[HiCache] Buffer mode support sidecar pool (#37424)

Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
This commit is contained in:
huangtingwei
2026-09-04 23:24:18 +08:00
committed by GitHub
co-authored by Zhiqiang Xie
parent 8b1d8c1703
commit 0d0e2f92be
3 changed files with 362 additions and 16 deletions
@@ -42,7 +42,12 @@ from sglang.srt.mem_cache.base_prefix_cache import (
InsertParams,
MatchPrefixParams,
)
from sglang.srt.mem_cache.hicache_storage import PoolHitPolicy, PoolName, PoolTransfer
from sglang.srt.mem_cache.hicache_storage import (
PoolHitPolicy,
PoolName,
PoolTransfer,
SidecarPoolSpec,
)
from sglang.srt.mem_cache.radix_cache import RadixKey
from sglang.srt.mem_cache.unified_cache.cache_action import RebuildFullToSWAMapping
from sglang.srt.mem_cache.unified_cache.components import (
@@ -56,6 +61,7 @@ from sglang.srt.mem_cache.unified_cache.unified_tree_core_interface import (
)
if TYPE_CHECKING:
from sglang.srt.mem_cache.pool_host import HostPoolGroup
from sglang.srt.mem_cache.unified_cache.components import SWAComponent
from sglang.srt.mem_cache.unified_radix_cache import UnifiedRadixCache
@@ -161,18 +167,35 @@ def staged_splice_tokens(f: _StagedPrefetch, device_prefix_len: int) -> int:
def validate_buffer_only_stack(
sidecar_pool_specs: list, swa_component: Optional[SWAComponent]
sidecar_pool_specs: list[SidecarPoolSpec],
host_pool_group: HostPoolGroup,
swa_component: Optional[SWAComponent],
) -> None:
"""Post-assembly buffer-mode fences.
Sidecar pools (DSv4 compressed regions) and unified_kv SWA (device-only
ring, never offloaded) have no per-pool staging path yet.
Sidecars reuse their source pool's transient slot ids, so every sidecar
host pool must expose the full source slot namespace. unified_kv SWA
(device-only ring, never offloaded) still has no staging path.
"""
if sidecar_pool_specs:
raise ValueError(
"--hicache-host-memory-mode buffer_only does not support "
"sidecar storage pools (DeepSeek-V4 compressed regions)."
)
entry_map = host_pool_group.entry_map
for spec in sidecar_pool_specs:
source = entry_map.get(spec.indices_from_pool)
sidecar = entry_map.get(spec.pool_name)
if source is None or sidecar is None:
raise ValueError(
"--hicache-host-memory-mode buffer_only sidecar pool mapping "
f"is incomplete: pool={spec.pool_name}, "
f"indices_from_pool={spec.indices_from_pool}."
)
source_size = source.host_pool.logical_size
sidecar_size = sidecar.host_pool.logical_size
if sidecar_size < source_size:
raise ValueError(
"--hicache-host-memory-mode buffer_only sidecar host pool is "
"smaller than its index source: "
f"pool={spec.pool_name}, host_slots={sidecar_size}, "
f"source={spec.indices_from_pool}, source_slots={source_size}."
)
swa = swa_component
if swa is not None and swa._swa_kv_pool_host is None:
# Only reachable on SWA models with the unified_kv layout (SWA as
@@ -547,6 +570,11 @@ class BufferModePipeline:
cc = cache.cache_controller
snapshot = intent.snapshot
aux_xfers = [x for xfers in comp_xfers.values() for x in xfers]
# Sidecars reuse the source pool's transient host/device indices. This
# includes both KV-derived pools and SWA-derived DSV4 state pools. They
# allocate no additional staging, but must ride the same D2H operation
# so their bytes are present when the storage write starts.
aux_xfers.extend(cache._build_backup_sidecar(device_value, comp_xfers))
host_indices = cc.write(
device_value,
node_id=snapshot.node_id,
@@ -625,20 +653,46 @@ class BufferModePipeline:
snapshot = intent.snapshot
self._cache.dec_lock_ref(snapshot.node_id, entry.lock_params)
# Every aux pool writes a trailing snapshot keyed by the last KV page
# hashes it covers: the SWA window spans page_size-sized pages, the
# Mamba state is a single slot (host pool page_size 1 -> one key).
# Every independently staged aux pool writes a trailing snapshot keyed
# by the last KV page hashes it covers. A derived sidecar writes the
# exact key span of its source pool while reusing that source's slots.
storage_xfers: list[PoolTransfer] = []
storage_sources = {
PoolName.KV: PoolTransfer(
name=PoolName.KV,
host_indices=entry.host_indices,
keys=snapshot.hash_values,
)
}
for staged in entry.aux_xfers:
if staged.indices_from_pool is not None:
continue
keys = self._aux_window_keys(snapshot.hash_values, staged)
if keys is None:
continue
transfer = PoolTransfer(
name=staged.name,
host_indices=staged.host_indices,
keys=keys,
hit_policy=PoolHitPolicy.TRAILING_PAGES,
)
storage_xfers.append(transfer)
storage_sources.setdefault(staged.name, transfer)
for staged in entry.aux_xfers:
if staged.indices_from_pool is None:
continue
source = storage_sources.get(staged.indices_from_pool)
if source is None:
raise AssertionError(
"Buffer-mode storage sidecar source missing: "
f"{staged.name} from {staged.indices_from_pool}."
)
storage_xfers.append(
PoolTransfer(
name=staged.name,
host_indices=staged.host_indices,
keys=keys,
hit_policy=PoolHitPolicy.TRAILING_PAGES,
keys=source.keys,
hit_policy=staged.hit_policy,
indices_from_pool=staged.indices_from_pool,
)
)
operation_id = self._cache.cache_controller.write_storage(
@@ -832,6 +886,14 @@ class BufferModePipeline:
prefix_ctx = self._prefetch_prefix_ctx.pop(req_id, None)
prefix_tokens = prefix_ctx[0] if prefix_ctx is not None else None
aux_xfers = [x for xfers in comp_xfers.values() for x in xfers]
# Component transfers are already present in comp_xfers. Preserve the
# derived sidecars from the storage operation as well; cc.load resolves
# them against the freshly allocated source host/device indices.
aux_xfers.extend(
transfer
for transfer in operation.pool_transfers or ()
if transfer.indices_from_pool is not None
)
if num_tokens == 0 or prefix_tokens is None:
# Nothing usable fetched: recompute.
@@ -439,7 +439,9 @@ class UnifiedRadixCache(BasePrefixCache):
if self.host_memory_mode == "buffer_only":
swa = self.components.get(ComponentType.SWA)
validate_buffer_only_stack(
sidecar_pool_specs=self.sidecar_pool_specs, swa_component=swa
sidecar_pool_specs=self.sidecar_pool_specs,
host_pool_group=self.host_pool_group,
swa_component=swa,
)
self.buffer_pipeline = BufferModePipeline(
cache=self,