Add Support for Page Size greater than 1 for Flashinfer MLA Backend (#8593)

Signed-off-by: Pavani Majety <pmajety@nvidia.com>

python/sglang/srt/server_args.py

@@ -636,6 +636,10 @@ class ServerArgs:
                     logger.warning(
                         "DeepSeek MTP does not require setting speculative_draft_model_path."
                     )
+                if self.page_size != 1 and self.attention_backend == "flashinfer":
+                    raise ValueError(
+                        "Speculative decoding with page_size != 1 is not supported. Please set page_size to 1."
+                    )
 
             # Auto choose parameters
             if self.speculative_num_steps is None: