[llvm] [AMDGPU] (xor (cmp_eq x, 1), -1) -> cmp_neq x, 1 (PR #133698)

[Ana Mihajlovic via llvm-commits]

mihajlovicana wrote:

> Shouldn't this be a DAG combine instead of a ISel pattern?

Because divergent i1 trunc is selected into and and cmp which is target specific

https://github.com/llvm/llvm-project/pull/133698