[llvm] [RISCV] Adjust select shuffle cost to reflect mask creation cost (PR #77963)

[Luke Lau via llvm-commits]

lukel97 wrote:

> The prior lowering scaled the cost of the bitmask with LMUL, which I don't understand. At m1 it did use the same base cost of 2. (@lukel97 You wrote the original code here, anything I'm missing here?)

Nope, looks like an oversight. AFAIK the vmerge.vvm is the only op that should be scaled by LMUL

https://github.com/llvm/llvm-project/pull/77963