[Kernel] Use flash-attn for decoding (#3648)

Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Co-authored-by: LiuXiaoxuanPKU <lilyliupku@gmail.com>

vllm/attention/selector.py

@@ -93,6 +93,20 @@ def _which_attn_to_use(
                     "torch.float16 or torch.bfloat16.")
         return _Backend.XFORMERS
 
+    if kv_cache_dtype is not None and kv_cache_dtype.startswith("fp8"):
+        logger.info("Cannot use FlashAttention-2 backend for FP8 KV cache.")
+        return _Backend.XFORMERS
+
+    if block_size % 16 != 0:
+        logger.info("Cannot use FlashAttention-2 backend for block size not "
+                    "divisible by 16.")
+        return _Backend.XFORMERS
+
+    if sliding_window is not None:
+        logger.info(
+            "Cannot use FlashAttention-2 backend due to sliding window.")
+        return _Backend.XFORMERS
+
     try:
         import vllm_flash_attn  # noqa: F401
     except ImportError: