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

[Simon Pilgrim via llvm-commits]

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

----------------
RKSimon wrote:

```
APInt MaxLen = CR.getUnsignedMax();
if (MaxLen == 0 || MaxLen.ugt(64))
  return false;
```
Also, why the 64 magic number?

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