Port tools/jtag/link_relay.py (Python, 7 functions) to specs/port/tools/jtag/link_relay.t27 - #4533
Merged
Merged
Conversation
The turn ended with these files edited and never committed. Uncommitted work is invisible to the review - it reads the branch - so the attempt would have been released as empty and the next bee would have started beside this work rather than from it. This commit is not a claim that the work is correct. It is the bee's work, committed on its behalf, and it is judged exactly like any other: the adversarial reviewer reads it, the compiler runs on it, and the issue's own criteria are measured against it. Issue: #4507 Turn: a8dbdad9-999c-4604-a2c7-dd0f9ade43df Ending: finished (the turn closed) Committed: 1 path(s) Left uncommitted: 0 path(s) outside the declared boundary
A pull request must add exactly one docs/now entry and a bee has no way to know that: its brief names a boundary file and acceptance criteria, and docs/now/ is neither. The publisher adds it rather than failing the gate. Closes #4507 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
gHashTag
enabled auto-merge (squash)
September 21, 2026 12:55
Contributor
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
This was referenced Sep 21, 2026
gHashTag
added a commit
that referenced
this pull request
Sep 21, 2026
specs/port/tools/jtag/link_relay.t27 landed in #4533 today with Zig bodies in a file named .t27: ?struct{} at line 104, 23 @intcast(x as T), 14 try std.testing.expect, m.tms(&[...]), while (i < n) : (i += 1), .{ .role = r }, null/.unwrap()/.or(0). It failed on gen-c, gen-rust, gen-verilog and gen. Rewritten against read_user1.t27 -- its neighbour in the same directory and the same JTAG/MPSSE domain, and a spec that does generate. t27c check now reports 0 errors and 0 warnings and all four backends exit 0. The optional is kept rather than dropped: DrWord { ok, value } preserves the distinction the Python drew with None, so a caller that ignores ok reads a zero instead of a word the hardware never sent. The chain-IR loop, written out four times between xfer and request, is one chain_ir(idx, chain, sel). One of the seven dead tests asserted 0x1F3DF where the arithmetic gives 0x1F39F. A test that cannot compile cannot be wrong out loud. adamw.t27 was fixed by #4540 and left in the baseline. check_specs_generate.py short-circuits on "baseline now generates" and never reaches "newly does not generate", so every run stopped at adamw and never looked at link_relay. The stale line is pruned here; that is what made this visible. Fixes #4559
gHashTag
added a commit
that referenced
this pull request
Sep 21, 2026
…ed the oldest problem instead Fixes #4560 specs/port/tools/run_conformance_vvp.t27 and check_fix_carries_source.t27 landed on 2026-09-21 and have never compiled on any backend. The first is transliterated Python, the second 565 lines of Zig with `let` against a 284 line source. The transliteration also changed the verdict. The emitted testbench increments `fails` and reports success on `if (folds == 0)`, a name declared nowhere, so the pass line is unreachable. `vvp_path` is bound to the compiled output and then rebound to the simulator binary, and the run becomes `vvp vvp`. Neither is a parse error. Both rewritten as specs: 0 errors, 0 warnings, four backends exit 0. check_fix_carries_source.t27 states the verdict lattice and the is_prose inversion but deliberately does not restate SOURCE_SCOPES or PROSE_EXT -- those have one definition in cli/tri/src/hooks.rs and the Python parses it rather than copying it. tools/check_specs_generate.py had four `return 1`s in a row, so the first class found was the only one reported, and the mildest was checked three lines before the most serious. One stale adamw ledger line left by #4540 blinded it for the whole window in which #4533, #4553 and #4556 landed. All four classes now report and the exit is taken once. The stale-ledger branch had no self-check case -- the only verdict in main() nothing executed -- and now has one, with a case that plants a stale line and a new break together.
