From c92135111edcaed3240d355ff6fd48ae4ca359ee Mon Sep 17 00:00:00 2001 From: root Date: Tue, 18 Aug 2026 06:44:01 +0000 Subject: [PATCH] =?UTF-8?q?[fix]=20reshape=5Fpaged=5Fcache=E5=8F=82?= =?UTF-8?q?=E6=95=B0=E9=A1=BA=E5=BA=8F?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ex_engine/python/xllm_ops.py | 4 ++-- qwen3_6_scripts/ex_engine/python/xllm_ops.py | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/ex_engine/python/xllm_ops.py b/ex_engine/python/xllm_ops.py index 118dead0..d8aa3d68 100644 --- a/ex_engine/python/xllm_ops.py +++ b/ex_engine/python/xllm_ops.py @@ -122,8 +122,8 @@ def gelu_and_mul(input, output=None): # --- Cache (xllm_cache.so: reshape_paged_cache) --- def reshape_and_cache(key, value, key_cache, value_cache, slot_mapping): """Write KV to paged cache. .so export: reshape_paged_cache.""" - return _get("xllm_cache").reshape_paged_cache(key, value, key_cache, - value_cache, slot_mapping) + return _get("xllm_cache").reshape_paged_cache(slot_mapping, key, value, + key_cache, value_cache) # --- Attention (ix_moe_bridge.so: ix_paged_attention) --- def paged_attention(out, query, key_cache, value_cache, diff --git a/qwen3_6_scripts/ex_engine/python/xllm_ops.py b/qwen3_6_scripts/ex_engine/python/xllm_ops.py index 118dead0..d8aa3d68 100644 --- a/qwen3_6_scripts/ex_engine/python/xllm_ops.py +++ b/qwen3_6_scripts/ex_engine/python/xllm_ops.py @@ -122,8 +122,8 @@ def gelu_and_mul(input, output=None): # --- Cache (xllm_cache.so: reshape_paged_cache) --- def reshape_and_cache(key, value, key_cache, value_cache, slot_mapping): """Write KV to paged cache. .so export: reshape_paged_cache.""" - return _get("xllm_cache").reshape_paged_cache(key, value, key_cache, - value_cache, slot_mapping) + return _get("xllm_cache").reshape_paged_cache(slot_mapping, key, value, + key_cache, value_cache) # --- Attention (ix_moe_bridge.so: ix_paged_attention) --- def paged_attention(out, query, key_cache, value_cache,