Capture CUDA logging output (#7298)

* logging: output capture in cuda module

* fix compile error

* fix: vsnprintf terminates with 0, string use not correct

* post review

* Update llama.cpp

Co-authored-by: slaren <slarengh@gmail.com>

* Update llama.cpp

Co-authored-by: slaren <slarengh@gmail.com>

---------

Co-authored-by: slaren <slarengh@gmail.com>

1 files changed, 4 insertions, 0 deletions

diff --git a/llama.cpp b/llama.cpp
index 2b91eec8..1409a05d 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -1697,6 +1697,8 @@ struct llama_state {
     llama_state() {
 #ifdef GGML_USE_METAL
         ggml_backend_metal_log_set_callback(log_callback, log_callback_user_data);
+#elif defined(GGML_USE_CUDA)
+        ggml_backend_cuda_log_set_callback(log_callback, log_callback_user_data);
 #endif
     }
 
@@ -18174,6 +18176,8 @@ void llama_log_set(ggml_log_callback log_callback, void * user_data) {
     g_state.log_callback_user_data = user_data;
 #ifdef GGML_USE_METAL
     ggml_backend_metal_log_set_callback(g_state.log_callback, g_state.log_callback_user_data);
+#elif defined(GGML_USE_CUDA)
+    ggml_backend_cuda_log_set_callback(g_state.log_callback, g_state.log_callback_user_data);
 #endif
 }