ssquare08 added a comment.

In D125669#3604452 <https://reviews.llvm.org/D125669#3604452>, @ABataev wrote:

> In D125669#3604302 <https://reviews.llvm.org/D125669#3604302>, @RitanyaB 
> wrote:
>
>> In D125669#3599134 <https://reviews.llvm.org/D125669#3599134>, @ABataev 
>> wrote:
>>
>>> What about the description? Shall we still emit ` if(0)`?
>>
>> Can you please elaborate for more clarity?
>
> The description of the patch 
> So, this
>
> #pragma omp target in_reduction(+:res)
>
>   for (int i=0; i<N; i++) {
>     res = res+i
>   }
>
> will become
>
> #pragma omp task in_reduction(+:res) if(0)
> #pragma omp target map(res)
> for (int i=0; i<N; i++) {
>
>   res = res+i
>
> }
>
> So, the question is is this still true?

The idea here is to wrap the target task with a host task and move in_reduction 
to that new host task with if(0) condition, making it an undeferred task. The 
approach here is basically using the framework that is in place for the host 
in_reduction.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D125669/new/

https://reviews.llvm.org/D125669

_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to