[LLVMdev] parallel loop metadata simplification

[Redmond, Paul]

I've started a new thread on cfe-commits.

On 2013-03-05, at 12:52 PM, Hal Finkel wrote:

> ----- Original Message -----
>> From: "Pekka Jääskeläinen" <pekka.jaaskelainen at tut.fi>
>> To: "Paul Redmond" <paul.redmond at intel.com>
>> Cc: "Hal Finkel" <hfinkel at anl.gov>, "Tobias Grosser" <tobias at grosser.es>, "<llvmdev at cs.uiuc.edu>"
>> <llvmdev at cs.uiuc.edu>
>> Sent: Tuesday, March 5, 2013 11:49:28 AM
>> Subject: Re: [LLVMdev] parallel loop metadata simplification
>> 
>> On 03/05/2013 07:12 PM, Redmond, Paul wrote:
>>> Attached is my most recent patch for clang. Maybe someone wants to
>>> play with
>>> it or has ideas on how to refine the llvm.mem metadata generation.
>> 
>> +    } else if (I->mayReadOrWriteMemory())
>> +      I->setMetadata("llvm.mem.parallel_loop_access",
>> +                     L.GetLoopID(I->getContext()));
>> +  }
>> 
>> Here we want to not add metadata to load/store
>> instructions which originate from unsafe allocas
>> like discussed.
>> 
>> I wonder if there's an easier way to find this out than to follow
>> the pointer argument until a global or a function argument is found?
>> Maybe that can be done via a helper method which can be later
>> augmented
>> with more "safe cases"?
> 
> If we're discussing a Clang patch, this discussion should be moved to the cfe-commits list.
> 
> -Hal
> 
>> 
>> --
>> --Pekka
>>