kv-cache : remove LLAMA_SET_ROWS checks (llama/15505)

ggml-ci

diff --git a/src/ggml-cann/common.h b/src/ggml-cann/common.h
index 33794062f565d5b7473f946a4206d6002c407463..88cc3f481ed3832c191d1a5e8f156677d23669f7 100755 (executable)
--- a/src/ggml-cann/common.h
+++ b/src/ggml-cann/common.h
@@ -374,7 +374,6 @@ struct ggml_backend_cann_context {
 #endif
     cann_task_queue task_queue;
     bool async_mode;
-    bool support_set_rows;
     // Rope Cache
     void* rope_init_ptr = nullptr;
     void* rope_sin_ptr = nullptr;
@@ -400,14 +399,6 @@ struct ggml_backend_cann_context {
         async_mode = parse_bool(get_env("GGML_CANN_ASYNC_MODE").value_or(""));
         GGML_LOG_INFO("%s: device %d async operator submission is %s\n", __func__,
             device, async_mode ? "ON" : "OFF");
-
-        support_set_rows = parse_bool(get_env("LLAMA_SET_ROWS").value_or(""));
-        GGML_LOG_INFO("%s: LLAMA_SET_ROWS is %s\n", __func__, support_set_rows ? "ON" : "OFF");
-
-        if (!support_set_rows) {
-            GGML_LOG_INFO("%s: CANN Graph currently only supports execution when LLAMA_SET_ROWS is ON. "
-                    "Falling back to eager mode.\n", __func__);
-        }
     }
 
     /**

diff --git a/src/ggml-cann/ggml-cann.cpp b/src/ggml-cann/ggml-cann.cpp
index 81215425618a35466430ce24d542db2a15c788e7..558121dff780b8e536e9ebce62c3a1a43e3c09ec 100755 (executable)
--- a/src/ggml-cann/ggml-cann.cpp
+++ b/src/ggml-cann/ggml-cann.cpp
@@ -2251,11 +2251,6 @@ static enum ggml_status ggml_backend_cann_graph_compute(
     bool use_cann_graph = true;
     bool cann_graph_update_required = false;
 
-    // check environment LLAMA_SET_ROWS
-    if (!cann_ctx->support_set_rows) {
-        use_cann_graph = false;
-    }
-
     if (use_cann_graph) {
         if (cann_ctx->cann_graph == nullptr) {
             cann_ctx->cann_graph.reset(new ggml_cann_graph());