[PATCH] Add HLE target feature

[Nadav Rotem]

Hi Jeffrey, 

> it's an extra argument passed through some layers and an extra level of arrays in
> some other layers.

This patch adds complexity to multiple layers of the compiler, and I would like to understand the motivation before we add this complexity (and compile time). 

> Your first message said you were judging by the
> number of patches, not the complexity of each patch, which is an
> incredibly short-sighted way to judge changes: you're explicitly
> discouraging contributors from splitting their patches into bite-sized
> pieces.

I agree. The number of patches is not a good indication of the complexity of the change. It only tells you how many different components are affected. 

> 
> I don't think we can judge whether the cost of HLE primitives as
> metadata is worth it unless there's a spec in terms of the LLVM
> instruction set, so that's what I asked Michael to provide. It's
> possible it will be better to ask him to implement x86 intrinsics
> (http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/IR/IntrinsicsX86.td?view=markup)
> instead, but we can't know without the IR-level proposal.

Thanks,
Nadav

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130227/f7d9bb50/attachment.html>