Codereview request for 7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])

[Alan Bateman]

On 16/07/2012 19:42, Xueming Shen wrote:
> Hi Alan,
>
> I gave the "continue ->else if " a try, it appears the server vm 
> obviously likes
> the "continue" approach (it is consistent with what we experienced in 
> the past
> when we did similar approach for ascii, in which we separate/singled 
> the ascii
> path out). So I guess we probably want to keep the continue here.
>
> Here are the data. dbcs2_new is to replace the continue with else if 
> and the
> dbcs2 is the "continue".
>
> http://cr.openjdk.java.net/~sherman/7183053/dbcs2_new
> http://cr.openjdk.java.net/~sherman/7183053/dbcs2
>
> -Sherman
Thanks for checking, I had expected it would ultimately get compiled to 
the same code but it seems not.

In that case the changes in the webrev look fine to me.

-Alan