llama : use n_swa + n_ubatch cells for SWA cache (#13833)

* llama : use n_swa + n_ubatch cells for SWA cache

ggml-ci

* llama : add warning about multi-sqeuence SWA contexts

src/llama-kv-cache.h

@@ -339,7 +339,7 @@ public:
                          bool   swa_full,
                      uint32_t   kv_size,
                      uint32_t   n_seq_max,
-                     uint32_t   n_batch,
+                     uint32_t   n_ubatch,
                      uint32_t   n_pad);
 
     ~llama_kv_cache_unified_iswa() = default;