[PATCH] D37051: Model cache size and associativity in TargetTransformInfo

[Roman via Phabricator via llvm-commits]

gareevroman added a comment.

In https://reviews.llvm.org/D37051#850439, @grosser wrote:

> In https://reviews.llvm.org/D37051#849798, @gareevroman wrote:
>
> > > I think throughput and latency of vector fma instructions are pretty constant across micro-architectures too. Can we also add them?
> >
> > Sorry, probably, it’d require to specify it for each architecture.
>
>
> We could do this. Some of the backend experts might be able to help you how to do this best. I think we should do this in a separate patch. Care to propose one?


Yes, I will propose it. Sometimes it can be difficult to find this information, since it's not always publicly available.


https://reviews.llvm.org/D37051