yams/tests/benchmarks
Trevon 211db15606
fix(daemon): bound semantic neighbor scores and report startup phase (#334)
Semantic-neighbor maintenance aborted the daemon during startup:
std::clamp<float> asserted '!(__hi < __lo)' because the adaptive cosine
threshold could exceed 1.0 for a legal pair of identical vectors. The same
work was also admitted while the daemon was still initializing, so the
300-second starvation guard forced a pass before readiness.

+ semanticInverseNorm/semanticCosineSimilarity/semanticExplicitThreshold in the
  existing embed policy header, shared by the streamed and corpus branches;
  non-finite vectors or scores produce no neighbor and legal scores are bounded
  before ranking and threshold selection
+ startupMaintenanceEligible() requires ServiceManagerState::Ready, not
  stopping and idle, and is evaluated before
  ResourceGovernor::admitDeferrable(), so an elapsed deferral deadline can no
  longer manufacture missing readiness
+ StartupPhaseTracker (two atomics, no mutex) plus a warn-level init-waiter
  line reporting state, phase and phase_elapsed_ms instead of a debug-level
  state number; a 53-minute startup previously logged nothing actionable
+ regression through the public backfill and corpus-rebuild APIs using a 64-D
  vector whose old float inverse norm rounds its self-cosine above 1
- removed the duplicated inverse-norm and cosine lambdas from EmbeddingService
- renamed __test_setModelProvider/__test_setKgStore to testing_* because they
  used reserved identifiers, across 11 files
- re-anchored 13 pre-existing raw-environment allowlist entries to their new
  line numbers; no new exception was added

Startup is still blocked by corpus-sized memory-sync scans; this change stops
the abort and names the blocking phase, it does not make startup fast.

