fix roi bar controller integration#80
Merged
sercanarga merged 10 commits intoJul 11, 2026
Merged
Conversation
NetVar1337
force-pushed
the
fix/roi-73-78-final
branch
from
4756661 to
de6d230
Compare
Owner
|
every audio build: bar_controller.sv.tmpl declares hda_tlp_tx_{tdata,tkeepdw,tvalid,tlast,tuser} but not hda_tlp_tx_tready. it's only used (~842) and assigned (~1183), so under `default_nettype none it errors "signal not found". one-line wire fix, verified locally, all 9 audio cells go green. worth pointing your ai at: go test and vet pass on both our machines, but only hdl-lint.sh caught it. the go verilator sims test blocks in isolation and never elaborate the full audio controller as --top-module. good one for your ai to reason about closing (a top-level elaboration per device class?) so full-controller structural bugs stop slipping past go test. fwiw on 5.048 one new test fails on a dotted tlps_ur.source modport ref its dut makes. 5.032 allows it. not firmware. please also ask for a comprehensive sv analysis and bug fix for this branch. |
sercanarga
added a commit
that referenced
this pull request
Jul 11, 2026
… reference Batch 1 from comparison with 16SalomonArs/Pcileech-DMA-NVMe-VMD (per-area, base-independent): - SMART: live Critical Warning byte (temp>=343 sets bit1, media errors bit2); temperature sensors at dword idx 50/51; cumulative warning/critical composite temp time at idx 48/49 (saturating). - I/O: MDTS cap (MAX_XFER_DW=8192 / 32KiB) rejects oversized read/write/write- zeroes with INVALID_FIELD; PRP list walk rejects >512 entries. - DSM: opcode 0x09 with Deallocate bit now invalidates backing-cache slots in the range (wires the existing S_DSM_* path) so post-TRIM reads return zero. - Admin: Set/Get Features persist + validate FIDs 0x01/0x02/0x04/0x06/0x07/ 0x08/0x09/0x0A/0x0B; unknown FIDs return INVALID_FIELD (was silent success). - MSI-X: vector-control dword write masked to bit 0 (bits 31:1 read 0, spec). - README: VMD / Intel RST host placement guidance (class 010802, VMD-mapped port). No port/PBA/interrupt-delivery changes; #75/#80 contracts preserved. go test -race ./... green (27 pkgs); hdl-lint: 170 total, 0 fail.
sercanarga
added a commit
that referenced
this pull request
Jul 11, 2026
… reference Batch 1 from comparison with 16SalomonArs/Pcileech-DMA-NVMe-VMD (per-area, base-independent): - SMART: live Critical Warning byte (temp>=343 sets bit1, media errors bit2); temperature sensors at dword idx 50/51; cumulative warning/critical composite temp time at idx 48/49 (saturating). - I/O: MDTS cap (MAX_XFER_DW=8192 / 32KiB) rejects oversized read/write/write- zeroes with INVALID_FIELD; PRP list walk rejects >512 entries. - DSM: opcode 0x09 with Deallocate bit now invalidates backing-cache slots in the range (wires the existing S_DSM_* path) so post-TRIM reads return zero. - Admin: Set/Get Features persist + validate FIDs 0x01/0x02/0x04/0x06/0x07/ 0x08/0x09/0x0A/0x0B; unknown FIDs return INVALID_FIELD (was silent success). - MSI-X: vector-control dword write masked to bit 0 (bits 31:1 read 0, spec). - README: VMD / Intel RST host placement guidance (class 010802, VMD-mapped port). No port/PBA/interrupt-delivery changes; #75/#80 contracts preserved. go test -race ./... green (27 pkgs); hdl-lint: 170 total, 0 fail.
sercanarga
added a commit
that referenced
this pull request
Jul 11, 2026
Batch 2 (generator-contained parts) from NVMe-VMD reference comparison: - AER (opcode 0x0C) is now asynchronous: latch cmd_cid + aer_registered, consume the SQE without posting a CQE; post the deferred completion from S_IDLE when an async event fires (SMART critical warning -> event_info 0x2). Duplicate AER while registered returns INVALID_FIELD. New states S_AER_POST/S_AER_FINISH; regs cleared on rst / cc_reset / shutdown. Fixes the prior behavior of bumping aer_pending and returning with no completion, which left the host's AER command pending forever. - cfgspace DWORD1 writemask 0xFFFFFFFF -> 0x7D00FFFF: Command fully writable; Status W1C error bits (8,10-14) stay writable; Status RO bits (Capabilities List, 66MHz, DEVSEL, FBTB, reserved) are now read-only so a host CfgWr of 0 can no longer permanently clear them (identity/detection risk). Full 3-way W1C shadow still needs the upstream pcileech-fpga submodule mem_wrap change (deferred). #75/#80 contracts preserved (no msix_vector_select; irq_delivery intact). go test -race ./... green (27 pkgs); hdl-lint: 170 total, 0 fail.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
summary
affected modules
go test ./...andgo vet ./...pass with Verilator 5.032.