[BugFix] Fix mla cpu - missing 3 required positional arguments (#17494)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

Author: LucasWilkinson

vllm/_ipex_ops.py

@@ -177,7 +177,7 @@ def varlen_attention(
         out: torch.Tensor,
         seqlen_q: torch.Tensor,
         seqlen_k: torch.Tensor,
-        alibi_slopes: torch.Tensor,
+        alibi_slopes: Optional[torch.Tensor],
         max_seqlen_q: int,
         max_seqlen_k: int,
         pdropout: float,
@@ -193,6 +193,8 @@ def varlen_attention(
         if ipex.__version__.endswith("cpu"):
             if logits_soft_cap != 0.0:
                 raise ValueError("IPEX CPU does not support logits_soft_cap")
+            assert alibi_slopes is None
+            assert window_size_left < 0 and window_size_right < 0
             ipex.llm.functional.varlen_attention(query.contiguous(),
                                                  key.contiguous(),
                                                  value.contiguous(), out,

vllm/attention/backends/cpu_mla.py

@@ -273,6 +273,9 @@ def _forward_prefill(
             return_softmax=False,
             gen_=None,
             logits_soft_cap=0.0,
+            window_size_left=-1,
+            window_size_right=-1,
+            alibi_slopes=None,
         )
 
         # remove padding