[Feature] Support KV Cache Storage (#5571)

* Support Mooncake Store

* up

* up

* add op

* fix conflict

* fix error

* up for comments

* avoid thread lock

* up

* fix unittest

* fix unittest

* remove debug info

* consider tp_size > 1

* add default rdma_nics

* add utils

* up

* fix error

---------

Co-authored-by: YuBaoku <49938469+EmmonsCurse@users.noreply.github.com>

custom_ops/setup_ops.py

@@ -288,6 +288,7 @@ elif paddle.is_compiled_with_cuda():
         "gpu_ops/tune_cublaslt_gemm.cu",
         "gpu_ops/swap_cache_batch.cu",
         "gpu_ops/swap_cache.cu",
+        "gpu_ops/swap_cache_layout.cu",
         "gpu_ops/step_system_cache.cu",
         "gpu_ops/cpp_extensions.cc",
         "gpu_ops/share_external_data.cu",