2b.3 fixes: restore two l2_pooling-adopt gaps surfaced by live validation

The flag-on 2b.3 bring-up (--enable-cp-shared-physical-l2-hicache, real GLM-5.1
PD traffic) surfaced two partial-adopt gaps from the piecemeal 2b.0/2b.1b ports
from l2_pooling — a caller brought in without its callee/declaration. Neither
was caught by the pool unit tests or import smoke (they don't exercise the
device-allocator load path or the flag-on capacity-snapshot path); only live
flag-on traffic hit them.

1. environ.py: declare SGLANG_CP_HICACHE_VERIFY_SNAPSHOT (EnvBool, default False).
   hiradix _cp_host_capacity_snapshot (:1477) reads it to gate a fail-loud
   running-count drift verifier (_cp_walk_capacity_counts vs _cp_counts), but the
   env field was never declared -> AttributeError crash on the first flag-on
   scheduler tick. The verifier IS fully implemented (it's the count-machinery
   analog of the placement_digest assert) -- keep it, default-off (expensive
   full-walk; on in bring-up/CI).

2. allocator.py: restore alloc_pages_for_shared_l2_load (thin wrapper over the
   adopted alloc_pages_with_owners) on the CP shared paged allocator. 2b.0 ported
   the callee but dropped this wrapper that cache_controller.load_cp_shared_l2
   calls -> RuntimeError on every L2->L1 cache-hit reload until restored.

Validated live: prefill boots with the flag, shared-L2 slab constructs over
hugetlbfs on all 8 ranks, write-through + load-back + cache hits + coherent
output all work, placement_digest assert green (no divergence), no crash.

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

python/sglang/srt/environ.py

@@ -393,6 +393,10 @@ class Envs:
     # is replicated (Theorem 1). When on, MIN/MAX-reduce placement_digest across the
     # CP group each tick and fail loud on divergence. On in CI/bring-up, off in prod.
     SGLANG_CP_HICACHE_PLACEMENT_ASSERT = EnvBool(False)
+    # CP HiCache debug drift-detector: when on, _cp_host_capacity_snapshot walks the
+    # full radix tree and asserts the running per-rank counts match (fail loud on
+    # drift). Expensive full-walk per snapshot -> default off; on in bring-up/CI.
+    SGLANG_CP_HICACHE_VERIFY_SNAPSHOT = EnvBool(False)
 
     # Mooncake KV Transfer
     SGLANG_MOONCAKE_CUSTOM_MEM_POOL = EnvStr(None)

python/sglang/srt/mem_cache/allocator.py

@@ -1045,6 +1045,20 @@ class CPSharedPagedTokenToKVPoolAllocator(PagedTokenToKVPoolAllocator):
 
         return out_indices
 
+    def alloc_pages_for_shared_l2_load(
+        self, current_page_owners: List[int]
+    ) -> Optional[torch.Tensor]:
+        """Allocate fresh L1 pages for shared-physical L2 load-back.
+
+        ``current_page_owners`` is the load-time owner plan recomputed
+        positionally ((page-1) % cp_size), NOT the write-time ``page_owners``
+        metadata used by legacy CP HiCache replay. Thin wrapper over
+        ``alloc_pages_with_owners``; restored here because the 2b.0 adopt
+        ported the callee but dropped this wrapper that
+        ``cache_controller.load_cp_shared_l2`` calls.
+        """
+        return self.alloc_pages_with_owners(current_page_owners)
+
     def alloc_pages_with_owners(
         self,
         page_compute_owners: List[int],