gHashTag
added a commit
that referenced
this pull request
Sep 21, 2026
…rted the oldest problem instead (#4562) * fix(specs): the JTAG relay spec was Zig, not t27, and had never compiled specs/port/tools/jtag/link_relay.t27 landed in #4533 today with Zig bodies in a file named .t27: ?struct{} at line 104, 23 @intcast(x as T), 14 try std.testing.expect, m.tms(&[...]), while (i < n) : (i += 1), .{ .role = r }, null/.unwrap()/.or(0). It failed on gen-c, gen-rust, gen-verilog and gen. Rewritten against read_user1.t27 -- its neighbour in the same directory and the same JTAG/MPSSE domain, and a spec that does generate. t27c check now reports 0 errors and 0 warnings and all four backends exit 0. The optional is kept rather than dropped: DrWord { ok, value } preserves the distinction the Python drew with None, so a caller that ignores ok reads a zero instead of a word the hardware never sent. The chain-IR loop, written out four times between xfer and request, is one chain_ir(idx, chain, sel). One of the seven dead tests asserted 0x1F3DF where the arithmetic gives 0x1F39F. A test that cannot compile cannot be wrong out loud. adamw.t27 was fixed by #4540 and left in the baseline. check_specs_generate.py short-circuits on "baseline now generates" and never reaches "newly does not generate", so every run stopped at adamw and never looked at link_relay. The stale line is pruned here; that is what made this visible. Fixes #4559 * fix(specs): two ported specs were Zig and Python, and the gate reported the oldest problem instead Fixes #4560 specs/port/tools/run_conformance_vvp.t27 and check_fix_carries_source.t27 landed on 2026-09-21 and have never compiled on any backend. The first is transliterated Python, the second 565 lines of Zig with `let` against a 284 line source. The transliteration also changed the verdict. The emitted testbench increments `fails` and reports success on `if (folds == 0)`, a name declared nowhere, so the pass line is unreachable. `vvp_path` is bound to the compiled output and then rebound to the simulator binary, and the run becomes `vvp vvp`. Neither is a parse error. Both rewritten as specs: 0 errors, 0 warnings, four backends exit 0. check_fix_carries_source.t27 states the verdict lattice and the is_prose inversion but deliberately does not restate SOURCE_SCOPES or PROSE_EXT -- those have one definition in cli/tri/src/hooks.rs and the Python parses it rather than copying it. tools/check_specs_generate.py had four `return 1`s in a row, so the first class found was the only one reported, and the mildest was checked three lines before the most serious. One stale adamw ledger line left by #4540 blinded it for the whole window in which #4533, #4553 and #4556 landed. All four classes now report and the exit is taken once. The stale-ledger branch had no self-check case -- the only verdict in main() nothing executed -- and now has one, with a case that plants a stale line and a new break together. * fix(specs): two type names that each meant two different things Refs #4560 `tri types ratchet` read 81 conflicted names against a ledger of 80. `Context` was mine. The rewritten check_fix_carries_source.t27 named its pull-request record `Context`, a name specs/neural/forward_pass.t27 and specs/queen/lotus.t27 had already given to two other things. Renamed to `PrEvent`, which is what it is; `context_incomplete` moved with it to `pr_event_incomplete` across 10 call sites. `Edge` came from check_graph_law8.t27 -- the same porting campaign, landed the same day. specs/tri/graph/bellman_ford.t27 has declared `Edge` as from/to/weight since long before it; the port's is from/to/kind. Two concepts, one name, and nothing for a cross-spec resolver to break the tie with. The newcomer takes the narrower name: `DepEdge`. Ratchet now reads `ledger 80, observed 80 -- CLEAN`. All four backends (gen-c, gen-rust, gen-js, gen-ts) exit 0 on both files. This is also how the ratchet is meant to be read. Resolving `Mpsse` while introducing `Context` left the count unchanged; only the identity-keyed list showed that two separate things had happened.
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.
Closes #4507
Written by a bee on
queen-4507and published bytools/queen/publish.py. The branch itself is the bee's; the second commit is the coordination entry every pull request must add, which a bee has no way to know about.🤖 Generated with Claude Code