[Openmp-dev] [LLVMdev] [cfe-dev] Provide LLVM IR and OpenMP LLVM IR as input in a Pass

Hal Finkel hfinkel at anl.gov
Tue Oct 21 13:47:01 PDT 2014


----- Original Message -----
> From: "Simone Atzeni" <simone.at at gmail.com>
> To: "Hal Finkel" <hfinkel at anl.gov>
> Cc: llvmdev at cs.uiuc.edu, cfe-dev at cs.uiuc.edu, openmp-dev at dcs-maillist2.engr.illinois.edu
> Sent: Tuesday, October 21, 2014 3:42:03 PM
> Subject: Re: [LLVMdev] [cfe-dev][Openmp-dev] Provide LLVM IR and OpenMP LLVM IR as input in a Pass
> 
> 
> Hi Hal,
> 
> 
> thanks for your answer.
> 
> 
> 
> 
> I think it would be better if you gave us some idea of what you're
> trying to do. clang's OpenMP implementation turns the OpenMP pragmas
> into function calls to the runtime library. You can run passes on
> this IR as you would any other IR. When you say, "Do some operation
> on it and get results" do you mean generating code, linking and
> executing that code?
> 
> -Hal
> 
> 
> 
> I want to use the pass Polly to find any data dependence in the code.
> Since it does not work on OpenMP,

Why not? The loops in the outlined OpenMP regions are just regular loops (with runtime bounds). Can polly not handle them?

> I was thinking to apply it on the
> not-OpenMP IR, then I need the results of Polly to apply other
> passes to the OpenMP IR.

I don't think this will work for the OpenMP'd loops. The IR will look completely different, be in different functions, etc.

 -Hal

> 
> 
> That is why I need both.
> 
> 
> Best,
> Simone
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> I hope this is clear and you can give me some suggestion.
> 
> 
> 
> 
> 
> Thanks.
> 
> 
> Best Regards,
> Simone
> 
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
> 
> 
> --
> Hal Finkel
> Assistant Computational Scientist
> Leadership Computing Facility
> Argonne National Laboratory
> 

-- 
Hal Finkel
Assistant Computational Scientist
Leadership Computing Facility
Argonne National Laboratory



More information about the Openmp-dev mailing list