Hi
I have two more questions
> 1) How to specify a customized CUDA codegen when building/compiling the model?
> 2) How to output/print the code generated by a customized CUDA codegen?
---
[Visit
Topic](https://discuss.tvm.apache.org/t/generating-gpu-code-in-tvm-and-check-out-the-generated-
OK I solve this problem, it is really stupid. It is caused by my unfamiliarity
of clion ide.
When I'm using Clion IDE, It will set up a doc called "/cmake-build-debug".
It's like our file "/build".
However , it lack of the file "config.cmake"
I guess it is my CLION IDE cmake-debug goes wrong?
It does not shows that llvm has been found!
when I use Ctest build/cpptest in terminal, it is totally ok! no bug, it proves
that is caused by IDE cmake-d
I want to confirm that I make my LLVM support indeed, so I add a random string
in src/target/llvm/llvm_module.cc, and remake TVM, it shows that I do compile
with this llvm_module.cc file.
pay attention to my compliation process. It recompile llvm_module.cc
How can I fix it?
My LLVM system path is quiet correct, and I do make config.cmake as tutorial.
