[llvm] [PreISelIntrinsicLowering] Lower bounded memcpy/memmove to masked load/store (PR #212710)

[via llvm-commits]

================

----------------
Harishankar14 wrote:

 @arsenm  `memcpy_bound_zero` covers it ,`assume(size < 1)` gives a range of [0,1), so getUnsignedMax() is zero and the first arm declines. I've reworded the comment there to name the branch, since it wasn't obvious from the test alone

https://github.com/llvm/llvm-project/pull/212710