[llvm] [Analysis][RISCV] More accurately estimate the cost of strided vector loads (PR #175135)

[Luke Lau via llvm-commits]

lukel97 wrote:

> @lukel97 this never got merged back when, any reason to not merge it in now?

Woops, sorry for missing all your pings. I think this PR still makes sense. I'll just check to make sure that making strides cheaper doesn't cause us to no longer form interleave groups in the LV

https://github.com/llvm/llvm-project/pull/175135