Signed-off-by: trvon <git@trevon.dev>
2026-10-22 16:48:49 -06:00
..
baseline docs: reduce documentation surface 2026-07-29 13:02:15 -04:00
configs feat: instrumentation improvements for benchmarks, shell completition improvements, packaging updates for completions and docs updates 2026-04-13 21:34:18 -04:00
data Search/grep UX, runtime hardening, and CI stabilization (#39) 2026-05-27 12:15:17 -04:00
relevance feat(search): make topology-assisted retrieval the default (#57) 2026-07-13 20:26:43 -04:00
scripts fix(memory-sync): stop re-walking the op store on every publish (#319) 2026-10-17 15:30:35 -06:00
search fix(daemon): bound semantic neighbor scores and report startup phase (#334) 2026-10-22 16:48:49 -06:00
api_benchmarks.cpp refactor(bench): type run configuration 2026-08-05 20:53:18 -04:00
beir_loader.h feat(topology): typed event time and session for topology inputs (#305) 2026-10-14 08:47:04 -06:00
bench_utils.h fix: harden CI and metadata stability (#48) 2026-06-11 16:42:54 -04:00
benchmark_base.h fix: small fixes for benchmarks and documentation update 2026-04-17 22:32:04 -04:00
benchmark_cli.h fix(bench): reject unknown arguments 2026-08-05 20:53:19 -04:00
core_benchmarks.cpp refactor(bench): type run configuration 2026-08-05 20:53:18 -04:00
daemon_socket_accept_bench.cpp fix: improving repair check 2026-04-17 18:47:24 -04:00
daemon_warm_latency_main.cpp test: isolate process environment 2026-08-04 21:31:39 -04:00
entity_extraction_bench.cpp build: standardize the Debug build dir on build/debug (#248) 2026-10-01 08:45:20 -06:00
fts5_indexing_bench.cpp perf: improve ingestion optimization and profiling infrastructure (#52) 2026-06-19 17:32:19 -04:00
grep_benchmarks.cpp fix(security): harden fold-flagged guards 2026-05-17 23:50:32 -04:00
hnsw_entity_search_bench.cpp refactor(symbols): remove in-tree tree-sitter symbol extractor plugin and tests 2026-09-22 07:37:30 -06:00
ingestion_throughput_bench.cpp fix(daemon): bound semantic neighbor scores and report startup phase (#334) 2026-10-22 16:48:49 -06:00
ipc_streaming_bench.cpp v0.7.3 staging 2025-10-07 11:18:35 -04:00
kg_edge_insert_bench.cpp docs: repair benchmark documentation references 2026-07-29 22:48:22 -04:00
large_scale_ingestion_bench.cpp fix(daemon): bound semantic neighbor scores and report startup phase (#334) 2026-10-22 16:48:49 -06:00
memory_sync_listing_bench.cpp fix(memory-sync): stop re-walking the op store on every publish (#319) 2026-10-17 15:30:35 -06:00
meson.build fix(memory-sync): stop re-walking the op store on every publish (#319) 2026-10-17 15:30:35 -06:00
metadata_contention_bench.cpp perf(metadata): profile write-path batching (#49) 2026-06-12 14:11:49 -04:00
metadata_path_query_bench.cpp feat(daemon): shutdown hardening + test helper linkage fixes (#44) 2026-06-08 16:29:44 -04:00
multi_client_ingestion_bench.cpp chore(warnings): clean up compiler warnings on GCC, Clang and Apple Clang (#279) 2026-10-08 14:30:28 -06:00
repair_ability_bench.cpp test(harness): relocate runtime dir and scrub model overrides under isolation 2026-10-01 05:54:16 -06:00
retrieval_contract.py fix: fence enrichment completion and reconcile promotion prerequisites (#157) 2026-09-12 18:53:12 -06:00
retrieval_service_benchmarks.cpp build: standardize the Debug build dir on build/debug (#248) 2026-10-01 08:45:20 -06:00
run_multi_client_optimization_loop.sh feat: expanding storage system to work with r2 backends via object storage plugin. Updating benchmarks and documentation 2026-03-07 19:43:14 -05:00
run_multi_client_sweep.sh feat: expanding storage system to work with r2 backends via object storage plugin. Updating benchmarks and documentation 2026-03-07 19:43:14 -05:00
search_bandit_bench.cpp feat(search): add Simeon bandit tuning (#33) 2026-05-17 04:08:50 -04:00
search_benchmarks.cpp refactor(bench): type run configuration 2026-08-05 20:53:18 -04:00
simeon_lexical_score_bench.cpp perf(search): reuse Simeon scores within a request 2026-09-05 11:54:16 -06:00
single_client_pipeline_bench.cpp fix: harden CI and metadata stability (#48) 2026-06-11 16:42:54 -04:00
storage_throughput_bench.cpp fix(build): restore clean test and benchmark links 2026-08-06 09:59:07 -04:00
symbol_ground_truth_v2.jsonl refactor: simplify runtime paths and harden retrieval (#61) 2026-07-31 14:16:45 -04:00
test_retrieval_contract.py fix: fence enrichment completion and reconcile promotion prerequisites (#157) 2026-09-12 18:53:12 -06:00
topology_ablation_bench.cpp chore(warnings): clean up compiler warnings on GCC, Clang and Apple Clang (#279) 2026-10-08 14:30:28 -06:00
topology_build_bench.cpp perf(topology): streamline artifact construction 2026-08-03 17:52:58 -04:00
topology_topic_vocab.h feat: expanding topology system and search for better testing and simeon integration 2026-04-19 09:25:28 -04:00
TRACY_PROFILING.md feat(profiling): Tracy for every library, zones on p2p and memory-sync, per-test CPU/RSS (#294) 2026-10-14 08:32:56 -06:00
tree_build_bench.cpp Remediate Fold SAST findings 2026-05-27 22:41:44 -04:00
tree_diff_benchmarks.cpp refactor(bench): type run configuration 2026-08-05 20:53:18 -04:00
tree_list_filter_bench.cpp experimental → main: MCP timeout fix + doctor integrity checks + search/perf improvements (#31) 2026-05-14 23:29:44 -04:00
vector_backend_engine_compare.cpp chore(warnings): clean up compiler warnings on GCC, Clang and Apple Clang (#279) 2026-10-08 14:30:28 -06:00
vector_memory_scaling_bench.cpp chore(warnings): clean up compiler warnings on GCC, Clang and Apple Clang (#279) 2026-10-08 14:30:28 -06:00
vector_persistence_warm_start_bench.cpp perf(vector): persist reusable warm starts 2026-08-03 17:52:58 -04:00
write_coordinator_bench.cpp refactor(plugins)!: remove the symbol_extractor_v1 plugin ABI 2026-09-22 13:30:35 -06:00
zyp_extraction_bench.cpp perf(zyp): true parallel PDF extraction + throughput benchmark 2026-05-15 12:13:33 -04:00