Repository navigation
theory(sieve): re-read the t27c and t27b cells after T741's fix issues closed, 11 FALSE -> 7 (Closes #7597) - #7599
Merged
Merged
Conversation
…s closed, 11 FALSE -> 7 (Closes #7597) sieve.t27 was written at master 682564f on 2026-10-05. T741 mapped its 11 FALSE t27 cells to seven fix issues, and all seven have since closed. Each cell was re-read against what landed. A cell moved only where the gap its own text named is now closed. FALSE -> TRUE: - t27c C2: per-test verdicts against t27b, #6441, PR #6530. - t27c C3: per test, plus generator #6442 and reducer #6445. - t27c C10: test-report counts executed asserts, #6509, PR #6788. - t27b C3: per test and fuzzer. Stays FALSE, with the reason now stated: - C1: #6446 proves nothing. New fix issue #7595. - t27c C4: DDC runs (T732), but its routes share compiler.rs's front end (T729). Fix issue #5980. - C6 and C7: #6444 gave native runners, not numbers. New fix issue #7596. The counting theorems follow the table: - T738: t27c is removed by 4 filters, t27b by 3. - T739: 5 filters fire; C2, C3 and C10 no longer fire. - T740: C10 is TRUE for both, and each has 4 TRUE cells. - T741: 7 FALSE cells, each mapped to an open issue. - T742: G10 moves MISSING -> PARTIAL, because the t27a table round-trips all 47 opcodes (T736, #6507). Totals: 1 present, 7 partial, 2 missing. The 682564f rows are kept as functions. A new test, t741_the_closed_fix_issues_moved_four_cells_and_none_back, checks the move: 11 -> 7 FALSE cells, 4 FALSE -> TRUE, and no other change. Verified on the t27c lab with t27c built from master 17e01d2 and zig 0.16.0: - test-report passes 10/10 with 0 vacuous passes; master passed 9/9. - Mutants are caught: - t27c C3 put back to FALSE fails 6 tests; - G10 put back to MISSING fails 1 test. - `t27c seal --save` was followed by `--verify`, which printed: all hashes MATCH. Closes #7597 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
gHashTag
enabled auto-merge (squash)
October 7, 2026 19:49
This was referenced Oct 7, 2026
Contributor
This was referenced Oct 7, 2026
3 of 6 tasks
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 #7597
Part of #6488.
What changes
specs/compiler/theory/sieve.t27was written at master 682564f (2026-10-05). T741 mapped its 11 FALSE t27 cells to seven fix issues: #6441, #6442, #6445, #6446, #6508, #6444 and #6509. All seven are now closed. This PR re-reads each cell against what landed. A cell moves only where the gap its own text named is closed.test-reportcounts executed asserts (#6509, PR #6788)The counting theorems follow the table:
The 682564f rows stay in the spec as functions. The new test
t741_the_closed_fix_issues_moved_four_cells_and_none_backchecks the move: 11 -> 7 FALSE cells, 4 FALSE -> TRUE, and nothing else.Verification
Run on the t27c lab with t27c built from master 17e01d2 and zig 0.16.0:
t27c test-reportpasses 10/10, with 0 vacuous passes. Master's version passed 9/9.t27c seal --save, then--verify, which printed "all hashes MATCH". The seal records 10/10 tests passed.check_all()passes on the diff. The added text is ASCII only.Only
.t27and its seal change. No hand-written code.🤖 Generated with Claude Code