Preserve decode suffix KV locations after cache hits

Decode queue compaction receives req_to_token rows after the prefill side has already populated cached prefix slots.  Cache-hit requests therefore need the extend/suffix slice, not the leading prefix slice, when building the prebuilt transfer chunk.

Constraint: Prefill/decode disaggregation shares req_to_token rows across cached prefix and new suffix positions.

Rejected: Keep slicing from zero | cache-hit requests would copy prefix KV locs into the prebuilt suffix chunk.

Confidence: medium

Scope-risk: narrow

Directive: Do not change prepare_for_prebuilt slicing without testing cache-hit req_to_token layouts.

Tested: python -m py_compile on changed runtime files.

Not-tested: Local pytest blocked before collection by missing orjson dependency.
(cherry picked from commit 416112b617fabe71e8cff7484794af73f3e84440)

python/sglang/srt/disaggregation/decode_schedule_batch_mixin.py

@@ -50,7 +50,7 @@ class ScheduleBatchDisaggregationDecodeMixin:
 
             pre_len = len(req.prefix_indices)
             chunk = self.req_to_token_pool.req_to_token[req.req_pool_idx][
-                : req.extend_input_len
+                pre_len : pre_len + req.extend_input_len
             ]
             assert (
                 offset + req.extend_input_len <= total_size
@@ -96,7 +96,7 @@ class ScheduleBatchDisaggregationDecodeMixin:
                         seq_len,
                         int(getattr(req, "cached_tokens", 0) or 0),
                         req.req_pool_idx,
-                        0,
+                        pre_len,
                         pre_len,
                         int(chunk.numel()),
                     )