_jtag_chain`): collects TAP pins by `spec.function`, maps each to its net, emits `JtagTapIncomplete` / `JtagBusUnbridged` (TMS or TCK not common to all TAP devices) / `JtagChainBreak` (dangling TDO/TDI, chain fan-out, or not a single head→tail daisy chain); (6) **source conflicts** (`check_source_conflicts`): a pin the BSDL declares power/ground (a must-connect rail) that the netlist leaves unconnected — a rail floated in the schematic (`SourceConflict`; the reverse, a BSDL no-connect that *is* wired, is the `NcWired` check). The BFS-reached `(module, signal)` set for any signal is shown live in `explore`'s detail pane when a signal entry is selected.
**`verify` (seven passes)**: (1) typed pins — local mismatch between each pin's `expected_signal_type()` (derived from its `PinSpec`) and the actual signal type; (2) bridged nets — Power↔GndShield inconsistencies; (3) orphan summary `N orphan pin(s) at import (X imported NC, Y dropped singleton)` (filters out pins bridged via any `Connection::pin_map` — typically `FillIdentityNCs`-materialised); (4) **model-driven pin checks** (`check_pin_specs`): `DriveContention` (≥2 push-pull `Out` on a net), `UndrivenNet` (a **fully-modelled** net with input(s) but no driver — nets with any Unknown-direction pin are skipped, so un-modelled drivers don't cause false positives), `NcWired` (a no-connect pin on a multi-pin net); (5) **JTAG chain** (`check_jtag_chain`): collects TAP pins by `spec.function`, maps each to its net, emits `JtagTapIncomplete` / `JtagBusUnbridged` (TMS or TCK not common to all TAP devices) / `JtagChainBreak` (dangling TDO/TDI, chain fan-out, or not a single head→tail daisy chain); (6) **source conflicts** (`check_source_conflicts`): a pin the BSDL declares power/ground (a must-connect rail) that the netlist leaves unconnected — a rail floated in the schematic (`SourceConflict`; the reverse, a BSDL no-connect that *is* wired, is the `NcWired` check); (7) **BSDL completeness** (`check_bsdl_completeness`): device power/ground ports (from the attached `.bsd`, re-parsed) with no matching pin on the netlist part — a rail the schematic symbol is missing (`BsdlPinMissing`, one aggregated finding per part). The BFS-reached `(module, signal)` set for any signal is shown live in `explore`'s detail pane when a signal entry is selected.
**`analyze` (post-processing pass)**: `analyze_system(System*) → AnalysisReport` (`src/system/analysis.{hpp,cpp}`) is a stateless read-only pass that detects structural signal groups and anomalies. Per-module (signals are module-scoped):
@@ -299,7 +299,7 @@ The analyze screen additionally surfaces two "verify-class" issues, computed the
- **pin-role mismatch** — a pin whose `expected_signal_type()` (derived from its `PinSpec`, set by `set-connector-type` via `pin_role(connector_type, pin_name)`) disagrees with the actual signal type.
- **net-mix** — a bridged net (BFS over `Connection::pin_map`, ≥ 2 members) where `net_type_consistent(net, &dominant)` returns false. Specifically, the net contains both `Power` and `GndShield` signals.
The `verify` command (not the analyze screen, yet) also emits the **model-driven `AnomalyKind`s** from `bsdl_check.{hpp,cpp}`: `DriveContention` / `UndrivenNet` / `NcWired` (`check_pin_specs`) and `JtagTapIncomplete` / `JtagChainBreak` / `JtagBusUnbridged` (`check_jtag_chain`); and `SourceConflict` (`check_source_conflicts` — a BSDL power/ground pin the netlist leaves unconnected). They consume the BSDL-populated `PinSpec` plus `compute_all_nets`, and are surfaced in three places: the `verify` command, the analyze screen's Issues pane (counted as `… N model`), and a `model:` health row on the dashboard. `check_pin_specs`/`check_jtag_chain` accept an optional precomputed net list, so verify, analyze and the dashboard each compute the nets once and reuse them across checks.
The `verify` command (not the analyze screen, yet) also emits the **model-driven `AnomalyKind`s** from `bsdl_check.{hpp,cpp}`: `DriveContention` / `UndrivenNet` / `NcWired` (`check_pin_specs`) and `JtagTapIncomplete` / `JtagChainBreak` / `JtagBusUnbridged` (`check_jtag_chain`); `SourceConflict` (`check_source_conflicts` — a BSDL power/ground pin the netlist leaves unconnected); and `BsdlPinMissing` (`check_bsdl_completeness` — a device power/ground port absent from the netlist part, re-parsing the attached `.bsd`). They consume the BSDL-populated `PinSpec` plus `compute_all_nets`, and are surfaced in three places: the `verify` command, the analyze screen's Issues pane (counted as `… N model`), and a `model:` health row on the dashboard. `check_pin_specs`/`check_jtag_chain` accept an optional precomputed net list, so verify, analyze and the dashboard each compute the nets once and reuse them across checks.
### Component kind