RFR(T): 8217782: Spill detection broken after JDK-8217716
Tobias Hartmann
tobias.hartmann at oracle.com
Fri Jan 25 12:54:43 UTC 2019
Hi Claes,
looks good.
Best regards,
Tobias
On 25.01.19 13:51, Claes Redestad wrote:
> Hi,
>
> in a recent code cleanup I accidentally removed setting of
> lrg._was_spilled1/2, missing a few cases where they were actually in
> use. This cause a severe performance degradation in some benchmarks like
> Octane Richards.
>
> Restore calculation of lrg._was_spilled1/2.
>
> Bug: https://bugs.openjdk.java.net/browse/JDK-8217716
> Webrev: http://cr.openjdk.java.net/~redestad/8217782/open.00/
>
> Testing: verified locally that performance in Octane Richards is back to
> normal levels, running tier1+2
>
> Thanks!
>
> /Claes
>
More information about the hotspot-compiler-dev
mailing list