[Mlir-commits] [mlir] [mlir][vector] add tensor.concat, bitcast, expand_shape, collapse_shape vectorization support (PR #97297)

[Andrzej Warzyński]

banach-space wrote:

Sorry, I've been travelling and didn't have the time to reply earlier. 

Basically, I agree with all the points from @hanhanW, thanks! I think that we should understand the motivation a bit better before proceeding with this.

> Is there a RFC or discussion about why we are adding vectorization patterns for these ops? 

I believe it is this one:
* https://discourse.llvm.org/t/rfc-vector-distribution-for-cpu-convert-vector-to-physical-register-size-vector/

We haven't really agreed on any specific path forward. 

https://github.com/llvm/llvm-project/pull/97297