Hi all. I’m wondering how I can hook a debugger into the generated LLVM code created by TVM, when TVM runs the code. Is there an easy way to do this?
Set breakpoints in codegen_llvm.cc
Insert a call to an external function, say dummy_foo in the LLVM IR that you want to debug (in codegen), then run it in a debugger with a breakpoint set on dummy_foo.