In function `MacroAssembler::cmpxchgw` and `MacroAssembler::cmpxchgptr`, when turn on `UseLSE`, there is a `dmb`
 instruction generated after `casal` instruction.

The `casal` means a CAS operate with both load-acquire and store-release semantics.

It looks like the subsequent dmb is redundant.

The code is bellow:
```
  if (UseLSE) {
    mov(tmp, oldv);
    casal(Assembler::xword, oldv, newv, addr);
    cmp(tmp, oldv);
    br(Assembler::EQ, succeed);
    membar(AnyAny);
  }
```