Repository navigation
floor spawn_width: derive from spec breadth (parallelize CI batch-2, 1→7) - #5356
Merged
Merged
Conversation
…minating witness Floor ran serially because the corpus WorkDemand hardcoded shard_count=1, so spawn_width = min(1, threads) = 1. Now the floor plan derives breadth from the single-authority spec (one unit per gate) and the fleet supplies the thread cap: spawn_width = min(breadth, hardware_threads). Verified by execution: 1 -> 7. Witnesses: spawn_width tracks breadth (the relationship), and floor-not-serial (discriminating — reverting to a constant shard_count goes RED). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
gunbai-bot
Bot
changed the title
claude-opus-4-7 review on #5356: the +1 fold is just length(). Verified length resolves on gunbc_ci_spec.gates across the v2->dsl bridge and still yields 7; witness stays green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
briansrls
force-pushed
the
session/warm-crane-135
branch
from
June 20, 2026 01:16
9bf04d6 to
412206f
Compare
briansrls
added a commit
that referenced
this pull request
Jun 20, 2026
…) into Arc 1 branch to resolve serial-floor OOM
briansrls
added a commit
that referenced
this pull request
Jun 20, 2026
2 of 3 tasks
7 of 8 tasks
briansrls
added a commit
that referenced
this pull request
Jun 20, 2026
… HEAD; main side == #5356 subset) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
briansrls
added a commit
that referenced
this pull request
Jun 20, 2026
…budget, not CPU only (#5375) * WIP: CI is extremely flakey * floor spawn_width: derive from spec breadth (no hand-tuning) + discriminating witness Floor ran serially because the corpus WorkDemand hardcoded shard_count=1, so spawn_width = min(1, threads) = 1. Now the floor plan derives breadth from the single-authority spec (one unit per gate) and the fleet supplies the thread cap: spawn_width = min(breadth, hardware_threads). Verified by execution: 1 -> 7. Witnesses: spawn_width tracks breadth (the relationship), and floor-not-serial (discriminating — reverting to a constant shard_count goes RED). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * review nit: use length(gates) instead of fold-count for breadth claude-opus-4-7 review on #5356: the +1 fold is just length(). Verified length resolves on gunbc_ci_spec.gates across the v2->dsl bridge and still yields 7; witness stays green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * WIP: CI is extremely flakey * WIP: CI is extremely flakey * fix(ci-floor): memory-aware spawn_width — bound batch-2 fan-out by cgroup budget, not CPU only placement_spawn_width was memory-blind (min of shard_count and hardware_threads). At width=7 the batch-2 corpus resolves OOM-killed (exit 137) deterministically as the corpus grows (eager-boar-790, 3 runs). Wire the already-modeled memory terms (PlacementSupplyRow.ram_bytes + ResourceEnvelope.memory) into the width: spawn_width = min(shard_count, hardware_threads, memory_budget / per_unit_peak). Verified by execution: floor spawn_width 7 -> 4 (4x14GiB=56GiB <= 0.50x125GiB budget, safe under srv1's ~65GiB cgroup cap); gunbc_ci_floor_spawn_fits_memory_budget -> true; ci_floor_plan_witnesses -> true (adds witness_floor_width_fits_memory_budget + witness_floor_width_memory_bounded, discriminating vs the memory-blind regression). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: resolve auto-committed merge conflict markers (take memory-aware HEAD; main side == #5356 subset) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansrls@gunb.ai> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> Co-authored-by: Brian Searls <11205878+briansrls@users.noreply.github.com>
briansrls
added a commit
that referenced
this pull request
Jun 20, 2026
