[llvm-dev] [RFC] Polly Status and Integration

[Renato Golin via llvm-dev]

On 20 October 2017 at 22:26, Saito, Hideki <hideki.saito at intel.com> wrote:
> Please continue to monitor our future patches as we expand VPInstruction usage (we'll be sure to include you) and provide valuable feedback.

Will do, thanks!


> FWIW, for example, at this point, we don't think it's feasible to derive VPInstruction from llvm::User, e.g., since we have a need to create VPUser class
> instances, but as we evolve VPInstruction usage in future patches, we'll look into such a possibility and come back to report our findings.

Indeed, and the VP* classes are, so far, simple and unique enough that
we can cope with it. It will be much easier to experiment and grow
around the current set of classes than try to map all possible routes
now and get a very stiff representation that not only we can't do
anything we want but involves cloning large portions of memory for no
benefit.

cheers,
--renato