[llvm] [AMDGPU][GlobalISel] Reduce instruction count for store in global addrspace for non-powers-of-2 (PR #206043)

[Rajveer Singh Bharadwaj via llvm-commits]

Rajveer100 wrote:

@arsenm 
I just realised there's already a splitting mechanism in `LegalizerHelper`.

So do we just need a filter here to also include `G_STORE` for `fewerElements`?

Let me know if I am in the right direction.

https://github.com/llvm/llvm-project/pull/206043