…errors (#5363) * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * Fix missing bash_lex_rule exposed by parse-resilient resolve. Slice 5d exit-code lex rules call bash_lex_rule but the helper was never defined; whole-tree discovery now reaches resolve and failed on bash.dag. Co-authored-by: Cursor <cursoragent@cursor.com> * Drop out-of-scope bash_lex_rule hotfix from parse-resilience PR. §3-correct lex authority fix lands separately on main (eager-boar-790); keep #5363 scoped to front_end_sources only (compile.dag + stage0 seed + test). Co-authored-by: Cursor <cursoragent@cursor.com> * hotfix(bash): bash_exit_code_lex projects from a fixed-span phrase, not the undefined bash_lex_rule (main-red fix) #5347 (slice 5d Exit) left `bash_exit_code_lex()` calling a `bash_lex_rule(...)` constructor that is defined nowhere in the tree. The whole-tree CI floor resolves every *_test.dag closure, so this dangling call failed `resolve` for bash_emit_command_test.dag and turned main RED for every open PR (since #5347 merged at 01:12 UTC; pre-#5347 5c493fc was green). Fix per §3 single authority (lexing.dag LexicalPhrase): the lex rule set PROJECTS from the construct's FIXED-spelling spans via `phrase_to_lex_rules(p: phrase([...]))`, exactly as the bound-token sibling `bash_env_prefixed_cargo_lex` does — NOT a direct rule constructor (the "authored three times that could silently disagree" anti-pattern). Only the `exit ` keyword is a fixed span; the exit code is an emit-time BOUND token (^bash_token_exit_code via binding_spellings), so it contributes no literal lex rule — preserving the original single-rule behavior bit-for-bit. NOTE: this is a tighter fix than the two-span suggestion (incl. the code span) — a LexicalPhrase is fixed-spelling-only by contract, so the bound code cannot be a span; a literal span for it would over-generate a spurious rule. Confirmed with the bash arc owner (warm-badger-46) who also cleared the colliding bash_lex_rule constructor in #5363. Verified by execution (prebuilt claim_batch, --source-root dsl --source-root src/v2): - bash_emit_command_test.dag now RESOLVES clean (43 modules, was resolve-FAIL) - bash_emit_command_accepts / bash_emit_bound_words: PASS - bash_emit_exit_code_{0,1}: PASS - bash_emit_exit_code_1_quoted_wrong_{ne_correct,model,rejects}: PASS (discriminating RED intact) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * test(#5363): discovery advisory classification + execution receipt. Unit test pins is_discovery_corpus_blocking_diagnostic boundary (typecheck advisory vs parse/resolve hard-fail). Ignored receipt proves 255 discovery entry resolves green with advisory gate (0 failures). Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * fix(#5363): scope advisory gate to discovery-only resolve entry points. Thread ResolveTypecheckGate through resolve paths: Strict default on resolve_entry_with_index / compile consumers; DiscoveryCorpusAdvisory only via resolve_entry_with_index_for_discovery_corpus and discover_owned_data_decls. Fixes compile error (CompilerDiagnostic import) from prior scope commit. Co-authored-by: Cursor <cursoragent@cursor.com> * chore(#5363): drop bash.dag from PR — fix already on main (#5356). Rebased onto green main; bash_exit_code_lex phrase projection is upstream. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * fix(#5363): satisfy clippy in parse-cache resilience test. rust_monolith_gate runs cargo clippy --all-targets -D warnings; use from_ref instead of cloned slice literal for build_multi_entry_index. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: front_end_sources PARSE-RESILIENCE root fix: a parse error in ONE file m * fix(#5363): discovery advisory gate must retain typed graph in merged resolve. resolved_graph_from_sources called compile_to_resolved, which nils graph on any typecheck error even when DiscoveryCorpusAdvisory would demote them — breaking discover_owned_data_decls. Thread the gate through a gated compile path so advisory-only typecheck debt still returns a ResolvedGraph. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Brian Searls <briansrls@gunb.ai> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Claude Opus 4.8 <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
1 participant
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.
Auto-opened by session-dashboard for session
warm-crane-135.Pushing to
session/warm-crane-135advances this PR.Worker attestation
Before flipping this PR to ready for review, confirm each item:
npm test,cargo test) and the result.Closes #Ndirective.Summary
TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.
Test plan