Remove the HiMambaRadixTree that is no longer in use (#33468)

This commit is contained in:
Zhangheng
2026-08-07 23:01:40 +08:00
committed by GitHub
parent d4be483efb
commit 12de7fb1f6
6 changed files with 5 additions and 2287 deletions
@@ -16,8 +16,7 @@ from sglang.srt.mem_cache.base_prefix_cache import (
)
from sglang.srt.mem_cache.cache_init_params import CacheInitParams
from sglang.srt.mem_cache.common import available_and_evictable_str
from sglang.srt.mem_cache.hi_mamba_radix_cache import HiMambaRadixCache
from sglang.srt.mem_cache.mamba_radix_cache import LRUList, MambaRadixCache, TreeNode
from sglang.srt.mem_cache.mamba_radix_cache import MambaRadixCache
from sglang.srt.mem_cache.memory_pool import (
HybridLinearKVPool,
HybridReqToTokenPool,
@@ -628,53 +627,6 @@ class TestMamba(unittest.TestCase):
return tree, allocator, req_to_token_pool, make_dummy_req
def test_hi_mamba_tombstone_cleanup_respects_host_ref(self):
tree = object.__new__(HiMambaRadixCache)
root = TreeNode()
parent = TreeNode()
deleted = TreeNode()
root.key = RadixKey(array("q", []))
parent.key = RadixKey(array("q", [1]))
deleted.key = RadixKey(array("q", [2]))
parent.parent = root
deleted.parent = parent
parent.value = torch.tensor([1], dtype=torch.int64)
parent.protect_host()
root.children[parent.key.child_key(1)] = parent
class RecordingCacheController:
def __init__(self):
self.device_evictions = []
self.host_evictions = []
def evict_device(self, value):
self.device_evictions.append(value)
def evict_host(self, value):
self.host_evictions.append(value)
tree.root_node = root
tree.page_size = 1
tree.full_lru_list = LRUList(mamba=False)
tree.full_lru_list.insert_mru(parent)
tree.cache_controller = RecordingCacheController()
tree.full_evictable_size_ = len(parent.value)
tree.evictable_full_device_leaves = {parent}
tree.evictable_full_host_leaves = set()
result_node, full_evicted, mamba_evicted = (
tree._iteratively_delete_tombstone_leaf(deleted)
)
self.assertIs(result_node, deleted)
self.assertEqual(full_evicted, 0)
self.assertEqual(mamba_evicted, 0)
self.assertIs(root.children[parent.key.child_key(1)], parent)
self.assertTrue(tree.full_lru_list.in_list(parent))
self.assertEqual(tree.cache_controller.device_evictions, [])
self.assertEqual(tree.cache_controller.host_evictions, [])
def test_mamba_pool_cpu_offload(self):
"""MambaPool.get_cpu_copy / load_cpu_copy round-trips conv and temporal state."""
_, _, req_to_token_pool, _ = self._setup_tree_and_allocator()