[v5 PATCH] arm64: mm: force write fault for atomic RMW instructions

[Ryan Roberts]

>> I don't think that really applies in your migration
>> case, because you are definitely going to write all the memory eventually, I
>> think?
> 
> Especially with memory ballooning and things like free-page-reporting we might
> be migrating a whole bunch of zero-memory and only have to make sure that the
> destination is actually zero. We don't want to consume memory.

Ahh you know if the contents of a page being migrated is all zero. In that case,
I get it. Thanks!