Kcpp triage for rowsplit: revert https://github.com/ggml-org/llama.cpp/pull/16715 until https://github.com/ggml-org/llama.cpp/issues/16799 is resolved

revert https://github.com/ggml-org/llama.cpp/pull/16715 (+2 squashed commit)

Squashed commit:

[289af2ee2] Revert "Hide latency of bias and gate-loading (#16847)"

This reverts commit 8b11deea46.

[a3e5c1e95] Revert "CUDA: add unused vars to mmvf and mmvq (#16807)"

This reverts commit 463bbf20bf.

src/llama-graph.cpp

@@ -810,9 +810,6 @@ ggml_tensor * llm_graph_context::build_ffn(
             GGML_ABORT("fatal error");
     }
 
-    //expand here so that we can fuse ffn gate
-    ggml_build_forward_expand(gf, cur);
-
     if (gate && type_gate == LLM_FFN_PAR) {
         cur = ggml_mul(ctx0, cur, tmp);
         cb(cur, "ffn_gate_par", il);
@@ -1093,9 +1090,6 @@ ggml_tensor * llm_graph_context::build_moe_ffn(
             GGML_ABORT("fatal error");
     }
 
-    //expand here so that we can fuse ffn gate
-    ggml_build_forward_expand(gf, cur);
-
     experts = build_lora_mm_id(down_exps, cur, selected_experts); // [n_embd, n_expert_used, n_tokens]
     cb(experts, "ffn_moe_down", il);