Repository navigation
fix(gen-c): a local nested array is declared as a C matrix (Closes #7441) - #7495
Merged
Merged
Conversation
) A local `[16][16]u8` kept its inner `[16]` in the element type and reached C as `[16]u8 m[16]`. Every inner fixed dimension now moves to the declarator, `uint8_t m[16][16]`, as a module-level array and a struct field already do. A slice or sentinel dimension stops the peel. The test in backend_behaviour.rs compiles and runs a 2-D, a 3-D with a named size, and an initialized local. Part of #5980. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…l matrix declarator Only gen_hash_c moves: the local declarators become `T m[A][B]`. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…rray # Conflicts: # bootstrap/stage0/FROZEN_HASH # bootstrap/tests/backend_behaviour.rs
gHashTag
added
the
owner-approved-foreign
gHashTag
enabled auto-merge (squash)
October 7, 2026 11:43
Owner
Author
|
Lab run on the merged head dad468e: https://t27c-lab-production.up.railway.app/runs/dad468e0574f23b35cc39537dca5bf7b51cd7a14.json. frozen-hash, build, suite (RATCHET CLEAN), seal-currency, seal-coverage (1489 seals, 6 stale, the same as master), specs-generate and misread are green. The one red, the lean gate ( |
This was referenced Oct 7, 2026
Merged
corpus: judge 26 new type conflicts from the port batches; types ledger 84 -> 110 (Refs #5906)
#7482
Merged
Merged
Closed
Merged
Merged
Contributor
This was referenced Oct 7, 2026
This was referenced Oct 8, 2026
#7331 landed and changed bootstrap/src/compiler.rs; compiler.rs merged cleanly (both sides' gen-c changes kept), and the derived files that conflicted are regenerated, not merged by hand: - bootstrap/stage0/FROZEN_HASH: the sha256 of the merged compiler.rs, 3c0ade9e73e425d5... - The three similarity_search seals (SimilaritySearch, VSASimilaritySearch, vsa_VSASimilaritySearch): resealed with `t27c seal --save`, t27c built from this tree on the t27c lab, zig 0.16 on PATH (8/8 tests pass). `check_seal_currency.py --stale-specs` then names only html, xml and terminal, master's ledgered debt. - check_seal_currency: exit 0; check_seal_coverage: OK (1641 seals, 1520 hold, 121 known-broken, as on the baseline); `tri seals twins`: 1636 / 548 / 31 disagree, the same as master eda79f0; backend_behaviour 24/24 (c_declares_a_local_nested_array_as_a_matrix ok), core_selfhost 1/1. gen-c of t27core.t27 is unchanged (the DDC pins of master and of #7436 both still hold). Closes #7441 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Owner
Author
|
Merge shepherd: #7331 landed, and this PR then conflicted, so I merged master eda79f0 in from master's side (9360cc3, a fast-forward). |
gHashTag
enabled auto-merge (squash)
October 8, 2026 17:57
This was referenced Oct 8, 2026
Contributor
PR DashboardGenerated at: 2026-10-08 18:34:18 UTC
Summary
Seal Status
|
This was referenced Oct 8, 2026
This was referenced Oct 8, 2026
Closed
gHashTag
added a commit
that referenced
this pull request
Oct 8, 2026
…loses #7929) (#7931) Part of #5980 (P1: `**` with a real meaning, numeric literal forms). The self-hosted core accepts a local `var` array as gen-c lowers it, a module array without `static`: a repeat (a GNU range, `{0}` for zero), a list, or a row list (a C matrix, #7495), in a function or a test. It refuses what gen-c lowers with loss: - E_NAME: a local array named like a module name, a parameter or another local of its function (C's shadowing array, which Zig refuses); - E_ARRAY: a local array read outside its block, a dimension or count that is not a number or a number-valued const (an initialized C VLA), and a repeat count that differs from the dimension, module arrays included (C zeros where Zig refuses the type). #7318 is fixed, so the `_` separators in a dimension and in a negated literal const are dropped as gen-c drops them; the three refusals move to accept/separator_dimensions.t27. Numeric `a ** b` stays refused (E_SYNTAX), now pinned by two fixtures on #7427. t27core_ddc.t27 restates the new module vars and is repinned: S 84235, E 78795 bytes. Co-authored-by: Claude <claude@anthropic.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
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 #7441
Part of #5980
What was wrong
gen-c declared a local nested fixed array with a t27 type left inside the C text:
var m: [16][16]u8became[16]u8 m[16], which no C compiler accepts.specs/depin/prove.t27andspecs/vsa/similarity_search.t27hit it, and the self-host needs it.The fix
bootstrap/src/compiler.rs(+24/-2): every inner fixed dimension moves from the element type to the declarator, the same way a struct field (c_array_field) and a module-level array are already declared:uint8_t m[16][16]. A slice, sentinel or inferred inner dimension stops the peel and keeps the old path.Regression test
c_declares_a_local_nested_array_as_a_matrixinbootstrap/tests/backend_behaviour.rscompiles three functions with cc and runs them: a[16][16]u8, a[2][K][4]i64with a named constant dimension, and an initialized[2][2]u8. It expects4 8 32.Run on the lab both ways:
Reseals
Only
gen_hash_cmoves:.trinity/seals/prove.jsonanddepin_depin.prove.json: 1fe0bdf2 -> 832862c5.trinity/seals/SimilaritySearch.json,VSASimilaritySearch.jsonandvsa_VSASimilaritySearch.json: fa36bdca -> 9b62b16eLab runs
Foreign-line budget (#7371)
Master's
check_budget()passes on this diff: compiler.rs +24/-2, backend_behaviour.rs +16. FROZEN_HASH and the seals do not count.🤖 Generated with Claude Code