[llvm] [SystemZ] Don't lower float/double ATOMIC_[LOAD|STORE] to [LOAD|STORE] (PR #75879)

[Jonas Paulsson via llvm-commits]

JonPsson1 wrote:

Patch updated:

- Don't return false for atomic memops in storeLoadIsAligned().
- Treat ATOMIC_LOADs the same as regular LOADs in isVectorElementLoad().
- The TernaryRXF instructions are atomic - remove the nonatomic_ld operator.


https://github.com/llvm/llvm-project/pull/75879