RFR [14]: 8231707: Improve Mutex inlining

Robbin Ehn robbin.ehn at oracle.com
Wed Oct 2 16:07:56 UTC 2019


Hi Claes,

On 2019-10-02 16:56, Claes Redestad wrote:
> Webrev: http://cr.openjdk.java.net/~redestad/8231707/open.00/

Looks good, thanks!

/Robbin

> Bug:    https://bugs.openjdk.java.net/browse/JDK-8231707
> 
> Robbin Ehn and I also did a more elaborate experiment[2] to make
> Mutex::lock/try_lock/... inlineable in full. This interestingly does
> slightly worse on the targetted micro, while increasing the size
> of the JVM more aggressively.
> 
> Test: tier1-3
> 
> Thanks!
> 
> /Claes
> 
>   [1]
>   Reference.hasReferencePendingList will use a mutex in the VM:
>   http://cr.openjdk.java.net/~redestad/8231707/ReferenceBench.java
> 
>             Mode  Cnt   Score   Error  Units
>   baseline  avgt   20  70.913 ± 2.381  ns/op
>   patch     avgt   20  61.014 ± 3.545  ns/op
> 
>   [2] http://cr.openjdk.java.net/~redestad/8231707/alternative.00
>             avgt   20  63.150 ± 4.135  ns/op


More information about the hotspot-runtime-dev mailing list