Error report for no GPU

Development
#1

I am learning How to optimize convolution on GPU — tvm 0.16.dev0 documentation

But If I don’t have GPU, TVM doesn’t report this situation. Even doesn’t report where is not implemented. target = "llvm" 

Traceback (most recent call last):
  File "try.py", line 83, in <module>
    func = tvm.build(s, [A,W,B], target=target, name="mmult")

    rt_mod_host = _driver_ffi.tir_to_runtime(annotated_mods, target_host)

    raise_last_ffi_error()

    raise py_err
tvm._ffi.base.TVMError: Traceback (most recent call last):
TVMError: not implemented