u99127 edited a comment on pull request #7060:
URL: https://github.com/apache/tvm/pull/7060#issuecomment-892130966


   are there any instructions on how to build mlir-hlo and llvm to experiment 
with this PR ? I'm personally curious to see how this works but I'm not 
observing any instructions as to how to configure mlir-hlo project alongside 
this work. 
   
   I've tried the build instructions from here 
https://github.com/tensorflow/mlir-hlo 
   
   > build_tools/build_mlir.sh ${PWD}/llvm-project/ ${PWD}/llvm-build
   
   Then to build this PR, I rebased the config.cmake with trivial rebasing 
issues and then set
   
   set(USE_MLIR 
/home/ramana/work/tvm/tvm-mlir-trial-with-hlo/mlir-hlo/mlir-hlo/llvm-build)
   # Whether to use MLIR HLO
   set(USE_MHLO ON)
   
   My config.cmake has : 
   
   set(USE_MLIR 
/home/ramana/work/tvm/tvm-mlir-trial-with-hlo/mlir-hlo/mlir-hlo/llvm-build)
   
   # Whether to use MLIR HLO
   set(USE_MHLO ON)
   
   I realised I also needed to set LLVM_PROJ_SRC and LLVM_PROJ_BUILD in my 
environment to try and build this which I've set to 
   
   
LLVM_PROJ_SRC=/home/ramana/work/tvm/tvm-mlir-trial-with-hlo/mlir-hlo/mlir-hlo/
   
LLVM_PROJ_BUILD=/home/ramana/work/tvm/tvm-mlir-trial-with-hlo/mlir-hlo/mlir-hlo/llvm-build/
   
   
   And then doing a cmake .. gives me the following error.
   
   
   CMake Error at cmake/modules/contrib/MLIR.cmake:73 (message):
     MLIREDSC not found, did you forget to build llvm-project?
   Call Stack (most recent call first):
     cmake/modules/contrib/MLIR.cmake:87 (find_mlir_lib)
     CMakeLists.txt:413 (include)
   
   
   regards
   Ramana
   
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to