RFR (XS) 8216049: stringTable::intern creates redundant String when looking up existing one

[gerard ziemski]

hi Aleksey,

The change looks reasonable and the code inspection seems to corroborate 
that an improvement should be expected, but I'd like to know whether we 
have any hard numbers from current benchmarks that validate the intended 
goal was achieved and by how much?


cheers

On 2/15/19 1:20 PM, Aleksey Shipilev wrote:
> RFE:
>   https://bugs.openjdk.java.net/browse/JDK-8216049
> 
> Fix:
>   http://cr.openjdk.java.net/~shade/8216049/webrev.01/
> 
> This is the fix for performance regression in 11 that shows up in stack trace generation benchmarks.
> See bug and linked issues for more info. The original patch is suggested by Claes. I have been
> looking how to improve it, but eventually backed off any improvements on top of it. The simplicity
> of this patch plays well with future backports to 12u and 11u to fix the performance regression there.
> 
> Testing: tier1, benchmarks, jdk-submit
> 
> Thanks,
> -Aleksey
> 
>