[llvm] [AMDGPU] (x or y) xor -1 -> x nor y (PR #130264)

[Petar Avramovic via llvm-commits]

petar-avramovic wrote:

Title should clarify that this is pattern for divergent i1/lane mask specifically,  pattern for uniform i32/i64 already exists. Also constant should not be `-1` but `true` instead since it its type is `i1`?

https://github.com/llvm/llvm-project/pull/130264