leavelet 144ae7acb1 Regression test: CP HiCache L3 reload same-tick admit->reserve race ...

Drives _cp_l3_reserve_reloads directly against a real CpSharedL2PageAllocator:
- same-tick state (l3rl object committed-and-live + node attached -> attach_ok False)
  no longer raises 'duplicate live reservation'; the candidate is dropped and the
  request left un-held (re-matches the inserted node).
- WITHOUT the recheck (attach_ok forced True = pre-fix path) the identical state
  raises the duplicate error -> proves the recheck is load-bearing.
- an in-flight reload still piggybacks (recheck sits after the dedup, not before).
- a genuine clean miss still reserves (no false-positive starvation).

CPU unit test (register_cpu_ci, stage-a-test-cpu); runs in the dev-cu13 container.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-06-24 14:52:30 +00:00

..

test_alloc_pages_with_owners.py

Fix CP HiCache load_cp owner-pattern mismatch (cache-hit corruption)

2026-05-27 00:01:39 +08:00

test_convert_to_bigram_key.py

L3 3.2: async disk->L2 reload + request hold (Model B), zero new collectives, EAGLE-correct

2026-06-23 00:09:38 +00:00

test_cp_hicache_chunked_prepare_repro.py

Fix CP HiCache catch_up_all_layers fallback on chunked-prefill final chunk

2026-06-18 11:54:24 +00:00

test_cp_hicache_load_back_owner_lanes.py

CP HiCache: redesign owner-lane eviction planner as a lazy-re-evaluation heap (O(n log n))

2026-06-24 01:35:37 +00:00

test_cp_hicache_metadata.py

Gate CP HiCache write-backup on hard deficit, not the free-room watermark

2026-06-19 12:14:36 +00:00

test_cp_l3_config.py

L3 3.0: config (disks + rank->disk mapping + PLP gate)

2026-06-22 21:31:12 +00:00

test_cp_l3_disk.py

CP HiCache L3 3.4: cold-start (restart durability) — load|clear via SGLANG_CP_L3_COLD_START

2026-06-23 18:09:37 +00:00

test_cp_l3_index.py

L3 3.0: shared per-node LMDB metadata index (CpL3MetaStore)

2026-06-22 21:31:12 +00:00

test_cp_l3_posix.py

CP HiCache L3 3.4: cold-start (restart durability) — load|clear via SGLANG_CP_L3_COLD_START

2026-06-23 18:09:37 +00:00

test_cp_l3_reload_reserve_race.py

Regression test: CP HiCache L3 reload same-tick admit->reserve race

2026-06-24 14:52:30 +00:00

test_cp_l3_slab_accessor.py

CP HiCache: multi-slab host KV cache to respect cudaHostRegister per-call ceiling

2026-06-23 15:33:10 +00:00

test_cp_l3_store.py

CP HiCache L3 3.4: cold-start (restart durability) — load|clear via SGLANG_CP_L3_COLD_START

2026-06-23 18:09:37 +00:00

test_cp_shared_kv_compose.py

Wire the symm current exchange into the CP shared-KV prefetchers

2026-06-11 22:39:47 +00:00

test_cp_shared_kv_layout.py

Keep CP HiCache free room from collapsing under cache reuse

2026-06-07 01:23:34 +08:00

test_cp_shared_kv_runtime.py

Avoid remap-cache aliasing across tensor views

2026-06-21 18:47:09 +00:00

test_cp_shared_l2_pool.py

CP HiCache: fix L3-reload partial-prefix node split crash (well-form the node)

2026-06-23 17:09:32 +00:00

test_cp_shared_l2_slab_cap.py

CP HiCache: multi-slab host KV cache to respect cudaHostRegister per-call ceiling

2026-06-23 15:33:10 +00:00

test_evict_policy.py

CP HiCache: replicated-clock SLRU eviction, collective-safe (Phase 1)

2026-06-19 19:35:21 +00:00

test_kv_pool_buf_infos.py

perf(disaggregation): cache static buffer metadata

2026-04-19 04:23:48 +08:00

test_mamba_unittest.py

ci: unify PR test suite naming (#21187)

2026-03-23 00:18:45 -07:00

test_nsa_pool_host_unit.py

Compact skipped NSA index-cache state safely

2026-06-10 05:37:06 +08:00

test_page_index_utils.py

feat: add page-aligned index validation for CP HiCache direct transfers

2026-05-10 02:02:54 +08:00

test_radix_cache_slru_accuracy.py

ci: unify PR test suite naming (#21187)

2026-03-23 00:18:45 -07:00

test_radix_cache_unit.py

CP HiCache: replicated-clock SLRU eviction, collective-safe (Phase 1)

2026-06-19 19:35:21 +00:00

test_radix_key_views.py

Speed up Python bigram key conversion with C-level zip

2026-06-10 04:56:24 +00:00

test_req_to_token_pool.py

Stabilize disaggregated decode burst handoff

2026-06-05 19:59:04 +08:00

test_swa_unittest.py

ci: unify PR test suite naming (#21187)

2026-03-23 00:18:45 -07:00