RFR: Don't require barriers for LoadKlass in C2

[Aleksey Shipilev]

On 5/22/19 1:09 PM, Roman Kennke wrote:
> Aleksey spotted a diff in sh/jdk vs jdk/jdk yesterday, which appears to
> be a left-over from earlier attempts to put the forwarding pointer in
> Klass* slot. Back then I needed to force barriers before loading Klass*.
> However that is not needed anymore, we should be good loading Klass*
> from from-space copy because it is immutable.
> 
> http://cr.openjdk.java.net/~rkennke/loadklass-no-barriers/webrev.00/

Looks good.

-Aleksey