llama.cpp

[CANN]: Fix ggml_backend_cann_buffer_get_tensor

#8871

Merged